Back to EPA - Global Climate Change [The Potential for Global Climate Change on the United States 12/89]

Ask the Scholar

Document scope · 1 page

doc

Scholar

Ask about this object, its catalog metadata, its source description, or the page inventory. For page-specific OCR and visual context, open one of the page chats.

Source Description

These records pertain to Global Climate Change.

Scholar Source Context

Document identity

localId

285411055

label

EPA - Global Climate Change [The Potential for Global Climate Change on the United States 12/89]

core

doc

dtoType

document

citationUrl

https://catalog.archives.gov/id/285411055

pageCount

Source metadata

285411055

sourceUrl

https://catalog.archives.gov/id/285411055

contentType

document

title

EPA - Global Climate Change [The Potential for Global Climate Change on the United States 12/89]

description

These records pertain to Global Climate Change.

citationUrl

https://catalog.archives.gov/id/285411055

identifierLocal

06683-004

collections

Records of the White House Office of Policy Development (George H. W. Bush Administration)

Ed Goldstein Subject Files

iiifBase

https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf

thumbnailUrl

https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf

largeImageUrl

https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf

imageCount

hasImages

yes

source

import

hasTranscription

Source extras

url

https://catalog.archives.gov/id/285411055

naId

285411055

levelOfDescription

fileUnit

recordType

description

ocrSource

nara-archive

Single page context

seq

pageIndex

type

document

url

https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf

mediaId

3ffdbb4b50d729a1

ocrText

Originally Processed With FOIA(s): FOIA Number: 2017-0310-F 2017-0310-F FOIA MARKER This is not a textual record. This is used as an administrative marker by the George Bush Presidential Library Staff. Record Group/Collection: George H.W. Bush Presidential Records Collection/Office of Origin: Policy Development, White House Office of Series: Goldstein, Ed, Files Subseries: OA/ID Number: 06683 Folder ID Number: 06683-004 Folder Title: EPA - Global Climate Change [The Potential for Global Climate Change on the United States 12/89] Stack: Row: Section: Shelf: Position: G 22 29 4 United States Policy, Planning, EPA-230-05-89-050 Environmental Protection And Evaluation December 1989 Agency (PM-221) EPA The Potential Effects Of Global Climate Change On The United States Printed on Recycled Paper THE POTENTIAL EFFECTS OF GLOBAL CLIMATE CHANGE ON THE UNITED STATES REPORT TO CONGRESS Editors: Joel B. Smith and Dennis Tirpak United States Environmental Protection Agency Office of Policy, Planning and Evaluation Office of Research and Development December 1989 TABLE OF CONTENTS Page Foreword xxi Acknowledgments xxiii EXECUTIVE SUMMARY XXV CHAPTER 1: INTRODUCTION 1 CONGRESSIONAL REQUEST FOR REPORTS 1 GOALS OF THIS REPORT 2 Sensitivities 2 Direction and Magnitude 2 Linkages 2 National Impacts 2 Regional Impacts 3 Uncertainties 3 Policy Implications 3 Research Needs 3 STRUCTURE OF THE ANALYSIS 3 Important Systems 3 Regional Case Studies 3 National Studies 4 ANALYTIC APPROACHES 4 PROCESS FOR CONDUCTING THIS REPORT 4 Step 1: Initial Scoping of the Report 4 Step 2: Preparatory Workshops 5 Step 3: Identification of Potential Projects 5 Step 4: Reviews of Proposals 5 Step 5: Planning and Integration 5 Step 6: Analysis 5 Step 7: Preliminary Project Review 5 Step 8: Project and Report Peer Review 5 STRUCTURE OF THIS REPORT 6 RELATIONSHIP TO CURRENT NATIONAL AND INTERNATIONAL ACTIVITIES 6 National Research and Policy Activities 6 International Activities 6 REFERENCES 7 CHAPTER 2: GLOBAL CLIMATE CHANGE 9 THE CLIMATE SYSTEM 10 CLIMATE FORCINGS 12 Greenhouse Gases 12 Carbon Dioxide (CO₂) 12 Methane (CH₄) 13 Chlorofluorocarbons (CFCs) 13 Nitrous Oxide (N₂O) 15 Ozone (O₃) 15 Solar Variations 15 Volcanoes 15 Tropospheric Aerosols 16 Surface Properties 16 V Quality of the Human Environment 87 Recreation 87 Wood Products 87 FOREST POLICY AND CLIMATE CHANGE 88 How Much Land Should Be Forested? 88 How Much Should Be Withdrawn From Timber Production? 88 How Should We Manage Federal Forests? 89 How Can We Ensure National Goals? 89 Reforestation 89 Who Should Pay? 90 RESEARCH NEEDS 90 Effects of Climate Change 90 Methods 90 Forest Management 91 Timing of Research 91 REFERENCES 91 CHAPTER 6: AGRICULTURE 93 FINDINGS 93 Crop Yields 93 Economic Impacts 93 Irrigation Demand 94 Agricultural Pests 94 Farm-Level Adjustments 94 Livestock Effects 94 Policy Implications 94 SENSITIVITY OF AGRICULTURE TO CHANGES IN CLIMATE 94 PREVIOUS STUDIES OF CLIMATE CHANGE AND AGRICULTURE 96 CLIMATE CHANGE STUDIES IN THIS REPORT 97 Structure of and Rationale for the Studies 97 Variability 99 Timing of Effects 99 RESULTS OF AGRICULTURAL STUDIES 100 Regional Crop Modeling Studies 100 Design of the Studies 100 Limitations 100 Results 100 Implications 102 Regional and National Economics Study 102 Study Design 103 Limitations 103 Results 104 Implications 106 Demand for Water for Irrigation 107 Irrigation Requirements in the Great Plains 107 Water Resources for Agriculture in California 108 Implications for Demand for Irrigation Water 109 Direct Effects of CO₂ on Crops 109 Climate Impacts on Pest-Plant Interactions 110 Study Design and Results 110 Limitations 111 Implications 111 Effects of Climate Change on Water Quality 111 Study Design 111 Limitations 111 viii Results 111 Implications 111 Climate Variability 112 Farm-Level Management Adjustments to Climate Change 113 Study Design 113 Results 113 Implications 113 Livestock 114 Design of Studies 114 Limitations 114 Results 114 Implications 115 ECONOMIC AND ECOLOGICAL IMPLICATIONS OF AGRICULTURAL STUDIES 116 Costs and Timing of Adjustment 116 Effects of CO, 116 Environmental Quality 116 Global Agriculture 117 POLICY IMPLICATIONS 117 Commodity Policies 117 Land-Use Programs 117 Water-Resource Management Programs 118 Water Quality Policy 118 Risk Management and Drought Policy 118 International Trade Agreements 118 Agricultural Contributions to the Greenhouse Effect 118 Agricultural Research 119 RESEARCH NEEDS 119 REFERENCES 120 CHAPTER 7: SEA LEVEL RISE 123 FINDINGS 123 Policy Implications 123 CAUSES, EFFECTS, AND RESPONSES 124 Causes 124 Effects 125 Destruction of Coastal Wetlands 125 Inundation and Erosion of Beaches and Barrier Islands 126 Flooding 127 Saltwater Intrusion 127 Responses 127 HOLDING BACK THE SEA: A NATIONAL ASSESSMENT 128 STRUCTURE OF STUDIES FOR THIS REPORT 129 SCENARIOS OF SEA LEVEL RISE 130 RESULTS OF SEA LEVEL STUDIES IN THIS REPORT 131 Loss of Coastal Wetlands and Dryland 131 Study Design 131 Limitations 132 Results 132 Costs of Defending Sheltered Shorelines 133 Study Design 133 Limitations 134 Results 134 Case Study of the Value of Threatened Coastal Property 136 Study Design 136 Limitations 136 ix CLIMATE CHANGE STUDY IN THIS REPORT 189 Study Design 189 Limitations 190 Results 191 SOCIOECONOMIC AND ENVIRONMENTAL IMPLICATIONS 194 POLICY IMPLICATIONS 196 RESEARCH NEEDS 196 REFERENCES 197 CHAPTER 11: AIR QUALITY 199 FINDINGS 199 RELATIONSHIP BETWEEN CLIMATE AND AIR QUALITY 200 Ventilation 200 Circulation 200 Precipitation 200 PATTERNS AND TRENDS IN AIR QUALITY 201 Total Suspended Particulates 201 Sulfur Dioxide 201 Ozone 202 Acid Deposition 202 STUDIES OF CLIMATE CHANGE AND AIR QUALITY 202 Climate Change and Its Interactions with Air Chemistry 205 Effect of Climate Change on Ozone Formation 205 MODELING STUDY OF CLIMATE AND AIR QUALITY 210 Study Design 210 Limitations 210 Results 211 Central California Study 211 Midwest and Southeast Study 211 Population Exposure 213 ECONOMIC, ENVIRONMENTAL, AND ECOLOGICAL IMPLICATIONS 213 Ozone 213 Acid Rain 215 POLICY IMPLICATIONS 216 RESEARCH NEEDS 216 REFERENCES 217 CHAPTER 12: HUMAN HEALTH 219 FINDINGS 219 CLIMATE-SENSITIVE ASPECTS OF HUMAN HEALTH 219 General Mortality and Illness 220 Cardiovascular, Cerebrovascular, and Respiratory Diseases 221 Vector-Borne Diseases 222 Human Reproduction 222 POTENTIAL HUMAN HEALTH EFFECTS OF CLIMATE CHANGE 222 General Mortality 223 Cardiovascular, Cerebrovascular, and Respiratory Diseases 225 Vector-Borne Diseases 226 Tick-Borne Diseases 227 Mosquito-Borne Diseases 227 Other Diseases 231 SOCIAL AND ECONOMIC IMPLICATIONS 232 POLICY IMPLICATIONS 233 xii RESEARCH NEEDS 233 REFERENCES 235 CHAPTER 13: URBAN INFRASTRUCTURE 237 FINDINGS 237 Northern and Southern Cities 237 Coastal Cities 237 Water Supply and Demand 237 Policy Implications 237 RELATIONSHIP BETWEEN URBAN INFRASTRUCTURE AND CLIMATE 238 PREVIOUS CLIMATE CHANGE STUDIES ON URBAN INFRASTRUCTURE 239 URBAN INFRASTRUCTURE STUDY IN THIS REPORT 239 RESULTS OF THE INFRASTRUCTURE STUDY 240 Impacts on Miami, Cleveland, and New York City 240 Study Design 240 Limitations 240 Results and Implications 240 Implications Arising From Other EPA Studies in This Report 245 RESULTS OF RELATED STUDIES 245 Metropolitan Water Supply 245 Washington, DC 245 New Orleans 246 New York City 246 Tucson 246 IMPLICATIONS FOR URBAN INFRASTRUCTURE 246 Water 246 Drainage and Wastewater Systems 246 Coastal Defenses 247 Roads 247 Bridges 247 Mass Transit 247 Electricity and Air-Conditioning 247 POLICY IMPLICATIONS 247 Investment Analysis Methods 247 Water Supply 248 Infrastructure Standards 248 RESEARCH NEEDS 248 REFERENCES 248 CHAPTER 14: CALIFORNIA 251 FINDINGS 251 Water Resources 251 Wetlands and Fisheries 251 Agriculture 252 Natural Vegetation 252 Air Quality 252 Electricity Demand 252 Policy Implications 252 CLIMATE-SENSITIVE RESOURCES OR CALIFORNIA 252 Current Climate 253 Water Resources 253 Water Distribution 253 Flood Control and Hydroelectric Power 256 Sacramento-San Joaquin River Delta 256 Commerce 256 xiii Agriculture 256 Forestry 256 Natural Vegetation 257 Wetlands 257 Wildlife and Fisheries 257 Recreation and Nature Preservation 257 PREVIOUS CLIMATE CHANGE STUDIES 257 Forests 257 Water Resources 258 CALIFORNIA STUDIES IN THIS REPORT 258 Analyses Performed for This Study 258 CALIFORNIA REGIONAL CLIMATE CHANGE SCENARIOS 260 RESULTS OF THE CALIFORNIA STUDIES 262 Hydrology of Catchments in the Central Valley Basin 262 Study Design 262 Limitations 262 Results 262 Implications 263 Water Resources in the Central Valley Basin 264 Study Design 264 Limitations 264 Results 264 Implications 265 Salinity in San Francisco Bay 266 Study Design 266 Limitations 266 Results 267 Implications 267 Wetlands in the San Francisco Bay Estuary 268 Study Design 268 Limitations 268 Results 268 Implications 269 California Agriculture 269 Study Design 270 Limitations 270 Results 270 Implications 271 Regional Implications of National Agriculture Changes 273 Results 273 Water Quality of Subalpine Lakes 273 Study Design 273 Limitations 273 Results 273 Implications 273 Summary of Effects on Water Resources 274 Vegetation of the Sierra Nevada 275 Study Design 275 Limitations 275 Results 276 Implications 276 Electricity Demand 277 Results 277 Implications 278 Air Pollution 278 Results 278 xiv Implications 278 POLICY IMPLICATIONS 278 Water Supply and Flood Control 278 Approaches for Modifying the Water Resource System 279 Options for Allocating Water Shortages 280 Sacramento-San Joaquin River Delta 280 Delta Island Land Use 280 Water Quality of the San Francisco Bay Estuary 281 Water Quality of Freshwater Systems 281 Terrestrial Vegetation and Wildlife 282 Agriculture 282 Wetland Vegetation and Fisheries 283 Shoreline Impacts of Sea Level Rise 283 Energy Demand 283 Air Quality 283 REFERENCES 284 CHAPTER 15: GREAT LAKES 287 FINDINGS 287 Lakes 287 Water Quality and Fisheries 287 Forests 288 Agriculture 288 Electricity Demand 288 Policy Implications 288 CLIMATE-SENSITIVE NATURAL RESOURCES IN THE GREAT LAKES REGION 289 Current Climate 289 The Lakes 290 Lake Regulation 290 Climate-Sensitive Uses of the Lakes 290 Climate and Water Quality 291 Fluctuating Lake Levels 291 Land Around the Lakes 292 Land Uses 292 PREVIOUS CLIMATE CHANGE STUDIES 292 GREAT LAKES STUDIES IN THIS REPORT 293 Direct Effects on Lakes 294 Impacts of Lake Changes on Infrastructure 294 Water Quality 294 Forests 294 Agriculture 295 Energy 295 Policy 295 GREAT LAKES REGIONAL CLIMATE CHANGE SCENARIOS 295 RESULTS OF THE GREAT LAKES STUDIES 296 Lakes 296 Lake Levels 296 Effects of Lower Lake Levels 299 Ice Cover 300 Shipping 302 Water Quality 304 Thermal Structure of Southern Lake Michigan 304 Eutrophication of the Lake Erie Central Basin 305 Fisheries 306 XV Forests 309 Potential Range Shifts 309 Transitional Effects 309 Forest Migration 311 Implications of Forest Studies 311 Agriculture 312 Crop Yields 313 Regional Shifts 314 Adjustments by Illinois Corn Producers 315 Electricity Demand 316 Study Design 316 Results 316 Implications 316 POLICY IMPLICATIONS 316 Water Supply Issues 317 Lake Regulation 317 Withdrawals 317 Shipping 317 Pollution Control 317 Fisheries 317 Land Use 318 Shorelines 318 Forestry 318 Agriculture 318 Demographic Shifts 318 REFERENCES 319 CHAPTER 16: SOUTHEAST 323 FINDINGS 323 Agriculture 323 Forests 323 Water Supplies 323 Sea Level Rise 324 Marine Fisheries 324 Electricity Demand 324 Policy Implications 324 CLIMATE AND THE SOUTHEAST 325 CLIMATE-SENSITIVE RESOURCES OF THE SOUTHEAST 325 Water Resources 325 Estuaries 326 Beach Erosion and Coastal Flooding 327 Agriculture 327 Forests 327 Indoor and Outdoor Comfort 328 PREVIOUS STUDIES OF THE IMPACTS OF CLIMATE CHANGE ON THE SOUTHEAST 328 Flooding 328 Wetlands 329 Infrastructure 329 CLIMATE CHANGE STUDIES IN THIS REPORT 329 SOUTHEAST REGIONAL CLIMATE CHANGE SCENARIOS 331 RESULTS OF SOUTHEASTERN STUDIES 333 Coastal Impacts 333 Coastal Wetlands 333 Total Coastal Land Loss 334 xvi Cost of Protecting Recreational Beaches 334 Cost of Protecting Calm-Water Shorelines 335 Tennessee Valley Authority Studies 335 TVA Modeling Study 335 Limitations 337 Results 337 Tennessee Valley Policy Study 338 Studies of the Impacts on Lake Lanier and Apalachicola Bay 339 Lake Lanier 340 Apalachicola Bay 342 Agriculture 346 Crop Modeling Study 346 Shifts in Production 348 Agricultural Pests 348 Implications of Agriculture Studies 348 Forests 350 Potential Range Shifts 350 Transitional Effects 350 Electric Utilities 352 COASTAL LOUISIANA 352 POLICY IMPLICATIONS 354 Agriculture and Forests 354 Water Resources 354 Impacts of Wetter Climate 354 Impacts of Drier Climate 354 Is Current Legislation Adequate? 355 Estuaries 355 Beach Erosion 356 REFERENCES 357 CHAPTER 17: GREAT PLAINS 359 FINDINGS 359 Agriculture 359 Ogallala Aquifer 359 Water Quality 359 Electricity Demand 359 Policy Implications 359 CLIMATE-SENSITIVE RESOURCES IN THE GREAT PLAINS 360 Dryland Agriculture 362 Irrigated Agriculture 362 Water Quality 363 Electricity Demand 363 PREVIOUS CLIMATE IMPACT STUDIES 363 GREAT PLAINS STUDIES IN THIS REPORT 364 GREAT PLAINS REGIONAL CLIMATE CHANGE SCENARIOS 364 RESULTS OF THE GREAT PLAINS STUDIES 366 Crop Production 366 Study Design 366 Limitations 366 Results 366 Implications 368 Agricultural Economics 368 Results 369 Implications 369 xvii Irrigation 370 Study Design 370 Limitations 370 Results 370 Implications 371 Water Quality 371 Study Design 372 Results 372 Implications 372 Livestock 373 Electricity Demand 373 Results 373 Implications 374 CLIMATE CHANGE AND THE OGALLALA AQUIFER 374 POLICY IMPLICATIONS 374 Land-Use Management 375 Water Resource Management 375 Risk Management 376 REFERENCES 376 CHAPTER 18: RESEARCH NEEDS 379 RELATIONSHIP BETWEEN POLICY AND SCIENCE 379 RESEARCH AND ASSESSMENT NEEDS IN THE SOCIAL SCIENCES 381 Institutional Response to Climate Variability and Climate Change 381 RESEARCH AND ASSESSMENT NEEDS IN THE NATURAL SCIENCES 382 Climate System 383 Research Scales 383 Socioeconomic Impacts 383 Data 383 Objectives of Federal Global Change Program 383 Three Major Scientific Objectives 385 THE ROLE OF EPA IN POLICY AND SCIENTIFIC RESEARCH 385 IMPACT ASSESSMENT METHODOLOGY 386 REFERENCES 388 CHAPTER 19: PREPARING FOR CLIMATE CHANGE 389 WHEN IS A RESPONSE WARRANTED? 389 Strategic Assessments 389 Decision-Oriented Assessments 390 Program-Oriented Assessments 390 Problem-Oriented Assessments 390 Criteria for Choosing a Strategy 390 EXAMPLE RESPONSES FOR ADAPTING TO GLOBAL WARMING 393 No Immediate Action 394 Reservoir Operating Rules 394 Choice of Crops 394 Anticipatory Action 394 Modifying Ongoing Projects to Consider Climate Change 394 Undertaking New Projects Primarily Because of Future Climate Change 395 Planning: Changing the Rules of the Game 396 Land Use 396 Water Allocation 398 xviii Research and Education: Increasing Our Understanding 398 Research and Development 398 Education 398 AUTHORS 401 CONTRIBUTING INVESTIGATORS AND PROJECTS 403 CONGRESSIONAL REQUEST FOR REPORT 411 APPENDICES A: WATER RESOURCES B: SEA LEVEL RISE C: AGRICULTURE D: FORESTS E: AQUATIC RESOURCES F: AIR QUALITY G: HEALTH H: INFRASTRUCTURE I: VARIABILITY J: POLICY xix FOREWORD I am pleased to transmit the attached Report to Congress: The Potential Effects of Global Climate Change on the United States. This report, written in response to a congressional request in the Fiscal Year 1987 Continuing Resolution Authority to prepare two reports on climate change, focuses on the health and environmental effects of climate change. A second draft report, Policy Options for Stabilizing Global Climate, is being revised in preparation for delivery to Congress. This report is one of the most comprehensive published studies of the potential impacts of the greenhouse effect. It examines national effects and, more specifically, impacts on four regions of the United Statees: California, the Great Lakes, the Southeast, and the Great Plains. Fifty studies conducted by government, academic, and consulting scientists to examine impacts are included. EPA provided common scenarios of climate change to the scientists for use in their analyses. This report is an overview of the results of those studies. I invite you to carefully read the Executive Summary and the chapters that follow. Although it is difficult to summarize such a large and comprehensive project in a few words, it is fair to say that climate change could lead to significant changes in many ecological and socioeconomic systems. The environmental impacts of a relatively rapid climate change may be particularly acute. Sea level rise could lead to the loss of many coastal wetlands, while a rapid warming could reduce the populations of many plants and animals and, in some cases, lead to extinction of species. The socioeconomic effects, especially on a regional scale, also may be quite important. Significant expenditures may be needed for such measures as protecting areas from sea level rise, building dams and reservoirs for flood and drought protection, modifying infrastructure, and adding electricity capacity. I urge caution in interpreting the results of these studies. Since we cannot predict regional climate change or extreme events such as hurricanes or droughts, we cannot predict impacts. The work done for this study was based on scenarios of climate change and is indicative of what could occur in the future. So, too, this work does not identify all of the impacts of climate change, the interactions, or the economic damages that could result. In examining a study such as this, there is often a temptation to identify "winners" and "losers." One must be careful in drawing such conclusions. The scenarios are based on a certain point in time (when carbon dioxide levels have doubled); and they assume that climate stops changing. If emissions are not stabilized, climate change will not stop at this carbon dioxide doubling, but will continue to warm. With continued warming, what was a positive effect could become negative. Responding to climate change would be a matter of keeping up with increasing rates of change. I feel this report is a significant contribution to our understanding of climate change impacts. More work needs to be done on understanding impacts on other systems and regions. Yet, this information will be helpful as we address the difficult problems associated with climate change. Terry Davies Assistant Administrator Office of Policy, Planning and Evaluation xxi ACKNOWLEDGMENTS This report was made possible because of the hundreds of people who participated in workshops, conducted research projects, reviewed draft manuscripts, and contributed ideas that shaped the final product. They shared a common belief that an objective analysis of global climate change could be undertaken despite the uncertainties in scientific information. We are grateful for their support and encouragement. It was the difference that sustained us through this effort. In particular, we wish to acknowledge the authors who organized and integrated the following chapters. The following people contributed to this report: Chapter 1: Introduction Joel B. Smith Chapter 2: Global Climate Change Alan Robock Chapter 3: Climate Variability Linda O. Mearns Chapter 4: Methodology Joel B. Smith Chapter 5: Forests Jack K. Winjum Ronald P. Neilson Chapter 6: Agriculture Cynthia Rosenzweig Margaret M. Daniel Chapter 7: Sea Level Rise James G. Titus Chapter 8: Biological Diversity Lauretta M. Burke Ross A. Kiester Chapter 9: Water Resources Mark W. Mugler Michael C. Rubino Chapter 10: Electricity Demand Kenneth P. Linder Chapter 11: Air Quality Joseph J. Bufalini Peter L. Finkelstein Eugene C. Durman Chapter 12: Human Health Janice A. Longstreth Chapter 13: Urban Infrastructure Ted R. Miller Chapter 14: California George A. King Robert L. DeVelice Ronald P. Neilson Robert C. Worrest Chapter 15: Great Lakes Joel B. Smith Chapter 16: Southeast James G. Titus xxiii Chapter 17: Great Plains Cynthia Rosenzweig William E. Riebsame Chapter 18: Research Needs Anthony Janetos Chapter 19: Preparing for Climate Change James G. Titus Special thanks are extended to Roy Jenne of the National Center for Atmospheric Research. Mr. Jenne and Dennis Joseph of his staff collected data from the GCMs and assembled them in a way that could be easily used by the effects researchers. In addition, he collected historic weather data for 1951-80 and distributed them to the researchers as needed. We wish to thank James Hansen, Syukuro Manabe, Richard Wetherald, and Michael Schlesinger for providing us with the results from their GCM runs. Special thanks are also necessary to Joan O'Callaghan and Karen Swetlow for editing; Roberta Wedge for assistance on production of the report; and Margaret Daniel, Michael Greene, and Chris Parker for research and administrative assistance. This work was conducted within EPA's Office of Policy Analysis, directed by Richard Morgenstern, within the Office of Policy, Planning and Evaluation, administered by Linda Fisher, and most recently by Terry Davies. Support was provided by EPA's Office of Environmental Processes and Effects Research, directed by Courtney Riordan, within the Office of Research and Development, administered by Eric Bretthauer. xxiv EXECUTIVE SUMMARY Scientific theory suggests that the addition of undertake two studies on the greenhouse effect: the greenhouse gases to the atmosphere will alter global first study was to address "The potential health and climate, increasing temperatures and changing environmental effects of climate change including, rainfall and other weather patterns. In 1979, the but not be limited to, the potential impacts on National Academy of Sciences estimated the most agriculture, forests, wetlands, human health, rivers, probable global warming from a doubling of carbon lakes, estuaries as well as societal impacts;" and the dioxide concentrations over preindustrial levels to be second study was to examine "policy options that if between 1.5 and 4.5°C. In 1985, the World implemented would stabilize current levels of Meteorological Organization (WMO), the United greenhouse gas concentrations." The second study, Nations Environment Programme (UNEP), and the "Policy Options for Stabilizing Global Climate," is a International Council of Scientific Unions (ICSU) companion report to this document. reaffirmed these estimates. Such a climate change could have significant implications for mankind and EPA responded to this request by first holding the environment: it could raise sea level, alter workshops with atmospheric scientists to discuss the patterns of water availability, and affect agriculture use of global climate change models for impact and global ecosystems. analyses and then meeting with ecologists, hydrologists, geographers, and forestry and Although there is consensus that increased agricultural specialists to identify topics for this greenhouse gas concentrations will change global study. A major purpose was to bridge the gap in climate, the rate and magnitude of change are not our ability to relate a rise in average annual surface certain (see box entitled "Climate Change"). temperatures to regional climate changes. Based on Uncertainties about climate feedbacks from clouds, these and other discussions, EPA decided to use vegetation, and other factors make it difficult to common scenarios of climate change to analyze the predict the exact amount of warming that a given sensitivities of coastal resources, water resources, level of greenhouse gases, such as doubled carbon agriculture, forests, biodiversity, health, air pollution, dioxide (CO₂) concentrations, would cause. How and electricity demand to climate change on quickly climate may change also is not known, regional and national scales (see Figure 1). These because scientists are uncertain both about how systems were chosen for analysis because they are rapidly heat will be taken up by the oceans and sensitive to climate and significantly affect our about some climate feedback processes. Generally, quality of life. EPA decided to conduct regional scientists assume that current trends in emissions analyses for the Southeast, the Great Plains, will continue and that climate will change gradually California, and the Great Lakes, because of their over the next century, although at a much faster climatological, ecological, hydrological, and pace than historically. At this rate, the full effect of economic diversity. Leading academic and the equivalent doubling of CO₂ concentrations government scientists in the relevant fields used probably would not be experienced until after 2050. published models to estimate the impacts on both It is possible, however, that sudden changes in the regional and national scales. As a common base ocean circulation could cause abrupt changes in for conducting these analyses, they used the global climate. Indeed, if climate changed more scenarios specified by EPA. rapidly than estimated, adapting to the effects would be more difficult and more costly. Furthermore, After consulting with scientific experts, EPA continued emissions of greenhouse gases could raise developed scenarios for use in effects analysis. atmospheric concentrations beyond doubled CO₂ Regional data from atmospheric models known as causing greater and more rapid climate changes, General Circulation Models (GCMs) were used as and larger effects. a basis for climate change scenarios (see box on "Scenarios and Methodology"). The GCMs are To explore the implications of climate change large models of the ocean-atmosphere system that and ways to control it, Congress asked the U.S. simulate the fundamental physical relationships in Environmental Protection Agency (EPA) to the system. GCMs provide the best scientific XXV Executive Summary CLIMATE CHANGE A panel of experts convened by the National Academy of Sciences (National Research Council, 1987) recently gave the following estimates of scientific confidence in predictions of the climate response to increased greenhouse gas concentrations. This table summarizes only their conclusions concerning "the possible climate responses to increased greenhouse gases." The full report should be consulted for the details. Large Stratospheric Cooling (virtually certain). The combination of increased cooling by additional CO₂ and other trace gases, and reduced heating by reduced ozone, "will lead to a major lowering of temperatures in the upper stratosphere." Global-Mean Surface Warming (very probable). For an equivalent doubling of CO2, "the long-term global-mean surface warming is expected to be in the range 1.5 to 4.5°C." Global-Mean Precipitation Increase (very probable). "Increased heating of the [Earth's] surface will lead to increased evaporation and, therefore, to greater global mean precipitation. Despite this increase in global average precipitation, some individual regions might well experience decreases in rainfall." Reduction of Sea Ice (very probable). This will be due to melting as the climate warms. Polar Winter Surface Warming (very probable). Due to the sea ice reduction, polar surface air may warm by as much as 3 times the global average. Summer Continental Dryness/Warming (likely in the long term). Found in several, but not all, studies, it is mainly caused by earlier termination of winter storms. "Of course, these simulations of long-term equilibrium conditions may not offer a reliable guide to trends over the next few decades of changing atmospheric composition and changing climate." Rise in Global Mean Sea Level (probable). This will be due to thermal expansion of seawater and melting or calving of land ice. Regional Case Studies Core Analytic California Outputs Areas Great Lakes Southeast Forests Great Plains Climate Agriculture Report Sea Level Rise Change to Congress Scenarios Biodiversity Water Resources Research Electricity Demand National Plan Air Quality Studies Human Health Models/ Urban Infrastructure Data Bases Policy Forests Agriculture Sea Level Rise Electricity Demand Health Figure 1. Elements of the effects report. xxvi Effects of Climate Change SCENARIOS AND METHODOLOGY A number of scenarios were specified by EPA to help identify the sensitivities of natural and manmade systems to climate change. Scenarios were used as inputs with models of natural resources. Most researchers used GCM-based scenarios. Some used analog scenarios or expert judgment. Regional outputs from three General Circulation Models (GCMs) were used: the Goddard Institute for Space Studies (GISS); the Geophysical Fluid Dynamics Laboratory (GFDL); and Oregon State University (OSU). All of these models estimate climate change caused by a doubling of CO₂ concentrations in the atmosphere. The regional estimates of doubled CO₂ changes were combined with 1951-80 climate observations to create doubled CO₂ scenarios. This GISS model has been used to estimate how climate may change between now and the middle of the next century. This is called a transient run, the outputs of which were used to create a transient scenario. Other approaches were used to supplement the GCMs. Weather observations from the 1930s were used as an analog for global warming, although greenhouse warming may raise temperatures much higher than they were in that decade. In some cases, paleoclimatic warmings were studied to provide evidence of how species respond to climate change. In addition, the use of scenarios was supplemented by expert judgment (gathered though literature reviews and workshops with scientific experts) to provide the best opinions on potential effects. Since we cannot predict the exact nature of climate change, we cannot predict its impacts. All these analytic approaches help us determine the potential sensitivities and vulnerabilities of systems to climate change. estimates of the impacts of increased greenhouse The GCM results should not be considered as gas concentrations on climate. Yet, they use predictions, but as plausible scenarios of future relatively simple models of oceans and clouds, both climate change. Ideally, one would like to use many of which will be very critical in influencing climate regional climate change scenarios to reflect the change. The GCMs generally agree concerning potential range of climate change. Resource global and latitudinal increases in temperature, but constraints allowed us to use only a limited number they disagree and are less reliable concerning other of regional climate scenarios. It would also be areas, such as regional changes in rainfall and soil useful to estimate the probabilities of occurrence for moisture. The GCM data were compared with each scenario. Given the state of knowledge, it is historic meteorologic data. In addition, the decade difficult to assign probabilities to regional climate of the 1930s was used as an analog for global change. Because the regional estimates of climate warming. change by GCMs vary considerably, the scenarios provide a range of possible changes in climate for In Figure 2, the temperature changes from the use in identifying the relative sensitivities of systems three GCMs used to create scenarios are shown for to higher temperatures and sea level rise. Hence, both the United States and four regions of the the results of the studies should not be considered United States for a doubling of carbon dioxide as predictions, but as indications of the impacts that levels. The GCMs agree on the direction of could occur as a result of global warming. temperature changes, but differ in the magnitude. Estimates of precipitation changes are shown in There are two other major limitations in the Figure 3. The GCMs agree that annual rainfall GCM scenarios. First, the scenarios assume that would increase across the country, but disagree climate variability does not change from recent about the direction of regional and seasonal decades. Second, the scenarios did not change the changes. All models show increased evaporation. frequency of events, such as heat waves, storms, hurricanes, and droughts in various regions, which xxvii Executive Summary 2xCO2 less 1xCO2 8 8 8 ANNUAL WINTER SUMMER 7 7 7 6 6 6 5 5 5 TEMPERATURE (Co) 4 TEMPERATURE (C) 4 TEMPERATURE (°C) 4 3 3 3 2 2 2 1 1 1 0 0 0 Great Lakes Southeast Great Plains California United States* Great Lakes Southeast Great Plains California United States* Great Lakes Southeast Great Plains California United States* Goddard Institute for Space Studies Geophysical Fluid Dynamics Laboratory Lower 48 States Oregon State University Figure 2. Temperature scenarios. 2xCO2 less 1xCO₂ 0.7 0.7 0.9 ANNUAL WINTER SUMMER 0.6 0.6 0.8 0.5 0.5 0.5 0.4 0.4 0.4 0.3 0.3 0.3 0.2 0.2 0.2 MILLIMETERS/DAY 0.1 8 MILLIMETERS/DAY 0.1 MILLIMETERS/DAY 0.1 NC 8 X 0 0 0 XX -0.1 -0.1 -0.1 -0.2 -0.2 -0.2 -0.3 -0.3 -0.3 -0.4 -0.4 -0.4 -0.5 -0.5 -0.5 -0.6 -0.6 -0.6 Great Lakes Southeast Great Plains California United States* Great Lakes Southeast Great Plains California United States* Great Lakes Southeast Great Plains California United States* Goddard Institute for Space Studies NC = No Change Geophysical Fluid Dynamics Laboratory * Lower 48 States Oregon State University Figure 3. Precipitation scenarios. xxviii Effects of Climate Change LIMITATIONS Climate Scenarios ** Differences Between Scenarios. The GCM and other scenarios do not provide consistent estimates of climate change. -- Variability. The scenarios assume no change in variability. -- Major Climate Events. The scenarios assume no changes in hurricanes, droughts, etc. Societal Changes. Most studies did not consider changes in population, technology, and other areas. There was only limited consideration of responses and adaptation measures, which could mitigate some of the results presented here. Linkages. Many indirect effects (e.g., effect of increased irrigation demand on water resources) were not quantitatively analyzed. Limited Effects Analyses. Many effects and regions in the United States were not analyzed. In addition, this report did not analyze the impacts of climate change on other countries. Compared to the United States, it may be much more difficult for poorer and less mobile societies to respond to climate change. It is not unreasonable to assume that climate change could have important geopolitical consequences, which could have subsequent impacts on the United States. Effects Models. These models were calibrated for historic climate conditions and may not accurately estimate future response to climate change. would have affected the results presented in this With some exceptions, we did not generally report (see "Limitations" box). Changes in examine human responses and adaptations to effects variability as estimated by GCMs were examined for of climate change. The report was intended to this report. We found that no firm conclusions can examine sensitivities and potential vulnerabilities of be drawn about how global warming could affect current systems to climate change. Many other variability. changes will also take place in the world at the same time that global climate is changing. We The methods used to estimate impacts (for cannot anticipate how changing technology, scientific example, how forests might change) also have advances, urban growth, and changing demographics limitations because our scientific understanding of will affect the world of the next century. These physiological processes is limited and subject to changes and many others may singularly, or in uncertainties. We have no experience with the rapid combination, exacerbate or ameliorate the impacts warming of 1.5 to 4.5°C projected to occur during of global climate change on society. the next century. Many of the effects are estimated based on knowledge of the response of systems to The results are also inherently limited by our known climate conditions. We cannot be certain imaginations. Until a severe event occurs, such as that a forest would be able to migrate, how higher the drought of 1988, we fail to recognize the close atmospheric concentrations of CO₂ would affect links between our society, the environment, and vegetation, whether fish would find new habitats, climate. For example, in this report we did not how agricultural pests would proliferate, or how analyze the reductions in barge shipments on the impacts would combine to create or reduce stress. Mississippi River due to lower river levels, the xxix Executive Summary increases in forest fires due to dry conditions, or the the United States to adapt managed systems in impacts of disappearing prairie potholes on ducks; response to gradual global warming. If change all these impacts were made vivid during 1988. The comes more quickly, adaptation by managed systems drought reminded us of our vulnerability as a will be more difficult and expensive. If it comes nation, but it cannot be viewed as a prediction of more slowly, the cost and difficulty of adaptation things to come. will be less. In many cases, the results of our analysis MAJOR FINDINGS appear to be consistent across scenarios, because either increasing temperatures or higher sea levels dominate the systems that were studied. For The findings collectively suggest a world example, higher temperatures would cause earlier different from the world that exists today, although snowmelt, a northward migration of forests, and a there are many uncertainties about specific effects. northward shift in crops, and higher sea levels could Global climate change could have significant inundate wetlands and low-lying areas. In other implications for natural ecosystems; for where and cases, however, only a range of values can be how we farm; for the availability of water to irrigate presented because uncertainties in an important crops, produce power, and support shipping; for variable, such as precipitation, make the direction of how we live in our cities; for the wetlands that change highly uncertain. spawn our fish; for the beaches we use for recreation; and for all levels of government and The main findings and policy implications of industry. this report are presented in national and regional chapters. They are summarized in the following The rate of global warming may be the most pages, but the reader is urged to explore the full important factor affecting both natural and managed report to understand the complete context of these systems. The faster the warming, the harder it will results. be to adapt. The ability of natural ecosystems (forests, wetlands, barrier islands, national parks) to adapt to a rapidly warming climate is limited. Rates NATIONAL FINDINGS of natural migration and adaptation could be much slower than the rate of climate change. Populations of many species and inhabited ranges could Natural Systems decrease, and many may face extinction. The ultimate effects could last for centuries and would The location and composition of various plants be virtually irreversible. Whether human and animals in the natural environment depend, to intervention could mitigate these effects was not a great extent, on climate. Trees grow in certain studied. areas and fish exist in streams and lakes because the local climate and other conditions are conducive to Managed systems may show more resilience. reproduction and growth. A major focus of this For example, although sea level rise may put report was to identify what may happen to plants additional stresses on coastal cities and although and animals, as a result of climate change changes in temperature and rainfall patterns may whether they would survive in their current locations require new strategies for managing water resources or be able to migrate to new habitats, and how soon and agriculture, we could adapt to changing climate these ecosystems could be affected. The following relatively quickly, if we have enough financial descriptions of impacts on natural systems are resources. We would expect that basic subject to uncertainties about climate change and requirements for food and water could be met in the responses of natural systems to such change. the United States (as crops are shifted and water management systems are modified), and that Natural Systems May Be Unable to Adapt Quickly developed areas with high economic value could be to a Rapid Warming protected against sea level rise (as bulkheads and levees are built). The total cost of adapting to If current trends continue, climate may change global climate change is beyond the scope of this too quickly for many natural systems to adapt. In report. It appears it could be expensive, but the past, plants and animals adapted to historic affordable, for a highly industrialized country like XXX Effects of Climate Change climate changes over many centuries. For example, forests were estimated for eastern North America since the last ice age 18,000 years ago, oak trees using temperature and precipitation correlations migrated northward from the southeastern United from pollen data. Changes in composition and States as the ice sheet receded. Temperatures abundance of particular forests were estimated for warmed about 5°C (9°F) over thousands of years, particular sites in the Great Lakes and Southeast but they rose slowly enough for forests to migrate at using site-specific models. These regions were the same rate as climate change. In the future, the chosen to represent a diversity of forest types and greenhouse effect may lead to similar changes in the uses. Finally, the ability of trees to migrate to new magnitude of warming, but the changes may take habitats was analyzed using shifts in climate zones place within a century. Climate zones may shift from GCMs and historic rates of tree migration. hundreds of miles northward, and animals and This study focused on several species that are widely especially plants may have difficulty migrating dispersed across the northeastern United States. northward that quickly. The direct effects of CO2, which could change water-use efficiency, pest interactions, and the Forests competitive balance among plants, were not modeled, nor were reforestation or the suitability of Forests occupy one-third of the land area of soils and sunlight considered. It is not clear how the United States. Temperature and precipitation these results would have been affected if such ranges are among the determinants of forest factors had been included. distributions. Forests are also sensitive to soils, light intensity, air pollution, pests and pathogens, The Range of Trees May Be Reduced disturbances such as fires and wind, and management practices. Figure 4 shows the potential shifts in forest ranges in response to climate change. The scenarios Several approaches were used to examine assume that climate change could move the geographic shifts in forests. Potential ranges of southern boundary northward by 600-700 km Hemlock Potential Range Inhabited Range Present Range Range After 2050: GISS Range After 2050: GFDL Sugar Maple Present Range Range After 2050: GISS Range After 2050: GFDL Scale 0 400Km Figure 4. Shifts in range of hemlock and sugar maple under alternative climate scenarios. xxxi Executive Summary (approximately 400 miles), while the northern dry soil conditions. In central Michigan, forests now boundary would move only as fast as the rate dominated by sugar maple and oak may be replaced ofmigration of forests. Assuming a migration rate by grasslands, with some sparse oak trees surviving. of 100 km (60 miles) per century, or double the These analyses did not consider the introduction of known historic rate, the inhabited ranges of forests species from areas south of these regions. In could be significantly reduced because the southern northern Minnesota, the mixed boreal and northern boundary may advance more quickly than the hardwood forests could become entirely northern northern boundary. Even if climate stabilizes, it hardwoods. Some areas might experience a decline could take centuries for migration to reverse this in productivity, while others (currently saturated effect. If climate continues to warm, migration soils) might have an increase. The process of would continue to lag behind shifts in climate zones. changes in species composition would most likely If elevated CO2 concentrations increase the water- continue for centuries. Other studies of the use efficiency of tree species and pest infestations potential effects of climate change in forests imply do not worsen, the declines of the southern ranges northward shifts in ranges and significant changes in could be partly alleviated. Reforestation could help composition, although specific results vary speed the migration of forests into new areas. depending on sites and scenarios used. Changes in Forest Composition Are Likely Changes May Begin in 30 to 80 Years Climate change may significantly alter forest Forest change may be visible in a few decades composition and reduce the land area of healthy from now. This would involve a faster rate of forests. Higher temperatures may lead to drier soils mortality among mature trees and a decline in in many parts of the country. Trees that need seedlings and growth of new species. The studies of wetter soils may die, and their seedlings could have forests in the Southeast and Great Lakes indicate difficulty surviving these conditions. A study of that these forests could begin to die back in 30 to 80 forests in northern Mississippi and northern Georgia years. Figure 5 displays possible reductions in indicated that seedlings currently in such areas balsam fir trees in northern Minnesota and forests would not grow because of high temperatures and in Mississippi in response to two different scenarios MISSISSIPPI FORESTS MINNESOTA BALSAM FIR 8000 180 NO CLIMATE CHANGE GISS A NO CLIMATE CHANGE 160 GISS A* 6000 140 GISS B WOODY BIOMASS (T/ha) 120 100 80 BASAL AREA (cm sq./100 m sq.) 4000 60 2000 40 20 0 0 1980 2000 2020 2040 2060 1980 2000 2020 2040 2060 2080 YEAR YEAR * Assumes constant exponential growth in emissions ** Assumes constant arithmetic growth in emissions Figure 5. Forest declines due to temperature increases. xxxii Effects of Climate Change of warming. At the same time in Minnesota, for recently, human activities, such as deforestation, example, sugar maple could become more abundant. have greatly accelerated the rate of species These forests appear to be very sensitive to small extinction. The faster rate of climate warming due changes in climate, because dieback starts to to the greenhouse effect, absent an active program become noticeable after an approximate 1 to 1.5°C to preserve species, would most likely lead to an warming. Once this process starts, major dieback even greater loss of species. The uncertainties may occur rapidly. The timing of a decline is surrounding the rate of warming, the response of sensitive to the rate of climate change; a warming individual species, and interspecies dynamics make slower than that assumed in the scenarios would it difficult to assess the probable impacts, although delay the dieback. natural ecosystems are likely to be destabilized in unpredictable ways. Other Factors Will Influence Forest Health As with trees, other plants and animals may The health of forests will not be determined by have difficulty migrating at the same rate as a climate change alone. The drier soils expected to rapidly changing climate, and many species may accompany climate change could lead to more become extinct or their populations may be reduced. frequent fires, warmer climates may cause changes The presence of urban areas, agricultural lands, and in forest pests and pathogens, and changes in air roads would restrict habitats and block many pollution levels could reduce the resilience of migratory pathways. These obstacles may make it forests. Continued depletion of stratospheric ozone harder for plants and wildlife to survive future would also further stress forests. None of these climate changes. On the other hand, some species outcomes was considered by the forest studies in may benefit from climate change as a result of this report, although they could speed forest increases in habitat size or reduction in population declines. of competitors. The extent to which society can mitigate negative impacts through such efforts as Biodiversity habitat restoration is not clear. Biological diversity can be defined as the Impacts on Fisheries Would Vary variety of species in ecosystems, and the genetic variability within each species and the variety of Freshwater fish populations may grow in some ecosystems around the world. Over 400 species of areas and decline in others. Fish in such large mammals, 460 species of reptiles, 660 species of water bodies as the Great Lakes may grow faster freshwater fishes, and tens of thousands of and may be able to migrate to new habitats. invertebrate species can be found in this country, in Increased amounts of plankton could provide more addition to some 22,000 plant species. About 650 forage for fish. However, higher temperatures may species of birds reside in or pass through the United lead to more aquatic growth, such as algal blooms, States annually. Biological diversity is needed to and decreased mixing of lakes (longer stratification), provide food, medicine, shelter, and other important which would deplete oxygen levels in shallow areas products. of the Great Lakes, for example Lake Erie, and make them less habitable for fish. Fish in small This report examined the impacts of climate lakes and streams may be unable to escape change on specific plants and animals by using temperatures beyond their tolerances, or their climate change scenarios and models of particular habitats may simply disappear. species or systems within a region. Analyses have been performed for impacts on finfish and shellfish Warmer temperatures could also exceed the in the Apalachicola Bay in the Florida Panhandle, thermal tolerance of many marine finfish and fish in the Great Lakes, and marine species in San shellfish in some southern locations, although some Francisco Bay. Additional information on potential marine species could benefit. The full impacts on impacts on biodiversity was gathered from the marine species are not known at this time. The loss published literature. of coastal wetlands could further reduce fish populations, especially shellfish. And while Extinction of Species Could Increase increased salinity in estuaries could reduce the abundance of freshwater species, it could increase Historic climate changes, such as the ice ages, the presence of marine species. Whether finfish and have led to extinction of many species. More xxxiii Executive Summary shellfish could migrate to new areas and the Protecting Developed Areas May Be Expensive effectiveness of restocking were not studied. Given the high property values of developed Effects on Migratory Birds Would Depend on coastlines in the United States, it is likely that Impacts on Habitats measures would be taken to hold back the sea along most developed shores. Preliminary estimates Migratory birds are likely to experience mixed suggest that the cumulative capital cost (including effects from climate change, with some arctic- response to current sea level rise) of protecting nesting herbivores benefiting, and continental currently developed areas would be $73 to $111 nesters and shorebirds suffering. Some winter billion (in 1988 dollars) through 2100 for a 1-meter habitats could experience increased productivity. global rise (compared with $4 to $6 billion to On the other hand, the loss of wintering grounds, protect developed areas from current trends in sea which may result from sea level rise and changing level rise). A 1-meter sea level rise would lead to climate, could harm many species, as would the loss a cumulative inundation of 7,000 square miles of of inland prairie potholes resulting from potentially dryland an area the size of Massachusetts (see increased midcontinental dryness. Table 1). If the oceans continue to rise at current rates, approximately 3,000 square miles of dryland Sea Level Rise would be lost. Most Coastal Wetlands Would Be Lost A rise in sea level is one of the more probable impacts of climate change. Higher global temperatures will expand ocean water and melt Historically, wetlands have kept pace with a slow rate of sea level rise. However, in the future, some mountain glaciers, and may eventually cause polar ice sheets to discharge ice. Over the last sea level will probably rise too fast for some century, global sea level has risen 10 to 15 cm (4 to marshes and swamps to keep pace. Although some 6 inches), and along the U.S. coastline, relative sea wetlands can survive by migrating inland, a study on level rise (which includes land subsidence) has coastal wetlands estimated that for a 1-meter rise, 26 to 66% of wetlands would be lost, even if averaged about 30 cm (1 foot). Published estimates of sea level rise due to global warming generally wetland migration were not blocked. A majority of range from 0.5 to 2.0 meters (1.5 to 7 feet) by 2100. these losses would be in the South (see Table 2). Sea level rise could be greater than or less than this Efforts to protect coastal development would range because uncertainties exist regarding the rate increase wetland losses, because bulkheads and of atmospheric warming, glacial processes, oceanic levees would prevent new wetlands from forming uptake of heat, precipitation in polar areas, and inland. If all shorelines are protected, 50 to 82% of wetlands would be lost. The different amounts of other variables. dryland lost for different regions and scenarios are The studies estimate the potential nationwide shown in Figure 6. loss of wetlands, and the cost of defending currently The loss of wetland area would have adverse developed areas from a rising sea, for three scenarios (50, 100, and 200 cm) of sea level rise by ecological impacts, with the ability of ecosystems to the year 2100. The scenarios are based on survive a rising sea level depending greatly on how quantitative estimates of sea level rise, but no shorelines are managed. For many fish and shellfish probabilities have been attributed to them. Wetland species, the fraction of shorelines along which loss estimates were based on remote-sensing data wetlands can be found is more important than the total area of wetlands. This fraction could remain and topographic maps for a sample of sites along the U.S. coast. The cost of holding back the sea at approximately present levels if people do not erect additional bulkheads and levees. In Louisiana, was based on (1) the quantity of sand necessary to elevate beaches and coastal barrier islands as sea with 40% of U.S. coastal wetlands, large areas of level rises; (2) rebuilding roads and elevating wetlands are already being converted to open water as a result of natural subsidence and the effects of structures; and (3) constructing levees and human activities, and most could be lost by 2030 if bulkheads to protect developed lowlands along current trends continue. sheltered waters. xxxiv Effects of Climate Change Table 1. Nationwide Impacts of Sea Level Rise Sea Level Rise by 2100 Alternative Baseline* 50 cm 100 cm 200 cm If Densely Developed Areas Are Protected Shore protection costs (billions of 1986 dollars) 4-6 32-43 73-111 169-309 Dryland lost (mi²) 1,500-4,700 2,200-6,100 4,100-9,200 6,400-13,500 Wetlands lost (%) 9-25 20-45 29-69 33-80 If No Shores Are Protected Dryland lost (mi²) N.C. 3,300-7,300 5,100-10,300 8,200-15,400 Wetlands lost (%) N.C. 17-43 26-66 29-76 If All Shores Are Protected Wetlands lost (%) N.C. 38-61 50-82 66-90 N.C. = Not calculated. *Baseline assumes current global sea level rise trend of 12 cm per century. Given coastal subsidence trends, this implies about a 1-foot rise in relative sea level along most of the U.S. coast. Source: Assembled by Titus and Greene. Table 2. Loss of Coastal Wetlands from a One-Meter Rise in Sea Level All Current Current dryland development No wetlands protected protected protection Region area (mi²) (% loss) (% loss) (% loss) Northeast 600 16 10 2 Mid-Atlantic 746 70 46 38 South Atlantic 3,813 64 44 39 South and West Florida 1,869 44 8 7 Louisiana 4,835 77 77 77 Other Gulf 1,218 85 76 75 West 64 56 gainᵇ gainᵇ United States 13,145 50-82 29-69 26-66 ᵃLouisiana projections do not consider potential benefits of restoring flow of sediment and freshwater. Potential gain in wetland acreage not shown because principal author suggested that no confidence could be attributed to those estimates. West Coast sites constituted less than 0.5% of wetlands in study sample. Source: Adapted from Park et al. XXXV Executive Summary A. DRYLAND LOSS BY 2100 WITHOUT SHORE PROTECTION 3.0 2.8 2.6 2.4 2.2 2.0 LOSS OF DRYLAND (THOUSANDS OF SQ. MILES) 1.8 1.6 1.4 1.2 1.0 0.8 0.6 SEA LEVEL RISE SCENARIO: 0.4 0.2 BASELINE 50 CM 0.0 Northeast Mid- South South Louisiana Other Gulf West Atlantic Atlantic & West Florida 100 CM 200 CM B. DRYLAND LOSS BY 2100 WITH PROTECTION OF DEVELOPED AREAS 3.0 2.8 2.6 2.4 2.2 2.0 LOSS OF DRYLAND (THOUSANDS OF SQ. MILES) 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Northeast Mid- South South Louisiana Other Gulf West Atlantic Atlantic & West Florida Figure 6. Dryland loss by 2100. Estuaries May Enlarge and Become More Saline of the gross national product in 1985, with farm assets totaling $771 billion. Crop production is Although future riverflows into estuaries are sensitive to climate, soils, management methods, uncertain, a rise in sea level would increase the size and many other factors. During the Dust Bowl and salinity of estuaries and would increase the years of the 1930s, wheat and corn yields dropped salinity of coastal aquifers. For example, sea level by up to 50%, and during the drought of 1988, corn rise may result in a more saline and enlarged yields declined about 40%. Sacramento-San Joaquin Delta, and Miami, New York, and other coastal communities would have to The agricultural analyses in this report step up current efforts to combat salinity increases examined potential impacts on crop yields and in surface water and groundwater supplies. productivity from changes in climate and direct effects of CO₂. (Higher CO₂ concentrations may Agriculture increase plant growth and water-use efficiency.) The studies used high estimates of the beneficial The temperate climate and rich soils in the effects of CO2 on crops. Changes in dryland and United States, especially in the Midwest, have irrigated corn, wheat, and soybean yields and in helped make this country the world's leading irrigation demand were estimated for the Southeast, agricultural producer. Agriculture, a critical Great Plains, and Great Lakes regions using widely component of the U.S. economy, contributed 17.5% validated crop growth models. Crop yield changes xxxvi Effects of Climate Change were estimated for California using a simple productivity of northern areas for the crops studied agroclimatic index. The studies did not examine was estimated to rise in comparison with that of effects on yields of introduction of crops, such as southern areas. In response to the shift in relative citrus, into new areas; changes in weed yields, grain crop acreage in Appalachia, the growthcaused by higher CO₂ concentrations; or new Southeast, and the southern Great Plains could technologies, such as biotechnology. Some of these decrease, and acreage in the northern Great Lakes changes could enhance the ability of agriculture to States, the northern Great Plains, and the Pacific adapt to global warming. Northwest could increase (see Figure 7). A change in agriculture would affect not only the livelihood of The estimated yield changes from the four farmers but also agricultural infrastructure and regional crop modeling studies and runoff changes other support services. The sustainability of crop from the GCMs were used in a nationwide production in northern areas was not studied. agricultural economic model to estimate regional Changes in foreign demand for U.S. crops, which and national changes in crop production, land use, would likely be altered as a result of global warming and demand for irrigation. The economic model and could significantly alter the magnitude of the did not consider the introduction of new crops, results, were not considered in this analysis. changes in government policies on agriculture, change in demand for water for nonagricultural The National Supply of Agricultural Commodities uses, and global agricultural changes. Both a May Be Sufficient to Meet Domestic Needs, But modeling study and a literature review were used to Exports May Be Reduced estimate changes in plant-pest interactions. An agricultural runoff and leaching model was used to Even under the more extreme climate change estimate potential changes in water quality in the scenarios, the production capacity of U.S. Great Plains. Some farm-level adjustments, agriculture was estimated to be adequate to meet including the effects of changed planting dates and domestic needs. Only small to moderate economic use of different varieties, were investigated in losses were estimated when climate change various studies, and the potential national scenarios were modeled without the beneficial implications on livestock were analyzed using effects of CO2 on crop yields. When the combined modeling studies and a literature review. effects of climate and CO2 were considered, results were positive with a relatively wetter climate change Yields Could Be Reduced, Although the Combined scenario and negative with the hotter, drier climate Effects of Climate and CO2 Would Depend on the change scenario. Thus, the severity of the economic Severity of Climate Change consequences could depend on the type of climate change that occurs and the ability of the direct In most regions of the country, climate change effects of CO2 to enhance yields. A decline in crop alone could reduce dryland yields of corn, wheat, production would reduce exports, which could have and soybeans, with site-to-site losses ranging from serious implications for food-importing nations. If negligible amounts to 80%. These decreases would climate change is severe, continued and substantial be primarily the result of higher temperatures, improvements in crop yields would be needed to which would shorten a crop's life cycle. In very fully offset the negative effects. Technological northern areas, such as Minnesota, dryland yields of improvements, such as improved crop varieties from corn and soybeans could increase as warmer bioengineering, could be helpful in keeping up with temperatures extend the frost-free growing season. climate change. These results could be affected by The combined effects of climate change and global changes in agriculture, which were not increased CO2 may result in net increases in yields considered in the analysis. in some cases, especially in northern areas or in areas where rainfall is abundant. In southern areas, Farmers Would Likely Change Many of Their however, where heat stress is already a problem, Practices and in areas where rainfall is reduced, crop yields could decline. Farm practices would likely change in response to different climate conditions. Most significantly, Productivity May Shift Northward in many regions, the demand for irrigation is likely to increase as a result of higher temperatures. If Under all of the scenarios (with and without national productivity declines, crop prices may rise, the direct effects of increased CO2), the relative making irrigation more economical and increasing xxxvii Executive Summary 20t 10 20 0 10 10 0 20- 10 20th Lake States 20 10- VI 30 0 40 10 50- 20- Northeast 20t Northern Plains ) 20t 10 10- 0 1) / 10 20+ :0 20- 10 20 20 Corn Belt 10 Mountain 0 10 0 20 10 30 20- 20 Pacific 40 10 20 20+ 50 0 10 10 60 -10 0 0 70 20 10 10 80L 30 20 20 Appalachia 40 30 30 GISS 50- 40 40 Southern Plains 50 50 GFDL 60 60 GISS + Direct Effects of CO₂ -70 .70 80L 80L GFDL + Direct Effects of CO₂ Delta States Southeast Figure 7. Percent change in regional agricultural acreage. 80 70 60 50 40 40 30 20 30 10 0 20 10 20- 10 Northern Plains 0 40t -10 30 20 20 Mountain 40 10 30 2.5 80t 0 2.0 70 20 1.5 60 1.0 -10 50 10 0.5 40 20 0 30 0 0.5 20 Pacific GISS 1.0 10 GFDL 10 Delta States 0 10 GISS Direct Effects of CO₂ 20 20- GFDL + Direct Effects of CO₂ Southern Plains Southeast Figure 8. Change in regional irrigation acreage (100,000 of acres). xxxviii Effects of Climate Change the use of it (see Figure 8). Irrigation equipment other needs, such as maintaining flow to preserve may be installed in many areas that are currently environmental quality. dryland farms, and farmers already irrigating may extract more water from surface and groundwater Although global precipitation is likely to sources. Changes in competing demands for water increase, it is not known how regional rainfall by municipal and industrial users, which could raise patterns will be affected. Some regions may have the cost of irrigation, were not considered. Farmers more rainfall, while others may have less. may also switch to more heat- and drought-resistant Furthermore, higher temperatures would most likely crop varieties, plant two crops during a growing increase evaporation. These changes would likely season, and plant and harvest earlier. Whether create new stresses for many water management these adjustments would compensate for climate systems. change depends on a number of factors, including the severity of the climate change. Under extreme To discuss the potential impacts of climate climate change conditions, some farms could be change on water resources, this report studied water abandoned. resources in California, the Great Lakes, and the Southeast, estimated the demand for irrigation in Ranges of Agricultural Pests May Extend the Great Plains, and drew on information from the Northward literature. These studies focused on changes in runoff and, for California and the Southeast, Warmer temperatures may result in the considered management responses. The studies northward extension of the range of diseases and examined the water management systems as they pests that now afflict livestock in the South, and are currently configured and did not examine new could make conditions more favorable for the construction. Among other factors not considered introduction of new livestock diseases into the were changes in demand for water resources (which southern United States. This extension could would most likely lead to greater changes in water reduce crop yields and affect livestock. management systems) and changes in vegetation due to climate change and increased CO2, which could Shifts in Agriculture May Harm the Environment in affect runoff. The studies did not estimate impacts Some Areas on groundwater. Expansion of irrigation and shifts in regional The Direction of Change in Some Water Bodies production patterns imply more competition for Can Be Estimated, but Total Impacts in the United water resources, greater potential for surface water States Cannot Be Determined and groundwater pollution, loss of some wildlife habitats, and increased soil erosion. A northward Results of hydrology studies indicate that it is migration of agriculture would increase the use of possible in some regions to identify the direction of irrigation and fertilizers on sandy soils, thus change in water supplies and quality due to global endangering the quality of underlying groundwater. warming. For example, in California, higher Chemical pesticide usage may change to control temperatures would reduce the snowpack and cause different crop and livestock pests. Thus, climate earlier melting. Earlier runoff from mountains change could exacerbate environmental pollution could increase winter flooding and reduce deliveries and increase resource use from agriculture in some to users. In the Great Lakes, reduced snowpack areas. combined with potentially higher evaporation could lower lake levels (although certain combinations of Water Resources conditions could lead to higher levels). In other areas, such as the South, little snowcover currently The United States is endowed with a bountiful exists, so riverflow and lake levels depend more on supply of water, but the water is not always in the rainfall patterns. Without better rainfall estimates, right place at the right time or of the right quality. we cannot determine whether riverflow and lake In some regions, such as the Great Basin and the levels in the South would rise or fall. Colorado River Basin, the gap between demand and supply of water is narrow. In these basins, such Water Quality in Many Basins Could Change offstream uses as irrigation and domestic consumption often conflict with each other and with Changes in water supply could significantly affect water quality. Where riverflow and lake xxxix Executive Summary levels decline, such as in the Great Lakes, there requirements in the United States. The demand for would be less water to dilute pollutants. On the electricity for summer cooling would increase, and other hand, where there is more water, water the demand for electricity for winter heating would quality may improve. Higher temperatures may decrease. Annual electricity generation in 2055 was enhance thermal stratification in some lakes and estimated under the transient scenarios to be 4 to increase algal production, degrading water quality. 6% greater than without climate change. The Changes in runoff and leaching from farms and annual costs of meeting the increase due to global potential increases in the use of irrigation for warming, assuming no change in technology or agriculture could affect surface and groundwater efficiency, was estimated to be $33-$73 billion (in quality in many areas. 1986 dollars). These results differ on a regional basis and are shown in Figure 9. States along the Water Use Conflicts May Increase northern tier of the United States could have net reductions in annual demand of up to 5%, because In some regions, decreased water availability decreased heating demand would exceed increased and increased demand for water, such as for demand for air-conditioning. In the South, where irrigation and powerplant cooling, may intensify heating needs are already low, net demand was conflicts among offstream uses. Conflicts between estimated to rise by 7 to 11% by 2055. these offstream uses and instream uses such as flood control and wildlife habitat also may be Generating capacity requirements are intensified. determined largely by peak demand, which occurs in the summer in all but the far northern areas of Electricity Demand the country. By 2010, generating requirements to meet increased demand could rise by 25 to 55 The demand for electricity is influenced by gigawatts (GW), or by 9 to 19% above new capacity economic growth, by changes in industrial and requirements, assuming no climate change. By residential/commercial technologies, and by climate. 2055, generating requirements could be up by 200 to The principal climate-sensitive electricity end uses 400 GW, or 14 to 23% above non-climate-related are space heating and cooling and, to a lesser growth. The cumulative cost of such an increase in degree, water heating and refrigeration. These uses capacity, assuming no change in technology or of electricity may account for up to a third of total improvements in energy efficiency, was estimated to sales for some utilities and may contribute an even be between $175 and $325 billion (in 1988 dollars). larger portion of seasonal and daily peak demands. The South would have a greater need than the North for additional capacity, as shown in Figure 10. This report analyzed potential changes in the Increases in capacity requirements could range from national demand for electricity in 2010 and 2055, 0 to 10% in the North, to 20 to 30% in the South using the relationship between demand and climate and Southwest. U.S. emissions of such greenhouse for several major utility systems. The study gases as CO₂ could increase substantially if estimated changes in demand due to nonclimate additional powerplants are built to meet these factors, such as increases in population and GNP. capacity requirements, especially if they burn coal. The impacts of climate change are expressed as an Improvement in the efficiency of energy production increase over non-climate growth, and results are and use would reduce these emissions. given on nationwide and regional bases. The study did not consider changes in technology and Air Quality improvements in energy efficiency; the impacts of higher temperatures on the demand for natural gas Air pollution caused by emissions from and oil for home heating, which will most likely industrial and transportation sources is a subject of decrease; changes in electricity supplies, such as concern in the United States. Over the last two hydropower; or changes in demand for electricity decades, considerable progress has been made in for such uses as irrigation. improving air quality by reducing emissions. Yet high temperatures in the summer of 1988 helped National Electricity Demand Would Rise raise tropospheric ozone levels to all-time highs in many U.S. cities. But air quality is also directly Global warming would increase annual demand affected by other weather variables, such as for electricity and total generating capacity xl Effects of Climate Change 2055 % CHANGE GENERATION 10 to 15 5 to 10 0 to 5 -5 to 0 Figure 9. Changes in electricity generation by state, induced by climate change scenarios in 2055. 2055 % CHANGE IN NEW CAPACITY 20 to 30 10 to 20 0 10 10 to 0 Figure 10. Changes in electricity capacity additions by state, induced by climate change scenarios in 2055. xli Executive Summary windspeed and direction, precipitation patterns, temperatures would speed the reaction rates among cloud cover, atmospheric water vapor, and global chemicals in the atmosphere, causing higher ozone circulation patterns. pollution in many urban areas than would occur otherwise. They would also increase the length of A literature review of the relationship between the summer season, usually a time of high air climate and air pollution was conducted for this pollution levels. As shown in Figure 11, preliminary report. In addition, air quality models were used analyses of a 4°C temperature increase in the San for a preliminary analysis of the changes in ozone Francisco Bay area (with no changes in other levels in several regions. The latter analysis did not meteorologic variables, such as mixing heights), consider reduction in emissions of air pollutants due assuming no change in emissions from current to enforcement of the Clean Air Act. levels, suggest that maximum ozone concentrations would increase by 20%, and that the area exceeding Climate Changes Could Increase Air Pollution, the National Ambient Air Quality Standards would Especially Smog almost double. Studies of the Southeast also show expansion of the areas violating the standards, but A rise in global temperatures would increase they show smaller changes in levels. Although the manmade and natural emissions of hydrocarbons impacts of higher temperatures on acid rain were and manmade emissions of sulfur and nitrogen not analyzed, it is likely that sulfur and nitrogen oxides over what they would be without climate would oxidize more rapidly under higher change. Natural emissions of sulfur would also temperatures. The ultimate effect on acid change, but the direction is uncertain. Although the deposition is difficult to assess because changes in potential magnitude of the impacts of the increased clouds, winds, and precipitation patterns are emissions on air quality is uncertain, higher uncertain. Exceeds Standard <6 >6 >8 >10 >12 >14 >16 Sacramento Sacramento San Fran. Oakland San Fran Oakland Stockton Stockton San Jose San Jose Modesto Crow's Landing Modesto Crow's Landing Castle AFB Castle AFB Yosemite Yosemite Salinas Salinas Base Case Climate Sensitivity Scenario No. 1 August 6, 1981 4°C Temperature Increase Figure 11. Changes in maximum daily ozone concentrations. xlii Effects of Climate Change Health Effects grasslands, thereby modifying the incidence of vector-borne diseases. Changes in summer rainfall Human illness and mortality are linked in could alter the amount of ragweed growing on many ways to weather patterns. Weather affects cultivated land, and changes in humidity may affect contagious diseases such as influenza and the incidence and severity of skin infections and pneumonia, and allergic diseases such as asthma. infestations such as ringworm, candidiasis, and Mortality rates, particularly for the elderly and the scabies. Increases in the persistence and level of air very ill, are influenced by the frequency and severity pollution episodes associated with climate change of extreme temperatures. The life cycles of disease- would have other harmful health effects. carrying insects, such as mosquitoes and ticks, are affected by changes in temperature and rainfall, as Urban Infrastructure well as by habitat, which is itself sensitive to climate. Finally, increased air pollution, which is related to The value of municipal infrastructure in the weather patterns, can heighten the incidence and United States, excluding buildings and electric severity of such respiratory diseases as emphysema and asthma. power production, probably approaches one trillion dollars. The majority of the nation's investments are in water supply, wastewater transport and Both expert judgment and modeling were used treatment facilities, drainage, roadways, airports, to study the potential impacts of climate change on and mass transit facilities. Like the regions studied human health. A literature review and workshop for this report, urban areas would feel a variety of were conducted to identify potential changes in impacts from climate change. This report examined vector-borne diseases caused by ticks, fleas, and the potential impacts of climate change on mosquitoes (such as dengue and malaria). Models Cleveland, New York City, and Miami. These areas were used to estimate potential geographic shifts in encompass a diversity of climates and uses of the prevalence of Rocky Mountain spotted fever and natural resources. malaria. Potential changes in mortality from heat and cold stress were quantitatively estimated, Much of the current inventory in urban although such estimates did not consider changes in infrastructure will most likely turn over in the next air pollution levels. The total impacts of climate 35 to 50 years. A warmer global climate would change on human health are difficult to assess; these require changes in the capital investment patterns of analyses looked at a limited number of potential effects and are only indicative of possible changes in cities for water supplies, peak electric generating capacity, and storm sewer capacity. Urbanized mortality and morbidity. coastal areas might have to invest additional billions of dollars into coastal protection to defend Summer Mortality Could Increase, While Winter developed areas from a rising sea. In Miami, for Mortality Could Decrease example, this could imply an increase of 1 to 2% in Global warming may lead to changes in the city's capital spending over the next 100 years. morbidity and increases in mortality, particularly for Generally, northern cities such as Cleveland may the elderly during the summer. Morbidity and fare better, since reductions in the operating and mortality may decrease because of milder winters, maintenance costs associated with heating public buildings, snow removal, and road maintenance although net mortality may increase. If the frequency or intensity of climate extremes increases, should offset increasing costs for air-conditioning and port dredging (see Table 3). mortality is likely to rise. If people acclimatize by using air-conditioning, changing their workplace habits, and altering the construction of their homes and cities, the impact on summer mortality rates REGIONAL IMPACTS may be substantially reduced. Studying the national impacts of climate change Regional Morbidity Patterns Could Change may disguise important differences in regional effects across the country. Shifting demands for Changes in climate as well as in habitat may economic and natural resources may cause stresses alter the regional prevalence of vector-borne that cannot be seen at a national level. diseases. For example, some forests may become Furthermore, changes in one system, such as water supply, may affect other systems such as irrigation xliii Executive Summary Table 3. Estimated Impacts of Doubled CO2 Scenarios on Cleveland's Annual Infrastructure Costs (millions of 1987 dollars) Annual Cost category operating costs Heating -2.3 Air-conditioning +6.6-9.3 Snow and ice control -4.5 Frost damage to roads -0.7 Road maintenance -0.5 Road reconstruction -0.2 Mass transit summer increase offsets winter savings River dredging less than $0.5 Water supply negligible Stormwater system negligible Total -1.6 to + 1.1 Source: Walker et al. for agriculture. These combined effects may be California. The discussion that follows should not most evident on a regional scale. The designs of be viewed as comprehensive, but rather as providing the regional studies on agriculture, forests, and examples of important issues for each region. electricity were described above. California The studies discussed below considered only some of the potential regional impacts. Many California contains a highly managed water potential impacts were not analyzed for example, resource system and one of the most productive demographic shifts into or out of the Southeast, agricultural regions in the world. The state recreational impacts in the Great Lakes, direct produces 14% of the nation's cash receipts for effects on such aquifers as the Ogallala in the Great agriculture. California's water resources are poorly Plains, and impacts on many specialty crops in distributed in relation to its needs. Precipitation is California. In addition, current GCMs often abundant in the north, with the highest levels in the disagree significantly about simulated regional winter, while water is needed in the south for changes, particularly about such key variables as agriculture and domestic consumption. The Central precipitation. Their spatial resolution is roughly of Valley Project (CVP) and State Water Project the same size as the regions of concern; for (SWP) were built basically to capture runoff from example, there are two simulation points in the north and deliver it to uses in the south. These xliv Effects of Climate Change projects also provide flood protection, hydroelectric the demand for carriage water may be affected. power, and freshwater flows to repel salinity (known The estimated changes in salinity and sea level rise as carriage water) in the Sacramento-San Joaquin were used to examine impacts on the ecology of the River Delta. Islands in the delta are highly bay. Yield changes for a number of crops grown in productive farmlands and are protected by levees. the state were estimated, as were changes in ozone levels in central California and changes in electricity The California case study focused on the demand (see Figure 12). Central Valley. First, changes in runoff in the valley were estimated. These results were then used to California's Water Management System Would estimate changes in deliveries from the CVP and Have to Be Modified SWP and in agricultural water use. These results were combined with sea level rise estimates and Warmer temperatures would change the were used to model how the salinity and shape of seasonality of runoff from the mountains the San Francisco Bay estuary may change and how surrounding the Central Valley. Runoff would be Water Resources Regional warming could cause: higher winter, lower summer runoff decreased deliveries from Central Valley Project and State Water Project decreased water quality in subalpine lakes Wetlands and Fisheries Sea level rise could cause: gradual inundation of wetlands increased salinity in and size of San Francisco Bay TEMPERATURE SCENARIOS shift from brackish and freshwater 2xCO2 less 1xCO2 species to marine species 6 5 Agriculture 4 TEMPERATURE (°C) Increases in temperature and CO₂ concentrations could cause: 3 variable crop responses 2 a northward shift in agricultural GISS production 1 GFDL increased irrigation demand resulting in groundwater extraction 0 OSU WINTER SPRING SUMMER FALL and decreased water quality PRECIPITATION SCENARIOS 2xCO2 less 1xCO2 0.6 0.5 Air Quality 0.4 Higher temperatures would increase 0.3 ambient ozone levels in central California MILLIMETERS/DAY 0.2 0.1 0 -0.1 0.2 Electricity -0.3 Higher temperatures could increase 0.4 -0.5 electricity demand -0.6 WINTER SPRING SUMMER FALL Figure 12. California. xlv Executive Summary higher in the winter months as a result of less bay. As a result of these changes, some wetlands snowpack and more precipitation in the form of would be lost, marine aquatic species would become rain. Consequently, runoff would be lower in the relatively more abundant, and freshwater species late spring and summer. Under these conditions, would decline. the current reservoir system in the Central Valley would not have the capacity to provide adequate Climate Change Could Degrade Air Quality in flood protection in the winter and store enough California water to meet deliveries in the summer. Thus, much of the earlier winter runoff would have to be Air quality is currently a major concern in released. This would leave less water in the system California. The area of central California in for late spring and summer deliveries, when runoff violation of ozone quality standards could increase would be lower. Under the three GCM scenarios, as a result of higher temperatures. Under one annual water deliveries from the SWP were climate scenario, with a 4°C rise and current estimated to decrease by 200,000 to 400,000 acre- emission levels, the maximum size of the area with feet (7 to 16% of supply). In contrast, the increase ozone levels in excess of the EPA standard of 0.12 in statewide demand for water from the SWP due to ppm could double. This scenario assumed that such non-climate factors, such as population growth, may climate variables as windspeed and mixing height total 1.4 million acre-feet by 2010. Reduced (the volume of air in which pollutants are diluted) snowpack and earlier runoff could occur throughout would not change. the West, exacerbating water management problems in a region that is currently short of water. Great Lakes Climate Change Is Likely to Increase Water The Great Lakes contain 18% of the world's Demand supply and 95% of the U.S. supply of surface freshwater, and they are an important source of On the whole, California's water demand could commerce and recreation for the region. In recent increase with a warmer climate. Twice as much years, reductions in pollutant loadings have carriage water may be needed to repel higher significantly improved the quality of such water salinity levels resulting from a 1-meter sea level rise. bodies as Lake Erie. The Great Lakes States In addition, consumptive uses may also increase. produce 59% of the country's corn and 40% of its Irrigation, which may come from groundwater, may soybeans, and their forests have important increase in some parts of the state. If new commercial, recreational, and conservation uses. powerplants are built, they will need water for cooling, which could come from surface water Models were used to estimate the potential supplies, depending on the location. Although it impacts of climate change on lake levels and ice was not studied, municipal demand for water may cover. Results from these studies were used to also rise. analyze impacts on navigation and shorelines. Changes in the thermal structure of the Central Sea Level Rise Would Affect the Size and Basin of Lake Erie and southern Lake Michigan Environment of San Francisco Bay were estimated. Output from these studies was used along with scenario temperatures to analyze A sea level rise would increase the salt potential impacts on fishes in the lakes. Changes in concentrations of San Francisco Bay. It is estimated crop yields were estimated for corn and soybean, that a 1-meter rise could cause the salt front in the and changes in forest composition were analyzed for Sacramento-San Joaquin River Delta to migrate Michigan and Minnesota (see Figure 13). upstream 4 to 10 km (2.5 to 6 miles). Sea level rise would also increase the difficulty of maintaining the Lake Levels Could Drop and Ice Cover Duration Sacramento-San Joaquin Delta islands. If the levees Could Decrease around the delta islands were strengthened and raised, a 1-meter rise could increase the volume of Higher temperatures would likely reduce the San Francisco Bay estuary by 15% and the area snowpack and could increase evaporation, which by 30%. If the levees were not maintained and the would lower lake levels. The level of Lake Superior islands were flooded, there would be a doubling and was estimated to be reduced under the climate tripling, respectively, of the volume and area of the scenarios by 0.4 to 0.5 meters (1.2 to 1.5 feet), and xlvi Effects of Climate Change Lakes Climate change could: cause average lake levels to fall by 0.5 to 2.5 meters reduce ice cover duration by 1-3 months Adjustments may be required, including: increased dredging of harbors and channels, or lower cargo capacities on ships Water Quality Changes in temperature and precipitation 0 could cause: greater stratification in lakes and TEMPERATURE SCENARIOS increased growth of algae, which in turn 2xCO2 less 1xCO2 could cause lower dissolved oxygen 8 levels in shallow areas an increase in pollutants resulting from 7 more dredging 6 TEMPERATURE (C) 5 Wetlands and Fisheries 4 Higher temperatures could cause: an increase in fish habitats in fall, winter, 3 GISS and spring, and a decrease in summer 2 accelerated growth for some fish species GFDL potential invasion by new species 1 OSU 0 WINTER SPRING SUMMER FALL Forests Higher temperatures could result in: PRECIPITATION SCENARIOS loss of mixed northern hardwood and 2xCO2 less 1xCO2 oak in southern areas 0.6 shifts of mixed northern hardwood and 0.5 boreal forests in northern areas to all 0.4 northern hardwood 0.3 forest declines evident in 30 to 60 years 0.2 MILLIMETERS/DAY 0.1 0 Agriculture -0.1 Higher temperatures could cause: -0.2 Not corn and soybean yields to increase in -0.3 Calculated North, decline in Cornbelt; mixed results 0.4 under climate change and CO₂ -0.5 acreage could expand in the North, -0.6 WINTER SPRING SUMMER FALL leading to increased erosion and runoff Figure 13. Great Lakes. that of Lake Michigan by 0.9 to 2.5 meters (3 to 8 duration by 1 to 3 months on Lake Superior and by feet). Diversions out of the lakes for irrigation or 2 to 3 months on Lake Erie, although ice still would to supply other basins would further lower lake form on both lakes. Changes in windspeed would levels, although these impacts were not analyzed. affect the reduction in duration of ice cover. In These results are very sensitive to assumptions response to lower lake levels, either ships would made about evaporation and under some have to sail with reduced cargoes or ports and circumstances, lake levels could rise. channels would have to be dredged. On the other hand, a shorter ice season would allow a longer Higher temperatures would also reduce ice shipping season. cover on the lakes. Specifically, they could cut ice xlvii Executive Summary Water Quality May Be Degraded in Some Areas sufficient oxygen is present, growth rates and productivity for such fish as bass and lake trout in Higher temperatures could lengthen open areas of large lakes may increase, provided stratification of the lakes (where summer that the forage base also increases. However, temperatures warm the upper part of lakes and reduced ice cover and decreased water quality could isolate the cooler lower layers of lakes). Analysis of harm some species in shallow basins of the Great the Central Basin of Lake Erie showed that longer Lakes. The effects of increased species interaction, stratification, combined with increased algal changes in spawning areas, and possible invasion of productivity, would most likely reduce dissolved exotic species were not analyzed. oxygen levels in the lower layers of the lake (see Figure 14). Reducing pollutant loadings in the lake Northern Agriculture May Benefit would likely result in less severe impacts. One study raised the possibility that the annual mixing of a As a result of the relative increase in northern lake such as Lake Michigan may be disrupted. If agricultural productivity, agriculture could be winds and storms increased, such outcomes would enhanced in Minnesota, Wisconsin, and northern be less likely. Disposal of contaminated dredge Michigan with additional opportunities for the soils could increase water pollution. agriculture support sector. The presence of relatively poor soils, however, could limit Fish Productivity in Open Areas May Increase agricultural expansion. Increased cultivation in northern areas could increase erosion and runoff, The average annual thermal habitat would with negative impacts on surface and groundwater increase with a warmer climate (see Figure 15). If quality. AUGUST 1970* AUGUST 1975* BASE CASE X 40.6% 0.0% GISS X 80.5% 0.0% GFDL W 94.4% 5.9% OSU W 100% 28.8% * Base Case Years ///// Area That Is Anoxic (Has No Oxygen) Figure 14. Area of central basin of Lake Erie that becomes anoxic under doubled CO₂ scenarios. xlviii Effects of Climate Change 0 50 BASE CLIMATE 100 0 50 OSU DEPTH (M) 100 HABITAT: 0 + 2°C OF OPTIMUM TEMPERATURE + 5°C OF OPTIMUM TEMPERATURE 50 GISS 100 0 50 GFDL 100 JAN MAR JUN SEP DEC MONTH Figure 15. Increases in thermal habitat for lake trout in southern Lake Michigan under alternative climate scenarios. Abundance and Composition of Forests Could the nation's softwood and hardwood timber, and Change tobacco, corn, and soybeans are among its major crops. Over 85% of the nation's coastal wetlands Northern hardwood forests in dry sites in are in the Southeast, and over 43% of the finfish Michigan may die back and could become oak and 70% of the shellfish harvested in the United savannas or grasslands. In northern Minnesota, States are caught in the region. mixed boreal and northern hardwood forests may become completely northern hardwoods. This report focused on two regions within the Productivity in some wet sites in Michigan could Southeast: the Tennessee Valley and the improve. Commercially important softwood species Chattahoochee and Apalachicola Rivers. The could be replaced by hardwoods used for different Tennessee Valley Authority examined the potential purposes. Changes in forests could be evident in 30 vulnerability of its water management system to to 60 years. Whether reforestation with southern high and low riverflow scenarios (based on runoff species not currently in the region and CO2 estimates from GCMs). Flow in the Chattahoochee fertilization would mitigate these impacts was not River Basin was estimated using hydrologic analysis studied. to study impacts on the management of Lake Lanier, which supplies water to Atlanta. The Southeast estimates of outflow from the lake, along with estimates of the flow in the Apalachicola River, The Southeast is distinguished from the other were combined with potential wetland losses regions in this study by its warm temperatures, attributable to sea level rise to identify impacts on abundant rainfall, large coastal plain, and productive finfish and shellfish in Apalachicola Bay. Sea level marine fisheries. The region supplies about half of rise impacts for the entire Southeast were derived xlix Executive Summary from the national studies. Crop yields were scenarios to lead to the abandonment of 10 to 50% estimated for corn and soybeans, and changes in of the agricultural acreage in the region. The forest composition were analyzed at several sites studies did not consider introduction of new crops, across the region (see Figure 16). such as citrus, or the use of new technologies, such as biotechnology. Adverse Impacts on Agriculture and Forests Could Hurt the Region Most forests in the Southeast were estimated to have difficulty surviving the assumed climate Decreases in the relative productivity of change. Dieback of existing forests in such areas as southeastern agriculture were estimated under the Georgia and Mississippi could be particularly large. Agriculture Climate change could: decrease corn and soybean yields in hotter areas and could have mixed results elsewhere decrease cultivated acreage increase need for irrigation increase pest infestations Forests Higher temperatures could result in: significant dieback of southern forests with declines evident in 30 to 80 years TEMPERATURE SCENARIOS regeneration of species becoming 2xCO2 less 1xCO₂ difficult 6 5 Water Resources Increased temperature and changes TEMPERATURE (C) in precipitation could: 4 produce uncertain effects for water resource availability 3 affect water quality and flood risks GISS lower levels in some recreational lakes 2 GFDL 1 Sea Level Rise OSU Rising sea level could: 0 WINTER SPRING SUMMER FALL inundate a significant proportion of the region's coastal wetlands PRECIPITATION SCENARIOS flood some dry land areas 2xCO2 less 1xCO2 create significant costs for protecting 0.7 coastal resources 0.6 0.5 Fisheries 0.4 Higher water temperatures and rising 0.3 MILLIMETERS/DAY sea level could reduce fish and 0.2 shellfish populations 0.1 0 -0.1 Electricity -0.2 Higher temperatures could increase -0.3 electricity demand -0.4 -0.5 WINTER SPRING SUMMER FALL Figure 16. The Southeast. 1 Effects of Climate Change These changes could be evident in 30 to 80 years. in the region, with potential long-term consequences The forest studies did not consider whether more for agriculture and the economy. southern species could be transplanted and survive in the region, nor did they account for higher CO2 The studies in this report focused on Nebraska, concentrations, which could mitigate some losses. Kansas, Oklahoma, and Texas, and concentrated The combined effects of reduced agriculture and mainly on agriculture-related impacts. They forestry could lead to significant economic losses in estimated changes in corn, wheat, and soybean the Southeast. yields and in the demand for irrigation. Changes in runoff and leaching of chemicals from farms were Some Coastal Fish Species Would Be Harmed also examined (see Figure 17). Sea level rise could inundate most of the Crop Acreage Could Decline coastal wetlands and raise salinity levels, which could reduce the populations of gulf coast fisheries. The crop yield and economic adjustment In addition, higher temperatures may exceed the studies indicate that grain crop acreage could thermal tolerances of many species of shellfish in diminish in the region. The direction of changes in gulf coast estuaries, further reducing fish wheat and corn yields depends on the direct effects populations. Whether these species would be able of CO₂ on crop growth and the severity of climate to migrate to cooler water was not considered. change. If climate becomes hotter and relatively Some species, however, could increase in drier, yields could decrease. Whatever the climate abundance, while others may migrate into the change, relative productivity may decline compared region. with northern areas. As a result, crop acreage was estimated to drop by 4 to 22%. Such a reduction in The Studies Were Unable to Determine Regionwide agriculture could adversely affect the economy of Impacts on Water Resources the region. These studies did not consider use of new technologies or introduction of new crops. The Southeast currently has little winter snowcover. Therefore, seasonal runoff depends Demand for Irrigated Acreage Would Increase much more on changes in rainfall than on changes in temperature that affect the size of snowpack. The demand for irrigation on the farms that Analysis of the rivers managed by the Tennessee continue to grow grain crops could increase. Valley Authority showed that increased runoff could Irrigated acreage, which currently makes up about lead to higher riverflow and higher flood 10% of the total acreage and is growing, could probabilities, while less runoff could reduce flood increase by 5 to 30%. This report did not examine probabilities, but could lead to lower riverflow and how this demand would be satisfied, although the problems maintaining adequate supplies for Ogallala Aquifer could be a candidate. Other industrial use, powerplants, and dilution of effluent. impacts of global warming could change ground and Use of climate change scenarios produced surface water supplies and, possibly, surface water inconclusive results concerning the potential change quality. Changes in precipitation could affect the in flow in the Chattahoochee River. A study of the leaching of pesticides into groundwater and runoff management of Lake Lanier concluded that changes to surface waters in some cases, although the in operating rules would be sufficient to handle direction of change cannot be determined because higher or lower flows estimated in the scenarios, runoff and leaching of pesticides and soils are very although some uses would be restricted. sensitive to rainfall variability. The Great Plains FINAL THOUGHTS AND POLICY Agriculture is one of the main sources of income in the Great Plains. The States of Kansas, IMPLICATIONS Nebraska, Oklahoma, and Texas produced 80% of the nation's sorghum and 30% of the wheat crop in Because this is the most comprehensive study 1982. In recent years, increased use of water from to address the issue of the environmental effects of the Ogallala Aquifer has reduced groundwater levels climate change in the United States, we expect that a sizable debate will follow its publication. li Executive Summary Agriculture Higher temperatures could: reduce corn and wheat yields, and could have mixed effects on yields when considering both climate change and increased CO₂ reduce crop acreage TEMPERATURE SCENARIOS 2xCO2 less 1xCO2 6 5 Irrigation Demand TEMPERATURE (C) 4 Changes in agriculture are likely to result in increased irrigation demand 3 and acreage 2 GISS 1 GFDL 0 OSU WINTER SPRING SUMMER FALL Water Quality PRECIPITATION SCENARIOS 0 Changes in rainfall, runoff, pesticide 2xCO2 less 1xCO₂ 0 loadings, erosion, and irrigation could affect water quality 0.2 0.1 0 MILLIMETERS/DAY -0.1 Electricity -0.2 Higher temperatures could increase -0.3 electricity demand -0.4 -0.5 -0.6 WINTER SPRING SUMMER FALL Figure 17. The Great Plains. Considerable additional research and analyses are identified in this report by delays in the onset of likely to amplify, improve, and challenge these climate change, and by the pressure to solve today's findings. We expect further research to develop problems. Many adaptations would undoubtedly new insights into the role of climate, but precise occur as climate changes, but some decisions being forecasts must await more advanced climate models, made today have a long enough lifetime and which may require many years to develop. For sufficient risk to support consideration of the some time to come, our ability to provide national potential range of impacts of the greenhouse effect. and local officials with guidance may be limited to These decisions should be made if they make effects driven primarily by temperature and sea level economic and environmental sense for today's changes. conditions and are sufficiently flexible to handle changing climate. Given the uncertainty about the Apart from strategies to limit emissions of timing, magnitude, and regional scope of climate greenhouse gases (discussed in the companion change, we cannot plan for specific climate report), policymakers should consider policy options conditions in the future, but we can strive to be for adapting to global warming. Consideration of ready to respond to significantly changed climate these options is complicated by the uncertainties conditions in the future. lii Effects of Climate Change Conversely, natural resource management The U.S. government is strongly supporting the should not assume that climate will not change. All Intergovernmental Panel on Climate Change managers of natural resources that are sensitive to (IPCC) under the auspices of the United Nations climate should consider the vulnerabilities of their Environment Programme and the World systems to climate change and whether anticipatory Meteorological Organization. The IPCC has steps are prudent. In some cases, no anticipatory established a process for governments to follow action would be needed -- the systems can be when reviewing scientific information and policy adjusted and adapted as climate changes. In other options. The federal government is conducting areas, where long-term decisions on sensitive other activities on global climate change. The systems may result in irreversible impacts, Global Climate Protection Act of 1987 calls for a anticipatory actions to mitigate these potential scientific assessment of climate change, which is to effects may be required. It may make sense in be completed by 1989. This work will be sponsored some instances to change the rules under which by EPA and other federal agencies such as the long-term planning is done, such as zoning laws, to National Aeronautics and Space Administration, the allow for consideration of climate change in private- National Oceanic and Atmospheric Administration, sector decisions. Finally, research and education and the National Science Foundation, and are needed in many areas to improve our ability to coordinated through the IPCC. Also, the respond to these changes. In any case, managers Department of Energy and EPA have been asked to should reexamine their systems to consider ways to report to Congress on policy options for reducing improve the flexibility and resiliency of the systems CO₂ emissions in the United States. In addition, to handle these and other changes. The criteria to various federal agencies conduct significant research guide decisions should include consideration of the programs on climate. These research efforts on following factors: climate change are coordinated by the National Climate Program Office and the Committee on the uncertainties in the magnitude and Earth Sciences. The latter has produced a plan timing of effects; called Our Changing Planet: A United States Strategy for Global Change Research, which whether the lifetime of the plan, project, outlines federal research activities. or policy is long enough to be affected by climate change; The federal government can also take the lead in pursuing prudent policies in anticipation of whether effects of climate change are climate change, and many agencies can play a role irreversible; in preparing the country for the impacts. These include the Departments of the Interior, Energy, whether the policy or project will increase Health and Human Services, and Agriculture; the flexibility and resilience or restrict future U.S. Environmental Protection Agency; and the options; U.S. Army Corps of Engineers (see box on "Federal Activities"). However, adaptation should not occur whether a policy or action makes just at the federal level, for there will likely be a economic or environmental sense, even need to involve other nations, state and local without climate change; governments, industry, and even individuals. The regional studies in this report demonstrate that the uniqueness of the ecosystems or climate change cuts across manmade and natural manmade structures that may need systems, geographic boundaries, and government protection; and agencies. Research, technical guidance, planning, and creative approaches to resource management whether the impacts would be greater if will be needed in the future to prepare for the no anticipatory action were taken. impacts of climate change on the United States. liii Executive Summary FEDERAL ACTIVITIES THAT SHOULD CONSIDER CLIMATE CHANGE Sample questions relating to climate change impacts that federal agencies should consider: Agency Policy Question U.S. Environmental How should current wetlands protection programs be modified to Protection Agency accommodate future sea level rise and precipitation changes? Should regulatory approaches to air pollution be supplemented with incentive systems, new chemicals, or relocation policies? U.S. Department of Should national parks and wildlife refuges purchase land to accommodate the the Interior migration necessitated by climate change? Should additional parks and refuges be created? Are current activities increasing the vulnerability of species that might be threatened by climate change? Should the U.S. Geological Survey produce coastal area maps with finer contour intervals? How will climate change alter projected groundwater levels? Will current water policies in the West prove to have been ill-advised if the climate changes? U.S. Department of Do price support programs help or hinder the adjustments that climate change Agriculture may necessitate? To what extent could irrigation be increased on a sustainable basis if climate became drier? What actions would be necessary to maintain national forests as the climate changes? U.S. Army Corps How does a consideration of future climate change alter the relative merits of Engineers of alternative approaches to coastal protection, flood control, and navigation? Will climate change affect the success of wetlands protection efforts in Louisiana as administered under Section 404 of the Clean Water Act? Federal Emergency Will current rate caps on premiums enable the National Flood Insurance Management Agency Program to remain solvent if climate changes? U.S. Department of Are current programs adequate to address potential changes in mortality and Health and Human shifts in diseases resulting from climate change? Services liv CHAPTER 1 INTRODUCTION Since the beginning of the Industrial Revolution, greenhouse effect may raise atmospheric human activities have led to increased temperatures several degrees in less than a century. concentrations of greenhouse gases in the atmosphere. Fossil fuel burning, which releases CO2, CO, N2O, and other pollutants, has expanded CONGRESSIONAL REQUEST many times over. Changes in agriculture have led to increased emissions of CH₄ and N2O. FOR REPORTS Population growth has contributed to deforestation in many areas of the globe, which in turn has The significant implications of the greenhouse affected the global carbon cycle. Atmospheric effect have been the subject of discussion within the concentrations of tropospheric ozone and scientific community for the past three decades. In chlorofluorocarbons have also increased, primarily recent years, Members of Congress have held because of industrial activity. hearings and have begun to explain the implications for public policy. This interest was accentuated Scientists have concluded that the increase in during a series of hearings held in June 1986 by the greenhouse gas concentrations will eventually Senate Subcommittee on Pollution of the change global climate. In 1979, the National Environment and Public Works Committee. Academy of Sciences stated that a doubling of Following the hearings, members of the Senate carbon dioxide levels would lead to an increase of Environment and Public Works Committee sent a 1.5 to 4.5°C (2 to 8°F) in global air temperatures. formal request to the EPA Administrator, which Since then, other researchers have examined the asked the Agency to undertake two studies on increase in all greenhouse gases and have concluded climate change due to the greenhouse effect. (The that a greenhouse gas increase equivalent to CO2 letter is reprinted in Appendix C of this report.) doubling could occur as early as the 2030s, with a hypothesized commensurate global warming lagging One of the studies we are requesting should by several decades. examine the potential health and environmental effects of climate change. The Earth's atmosphere has undergone many This study should include, but not be limited cycles of warming and cooling in the past. to, the potential impacts on agriculture, Paleoclimatologists have estimated that at the glacial forests, wetlands, human health, rivers, lakes, maximum of the last ice age, which was about and estuaries, as well as other ecosystems 18,000 years ago, the Earth was approximately 5°C and societal impacts. This study should be (9°F) cooler than at present. This is generally designed to include original analyses, to attributed to changes in orbital characteristics identify and fill in where important research combined with lower trace gas concentrations and gaps exist, and to solicit the opinions of different climate feedbacks. knowledgeable people throughout the country through a process of public hearings Two aspects may make the current greenhouse and meetings. warming different from past climate changes. First, it will raise temperatures higher than the planet has Congress also requested that EPA prepare a experienced in the last 125,000 years. (During the study on policy options to stabilize current levels of Pliocene Epoch (2 to 5 million years ago), global atmospheric greenhouse gas concentrations. That temperatures were several degrees higher than they study analyzes policy options for limiting gas are now.) Second, past climate changes of concentrations including energy efficiency, comparable magnitude have generally occurred over alternative technologies, reforestation options, tens of thousands of years. Estimates are that the chlorofluorocarbon (CFC) reductions, and other options for limiting CH4 and N2O. It is entitled 1 Chapter 1 Policy Options for Stabilizing Global Climate and is a companion to this report. Congress requested the Direction and Magnitude studies in the Fiscal Year 1987 Continuing Resolution. Since the scenarios do not encompass all possible combinations of climate change due to increased greenhouse gases, the results do not GOALS OF THIS REPORT represent the entire range of possible effects. For example, there could be more or less rainfall, or higher or lower temperatures than estimated by This report builds on the past contributions of climate models. Yet, the results from various many scientists throughout the world, most notably scenarios help define the direction and magnitude of the reports by the National Academy of Sciences effects. First, we examined them to see if a (1979, 1983, 1987), the World Meteorological direction of change (e.g., more water, lower crop Organization and the International Council of yields) is evident. Second, we attempted to Scientific Unions (1986), the United Nations determine if the magnitude of change is significant. Environment Programme (1986), Scope 29 (1986), Third, we asked whether the results are consistent and the U.S. Department of Energy (1985a,b). It is with scientific theory. Outcomes outside the bounds an attempt to identify some of the sensitivities, of our results cannot be ruled out at this time. direction and magnitude, linkages, regional differences, national impacts, policy implications, Linkages and uncertainties associated with the effects of global climate warming. Individual environmental systems will not be affected by climate change in isolation. Water We hope it will provide useful information to climate modelers and effects researchers. We also resources, for example, may be affected not only by changes in water supply but also by changes in hope that officials, at all levels of government, will demand for water for such purposes as irrigation. be encouraged to examine the implications of Wildlife may be directly affected by changes in climate change for long-term policies. Since this is climate and indirectly affected by changes in habitat the first study of this type, we expect that a great due to climate change. This report attempts to deal more research, analysis, and planning will be identify linkages among effects, quantitatively where needed in the future. We do not pretend to have all possible and qualitatively elsewhere. Linkages are the answers. identified mainly in regions. Quantitative analysis of all linkages would change the numerical results of This report has been designed to identify the this report, in many cases exacerbating impacts. following: Sensitivities National Impacts Since the rate and extent of climate change on Impacts were analyzed on a national scale to see a regional level are uncertain, we cannot predict how the country as a whole may be affected by effects. However, we can identify the sensitivities of climate change and to see if latitudinal patterns systems to climate change. Our goal was to use a (such as northward shifts in species) are detectable. variety of scenarios to determine what climate Some analyses, such as coastal wetland impacts and variables are important in causing impacts and the changes in electricity demand, were conducted on a degree to which systems are sensitive to changes in national basis. Other national analyses, such as these variables. Specifically, we were interested in forests, were based on results from regional studies. identifying the sensitivity of systems to higher In some cases, national analyses estimated total temperatures and sea level, which are among the costs over the next century. No attempt was made changes most likely to occur following increased to assess the total national impact from climate greenhouse gas concentrations. (For further change, and conclusions about the total costs and discussion, see Chapter 2: Climate Change.) benefits of climate change should not be made. 2 Introduction Regional Impacts Fundamentally, these goals center on the identification of important issues and state-of-the- Effects were examined in several regions of the science investigations in each environmental system. United States for a number of reasons. As pointed Because each component of science and policy out above, linkages exist among many of the effects, development is at an early stage, the goals of the and these are likely to be seen on a regional scale. report are to develop insights and estimates of the For example, the supply of water in a river basin ranges of possible future effects and to use that may change as a result of climate change. The information for identifying where the policies and water resource in that basin may also be affected by research programs of EPA and other agencies changes in the demand for water for irrigation, should be reexamined. powerplant cooling, and other uses. Analysis of similar systems in different regions allows for comparison of impacts among regions. This report, STRUCTURE OF THE ANALYSIS however, does not attempt to identify "winners and losers." Important Systems Uncertainties This report focuses on the following systems, Many uncertainties are related to our knowledge which are important, are sensitive to climate, and about the rate and magnitude of warming and may be particularly affected by climate change: changes in regional weather patterns. As discussed Forests in Chapter 2: Climate Change, we do not know how much and how quickly climate may change and how Agriculture Sea Level Rise regional climates may change. Uncertainties also exist about how ecological and other systems will be Biodiversity Water Resources affected by climate change. We do not have empirical evidence on how these systems will Electricity Demand respond to higher temperatures and CO₂ levels, as Air Quality well as to different rainfall amounts. These Human Health uncertainties are reflected in the models used to Urban Infrastructure estimate climate change and impacts. This report attempts to clearly state these limitations. Regional Case Studies Policy Implications Four regional case studies were selected: the Southeast, the Great Lakes, California, and the The management of most natural resources has southern Great Plains. These regions were picked generally been undertaken assuming that climates because each is important for economic, social, and will not change. A change in climate could affect environmental reasons, and each offers some unique many of these resources and raise implications for current characteristics that make it an interesting resource management. This report discusses some example of the range of possible environmental policy implications of climate change, but it does not issues that may need to be considered. The lay out a prescriptive policy agenda. Southeast depends heavily on forestry and agriculture, and has extensive and fragile wetlands Research Needs and coastal ecosystems. The Great Lakes are the dominant natural resource in their region, supplying freshwater, fishery resources, and a pathway for The analysis in this report should provide climate shipping and transportation, and providing a natural modelers with information concerning how general laboratory for environmental issues that affect both circulation models could be improved. It should the United States and Canada. California already also help define research needs for future analysis must carefully manage its water supplies, and its of the potential impacts of climate change. agricultural industry provides many crops for the United States and a large share of the international 3 Chapter 1 market; it is among the most productive agricultural climate scenarios we used were based on outputs regions in the world. The Great Plains is one of the from general circulation models (GCMs) (see largest producers of grain crops in the world. Chapter 4: Methodology). Where possible, we tried Although these regions are diverse, they do not to obtain quantitative estimates of effects. encompass the entire range of regional differences However, the development of quantitative estimates in the United States. The analysis of effects in was constrained by the availability of well- these regions does not cover all potential impacts in documented models that included some interaction the United States. of the particular effect in question and climatic variability. We obtained additional information on National Studies sensitivities by reviewing the literature and by gathering expert judgment. The approach of using The effects on a number of systems were existing models, all of which were originally quantitatively analyzed on a national scale. National constructed for other purposes, makes the agricultural markets were analyzed with respect to interpretation of results instructive but somewhat their sensitivities to changes in yield derived from limited with respect to the full range of climatically our agricultural models. Options for adapting to a relevant questions that could be asked. sea level rise were examined on a national scale, as were possible health impacts. Forestry, water management, air quality, and biodiversity issues PROCESS FOR CONDUCTING were explored by analyzing the results of several of THIS REPORT the regional case studies with a broader perspective. In each case, the national-level analyses provide an additional level of qualitative integration that a We used an eight-stage process to define the purely regional analysis could not. The structure of scope of this report, select the projects, write the the regional and national studies is displayed in chapters, and review the results. Figure 1-1. Step 1: Initial Scoping of the Report ANALYTIC APPROACHES This stage immediately followed the request from the Senate Environment and Public Works Committee. We agreed on using the regional case Since we do not know how climate will change, study approach, on the four regions to be this report used scenarios of possible climate change investigated, and on using climatic scenarios. We to identify sensitivities of systems to climate. The Regional Case Studies Core Analytic California Outputs Areas Great Lakes Southeast Forests Great Plains Climate Agriculture Report Sea Level Rise Change to Congress Scenarios Biodiversity Water Resources Research Electricity Demand National Plan Air Quality Studies Human Health Models/ Urban Infrastructure Data Bases Policy Forests Agriculture Sea Level Rise Electricity Demand Health Figure 1-1. Elements of the effects report. 4 Introduction also decided not to attempt to analyze to all comments and modified proposals as environmental effects outside the United States in appropriate. EPA used a combination of this report. Our rationale for this decision was cooperative agreements, existing contracts, and based on available time and funds, and on the lack interagency agreements to fund projects for this of suitable models that would be immediately report. accessible to us. Step 5: Planning and Integration Step 2: Preparatory Workshops All the researchers met with EPA staff in We held two workshops in February and April October 1987 to discuss scenarios, goals, and 1987 in Boulder, Colorado, to prepare the report. approaches for the studies. Researchers discussed In the February workshop, sponsored and organized integration of projects within regions as well as the by the National Center for Atmospheric Research, commonality of approaches within disciplines. general circulation modelers convened to discuss some of the problems inherent in attempting to Step 6: Analysis understand the regional results from global models. Several major topics were discussed from the The National Center for Atmospheric Research standpoint of how the results from GCMs should be assembled the scenarios and distributed them to used in impact studies. A list of variables that researchers in the fall of 1987. Researchers would be available for use by effects researchers conducted their analysis over the winter and was produced at the end of the workshop. In prepared draft reports in March and April 1988. addition, several potential studies on aspects of the frequency of extreme weather events were identified. Step 7: Preliminary Project Review In April 1988, EPA assembled panels of The April workshop was organized with the scientists to provide a preliminary review of most of assistance of the University of Colorado. Approximately 100 scientists explored the major the agriculture, forestry, and hydrology projects. climate change-related issues in agriculture, forest The principal investigators of the appropriate effects, water resources, and sea level rise. Working projects were asked to present their work orally and in written drafts. EPA project managers used the groups in each discipline discussed the potential comments from the review panels to make impacts that climate change might have and the most important uncertainties to explore to arrive at corrections in the conduct of a few projects, and as a guide to interpreting the results of individual better predictions. The working groups were then projects and to writing this report. rearranged into regionally oriented groups. They identified a series of studies that would address the major scientific issues in each region. Step 8: Project and Report Peer Review Step 3: Identification of Potential Projects At least two to three peer reviewers examined the final reports from all principal investigators From the lists identified in the two Boulder before the EPA project managers accepted them. workshops, and from additional studies on urban During this time, EPA staff on the report project team wrote the overviews that are reflected in this and regional air quality subsequently identified internally by EPA, we arrived at list of investigators final report. In November 1988, a special from whom we would solicit proposals. The subcommittee of EPA's Science Advisory Board (SAB) was convened and asked to review the entire decision to solicit proposals was based primarily on the potential coverage of environmental issues in report. Following the SAB's written review, the each region. EPA project team responded to comments and produced the final version of the Effects Report. The draft of the report was sent to other federal Step 4: Reviews of Proposals agencies and the Office of Management and Budget for review and comment, and these comments were At least one intramural and two extramural also taken into account in the final version. reviewers examined each proposal. We responded 5 Chapter 1 STRUCTURE OF THIS REPORT RELATIONSHIP TO CURRENT NATIONAL AND This report is divided into several sections. INTERNATIONAL ACTIVITIES Section I consists of Chapter 2 on trends in emissions of greenhouse gases and potential impacts on climate; Chapter 3 on changes in variability; and National Research and Policy Activities Chapter 4 on the choice of scenarios and effects modeling. In Section II, the results of national The Global Climate Protection Act of 1987 analyses are presented. Each chapter covers a requested EPA to develop a national policy on different system. The chapters include an overview global climate change and to prepare an assessment of relevant regional studies, and they present results of scientific information. The very scope of this from national analyses. Each chapter discusses the issue suggests that this request can be fulfilled only current state of resource, reviews previous literature in cooperation with other federal agencies; hence, on climate change and the resource, discusses EPA is working with these agencies to formulate a studies used for this report, presents national results process to achieve this goal. The scientific from regional and national studies, and discusses assessment will be conducted in coordination with broader socioeconomic and policy implications. The the National Aeronautics and Space Administration, design and limitations for each study are presented the National Oceanic and Atmospheric only once -- in a regional chapter if it is a regional Administration, the National Science Foundation, study or in a national chapter if it is a national and other agencies. To the extent possible, this study. Section III contains results from the regional scientific assessment will also be developed on an case studies, with each chapter devoted to different international basis and should be available in 1990. regions. Each regional chapter describes the climate-sensitive systems in the region; reviews The development of a national policy will be previous studies on impacts of climate change on coordinated with the Department of Energy and the region; describes the structure of regional other natural resource departments. The goal will studies for the report; discusses regional climate be to build on this report and others under change scenarios; reviews the design, results, and development by federal agencies to identify the limitations of the studies; and discusses the broader adoptive policies and other measures that may be socioeconomic and policy implications of climate appropriate to deal with this issue. The nature of change for the region. The regional chapters this issue suggests that a continuous review of include relevant regional results from national domestic policy will be required for many years. studies. Not all regionally relevant results are presented in the appropriate regional chapters. International Activities Results for health are presented only in the health chapter in Section II. Section IV includes In 1987, the United Nations Environment conclusion chapters. Chapter 18 discusses directions for future research on climate change effects, and Programme (UNEP) and the World Meteorological Chapter 19 discusses policy implications and Organization (WMO) were asked by member recommendations. governments to establish an Intergovernmental Panel on Climate Change (IPCC) for the specific This report is designed to be an overview of the purpose of reviewing the scientific information and individual studies. Those studies are printed in potential response strategies. The WMO has appendix volumes. In this report, the studies are primary responsibility for the World Climate referenced by the author's name or names in Research Programme, and UNEP has responsibility parentheses and volume letter. Previously published for the World Climate Impacts Programme. The work is referenced by the author's name and the UNEP was the primary international agency responsible for negotiations leading to the Montreal year of publication. 6 Introduction Protocol To Protect the Ozone Layer. The first Scope 29. 1986. The Greenhouse Effect, Climatic meeting was held in November 1988, and Change, and Ecosystems (SCOPE 29). Bolin, B., B. subsequent meetings have been held in 1989 to Doos, J. Hager, and R. Warrick (eds). Chichester, organize activities. It is expected that the IPCC will England: John Wiley and Sons. be the primary forum for multilateral discussions between governments on this issue. United Nations Environment Programme. 1986. Effects of Changes in Stratospheric Ozone and Other governments and international agencies Global Climate. Titus, J.G. (ed). Washington, DC: are also examining this issue. Italy, Japan, and the U.S. EPA and United Nations Environment Netherlands held conferences in 1989. The United Programme. States has bilateral activities with the Soviet Union and China. The Organization for Economic U.S. Department of Energy. 1985a. Atmospheric Cooperation and Development and the International Carbon Dioxide and the Global Carbon Cycle. Energy Agency are examining their potential Trabalka, J.R. (ed). Washington, DC: Government contributions. Printing Office (DOE/ER-0239). U.S. Department of Energy. 1985b. Projecting the REFERENCES Climatic Effects of Increasing Carbon Dioxide. MacCracken, M.C., and F.M. Luther (eds). National Academy of Sciences. 1979. Carbon Washington, DC: Government Printing Office Dioxide and Climate: A Scientific Assessment. (DOE/ER-0237). Washington, DC: National Academy Press. World Meteorological Organization. 1986. Report National Academy of Sciences. 1983. Changing of the International Conference on the Assessment Climate. Washington, DC: National Academy of the Role of Carbon Dioxide and of Other Press. Greenhouse Gases in Climate Variations and Associated Impacts, Villach, Austria, 9-15 October National Academy of Sciences. 1987. Current 1985. World Climate Programme Report No. 661. Issues in Atmospheric Change. Washington, DC: Geneve, Switzerland: World Meteorological National Academy Press. Organization, The International Council of Scientific Unions, and the United Nations Environment Programme. 7 CHAPTER 2 GLOBAL CLIMATE CHANGE The Earth's climate has changed continuously These studies should be consulted for more detailed over the entire lifetime of our planet as a result of information. various natural causes. Recently, we have come to the realize that human activities may, in the near This chapter describes the climate system, the future, produce effects powerful enough to important causes of climate change for the next overwhelm these natural mechanisms and dominate century, and the so-called climate forcings, and it the changes of climate. By early in the next century, summarizes the various trace gases that human the planet's temperature may rise to a range never before experienced by our species, at a rate faster and to temperatures warmer than the Earth has experienced in the past million years. This The Greenhouse Effect anticipated temperature increase would be caused by an enhancement of the greenhouse effect. Gases in the atmosphere are virtually transparent to sunlight (shortwave Although the overall effect of increased radiation), allowing it to pass through the greenhouse gases is understood, many details are air and to heat the Earth's surface. The less clear, including both the timing of the predicted surface absorbs the sunlight and emits warming and its spatial distribution. This is because thermal radiation (longwave radiation) the response of the climate system to the additional back to the atmosphere. Because several greenhouse gases, including all the feedbacks and gases in the atmosphere, particularly interactions that would take place, is very water vapor (H₂O) and carbon dioxide complicated and not completely understood. In (CO₂), are not transparent to the addition, while the human-induced component of outgoing thermal radiation, they absorb the greenhouse effect increases in magnitude, other some of it and heat the atmosphere. The causes of climate change remain important, such as atmosphere emits thermal radiation, both changes in the amount of energy emitted by the sun, upward to outer space and downward to changes in the atmospheric composition due to the surface, further warming the surface. volcanic eruptions and human input of aerosols, internal redistributions of energy by El Niños, and This phenomenon is called the random, unpredictable variations. Thus, the task of greenhouse effect because in some predicting the future evolution of climate involves respects it describes how an actual not only understanding the response of the climate greenhouse works. Even without any system to increased concentrations of greenhouse human impacts, this natural greenhouse gases but also predicting the concentrations of these makes the Earth's surface about 33°C gases and the effects of other causes of climate (59°F) warmer than it would be without change. the atmosphere. Gases that are transparent to sunlight, but not to thermal Several detailed assessments of the current state radiation, are called greenhouse gases. of our knowledge of these projected climate changes have been conducted recently. These include If either the concentration of existing studies by the National Research Council (NRC, greenhouse gases increases or greenhouse 1979, 1983, 1987), the World Meteorological gases that were not there before are Organization (1986a,b), and the "state-of-the-art" added to the atmosphere, more thermal reports of the Department of Energy (MacCracken radiation will be absorbed and re-emitted and Luther, 1985a,b; NRC, 1985; Trabalka, 1985; downward, making the surface warmer Strain and Cure, 1985; White, 1985). Excellent than before. shorter summaries include Ramanathan (1988) and Chapters 2 and 3 of Lashof and Tirpak (1989). 9 Chapter 2 activities put into the atmosphere. It then describes and direction, and sea level, also have important important feedbacks in the climate system that act impacts on human activities. to amplify or dampen the climate change induced by the forcings. Uncertainties in our understanding of Figure 2-1 shows a schematic representation of these feedbacks are an important component of our the climate system. Changes in the amount of current uncertainty of the timing and amount of energy emitted by the sun, changes in the future climate change. Next, it discusses the recent atmospheric composition (such as from volcanic history of climate change, compares these eruptions and human input of aerosols and observations with theory, and presents theoretical greenhouse gases), and changes in the Earth's models of the climate and their projections of future surface (such as deforestation) can affect the climate change. Finally, the concluding section Earth's energy balance. Atmospheric and oceanic summarizes the extent of our knowledge about the circulation can redistribute the energy. future climate and discusses future research needs. The radiative balance of the planet, as shown in Figure 2-2, determines the global average vertical distribution of temperature. If the concentration of THE CLIMATE SYSTEM certain trace gases (carbon dioxide (CO₂), water vapor (H₂O), methane (CH₄), nitrous oxide (N₂O), The climate system includes all the interactive tropospheric ozone (O₃), and chlorofluorocarbons components of our planet that determine the (CFCs)) increases, the atmosphere's absorption of climate. This includes the atmosphere, oceans, land longwave radiation (thermal radiation from the surface, sea ice, snow, glaciers, and biosphere. Earth's surface) will increase. Some of this energy Climate change can be measured in terms of any will be radiated downward, heating the surface and part of the system, but it is most convenient to use increasing the surface temperature. Because the surface air temperature as a measure of climate, concentrations of all these gases are projected to since it is the parameter for which we have the best increase in the future, this effect and its timing must record, and it is measured where the most be compared to the other projected causes of important component of the biosphere humans climate change (forcings), and the response of the lives. Other components of the climate system, such climate system, to project the future climate. as precipitation, cloudiness, evaporation, windspeed Uncertainties are associated with all these factors. Changes of Solar Radiation SPACE ATMOSPHERE Terrestrial Radiation Clouds H₂O, N₂, O₂, CO₂, O3, etc. Air-Biomass Precipitation, -Land Aerosols Air-Ice Coupling Evaporation Coupling Heat Exchange Wind Stress BIOMASS ICE Ice-Ocean Changes of Coupling Atmosphere-Ocean Coupling LAND Atmospheric Composition OCEAN Changes of Land Features, Orography, Vegetation, Changes of Ocean Basin, Albedo, etc. Shape, Salinity, etc. Figure 2-1. The climate system. The principal interactions among components of the atmosphere, ocean, ice, and land surface, and some examples of external forcings are indicated (Gates, 1979). 10 Global Climate Change Climate Terminology Although this report avoids most technical jargon, some specialized terminology is inevitable. These terms are defined below. aerosols Tiny solid or liquid particles suspended in the atmosphere. Volcanic dust, forest fire smoke, and cloud droplets are examples. albedo Fraction of incoming solar radiation that is reflected. The fraction of energy absorbed is equal to 1 minus albedo. Thus, if the albedo of the earth's surface goes down, e.g, by snow melting that uncovers darker land, then the amount of energy absorbed would go up, raising the temperature. energy [also called heat balance] The process by which climate is determined. At any point balance on Earth, the incoming solar energy is balanced by outgoing thermal radiation, storage or release of heat in the surface, and redistribution of heat by wind and ocean currents. longwave [also called infrared radiation or thermal radiation] Electromagnetic radiation, like radiation light (solar radiation), radio waves and x-rays (microwaves), but of the wavelength that every object emits in order to cool itself. The Earth's surface emits longwave radiation in the wavelength region that is absorbed by CO2, H2O, and other greenhouse gases, producing the greenhouse effect, since these gases are much more transparent to sunlight. ppmv, ppbv Parts per million by volume, parts per billion by volume; units of concentration of gases. The 1989 concentration of CO₂ in the atmosphere is about 0.035% = 350 ppmv = 350,000 ppbv. The 1989 concentration of CFC-11 is about 0.000000026% = 0.00026 ppmv = 0.26 ppbv. sink Mechanism that removes a gas from the atmosphere. For example, oceans serve as a sink for CO₂, which dissolves in the surface waters. source Mechanism that adds a gas to the atmosphere. For example, foam blowing, leaky automobile air conditioners, and cleaning computer chips are all sources of CFCs. stratosphere The atmospheric layer above the troposphere, extending from the tropopause (the top of the troposphere) to about 50 kilometers (31 miles). The troposphere and stratosphere together contain more than 99.9% of the mass of the atmosphere. thermal Resistance to temperature change. Oceans have a much larger thermal inertia than inertia land because heat added or subtracted must come or go from a thick layer of well- mixed water rather than a thin immobile layer of soil. trace gas A gas with a very low concentration in the atmosphere. The important greenhouse trace gases are discussed in this chapter in the section on climate forcings. troposphere The lowest atmospheric layer, which extends from the Earth's surface to a height of about 8 kilometers (5 miles) in the polar regions, 12 kilometers (7 miles) in the midlatitudes, and 18 kilometers (11 miles) in the tropical regions. All weather and precipitation take place in the troposphere, which contains about 80% of the mass of the atmosphere. 11 Chapter 2 SPACE INCOMING OUTGOING RADIATION SOLAR RADIATION Shortwave Longwave 100 8 17 6 9 40 20 ATMOSPHERE Backscattered Net Emission by Air by Absorbed by Water Vapor, Emission Water Vapor, 19 CO2,O3 by Clouds Reflected Dust, O3 by Clouds Absorption by Clouds Water Vapor, 106 CO₂,O₃ 4 Latent Heat Flux Absorbed by Reflected Clouds Sensible by Surface Absorbed Heat Flux LONGWAVE RADIATION 46 115 100 7 24 OCEAN, LAND Figure 2-2. The Earth's energy balance. If the average amount of solar radiation received by the Earth (342 watts per meter²) is represented as 100 units, then the amplitudes of the various components of the energy flux are shown proportionately (MacCracken, 1985). CLIMATE FORCINGS and is a region where longwave radiation can escape relatively unimpeded to space. Both the past and future courses of climate change are determined by a combination of external The concentration of a number of trace gases in forcings, unforced internal fluctuations, and the the atmosphere is increasing as a result of human response of the climate system. This section briefly activities. Because the trace gases are very effective discusses the forcings that will be important in the absorbers of longwave radiation in the atmospheric window region, small (trace) amounts can have next century. large effects on the radiation balance, in effect Greenhouse Gases "dirtying" the atmospheric window. Trends and concentrations of some of these gases are shown in Table 2-1 and Figure 2-3. The projected relative If the Earth had no atmosphere, its average effects of these gases are shown in Figure 2-4. Each surface temperature, determined by the balance of the gases is discussed in more detail below. between incoming solar radiation and emitted longwave radiation at the surface, would be about Carbon Dioxide (CO₂) 0°F (-18°C), the same as the current temperature of the moon. The average temperature is actually a Combustion of fossil fuels and deforestation are hospitable 59°F (15°C) because of the natural increasing the concentration of CO2. Since Keeling greenhouse effect of H2O, CO₂, and O₃. Because began detailed measurements during the a large amount of the radiation in the wavelength International Geophysical Year in 1958 at Mauna band 7 to 13 micrometers is not absorbed by these Loa, Hawaii, the atmospheric concentration of CO₂ gases, it is referred to as the "atmospheric window," has risen from 315 ppmv (0.0315%) to a 12 Global Climate Change Table 2-1. Trace Gas Concentrations and Trends Concentrations Current annual Mid-21st Gas Pre-1850 1987 observed trends (%) century CO2 275.00 ppmvᵃ 348.00 ppmv 0.3 400.00-550.00 ppmv CH 0.70 ppmv 1.70 ppmv 0.8-1.0 1.80-3.20 ppmv N,O 0.29 ppmv 0.34 ppmv 0.2 0.35-0.40 ppmv CFC-11 0 0.22 ppbvᵇ 4.0 0.20-0.60 ppbv CFC-12 0 0.39 ppbvᵇ 4.0 0.50-1.10 ppbv CH₂CCl₃ 0 0.13 ppbvᵇ 7.0 CCI 0 0.00 0.08-0.10 ppbvᵇ 10.00-100.00 ppbvᵈ a Units of ppmv are parts per million by volume; 1 ppmv = 0.0001% of the atmosphere. Units of ppbv are parts per billion by volume; 1 ppbv = 0.001 ppmv. Value given is for 1986. ᶜTropospheric ozone only (below 12 kilometers). Values (below 9 km) for before 1850 are 0 to 25% less than present-day; values (12 kilometers) for mid-21st century are 15 to 50% higher. Value given is for 1985. Source: Ramanathan (1988), Lashof and Tirpak (draft 1989). current level of 350 ppmv. About half of the CO2 Chlorofluorocarbons (CFCs) put into the atmosphere each year remains in the atmosphere, with the rest absorbed in the ocean. These completely anthropogenic gases, the most Because society's basic energy sources (combustion important of which are known by the trade name of coal, oil, and natural gas) produce CO2, unless Freon, have been implicated not only in greenhouse strong energy conservation measures and shifts to warming but also in chemical destruction of other energy sources take place, it is projected that stratospheric ozone (O₃). Because of this, nations the atmospheric concentration of CO₂ will continue agreed to limit production of these gases in an to increase. As climate changes, the effectiveness of international agreement signed in Montreal in 1987. the oceanic sink for CO2 may also change, The most important of these gases are CFC-11 increasing or decreasing the fraction of CO2 that (CFCl₃) and CFC-12 (CF₂Cl₂). CFCs are used in remains in the atmosphere. CO2 contributes about refrigerants, aerosol propellants, foam-blowing half of the total anthropogenic greenhouse forcing. agents, and solvents. Substitutes for CFCs are being developed that are not as stable chemically and, Methane (CH₄) therefore, would not accumulate as fast in the atmosphere. The resulting lower concentration Although the methane concentration is now would produce a smaller greenhouse effect and increasing at a rate of about 1% per year and was would be less effective at destroying O₃. The much lower during the ice ages, the basic cycle is current fractional greenhouse contribution of not completely understood. Sources include rice CFC-11 and CFC-12 of 14% would probably paddies, cows, termites, natural gas leakage, decrease in the future, but the total CFC biomass burning, landfills, and wetlands. Although greenhouse effect would most likely increase for methane has a much lower atmospheric some time because of the long lifetime of these concentration than CO2 (currently 1.7 ppmv), it is gases. more effective at dirtying the atmospheric window and accounts for about 18% of current anthropogenic greenhouse forcing. 13 Chapter 2 CONCENTRATIONS OF TRACE GASES FROM ICE CORE AND ATMOSPHERIC SOURCES ICE CORE DATA ATMOSPHERIC DATA 1.7 1.6 1.4 CH4 (ppm) 1.2 CH4 CH4 (ppmv) 1.6 1.0 0.8 & 0.6 1.5 1850 1950 1978 1983 1988 1750 Source: Stauffer et al., 1985 Source: Blake & Rowland, 1988 355 Mauna Loa and Ice Core Data Mauna Loa 350 350 Ice Core 345 Monthly Concentrations of Carbon Dioxide at Mauna Loa, Hawaii 330 340 CONCENTRATION (ppm) 310 CO₂ CO2 CONCENTRATION (Parts Per Million by Volume) 335 330 7 290 4 325 V 320 270 315 250 310 1740 1860 1980 1955 1970 1985 Source: Neftel et al., 1985; Keeling et al., 1982 Source: Keeling, 1984; Keeling, unpublished, 1988 310 N₂O 308 350 N o CONCENTRATION (ppbv) 300 CONCENTRATIONS (ppbv) 306 N₂O 304 302 250 300 1600 1800 2000 1979 1983 1986 Source: Pearman et al., 1986 Source: Khahil, 1987 Figure 2-3. Greenhouse gas trends in ice cores and atmospheric instrument data. 14 Global Climate Change Although the ozone concentration is believed to be increasing in the troposphere, it is active chemically and has highly variable concentrations in time and 1880-1980 space. Responding to local air pollutants, such as Other (8%) nitrogen oxides (NOx) and hydrocarbons, ozone CFC-11 & -12 (8%) provides a complex link between local air pollution and global climate change. Other gases, such as N₂O (3%) carbon monoxide (CO) and volatile organic CO2 (66%) compounds, also play important roles in CH4 (15%) atmospheric chemistry and hence affect the greenhouse problem. Solar Variations 1980s The sun provides the energy source for all weather on the Earth, and the balance between Other (13%) incoming sunlight and outgoing longwave radiation determines the climate. Small variations in solar CO₂ (49%) radiation have the potential for causing climate CFC-11 & -12 (14%) changes as large as those caused by projected increases of greenhouse gases. Precise observations N₂O (6%) of the sun have been taken only for the past decade (Willson and Hudson, 1988). They show, however, CH4 (18%) that solar variations during this period have been so small that they would not be important compared with the other forcings discussed in this section. Since these high-quality observations have been Figure 2-4. Greenhouse gas contributions to global taken only for a short period, they do not rule out warming; estimated values based on concentration past or future variations of the sun that would be changes (1880-1980: Ramanathan et al., 1985; 1985, larger. But on the time scale of centuries, solar 1980s: Hansen et al., 1988). variations do not now seem to be an important factor. Nitrous Oxide (N2O) Volcanoes This gas, with both natural and anthropogenic Large volcanoes can significantly increase the sources, contributes about 6% to the enhanced concentration of stratospheric aerosols, decreasing greenhouse effect, although its concentration is only the amount of sunlight reaching the surface and about 0.31 ppmv. Its concentration is increasing at reducing surface temperatures by several tenths of a rate of about 1 ppbv per year, and sources include degrees for several years (Hansen et al., 1978, 1988; oceans, fossil fuel and biomass combustion, Robock, 1978, 1979, 1981, 1984). Because of the agricultural fertilizers, and land disturbances. thermal inertia of the climate system (discussed below), volcanoes can even be responsible for Ozone (O₃) climate changes over decades. It has been suggested that a significant part of the observed In addition to its role in the stratosphere as an global climate change of the past 100 years can be absorber of ultraviolet shortwave radiation, O₃ has attributed to the effects of volcanic eruptions an important impact on climate. This role is (Robock, 1979). Since large eruptions occur fairly complicated by its dependence on the altitude where frequently, this component of climate change will O₃ occurs. Both ozone increases in the troposphere have to be considered when searching past climate and lower stratosphere and ozone decreases in the for a greenhouse signal and when projecting future upper stratosphere would tend to warm the surface. climate change. 15 Chapter 2 Tropospheric Aerosols CLIMATE FEEDBACKS Natural sources, such as forest fires and sea Any imposed imbalance in the Earth's radiative spray, and human activities generate atmospheric balance, such as discussed above, will be translated aerosols in the troposphere. The concentrations into a changed climate through feedback vary greatly in space and time, and local sources are mechanisms that can amplify or decrease the initial important. Furthermore, these aerosols can imposed forcing. A feedback in which the final produce either warming or cooling, depending on temperature is higher than what it would have been their concentration, color, size, and vertical without the feedback is termed a "positive distribution. It is not now possible to definitively feedback." If the effect of the initially imposed determine their role in global climate. forcing is reduced, it is termed a "negative feedback." This section describes several of these Surface Properties mechanisms that are internal to the physical climate system and that involve the planet's biology and The Earth's radiative balance can also be chemistry. changed by variations of surface properties. While interactions with the oceans which cover 70% of the Although important climate feedback Earth's surface, are considered internal to the mechanisms have been identified, we may not climate system, land surfaces can exert a strong understand or even know about all the mechanisms influence on the climate. Human activities, such as involved in climate feedbacks. Figure 2-5 shows deforestation, not only provide a source of CO₂ and that even with the known physical climate feedbacks CH₄ to the atmosphere but also change the surface involved in changing surface temperature, the albedo and rate of evaporation of moisture into the potential interactions are complex. Current atmosphere. Detailed land surface models, state-of-the-art climate models attempt to incorporating the effects of plants, are now being incorporate most of the physical feedbacks that have developed and incorporated into climate model been identified but are forced, for example, to studies (Dickinson, 1984; Sellers et al., 1986). provide a very crude treatment for one of the most important ocean circulation because of large Internal Variations computer demands and inadequate ocean climate models. Another important and inadequately Even with no changes in the external forcings understood feedback clouds -- has been the discussed above, climate exhibits variations due to subject of recent climate calculations but, as internal rearrangements of energy both within the described below, is also treated crudely owing to atmosphere and between the atmosphere and the inadequate understanding of cloud physics and the ocean. The total amplitude and time scales of these small spatial scale on which clouds form as variations are not well understood; this contributes compared with the resolution of the climate models. to the difficulty of interpreting the past record and projecting the level of future climate change. Water Vapor Greenhouse Effect Some studies suggest that these random When the climate warms, more water (H₂O) variations can have amplitudes and time scales evaporates into the atmosphere from the warmed comparable to climate changes expected to be surface. This enhances the warming because it caused by greenhouse warming in the coming increases the greenhouse effect of the water vapor, decades (Lorenz, 1968; Hasselmann, 1976; Robock, producing still more evaporation. This positive 1978; Hansen et al., 1988). A large El Niño, such as feedback acts to approximately double imposed that observed in 1982-83, can take large amounts of forcings. Thus, an important greenhouse gas, H2O energy out of the oceans and warm the surface vapor, is controlled by the climate system itself. climate for a few years; this warming is then Transformations of H₂O between vapor and other superimposed on any warming due to the phases, liquid and solid, provide other important greenhouse effect. Our understanding of these El climate feedbacks discussed below. Niño/Southern Oscillation variations is improving, allowing us to account for this factor in interpreting past global climate change (Angell, 1988). 16 Global Climate Change OUTGOING LONGWAVE LATENT HEAT RADIATION FLUX SENSIBLE HEAT AND Volcanoes ABSORBED SOLAR POTENTIAL ENERGY FLUX Atmospheric RADIATION NET ENERGY BALANCE composition SUBSURFACE HEAT STORAGE Human Solar activities radiation THERMAL INERTIA OCEAN HEAT FLUX Planetary albedo Current Ice area TEMPERATURE Mixing depth Surface Meltwater Latitude albedo Temperature gradient Topography Snow area Geography Ocean albedo Atmospheric moisture Pressure Land albedo Vegetation capacity Horizontal wind gradient Soil properties Surface Evapotranspiration roughness Precipitation Vertical wind Soil moisture Atmospheric moisture Cloud content cover Relative humidity Figure 2-5. Physical climate feedback relationships. External forcings are indicated in underlined italics (Robock, 1985). Snow and Ice Clouds When climate warms, snow and ice cover are reduced, exposing land or ocean with a lower albedo Clouds respond directly and immediately to than the snow or ice. In addition, the albedo of the changes in climate and may represent the most remaining snow and ice is reduced owing to important uncertainty in determining the sensitivity meltwater puddles and debris on the surface. This of the climate system to the buildup of greenhouse acts to absorb more energy at the surface, further gases. Fractional cover, altitude, and optical depth enhancing the warming. This albedo feedback was of clouds can change when climate changes originally thought to be the dominant positive (Schlesinger, 1985). At the present time, clouds feedback effect of snow and ice, but we now have a large greenhouse effect, but this is offset understand that the thermal inertia feedback of sea (averaged over the globe) by their even stronger ice plays a much more important role (Manabe and cooling effect, because clouds reflect sunlight back Stouffer, 1980; Robock, 1983). to space (Ramanathan et al., 1989). Since the current greenhouse effect of clouds is larger than The thermal inertia feedback acts to increase the effect of an increase of CO₂ by a factor of 100, the thermal inertia of the oceans when climate small changes in clouds as climate changes can be warms by melting sea ice and exposing ocean waters very important in affecting the overall climate to the atmosphere. Since imposed climate change response to increases in trace gases. must then affect the ocean and atmosphere together rather than the atmosphere alone, this acts to If climate becomes warmer, more water will reduce the seasonal cycle of surface temperature evaporate into the atmosphere. Coupled with and is the prime reason for the enhancement of warmer surface temperatures, this may produce imposed climate change in the polar regions in the more upward motion of air, which would produce winter (Robock, 1983). more clouds. One way clouds could increase is to 17 Chapter 2 increase in area. This would raise the albedo of the of these gases induced by climate change, and they planet (except over polar snow and ice fields, which can influence the climate change itself through have an albedo larger than clouds), reflecting more changes in vegetation, and hence the surface heat sunlight back to space and having a cooling effect. and moisture balance. Such processes include Thus, the initially imposed warming is reduced, changes in releases of methane hydrates from ocean producing a negative feedback. Clouds already sediments, changes of land albedo due to shifting increase the planetary albedo from about 17% (if ecosystems, and changes in the ability of the oceans there were no clouds) to 30% (Ramanathan et al., to absorb CO₂ (this process is discussed in the next 1989). An increase of planetary albedo of only 0.5% section). would cut in half the warming imposed by doubled CO2 (Ramanathan, 1988). Methane hydrates are combinations of a methane molecule trapped in a lattice of water Other studies suggest that, especially in the molecules. They are found in ocean sediments and tropical regions, convection could increase, are stable under current pressure and temperature producing taller but narrower clouds. This would conditions in many ocean shelf regions. As the produce additional warming in two ways: (1) by climate warms, these conditions may change, reducing the cloud area, thus decreasing the releasing more methane into the atmosphere and planetary albedo; and (2) by decreasing the cloud enhancing the greenhouse effect. top temperature and reducing longwave radiation to space. This mechanism would be a positive As the climate warms, forests may shift closer feedback. In addition, convective clouds in the to the pole, producing a region with a lower albedo. tropical regions (thunderstorms) tend to produce The surface will thus absorb a larger fraction of large shields of high cirrus clouds, which have a sunlight, warming the Earth and producing a large greenhouse effect further enhancing the positive feedback, further enhancing the warming. warming. Cirrus clouds allow much sunlight to penetrate because they are so thin, but the cloud Oceans particles absorb the outgoing longwave radiation from the surface, efficiently trapping much of it Oceans play an important role in the climatic (Ramanathan, 1988). response to changed forcings because they absorb and emit both heat and CO₂, and because changing In the latest climate model simulations, it was ocean circulation can change the redistribution of found that clouds have a net positive feedback on energy internal to the climate system, as discussed global climate (Schlesinger, 1988), but the final above. When any of the above climate forcings are answer will be known only after more research. It applied to the climate system, the climate will start is not possible to be certain of the net effect of to change. Since both the climate forcings and the cloud feedbacks because of the complexity of clouds climatic response are time-dependent, and since the and their response to climate change. The climate system has a certain amount of inertia built complexity is because all the above properties of in owing to the response times of the ocean, the clouds can change simultaneously, because clouds exact relationship between the timing of the forcings affect both longwave and shortwave radiation, and the timing of the response is complex. Much of because clouds affect precipitation (which affects the lag between the imposed forcing and the land temperatures), and because the net effect climatic response depends on the oceans. The depends on the location of the cloud, surface upper 50 to 100 meters (164 to 328 feet) of the albedo, time of day, and time of year. ocean, called the mixed layer, responds relatively rapidly to imposed forcings. The deep ocean is also Biogeochemical Feedbacks important because its interactions could impose lags of as much as 100 years. In addition to the physical climate feedbacks discussed above, a number of positive The relative depth and role of the mixed layer, biogeochemical feedbacks may be important as well as the circulation of the ocean, will change (Lashof, 1989). These feedbacks can influence in a complex way in response to changed climate. future concentrations of greenhouse gases, especially Broecker (1987) has suggested that a rapid shift in CO2 and CH₄, through changes in sources and sinks ocean currents, such as the Gulf Stream, may occur 18 Global Climate Change as the climate warms, producing large regional and While the gradual warming seen in Figure 2-6 relatively rapid global climate changes. In during the past century is consistent with the preliminary tests with the Geophysical Fluid increasing greenhouse gases during this period, most Dynamics Laboratory models, when CO₂ is doubled, scientists suggest that a clear link has not yet been the oceanic circulation around Antarctica changes so established between observed temperatures and the as to increase the upwelling of cold bottom water. greenhouse effect. The large interannual variations As a result, cooling occurs in the Southern and the relatively flat curve from 1940 to 1975 show Hemisphere high latitudes for a period of several that there are also other important causes of climate decades as the rest of the globe warms! These two change. For example, large volcanic eruptions, such examples suggest that unforeseen climate events as Hekla in 1947 and Agung in 1963, and El Niños may be possible in the future and that until the certainly have produced some of the variations ocean response is well understood, the timing and shown in this record. Because of the projected amplitude of the climatic response to increased future emissions of greenhouse gases, global greenhouse gases and the other forcings will need to warming is likely to dominate these factors during remain the subject of additional research. the next century. Oceans are also the dominant sink of The global temperature record shown in Figure atmospheric CO2, absorbing about half of all CO₂ 2-6 can also be compared with the record for the that is put into the atmosphere each year by the United States for the same period shown in Figure combustion of fossil fuels and deforestation. The 2-7 (Hanson et al., 1989). While the globe as a amount of absorption is a strong function of oceanic whole has been generally warming, the lower 48 temperature, and shifts in oceanic circulation and states of the United States have actually been temperature may shift the fraction of CO₂ absorbed cooling for the past 40 to 50 years, although the in the future and, hence, change the rate of CO₂ high temperatures in the 1980s are among the accumulation in the atmosphere. As the oceans warmest on record. Since the lower 48 states of the warm, they may absorb a smaller fraction of the United States cover only 1.5% of the planet, this excess CO₂ in the atmosphere, thereby enhancing indicates that regional climatic variations, which may the warming (Lashof, 1989). In addition, oceanic be caused by changes in sea surface temperature chemical reactions change as climate changes. and wind circulation patterns, can be an important Oceanic production of dimethyl sulfide particles factor in the climate of small regions of the Earth. could also change as climate changes (Charlson et These factors will continue to be important as al., 1987). These particles serve as cloud global climate warms. For example, such regional condensation nuclei and may change the reflectivity events as the midwestern drought of 1988 may be of marine clouds by changing the number of related to changes in ocean temperature (Trenberth droplets in the clouds. et al., 1988) and can be greater than the effect of greenhouse gases on a national or larger scale. Observational Evidence of Climate Change On a longer time scale, proxy climate variables can indicate how climate has changed. An Thermometers have been used to actually intriguing record comes from a core drilled in the measure global climate change for more than 100 antarctic icecap at Vostok and is shown in Figure years in enough locations to provide an estimate of 2-8 (Barnola et al., 1987). The temperature record how the planet's climate has changed during this is deduced from the deuterium isotope ratio. The period. The most complete and up-to-date global past CO₂ concentration is actually measured from surface air temperature record available is shown in bubbles of ancient air trapped in the ice. The warm Figure 2-6 (Wigley et al., 1989). Other analyses, period of the past 10,000 years is called the including Hansen and Lebedeff (1988) and Vinnikov Interglacial and represents an anomalously warm et al. (1987), give similar results. Problems period compared with the climate of the past common to all these data sets include possible 100,000 years. It is projected that because of the contamination from urban heat islands, inadequate greenhouse effect, our climate will warm to a level spatial coverage of the Earth, and corrections much above even the level of the Interglacial, necessary to counteract the effects of changing the warmer in fact than the Earth has experienced for methods used to take observations from ships. the past million years. The rate of warming will 19 Chapter 2 0.5 0.0 -0.5 Northern Hemisphere 0.5 Temperature (°C) 0.0 -0.5 Southern Hemisphere 0.5 0.0 -0.5 Global 1860 1880 1900 1920 1940 1960 1980 2000 Year Figure 2-6. Hemispheric and global surface air temperatures, 1861-1988. The 1988 value is preliminary and includes data only through November. This record incorporates measurements made both over land and from ships. The smooth curve shows 10-year Gaussian filtered values. The gradual warming during this period is not inconsistent with the increasing greenhouse gases during this period, but the large interannual variations and the relatively flat curve from 1940 to 1975 show that there are also other important causes of climate change (Wigley et al., 1989). 13 Temperature (°C) 12 11 10 850 800 750 700 (mm) Pecipitation 650 600 1895 1905 1915 1925 1935 1945 1955 1965 1975 1985 Year Figure 2-7. Annual average surface air temperature (solid) and precipitation for the contiguous United States, 1895-1987. Note that the United States has been cooling for the past 50 years (Hansen et al., 1988). 20 Global Climate Change also be unprecedented. From Figure 2-8, it appears Two recent studies of CH₄ concentration in that the warming from the chill of the ice age 18,000 ancient air found in Greenland and Antarctic ice years ago to the Interglacial was very rapid, but in cores also have shown that CH4 concentration fact a warming of even 2°C in one century would varied with climate in prehistoric times (Stauffer et be much faster than this warming. al., 1988; Raynaud et al., 1988). Although the CH4 concentrations were not large enough to have an appreciable impact on the greenhouse effect, the CH₄ did vary in the same sense as CO2 and climate (see Figure 2-8). The CH₄ variations indicate that 280 sources of CH₄ increased in a warmer climate, which suggests that natural sources of CH4 may also 260 increase in the future as global climate warms, 240 220 CO2 (p.p.m.v.) further amplifying the greenhouse effect. 2.5 0 200 CLIMATE MODELS Temperature Change (°C) -2.5 180 In many sciences, such as biology, chemistry, or -5.0 physics, it is possible to investigate new phenomena -7.5 by doing research in a laboratory. In the field of climate, this is not possible. One cannot bring the -10.0 Earth's climate system into a room and perform experiments on it, changing the trace gas 0 40 80 120 160 Age concentration or increasing the amount of sea ice. (Thousands of years before present) It is not possible to have two identical systems, one a control and one that is changed to compare the Figure 2-8. Temperatures and carbon dioxide outcomes. There is only one climate system, and concentrations for the past 160,000 years at Vostok, humans are now performing an uncontrolled Antarctica. Since these observations were taken experiment on it by polluting it with CO2, CH4, near the South Pole, they show larger temperature CFCs, and other trace gases. variations (by a factor of 2 or 3) than took place averaged over the whole globe (Barnola et al., To try to understand how the global climate 1987). will change in response to human activities, researchers have applied various approaches. The climates of other planets, particularly Venus and Figure 2-8 shows that during the entire Mars which are the most Earth-like, can give us 160,000-year period, the atmospheric CO₂ some ideas about climate under very different concentration varied along with the temperature. conditions. However, their atmospheres are not When it was warmer, the CO2 concentration was similar enough to Earth's to give us definitive higher, although it never approached the current answers about the next 100 years here. The history of the Earth's climate is another area we could level of 350 ppmv. It is not known whether the climate change preceded the increase in CO2, study, but since many different forcings of similar whether the increase in CO₂ preceded the warming, strengths have been acting, and since the data or whether they both happened simultaneously. It coverage is imperfect, it has not been possible to is well accepted that the changing orbit of the Earth definitively isolate the roles of the different forcings. produced the ice ages (the Milankovitch Attempts have been made to use rotating tanks of Hypothesis), and this recently discovered variation water or other fluids (called dishpan experiments) of CO2 certainly worked to enhance the climate as models for the atmosphere, but these are changes caused by the changing orbit. These imperfect as they cannot simulate realistic heating natural processes are now being overwhelmed by profiles or the detail of the real climate system. the human impact of fossil fuel burning and deforestation. 21 Chapter 2 The most useful tool to investigate future They generally use different schemes for computing climate is the computer model of the climate cloud height, cloud cover, and optical properties. system. In a climate model, the various physical The models also differ in their treatment of ground laws that determine the climate, such as hydrology, sea ice, surface albedo, and diurnal and conservation of energy, conservation of mass, and seasonal cycles (Schlesinger and Mitchell, 1985). the gas law, are expressed as mathematical Perhaps the most important differences lie in the equations that specify the relationship between treatment of oceans, ranging from prescribed sea different variables, such as temperature, pressure, surface temperatures to "swamp" oceans with mixed- wind, and precipitation. By specifying the various layer thermal capacity but no heat transport, to climate forcings, it is possible to calculate the mixed layers with specified heat transport, to full climate. An experiment can be performed by oceanic GCMs. Models are constantly becoming doubling CO2, for instance, and comparing the more complex and sophisticated as new resulting climate to the current CO₂ concentration. understanding of the physics evolves and faster Many theoretical calculations can be made to test computers become available. the importance of various assumptions and various proposed feedback mechanisms. One of the first experiments used to test any climate model is its ability to simulate the current The simplest climate model is the climate. In these tests, the various state-of-the-art zero-dimensional global average model, which can climate models have differences. Grotch (1988) has be used to give a global-average measure of climate recently compared the simulations of surface air but cannot consider many important processes and temperature and precipitation of four recent GCM cannot give regional distribution of climate changes. simulations and found that although they do a Models that are one-dimensional in the vertical, reasonable job of simulating global values, the called radiative-convective models, or in the simulations at the regional scale are poor. He horizontal, called energy-balance models, are very compared model simulations and observations on useful for quickly and inexpensively testing various gridpoints, where each gridpoint "represents a components of the climate system. However, to region of about 400 kilometers (250 miles) by 400 calculate the location of future climate change, and kilometers or larger, or roughly the size of to incorporate all the important physical Colorado, even though regions of this size may have interactions, especially with atmospheric circulation, very diverse local climates" (Grotch, 1988). He fully three-dimensional general circulation models found differences between models and observations (GCMs) are necessary. These sophisticated models (see Table 2-2), and between models, particularly solve simultaneous equations for all the important for smaller regions. Grotch concluded that GCMs climate variables in three dimensions. The world is cannot currently project regional changes of broken up into a discrete grid of boxes placed side precipitation or temperature. by side and stacked to cover the globe. The biggest and fastest supercomputers available are used, but Given the current state of the art, how can computer speed and size constraints limit the size of these models be used? As discussed in Chapter 4, these grid boxes to 250 to 1,000 kilometers (150 to model simulations can be of use even in their crude 600 miles) on a side and to a height of 1 to 5 state. In the first place, even if the models do not kilometers (0.6 to 3 miles). Thus, in one of these exactly reproduce the current climate, perhaps the grid boxes, all the complexity of weather and differences between their simulations of current and horizontal variation is reduced to one number for future climates provide an estimate of potential temperature, one for cloudiness, and so forth. The future changes. In addition, the models produce a equations used to represent the physical and data set of all the variables needed for impact chemical processes involved are also simplifications assessment that are physically consistent within the of real-world processes. physics of the model. Thus, although the actual model projections can not be taken as predictions of Different climate modelers represent physical the future, they are useful in providing scenarios for processes in different ways. In all the models, the impact assessment. As model projections become radiation schemes attempt to account for the more accurate in the future, the scenarios they radiatively significant gases, aerosols, and clouds. generate will become more accurate. 22 Global Climate Change Table 2-2. Differences Between Winter and Summer Temperature Estimates for Four GCMs and Observed Temperatures Domain of comparison Variable and model Global North America Contiguous U.S. Midwestern U.S. December-January-February Observed median temperature (°C) 8.5 -5.8 0.9 -1.5 Difference in median temperatures (GCM Observation) CCM -1.6 -0.3 -2.1 -0.5 GFDL 1.5 -1.8 -0.8 -1.3 GISS 0.8 -0.5 0.0 1.1 OSU 0.3 0.5 -0.6 -1.0 June-July-August Observed median temperature (°C) 13.9 18.9 23.0 23.0 Difference in median temperatures (GCM Observation) CCM 1.3 6.0 6.3 6.8 GFDL -0.2 0.6 0.1 3.7 GISS 0.4 -3.1 -4.5 -4.8 OSU -0.6 -2.2 -2.2 -1.6 CCM = Community Climate Model (National Center for Atmospheric Research). This is the Washington version discussed in Chapter 3: Variability. Source: Grotch (1988). In generating scenarios, an important convenient to refer to an "equivalent doubling of component is the timing of future climate changes. CO2," which means the effect of all the greenhouse This depends not only on the timing of the changes gases together that would have the same effect as in the forcing (how rapidly trace gas concentrations doubling CO₂. This would occur with less than a increase) but on the sensitivity of the climate system doubling of CO2 itself, since the other to these forcings. A simpler question to ask is, anthropogenic greenhouse gases currently contribute "What would be the change in global average approximately the same amount of warming as does surface air temperature if the CO₂ concentration in CO₂. While it is reasonable to lump all the the atmosphere were doubled from the preindustrial greenhouse gases together for the purposes of level, all other climate forcings were held constant, calculating the radiative effect, the other effects of and the climate became completely adjusted to the these gases, such as fertilization of plants by CO2 new radiative forcing?" This is referred to as the or chemical reactions, must be determined based on equilibrium climate sensitivity to a CO₂ doubling. the actual concentrations of each gas. When discussing climate change, it is sometimes 23 Chapter 2 Model Projections of a Doubled-CO₂ World all of the warming of the past 100 years were due to greenhouse gases, a doubling of CO2 would warm Several climate modeling groups have climate by about 2°C. If, however, we allow for conducted GCM experiments to calculate the other possible forcings (including natural equilibrium climate response to doubled CO2. variability), for uncertainties in ocean heat uptake These include researchers at the National Center and the timing of the climate response, and for for Atmospheric Research (NCAR), Oregon State uncertainties in preindustrial greenhouse gas University (OSU), NOAA's Geophysical Fluid concentrations (Hansen et al., 1985; Wigley and Dynamics Laboratory (GFDL), NASA's Goddard Schlesinger, 1985; Wigley et al., 1986), then from Institute for Space Studies (GISS), and the United past data we can only say that a CO2 doubling Kingdom Meteorological Office (UKMO). The might produce a global climate change anywhere in results from the different experiments depend on the range of 0 to 6°C (Wigley, personal the assumptions made, especially on the treatment communication). Wigley et al. (1989) point out that of clouds and of oceans. The models predicted while the global warming of the past 137 years is global temperature increases of 2.8 to 5.2°C and highly significant statistically, it is not possible to global precipitation increases of 7 to 16% (see definitively attribute this warming to a specific Table 2-3). cause. Attempts have also been made to determine The actual path that the climate system would climate sensitivity from past data. If we could take to approach the equilibrium climate would be accurately determine the strength and timing of all determined by the time scales of the forcings and the climate forcings that have competed with the the various elements of the climate system and is greenhouse effect in the past, we could account for referred to as the transient response. Because the them, and the residual warming would be a measure climate system response lags behind the forcing, a of the greenhouse effect to date. Unfortunately, our built-in unrealized warming will always occur in the knowledge of both past climate change and the future, even if no more greenhouse gases are added. responsible forcings is too poor to reliably Thus, some future climate response to the determine the sensitivity of climate to greenhouse greenhouse gases that were put into the atmosphere warming. Wigley and Raper (1987) estimate that if in the past will certainly occur, even if emissions were stopped today. Table 2-3. General Circulation Model Predictions of Globally Averaged Climate Change Due to Doubled CO2 Surface air temperature Precipitation Model increase (°C) increase (%) GFDL 4.0 8.7 GISS 4.2 11.0 NCAR 3.5 7.1 OSU 2.8 7.8 UKMO 5.2 15.8 Source: Karl et al. (1989). 24 Global Climate Change What We Know About Future Climate A panel of experts convened by the National Academy of Sciences (National Research Council, 1987) recently considered the climatic response to increasing greenhouse gases and gave the following assessment, including their estimate of scientific confidence in the predictions. This table is limited to a summary of their conclusions about "the possible climate response to increased greenhouse gases" only; the full report should be consulted for the details. Large Stratospheric Cooling (virtually certain). The combination of increased cooling by additional CO₂ and other trace gases, and reduced heating by reduced O3," will lead to a major lowering of temperature in the upper stratosphere." Global-Mean Surface Warming (very probable). For an equivalent doubling of CO₂, "the long-term global-mean surface warming is expected to be in the range of 1.5 to 4.5°C." Global-Mean Precipitation Increase (very probable). "Increased heating of the surface will lead to increased evaporation and, therefore, to greater global mean precipitation. Despite this increase in global average precipitation, some individual regions might well experience decreases in rainfall." Reduction of Sea Ice (very probable). This will be due to melting as the climate warms. Polar Winter Surface Warming (very probable). As a result of sea ice reduction, polar surface air may warm by as much as three times the global average. Summer Continental Dryness/Warming (likely in the long term). Found in several but not all studies, it is mainly caused by earlier termination of winter storminess. "Of course, these simulations of long-term equilibrium conditions may not offer a reliable guide to trends over the next few decades of changing atmospheric composition and changing climate." Rise in Global Mean Sea Level (probable). This will be because of thermal expansion of seawater and melting or calving of land ice. REFERENCES Detecting the Climatic Effects of Increasing Carbon Dioxide. Washington, DC: U.S. Department of Angell, J.K. 1988. Impact of El Niño on the Energy, (DOE/ER-0235), pp. 91-107. delineation of tropospheric cooling due to volcanic eruptions. Journal of Geophysical Research Barnola, J.M., D. Raynaud, Y.S. Korotkevich, and 93:3697-3704. C. Lorius. 1987. Vostok ice core provides 160,000-year record of atmospheric CO2. Nature Barnett, T.P. 1985. On long-term climate change 329:408-414. in observed physical properties of the ocean. In: MacCracken, M.C., and F.M. Luther, eds. 25 Chapter 2 Blake, D.R., and F.S. Rowland. 1988. Continuing Hansen, K., G.A. Maul, and T.R. Karl. 1989. Are worldwide increase in tropospheric methane, 1978- atmospheric "greenhouse" effects apparent in the 1987. Science 239:1129-1131. climatic record of the contiguous U.S. (1895-1987)? Geophysical Research Letters 16:49-52. Broecker, W. 1987. Unpleasant surprises in the greenhouse? Nature 328:123-126. Hasselmann, K. 1976. Stochastic climate models. I. Theory. Tellus 28:473-485. Charlson, R., J. Lovelock, M. Andreae, and S. Warren. 1987. Oceanic phytoplankton, atmospheric Karl, T.R., H. Diaz, and T. Barnett. 1989. Climate sulphur, cloud albedo and climate. Nature variations of the past century and the greenhouse 326:655-661. effect (A report based on the First Climate Trends Workshop). Rockville, MD: National Climate Dickinson, R.E. 1984. Modeling evapotranspiration Program Office/NOAA. for three-dimensional global climate models. In: Hansen, J.E., and T. Takahashi, eds. Climate Keeling, C.D. 1984. The global carbon cycle: What Processes and Climate Sensitivity. Washington, DC: we know and could know from atmospheric, American Geophysical Union, pp. 58-72. biospheric, and oceanic observations. In: Institute for Energy Analysis. Proceedings: Carbon Dioxide Dickinson, R.E. 1986. The climate system and the Research Conference: Carbon Dioxide, Science and modelling of future climate. In: Bolin, B., B. Doos, Consensus. Washington, DC: U.S. Department of J. Hager, and R. Warrick, eds. The Greenhouse Energy, (DOE CONF-820970). Effect, Climatic Change, and Ecosystems (SCOPE 29). Chichester, England: John Wiley and Sons, Khalil, M.A.K., and R.A. Rasmussen. 1987. pp. 207-270. Nitrous oxide: Trends and global mass balance over the last 3,000 years. Annals of Glaciology 10:73-79. Gates, W.L. 1979. The physical basis of climate. In: Proceedings of the World Climate Conference. Lashof, D. 1989. The dynamic greenhouse: WMO No. 421. Geneva, Switzerland: World Feedback processes that may influence future Meteorological Organization, pp. 112-131. concentrations of greenhouse gases. Climatic Change. Grotch, S.L. 1988. Regional Intercomparisons of General Circulation Model Predictions and Lashof, D., and D. Tirpak, eds. 1989. U.S. Historical Climate Data. Washington, DC: U.S. Environmental Protection Agency. Policy Options Department of Energy. (DOE/NBB-0084). for Stabilizing Global Climate. Draft Report to Congress. Washington, DC: U.S. Environmental Hansen, J., I. Fung, A. Lacis, S. Lebedeff, D. Rind, Protection Agency, Office of Policy, Planning and R. Ruedy, and G. Russell. 1988. Global climate Evaluation. February. changes as forecast by the GISS 3-D model. Journal of Geophysical Research 93:9341-9364. Lorenz, E.N. 1968. Climatic determinism. Meteorological Monographs 30:1-30. Hansen, J., and S. Lebedeff. 1988. Global surface air temperatures: update through 1987. Geophysical MacCracken, M.C. 1985. Carbon dioxide and Research Letters 15:323-326. climate change: Background and overview. In: MacCracken, M.C., and F.M. Luther, eds. Hansen, J., G. Russell, A. Lacis, I. Fung, and D. Projecting the Climatic Effects of Increasing Carbon Rind. 1985. Climate response times: Dependence Dioxide. Washington, DC: U.S. Department of on climate sensitivity and ocean mixing. Science Energy, (DOE/ER-0237), pp. 1-23. 229:857-859. MacCracken, M.C., and F.M. Luther, eds. 1985a. Hansen, J.E., W.C. Wang, and A.A. Lacis. 1978. Projecting the Climatic Effects of Increasing Carbon Mount Agung eruption provides test of a global Dioxide. Washington, DC: U.S. Department of climatic perturbation. Science 199:1065-1068. Energy. (DOE/ER-0237). 26 Global Climate Change MacCracken, M.C., and F.M. Luther, eds. 1985b. Raynaud, D., J. Chappellaz, J.M. Barnola, Y.S. Detecting the Climatic Effects of Increasing Carbon Korotkevich, and C. Lorious. 1988. Climatic and Dioxide. Washington, DC: U.S. Department of CH4 cycle implications of glacial-interglacial CH₄ Energy. (DOE/ER-0235). change in the Vostok ice core. Nature 333:655-657. Manabe, S., and R.J. Stouffer. 1980. Sensitivity of Robock, A. 1978. Internally and externally caused a global climate model to an increase of CO₂ climate change. Journal of the Atmospheric concentration in the atmosphere. Journal of Sciences 35:1111-1122. Geophysical Research 85:5529-5554. Robock, A. 1979. The "Little Ice Age": Northern National Research Council. 1979. Carbon Dioxide Hemisphere average observations and model and Climate: A Scientific Assessment. Washington, calculations. Science 206:1402-1404. DC: National Academy Press. Robock, A. 1981. A latitudinally dependent National Research Council. 1983. Changing volcanic dust veil index, and its effect on climate Climate. Washington, DC: National Academy simulations. Journal of Volcanology and Press. Geothermal Research 11:67-80. National Research Council. 1985. Glaciers, Ice Robock, A. 1983. Ice and snow feedbacks and the Sheets, and Sea Level: Effect of a CO,-Induced latitudinal and seasonal distribution of climate Climatic Change. Washington, DC: U.S. sensitivity. Journal of the Atmospheric Sciences Department of Energy. (DOE/EV/60235-1). 40:986-997. National Research Council. 1987. Current Issues Robock, A. 1984. Climate model simulations of in Atmospheric Change. Washington, DC: the effects of the El Chichón eruption. Geofísica National Academy Press. Internacional 23:403-414. Neftel, A., E. Moor, H. Oeschger, and B. Stauffer. Robock, A. 1985. An updated climate feedback 1985. Evidence from polar ice cores for the diagram. Bulletin of the American Meteorological increase in atmospheric CO2 in the past two Society 66:786-787. centuries. Nature 315:45-47. Schlesinger, M.E. 1985. Analysis of results from Pearman, G.I., D. Etheridge, F. deSilva, and P.J. energy balance and radiative-convective models. In: Fraser. 1986. Evidence of changing concentrations MacCracken, M.C., and F.M. Luther, eds. of atmospheric CO2, N2O, and CH₄ from air Projecting the Climatic Effects of Increasing Carbon bubbles in Antarctic ice. Nature 320:248-250. Dioxide. Washington, DC: U.S. Department of Energy, (DOE/ER-0237), pp. 280-319. Ramanathan, V. 1988. The greenhouse theory of climate change: A test by an inadvertent global Schlesinger, M.E. 1988. How to make models for experiment. Science 240:293-299. behaviour of clouds. Nature 336:315-316. Ramanathan, V., R.D. Cess, E.F. Harrison, P. Schlesinger, M.E., and J.F.B. Mitchell. 1985. Model Minnis, B.R. Barkstrom, E. Ahmad, and D. projections of the equilibrium climatic response to Hartmann. 1989. Cloud-radiative forcing and increased carbon dioxide. In: MacCracken, M.C., climate: Results from the Earth Radiation Budget and F.M. Luther, eds. Projecting the Climatic Experiment. Science 243:57-63. Effects of Increasing Carbon Dioxide. Washington, DC: U.S. Department of Energy, (DOE/ER-0237), Ramanathan, V., R.J. Cicerone, H.B. Singh, and pp. 81-147. J.T. Kiehl. 1985. Trace gas trends and their potential role in climate change. Journal of Sellers, P.J., Y. Mintz, Y.C. Sud, and A. Dalcher. Geophysical Research 90:5557-5566. 1986. A simple biosphere model (SiB) for use within general circulation models. Journal of the Atmospheric Sciences 43:505-531. 27 Chapter 2 Somerville, R.C.J., and L.A. Remer. 1984. Cloud Wigley, T.M.L., P.D. Jones, and P.M. Kelly. 1986. optical thickness feedbacks in the CO₂ climate Empirical climate studies. In: Bolin, B., B. Doos, problem. Journal of Geophysical Research J. Hager, and R. Warrick, eds. The Greenhouse 89:9668-9672. Effect, Climatic Change, and Ecosystems (SCOPE 29). Chichester, England: John Wiley and Sons, Stauffer, B., G. Fischer, A. Neftel, and H. Oeschger. pp. 271-322. 1985. Increase of atmospheric methane recorded in Antarctic ice cores. Science 229:1386-1389. Wigley, T.M.L., P.D. Jones, P.M. Kelly, and S.C.B. Raper. 1989. Statistical significance of global Stauffer, B., E. Lochbronner, H. Oeschger, and J. warming. Proceedings of the Thirteenth Annual Schwander. 1988. Methane concentration in the Climate Diagnostics Workshop. Washington, DC: glacial atmosphere was only half that of the National Oceanographic and Atmospheric preindustrial Holocene. Nature 332:812-814. Administration. pp. A-1 -A-8. Strain, B.R., and J.D. Cure, eds. 1985. Direct Wigley, T.M.L., and S.C.B. Raper. 1987. Thermal Effects of Increasing Carbon Dioxide on Vegetation. expansion of sea water associated with global Washington, DC: U.S. Department of Energy. warming. Nature 330:127-131. (DOE/ER-0238). Wigley, T.M.L., and M.E. Schlesinger. 1985. Trabalka, J.R., ed. 1985. Atmospheric Carbon Analytical solution for the effect of increasing CO₂ Dioxide and the Global Carbon Cycle. Washington, on global mean temperature. Nature 315:649-652. DC: U.S. Department of Energy. (DOE/ER-0239). Willson, R.C., and H.S. Hudson. 1988. Solar luminosity variations in solar cycle 21. Nature Trenberth, K.E., G.W. Branstator, and P.A. Arkin. 332:810-812. 1988. Origins of the 1988 North American drought. Science 242:1640-1645. World Meteorological Organization. 1986a. Atmospheric Ozone 1985. Assessment of Our Vinnikov, K.Ya., P.Ya. Groisman, K.M. Lugina, and Understanding of the Processes Controlling Its A.A. Golubev. 1987. Variations in Northern Present Distribution and Change. Global Ozone Hemisphere mean surface air temperature over Research and Monitoring Project - Report No. 16. 1841-1985. Meteorology and Hydrology No. 1: Geneva, Switzerland: World Meteorological 45-55. In Russian. Organization. White, M.R., ed. 1985. Characterization of World Meteorological Organization. 1986b. Information Requirements for Studies of CO2 Report of the International Conference on the Effects: Water Resources, Agriculture, Fisheries, Assessment of the Role of Carbon Dioxide and of Forests and Human Health. Washington, DC: U.S. Other Greenhouse Gases in Climate Variations and Department of Energy. (DOE/ER-0239). Associated Impacts, Villach, Austria, 9-15 October 1985. World Climate Programme Report No. 661. Geneva, Switzerland: World Meteorological Organization. 28 CHAPTER 3 VARIABILITY FINDINGS current climate for the two GCMs for selected U.S. regions reveals interesting contrasts and similarities regarding the reproduction of climate variability. A changed climate variability (defined in the Simulation of variability is reasonably good in several cases. following section of this chapter) associated with climate change could significantly affect natural resources. However, lack of information on Although some discrepancies exist between potential changes in climate variability has limited actual and estimated temperature and the completeness of climate change impact studies precipitation values, the models simulate the presented in this report. It is not possible to seasonal cycles of temperature and precipitation definitively state how climate variability will change reasonably well in the four regions investigated. with a changed climate because model results are mixed. At this time, there is not a strong case for The models make errors (generally altering the assumption of no change in variability overpredictions) in predicting daily and year-to- used in the scenarios for this report. year temperature and precipitation variability. Analyses of changes in climate variability for a Explanations for some discrepancies, such as CO₂ doubling estimated by two general circulation why the daily temperature variances are too high, models (GCMs) -- Goddard Institute for Space relate to how the surface hydrology is modeled in Studies (GISS) and National Center for both GCMs (NCAR and GISS). More Atmospheric Research (NCAR) -- are not investigations of model results are necessary to conclusive. Some overall trends, but also some improve understanding of future climate variability inconsistencies, are obtained when comparing the changes. changes in climate variability associated with a changing climate calculated by the two GCMs for four U.S. regions. NATURE OF CLIMATE VARIABILITY The model results suggest that daily and year- to-year temperature variability could decrease and precipitation variability could increase. Global warming can change the variability of However, the results for temperature are not climate. Although less is known about variability statistically significant. Furthermore, the two than about most other aspects of climate change, it models produce some inconsistent results. may have greater impacts on some systems than changes in average climate conditions. Results indicate that the diurnal (day and night) cycle may be reduced in the summer, Variability is an inherent characteristic of although results for the other seasons are climate (Gibbs et al., 1975) and is closely related to inconclusive. the concept of climate change. However, no clear universally accepted distinction is made between the To determine the validity of the variability terms "climate variability" and "climate change." statistics of greenhouse gas-perturbed experiments, Both terms refer to fluctuations in climate from investigators examined how well the GCMs some expected or previously defined mean climate reproduce present-day climate variability. A state. Berger (1980) makes the distinction that comparison of observed and model results for the climate change refers to a secular trend that 29 Chapter 3 produces a change in the average, whereas variability refers to the oscillations about that mean. 100 Distinctions can only be made relative to the time 0.0 (a) 95 scales of concern. The climate change discussed in this report refers to a change from the mean global 90 climate conditions we have experienced in roughly the past few centuries. On a longer time scale (i.e., 85 thousands of years), however, this climate "change" 80 would be viewed as an instance of climate variability (i.e., as one of many fluctuations around mean 75 conditions prevailing over several thousand years). 70 100 For the purpose of this report, climate 0.583 (b) variability is defined as the pattern of fluctuations 95 about some specified mean value (i.e., a time DAILY MAXIMUM TEMPERATURE (PF) 90 average) of a climate element. Hence, in regard to the climate change considered here, climate 85 variability refers to fluctuations of climate around the new mean condition that constitutes the climate 80 change, and is expressed on time scales shorter than 75 the time scale of the climate change. For example, if it is assumed that the average annual global 70 surface temperature will be 3°C warmer than it is 95 0=0.9 currently, then the climate variability on a year-to- (c) 90 pattern of departures from this mean increase. 85 One of the main concerns regarding climate 80 change is whether and how climate variability will change (i.e., will the pattern of fluctuations around 75 the new mean at any given location be the same as that around the "old mean"). This concept of 70 1 5 9 13 17 21 25 29 33 changing climate variabilities is illustrated in Figure DAY 3-1, which displays three simulated time series of daily maximum July temperature for Des Moines, Figure 3-1. Simulated July daily maximum Iowa. In all three cases, the mean maximum temperature time series at Des Moines, Iowa. All monthly temperature is the same (i.e., 86.2°F), but assume the same average temperature but use the patterns of daily fluctuations about this mean different statistical estimates (first-order differ significantly. Changes in climate variability autocorrelation coefficient Φ) of variability (Mearns refer to the differences in these patterns. et al., 1984). The causes of climate variability depend largely on time scales and may be divided into two major Variations of climate on a year-to-year basis categories: (1) those arising from internal dynamics (interannual variability) can arise from external that produce stochastic (random) fluctuations (and forcings, such as volcanic eruptions, or from slowly possibly chaotic behavior) within the climate system, varying internal processes including, as part of the and (2) those arising through external forcing of the internal system, interactions between the system. Table 3-1 summarizes different causes of atmosphere and oceans, soils, and sea ice fields. climate variability on different time scales. On very These interactions can result in shifts in locations of long time scales (e.g., 100,000 years), astronomical major circulation features or changes in their factors account for much variability (orbital intensity (Pittock, 1980). The largest effect, parameters in Table 3-1). presumably, is due to variations in sea surface temperatures, such as those occurring in El Niño Southern Oscillation (ENSO) events. 30 Climate Variability Table 3-1. Major Processes Involved in Climate Fluctuations for Different Time Scales TIME SCALES related to Earth's history Quaternary History Instruments ice ages Years 10¹⁰ 109 10⁸ 10⁷ 10⁶ 10⁵ 10⁴ 10³ 100 10 1 galactic dust EXTERNAL affecting available incoming radiation Sun's evolution solar variability orbital parameters NATURAL POTENTIAL CAUSAL MECHANISMS geophysical boundary conditions plate tectonics epeirogeny, orogeny isostasy atmospheric evolution volcanic activity INTERNAL (related to) net radiation tropospheric dust surface cover: vegetal snow sea ice glacier ice sheet feedbacks atmosphere-cryosphere-lithosphere atmosphere-ocean atmosphere autovariation land use HUMAN ACTIVITIES traces gases aerosols heat pollution Source: Berger (1980). Daily variability of a nonperiodic nature largely NATURE AND IMPORTANCE OF results from variations in synoptic scale weather CLIMATE EXTREMES processes, such as high- and low-pressure cells and upper-atmosphere wind streams, which direct the movement of such features (atmosphere Climate variability is experienced on an impact autovariation in Table 3-1) (Mitchell, 1976). These level mainly through the occurrence of extreme features interact with local topography to provide climate events. The impact of extreme variability location-specific variability. (Variations caused by may be the first indication of climate change. It is these weather processes are largely stochastic and important to note, however, that change in the internal to the climate system.) frequencies of extreme events (e.g., heat waves, drought) is not synonymous with change in climate This report mainly discusses variations on time variability. scales of several years or less -- that is, from interannual to daily variability. Climate variability To illustrate this point, an example is presented does not have a specific operational statistical of a change in the frequency of heat waves in Des definition, but can be described by a constellation of Moines in July, defined as 5 consecutive days in the statistical properties other than the mean. The most month with maximum temperatures exceeding 95°F. commonly used measure is the variance (which is Just changing the monthly mean of the series by the mean of the sum of squared deviations from the 3°F, without changing variability (as measured by mean of a time series) or its positive square root, the standard deviation and/or autocorrelation), the standard deviation. increases the probability of experiencing a heat wave 31 Chapter 3 in July from the current level of 6% to 21%. (Policansky, 1977), to that of heat waves However, the increase can be even more dramatic (temperatures above 100°F for 5 consecutive days) if the variability is altered as well as the mean. By in Dallas, Texas (P = 0.38). increasing the persistence in the time series (i.e., the day to day dependence of the daily temperatures) as What defines an event as extreme is not only well as the mean, the probability of a heat wave a certain statistical property (for example, likelihood increases from 6% to 37% (see Mearns et al., 1984, of occurring less than 5% of the time), but also how for further details). Hence, changes in the prepared a particular system is to cope with an frequencies of extreme events will occur with event of such magnitude. Hence, very few extreme changes in the mean climate conditions, but this events have a fixed absolute value independent of change can be reduced or rendered more extreme particular response systems at a particular location. by changes in variability.* This implies that what constitutes an extreme event can also change over time because of changes in the The impacts of climate change on society relevant response system (Heathcote, 1985). accrue not necessarily from the relatively slow trends in the mean of a climate variable, but rather It is thus very difficult to comprehensively from the attending shifts in the frequency of review all climate extremes of importance to society, extreme events. This issue has already received and what is presented here is far from an exhaustive some attention in the literature where the nonlinear catalog. Because one of the purposes of this review relationship between changes in the mean and is to highlight the extreme events of importance that extreme events has been examined (e.g., Schwarz, can serve as guides for choosing what extreme 1977; Parry, 1978; Mearns et al., 1984). However, events should be quantitatively analyzed in GCM less is known about this factor than about most experiments, priority is given to events related to other aspects of climate change. variables that can be relatively easily analyzed. For the purposes of climate impact analysis, This review considers the two most important extreme climate events may be considered climate variables temperature and precipitation perturbations of climate that result in conditions - and their extremes (maxima and minima), and one outside normal ranges that exceed some critical type of major meteorological disturbance severe threshold. What constitutes "normal" (i.e., the storm effects. Extremes in these variables affect the averaging period) is, of course, a central issue in areas of energy use and production, human defining extremes. mortality and morbidity, agriculture, water resources, and unmanaged ecosystems (although not Extreme events relevant to climate impacts all areas are discussed under each climate extreme). function on different time scales, depending upon the climate variable involved and the impact area of Temperature interest. Thus, events can range from the length of time (in minutes and hours) that minimum temperatures in Florida remain below a critical Given the scientific consensus that higher value, resulting in damage to citrus crops, to the atmospheric concentration of greenhouse gases will length of time (in months and years) that raise average global temperatures, extreme precipitation is particularly low in California, temperature effects are given priority in this resulting in serious water shortages for industry and analysis. agriculture. The probability of extreme events can also vary considerably - for example, from that of Maximum Temperatures extreme snowfall in the Buffalo, New York area such as that of the 1976-77 winter (P = 0.0002) Extreme temperature effects on human mortality and morbidity have received the most attention in the scientific literature (e.g., Kalkstein, *Although the scenarios created for this study assume no Volume G; Becker and Wood, 1986; Jones et al., change in variability (see Chapter 4: Methodology) they do 1982; Bridger et al., 1976; Ellis, 1972). This is partly assume, for example, increases in heat waves and decreases in because the relevant climate factors (i.e., maximum cold waves that result from changes in mean climate conditions. 32 Climate Variability daily temperatures and relative humidity) are readily on natural ecosystems, some research has been available for analysis. done on forest responses to temperature extremes. Solomon and West (1985) indicate in their summary A heat wave is defined as a series of days with of climate effects on forests that the frequency, abnormally high temperatures (i.e., temperatures intensity, and lengths of heat waves under climate exceeding some critical threshold). Examples change conditions are important factors influencing include the 1980 heat wave in the United States seedling survival and can contribute to the loss of a when Kansas City had 17 consecutive days above species from an ecosystem. A run of warm years 39°C (102°F) (Jones et al., 1982), and Dallas had can affect the location of tree lines. Shugart et al. 42 consecutive days with temperatures above 38°C (1986) established that a period of warm summers (100°F) (Becker and Wood, 1986). The death toll at high altitudes during the 1930s, when the mean that year was several times above normal (1,265 annual temperature was no more than 1°C higher lives). than average, resulted in a burst of regeneration in boreal forest trees near polar and altitudinal limits Studies have specifically tried to pinpoint the in North America. most significant meteorological factors associated with heat-related death and illness. Jones et al. High temperatures have their most immediate (1982) determined that high maximum impact on energy by causing increased electricity temperatures, the number of days that the demand for air-conditioning. Using climate temperature is elevated, high humidity, and low scenarios similar to those in this report (see wind velocity contributed to excess mortality in Chapter 4: Methodology), Linder et al. (1987) found Kansas City and St. Louis in the 1980 heat wave. that energy demand in New York would significantly Kalkstein et al. (1987) established that runs of days increase in summer (on the order of 3% for an with high minimum temperatures, low relative average August day in 2015 for the downstate area). humidities, and maximum temperatures above 33°C (92°F) contributed to heat-related deaths in New Minimum Temperatures York City. Extreme minimum temperatures will not Increases in heat waves are virtually certain, necessarily be less of a problem with CO2-induced assuming global warming. But how they increase climate warming. For example, changes will most (longer or greater departure from the mean) very likely occur in the growing areas of certain crops, much depends on changes in variability that would where risks of frost damage may not be clearly affect the persistence of high temperatures. known. Such crops as corn, soybeans, wheat, and The best example of frost damage to crops is sorghum are sensitive to high temperatures during the effect of low minimum temperatures on citrus their bloom phases. For example, Shaw (1983) trees. This problem has been studied in depth for reported that severe temperature stress during a the citrus crop in Florida. (See Glantz, Volume J, 10-day period around silking (a critical period for a discussion of the Florida citrus industry's during which the number of kernels on the ear is responses to freezes in the early 1980s.) The most determined) will result in crop failure. McQuigg striking aspect of these freezes is the very short (1981) reported that the corn crop was severely freezing time necessary for damage to occur. New damaged in July 1980 as a result of temperatures citrus growth (i.e., bloom buds) can be completely exceeding 38°C (100°F). The destructive effects of killed during a 30-minute exposure to -3.3°C (26°F) runs of hot days on corn yields were particularly or a 3-hour exposure to -2.2°C (28°F). The effect apparent during 1983 in the U.S. Corn Belt. of freezes is exacerbated if the crops have not Although the damage from high temperatures is hardened with the cold. Thus, if a freeze follows a best documented for corn, it has also been noted in warm period (i.e., indicating high daily temperature wheat and soybean yields (e.g., Neild, 1982; variability) when dormancy has been broken, more Mederski, 1983). damage will occur at less extreme temperatures. For example, the December 24-26, 1983, freeze Although not as much research has been caused the Florida citrus yield to be 30% lower than performed on the effects of temperature extremes it had been the previous year (Mogil et al., 1984). 33 Chapter 3 Extreme lows on a seasonal basis tend to most parameters, such as streamflow). These "types" of directly affect winter energy use for heating. In the drought are not completely independent, but can United States, the difference in heating fuel use show up at different time lags one from the other. for a warm as compared with a cold winter can vary by as much as 400 million gallons of oil. During the Drought of any kind is anomalous as an extremely cold winter of 1976-77, heating degree extreme climatological event in that it is a "creeping" days (calculated on a base of 18°C (65°F)) were phenomenon; neither its onset nor its end is clearly 10% greater than normal for the nation as a whole punctuated in time. It is difficult to measure (Dare, 1981). drought severity, since drought is a combination of factors: duration, intensity, and areal extent. Precipitation Drought also can be one of the longer-lived extreme events in that it can be measured in terms of Anticipated changes in precipitation resulting seasons or, more frequently, years. from climate change are not well known at this point. However, geographic shifts in rainfall In the United States, major droughts have patterns will likely occur. Changes in the usually been defined in terms of several years, and frequencies of extremes of both droughts and floods the rate of occurrence is most strongly influenced by must be considered. interannual variability of precipitation. Drought is of particular interest at the time of The effect of drought on crop production is this writing because of the 1988 drought in the perhaps the impact of drought that has received the United States and the energetic speculations being most research attention. The occurrence of made concerning its possible connection with droughts has been a major cause for yearly CO2-induced climate change (Wilford, 1988). It variability in crop production in the United States cannot be said that the summer 1988 drought was (Newman, 1978). During the 1930s, drought yields caused by CO2-induced climate warming, but rather of wheat and corn in the Great Plains dropped to as that such droughts would be possible and perhaps much as 50% below normal, whereas the drought in more frequent with such a warming. (In fact, most the 1950s brought less dramatic declines in yields recent evidence presented by Trenberth et al. (1988) (Warrick et al., 1975). In 1988, national corn yields indicates that the cause of the drought was primarily were 40% below normal (see Chapter 6: temperature anomalies in the Pacific (i.e., cool Agriculture). temperatures along the Equator and warmer temperatures to the North), which led eventually to Soil moisture deficits affect natural vegetation the anomalous displacement of the jet stream as well as crops. Much of the research in natural northward. These causes are considered to be ecosystems has been on forests. Soloman and West natural variations in the coupled atmosphere-ocean (1985) identify drought as the cause for death of system.) seedlings and for slowed or stopped growth of mature trees. Droughts Aside from the direct effects of insufficient The most basic, general definition of drought moisture on unmanaged ecosystems, indirect effects may be lack of sufficient water to meet essential also result from increased incidence of fires. needs (Gibbs, 1984). From a more strictly During the drought of 1988, forest fires broke out climatological point of view, it may be considered a across the country; the most notable was the condition determined relative to some long-term devastating August fire in Yellowstone National average condition of balance between rainfall and Park, which blackened 60% of its land area. evapotranspiration in a particular region (Wilhite and Glantz, 1987). Different types of drought are The effects of drought on U.S. energy recognized, such as meteorological drought (a resources are most apparent with regard to departure of precipitation from normal), agricultural hydroelectric power generation. Linder et al. (1987) drought (insufficient soil moisture based on crop discussed the effect of decreased streamflow due to growth needs), or hydrological drought (based on drought on the production of hydroelectric power in departures from normal or relevant hydrologic New York (see Chapter 10: Electricity Demand). 34 Climate Variability The possibility of combined effects of higher The recurrence interval of flooding is most maximum temperatures and drought on electricity important in applying effective control and demand and supply should be noted. Increased protection mechanisms. These include building demand (due primarily to increased temperature) dams, reservoirs, and levees, and improving would very likely occur when drought would limit channels and floodways (White et al., 1975). For generating capacity in regions such as New York example, flood control reservoirs are designed to and the Pacific Northwest. operate at a certain level of reliability, and the reliability is determined by a certain flood Floods magnitude that the reservoir can handle, such as a 100-year flood. The statistics of flooding are vital On average, 200 people die each year from for designing for protection and are based on a flooding; flash floods account for most of these certain climate variability determined from the deaths (AMS, 1985). Floods also destroy property, historical record. As that variability changes, the crops, and natural vegetation, and disrupt organized reliability of the protection system will change. social systems. Floods in the 1980s have been less serious in Floods result from a combination of terms of loss of life, but changing frequencies of meteorological extremes (heavy precipitation from severe storms, such as thunderstorms and severe storms, such as hurricanes and hurricanes, as well as general shifting of thunderstorms), the physical characteristics of precipitation patterns could result in unprecedented particular drainage basins, and modifications in losses from floods in a climate-changed world. drainage basin characteristics made by urban development. Loss of life and property is increasing Severe Storms - Hurricanes as use of vulnerable floodplains increases. Three important kinds of weather extremes are present in hurricanes: strong winds, intense and Major recent floods include the following: high precipitation amounts, and extreme storm surges. A hurricane is an extreme form of a 1. Rapid City, South Dakota (June 1972), tropical cyclone, characterized by torrential rains, 231 deaths and more than $100 million typically as much as 127 to 254 millimeters (5 to 10 in property damage: inches) in one storm; high windspeeds, which can exceed 160 kilometers per hour (100 miles per 2. Northeastern United States (June 1972), hour); very steep pressure gradients, with pressure 120 deaths and about $4 billion in at the center as low as 915 millibars (27 inches) and property damage inundation from diameters of 160 to 640 kilometers (100 to 400 Hurricane Agnes; miles). 3. Big Thompson Canyon, Colorado (July Hurricanes are classified according to their 1976), 139 deaths and $50 million in severity on the Saffir/Simpson Scale (categories 1 property damage - a result of a stalled through 5), taking into account the central pressure, thunderstorm system that delivered 12 windspeed, and surge. Major hurricanes are inches (305 millimeters) of rain in less considered to be all those of categories 3 through 5 than 6 hours (Henz and Sheetz, 1976); wherein central pressure is less than 945 millibars and (27.9 inches), windspeeds exceed 176 kilometers per 4. Johnstown, Pennsylvania (July 1977), 76 hour (110 miles per hour), and the surge is greater deaths and $200 million in property than 2.4 meters (8 feet) (Herbert and Taylor, 1979). damage a result of slowly moving thunderstorms that deposited 11 inches From 1900 through 1978, 53 major hurricanes (279 millimeters) of rain in 9 hours. (averaging two major hurricanes every 3 years) directly hit the United States. Overall, 129 hurricanes of any strength hit the United States (averaging approximately two each year). In recent 35 Chapter 3 decades, the number of major hurricanes has STUDIES OF CHANGING declined. From 1970 to 1978, only three hurricanes occurred, compared with six or more in earlier CLIMATE VARIABILITY decades. The last hurricane of category 4 or 5 to strike the United States was Hurricane Camille in Empirical Studies 1969. In 1980, Hurricane Allen, which at one time reached force 5, weakened before it struck a relatively unpopulated segment of the Texas coast One of the methods available for gaining some (Oliver, 1981). Since then, the population of the insight into how climate variability may change in a south coastal regions of the United States has grown generally warmer climate is to investigate the tremendously, and most inhabitants have never climate record for past relationships between mean experienced a major-force hurricane. Building in climate change and changes in variability. However, coastal areas has also increased with population, past research efforts to determine changes in which raises the potential for high property damage. climate variability and relationships with changes in mean climate conditions have not resulted in a clear Thus, the population may be more vulnerable and consensus. less prepared to handle this particularly devastating extreme event (Sanders, 1982). Van Loon and Williams (1978) found Any increase in the frequency and/or intensity significant differences in interannual temperature of these storms, which could result from climate variability in North America during two different 51- change, would be of great concern to southern year periods. However, they found no single coastal regions of the United States. Hurricane connection between trend in temperature and trend Gilbert, which occurred in September 1988, in its interannual variability. Specifically, they assert reinforced this concern, even though it did not cause that their results do not support the postulated major damage to the coastal United States. association between cold periods and high variability Hurricane Gilbert may well prove to be the most of temperature. Diaz and Quayle (1980), in a powerful hurricane of the 20th century; its lowest thorough analysis of the U.S. climate (temperature central pressure (883 millibars or 26.13 inches) was and precipitation), found no systematic relationship the lowest ever measured in the Atlantic Gulf and between changes in mean temperature and Caribbean regions of tropical storm activity. precipitation and their corresponding variances. Serious damage did occur primarily in Jamaica, the Cayman Islands, and the northern tip of the Brinkmann (1983) analyzed the relationship Yucatan Peninsula (Ludlum, 1988). between mean temperature and variability in Wisconsin using climate data for three stations. She Coleman (1988) has found in the historical found no relationship between mean temperature record some limited evidence for increased and interannual variability, but did find a negative frequency for the number of storms formed in the correlation between winter mean temperatures and the day-to-day variability, and a corresponding North Atlantic during years of warmer-than-average positive relationship for summer conditions. What sea surface temperatures. Emmanuel (1987) has this means is that cold winters are more variable found through a hurricane modeling experiment that than warm winters, but that cool summers are less the intensity of hurricanes increases under warmer variable than warm ones. Brinkmann explains these conditions. The extreme intensity of Hurricane relationships on the basis of Wisconsin's location Gilbert in September 1988 is consistent with the findings. Emmanuel (1988) also asserts the with respect to general circulation patterns. importance of establishing a general theory of Lough et al. (1983) analyzed the association hurricane development independent of current between mean temperature and precipitation and atmospheric conditions, so that scientists can predict variability in Europe by using the analog approach changes in frequency and intensity of storms with climate change. to create climate change scenarios (the analog 36 Climate Variability approach is further discussed in Chapter 4: variability and climate prediction. He focused on Methodology). They selected two periods when the additional variability attributed to external arctic temperatures were particularly warm and cold boundary conditions (i.e., in this modeling context, (1934-53 and 1901-20). Results indicate that the external boundary conditions refer to important regions of lower winter temperatures roughly conditions outside the atmosphere that cause coincide with the region of increased variability, but changes to the atmosphere but are not in turn the coincidence is far from perfect. affected by it, such as sea surface temperatures). He eliminated sources of external variability in the These studies indicate that significant changes model, such that discrepancies between modeled have occurred in both interannual and day-to-day and observed variability would reflect this external climate variability in historical times, but that simple component. The variability of mean sea level or distinct relationships between changes in mean pressure and 700-millibar geopotential height (which climate conditions and changes in variability have roughly corresponds to the height above the surface not been established. Moreover, the value of where the atmospheric pressure equals 700 seeking such relationships in the past as a key to the millibars, and is related to large-scale wind patterns) future is potentially limited, since the causes of very were analyzed for the Northern Hemisphere, with short-term warming or cooling in the past are not particular focus on the United States. Results, known, but in any event, are not caused by increases however, indicated no significant differences in greenhouse gases. between modeled and observed variabilities of mean sea level pressure over the United States and only The failure of the analog approach to provide limited areas of differences in the variability of an empirically consistent and causally coherent 700-millibar geopotential height. scenario of possible changes in climate variability contributes to the necessity of examining climate Bates and Meehl (1986) also used the CCM to variability in climate modeling experiments. As investigate changes in the frequency of blocking discussed in Chapters 2 and 4, GCMs have events (stationary pressure systems that block the limitations, but they have one clear strength over flow of upper air currents in the atmosphere) on a empirical attempts to analyze future climate change: global scale under doubled CO₂ conditions. the modeling experiments are constructed such that Blocking events are strongly related to persistent the response of the climate system to the true cause surface temperature anomalies, such as heat waves of the change (increased greenhouse gases in the in the summer. They found that the model atmosphere) is simulated. generally produces too few extreme blocking events. Under doubled CO₂ conditions, standard deviations Modeling Studies of blocking activity were found to mainly decrease in all seasons (i.e., the variability of blocking events Studies comparing variability statistics of decreased). observed time series with variability statistics of GCM-generated time series of climate variables Two studies were recently conducted on local or relevant to climate impacts are not numerous in the regional scales using the U.K. Meteorological Office atmospheric sciences literature, although studies five-layer GCM. Reed (1986) analyzed observed first appeared in the early 1980s (e.g., Manabe and versus model control run results for one gridpoint in Hahn, 1981; Chervin, 1981). Such studies are eastern England. Compared with observations, the critical if climate change research is to determine model tended to produce temperatures that were whether the variability statistics of doubled CO₂ too cool and variability that was too high as experiments with GCMs are valid. To accomplish measured by the standard deviation. For this, the ability of GCMs to reproduce present-day precipitation, the model produced too many rain climate variability statistics must be examined, and days but did not successfully simulate extreme rain a thorough understanding of discrepancies must be events of greater than 20 millimeters per day. attained. More recently, Wilson and Mitchell (1987) Chervin (1986) used the National Center for examined the modeled distribution of extreme daily Atmospheric Research Community Climate Model climate events over Western Europe, using the same (NCAR CCM) to investigate interannual climate model. Again, the model produced temperatures 37 Chapter 3 that were too cold, and hence, extreme minimum temperatures were overestimated. This problem STUDIES FOR THIS REPORT was most pronounced in grid boxes away from the coasts. The model also produced too much Two research efforts were undertaken for this precipitation in general, did not successfully report to attempt to increase knowled~ concerning reproduce observed highest daily totals, and how climate variability may change. The climate overestimated the number of rain days. Wilson and change scenarios used in the climate change impact Mitchell examined changes under quadrupled CO₂ studies reviewed in this report excluded conditions and found that variability of temperature consideration of changes in variability (see Chapter generally decreased. 4: Methodology). The following two studies on GCM estimates of current and future variability Hansen et al. (1988) used the Goddard Institute were performed for this report: for Space Studies (GISS) general circulation model to simulate the global climate effects of Variability and the GISS Model - Rind, time-dependent variations of atmospheric trace Goldberg, and Ruedy, Goddard Institute for gases and aerosols. It was determined that the Space Studies (Volume I); and model only slightly underestimates the observed interannual variability across the globe. However, Variability and the NCAR Model - Mearns, the model's variability tends to be larger than that Schneider, Thompson, and McDaniel, National observed over land (i.e., only considering land areas, Center for Atmospheric Research (Volume I). not ocean areas). It should be recalled that scenarios of climate Among the calculations made with output from change generated by the GISS GCM are used in the transient run were changes in the frequencies of most of the impact studies summarized in this extreme temperature events. This was accomplished report. The results of these two studies are directly by adding the model-induced temperature change compared in a later section. with climate warming to observed local daily temperatures, assuming no change in variability. The GISS Study Results indicate that predicted changes in the frequency of extremes beyond the 1900s at locations Rind et al. (1989) examined how well the GISS such as New York, Washington, and Memphis GCM simulates the observed variability of climate become quite large and would have serious impacts. by comparing the model and the observed interannual and daily variations of temperature and The studies reviewed above indicate some precipitation. They described the model assessment important shortcomings of GCMs with regard to of changes in variability for these two major climate their ability to faithfully reproduce observed variables, under climate change using the GISS variability statistics. More research is clearly doubled CO₂ run (8° X 10° resolution) and the needed to further determine the sensitivity of the transient climate change experiment in which trace models to changes in physics, resolution, and so gases were increased gradually. The analysis was forth, with regard to the determination of variability. conducted for the Great Plains, the Southeast, the Moreover, only one of these studies explicitly Great Lakes region, and California (see Figure 3- concerns variables of importance to climate impact 2). Observed data consist of the average of analysis. Studying the higher moments (e.g., observations at nine different stations per grid box. variance) of climate variable statistics, and carefully verifying the models' ability to reproduce observed First, mean conditions were compared for actual variability on regional scales, are the necessary weather observations with the GCM control run (or prerequisites to rigorously analyzing possible single CO₂), the doubled CO2 run, and the transient changes in these statistics under doubled CO₂ run. The model values for mean temperatures for conditions. four months in the four regions are generally cooler than observations (particularly in summer and fall), 38 Climate Variability 47.0 as 39.1 31.3 23.5 -135 -125 -115 -105 -95 -85 -75 -65 Figure 3-2. The locations of the four GISS model grids. but only by a few degrees Celsius. Model compared for all months. In most months, the precipitation values are fairly close to observed model year-to-year temperature variability is similar values in the Great Lakes and Southeast grid boxes, to the observed variability in the four regions, but in but model values are higher than observed for the summer the variability was overestimated by 0.3 to other two regions (e.g., January in the southern 0.6°C (0.5 to 1.1°F). Precipitation variability is Great Plains: model = 2.1 millimeters per day, overestimated in half the cases where precipitation observed = 0.46 millimeters per day). Under the amount is also overestimated. The relative annual doubled CO2 scenarios, temperatures increase over variability of precipitation (that is, the standard the control run by 4 to 6°C (7 to 11°F) in the winter deviation relative to the mean) of the model is and 3 to 4°C (5 to 7°F) in the summer. Warming generally in agreement with observations. in the transient scenarios is progressive, but temperature changes more gradually than with Under conditions of climate change (doubled simply doubling the CO₂ amount. Winter warms CO₂), comparing control versus climate change, more than summer, and so the annual seasonal there is generally reduced variability of temperature cycle is reduced under climate change. Precipitation from January through April. Results for other changes are not statistically significant at individual seasons of the year are more ambiguous. For grids, but there is an overall tendency for increased precipitation, the doubled CO2 climate resulted in precipitation. increased variability in most months at the four grids (in 31 of 48 cases), but was particularly Interannual Variability striking at the Southeast grid. These changes, however, were often of the same order as the Standard deviations of temperature and model's natural variability (from examination of the precipitation of observed and modeled data were 100-year control run). The sign of the change in 39 Chapter 3 mean value and the sign of change in interannual warmest time period exhibits increases in standard variability are highly correlated. deviations in half of the cases. These results are again consistent with those for interannual Daily Variability variability. Daily variability of temperature was analyzed by Variability of the Diurnal Cycle taking the daily departures from monthly means and comparing the resulting model distribution with the It would be expected that the diurnal cycle distribution formed in the same manner from the would decrease under changed climate as the observational data. additional greenhouse gases could limit nighttime cooling. Comparisons of control model results with Ten years of control run for the transient observations are reasonable in the four regions. experiment for four months (January, April, July, Under doubled CO₂ conditions, it was found that and October) were compared with 30 years of the amplitude of the diurnal cycle very definitely observations. Distributions of observed versus decreases in summer but changes inconsistently in modeled daily temperature data were, in general, the other seasons. The reason for this is the not significantly different. Comparisons were also dominance of radiative heating in the summer and made by calculating the standard deviations of the of other forms of heating and cloud cover change departures from the mean for the four months in other seasons. (Table 3-2). These results indicate that the model's values are significantly greater than the observed The NCAR Study values, which demonstrates that the model is producing too many extremes. In this study, Mearns et al. (1989) analyzed mean and variance of climate variable time series Results in Table 3-2, comparing standard from selected empirical stations and those produced deviations, indicate that although changes with time by general circulation model control and doubled are not strictly progressive, most cases by the end of CO₂ runs. They attempted first to determine how the climate change experiment show reductions in faithfully the GCMs reproduce these measures of the standard deviation although these reductions are the present variability and then to examine how the not statistically significant. (Note in Table 3-2 that variability is estimated to change in CO2-perturbed standard deviations for the future decades are cases. By comparing the relative performance (i.e., changes in standard deviation (SD): model current model versus observations) of various versions of SD minus future decade SD) Since the results are the NCAR CCM (i.e., versions with different not statistically significant, a decrease of daily physical parameterizations or formulations), Mearns temperature variability is not demonstrated. et al. helped to determine what formulations may be needed for forecasting certain measures of For precipitation, comparisons are more variability and how much credibility to assign to complex. For example, the number of observation those forecasts. stations used to represent a grid box does affect the results. Model rainfall distributions differ Methods significantly from observed distributions in half the cases (in three seasons for California and the This study used the output from control runs of southern Great Plains). The model also produces three different versions of the NCAR Community fewer days of light rain in general and more Climate Model (CCM). These versions use extreme values in the winter in all four regions different parameterizations of important physical (Table 3-3). processes in the model, such as surface hydrology. The Chervin version (Chervin, 1986) is the primary In the transient experiment, the precipitation one used for comparison of observed and model distributions differ from the control climate about control output (i.e., model runs to simulate the one-fourth of the time with no general progression actual present-day climate), since it has the longest over the decades. Figure 3-3 presents a sample set time integration (20 years). of distributions for precipitation during several decades of warming for the West Coast in April. In The CCM is a spectral general circulation comparing standard deviations (Table 3-3), the model originally developed by Bourke and 40 Climate Variability Table 3-2. Daily Temperature Standard Deviations (SD) (°C) Model Observed Current 2010s 2030s ~2060 Month Location SD SD *ASD SD SD January Southern Great Plains 4.81 8.15 0.61 -1.19 -0.83 Southeast 4.53 6.90 -0.14 -1.14 -0.23 West Coast 3.63 5.86 -0.61 0.05 -0.16 Great Lakes 4.97 5.79 0.44 -0.33 -0.44 April Southern Great Plains 3.72 5.77 -0.57 -0.27 -0.80 Southeast 3.71 5.50 -0.65 -1.61 -1.24 West Coast 2.59 4.29 0.77 0.60 0.33 Great Lakes 4.65 6.15 -0.51 -0.26 -1.39 July Southern Great Plains 1.74 2.56 0.54 -0.19 0.18 Southeast 1.50 2.34 0.14 -0.22 -0.24 West Coast 2.40 3.56 0.03 0.54 0.28 Great Lakes 2.38 3.02 -0.48 -0.84 -0.14 October Southern Great Plains 3.79 5.16 1.16 0.97 1.35 Southeast 3.59 5.21 -0.54 -0.25 -0.73 West Coast 3.15 6.51 -0.55 -0.30 -0.80 Great Lakes 4.09 5.46 -0.37 0.91 -0.06 *ASD = Change in standard deviation (model current - future decade). Source: Rind et al. (Volume I). collaborators (Bourke, 1974; Bourke et al., 1977), latitude and 7.5 degrees in longitude, and has nine which has been modified by the incorporation of levels in the vertical. radiation and cloud parameterization schemes. The model has a resolution for physical processes (i.e., The other two versions of the CCM used are grid box size) of approximately 4.5 degrees in the Washington version (Washington and Meehl, 41 Chapter 3 Table 3-3. Daily Precipitation Standard Deviations (SD) (mm/day) Model Observed Current 2010s 2030s ~2060 Month Location SD SD * SD SD ASD January Southern Great Plains 1.08 2.80 0.05 0.05 1.68 Southeast 4.35 4.62 -1.20 -1.35 -0.85 West Coast 3.23 4.55 -0.18 0.34 0.13 Great Lakes 2.23 4.06 -1.07 -0.94 -0.50 April Southern Great Plains 2.51 3.26 0.94 1.99 1.17 Southeast 4.35 3.85 0.95 -0.15 0.81 West Coast 1.41 2.76 0.07 1.02 -0.12 Great Lakes 3.85 3.29 -0.43 -0.31 0.44 July Southern Great Plains 2.79 3.08 -0.10 -0.09 0.36 Southeast 4.13 3.31 0.28 0.29 0.11 West Coast 0.57 1.53 0.44 0.24 0.71 Great Lakes 3.68 2.48 -0.06 0.72 0.35 October Southern Great Plains 2.75 1.79 0.52 0.34 0.00 Southeast 3.77 3.88 0.72 -0.15 -0.28 West Coast 1.86 2.69 1.20 -0.63 1.34 Great Lakes 3.58 2.26 0.52 0.76 0.95 *ASD = Change in standard deviation (model current - future decade). Source: Rind et al. (Volume I). 1984), which includes an interactive thermodynamic This model calculates the transfer of momentum, ocean and surface hydrology; and the Dickinson heat, and moisture between the Earth's surface and version (Dickinson et al., 1986), a version of the atmospheric layers, and includes a very detailed more sophisticated CCM1 containing a diurnal cycle surface hydrology scheme that accounts for and a very sophisticated land surface package, the vegetation type and amount, and water use by the Biosphere-Atmosphere Transfer Scheme (BATS). vegetation. 42 Climate Variability 70 70 CONTROL 2030 60 60 50 50 FREQUENCY (Percent) 40 30 FREQUENCY (Percent) 40 30 20 20 10 10 0 0 0.0 3.0 6.0 9.0 12.0 15.0 18.0 21.0 0.0 4.0 8.0 12.0 16.0 20.0 24.0 28.0 PRECIPITATION (mm/Day) PRECIPITATION (mm/Day) 70 70 2010 2060 60 60 50 50 FREQUENCY (Percent) 40 30 FREQUENCY (Percent) 40 30 20 20 10 10 0 0 0.0 3.0 6.0 9.0 12.0 15.0 18.0 21.0 0.0 3.0 6.0 9.0 12.0 15.0 18.0 21.0 PRECIPITATION (mm/Day) PRECIPITATION (mm/Day) Figure 3-3. Sample set of precipitation distributions for the West Coast in April for specified years of the transient run (Rind et al., Volume I). The four regions of the United States chosen Temperature for investigation were roughly the same as those chosen for the GISS study: the Great Plains Figure 3-5 displays the time series of daily (GP; represented by three grid boxes), the average temperature for modeled and observed data Southeast (SE), the Great Lakes (GL), and the for the four regions investigated. The model West Coast (WC). The locations of the grid boxes successfully simulates the annual cycle for the four and observation stations are indicated on Figure 3-4. regions, which represents the seasonal variability. Comparison of Observed versus Chervin Control Solar Radiation and Relative Humidity Run Simulation of solar radiation ranges from very Four variables deemed particularly relevant to good (the Great Plains region) to only fair at the climate impact analysis were chosen for this Southeast, where the model consistently analysis: daily mean temperature, daily total overestimated absorbed solar radiation during all precipitation, mean daily relative humidity, and months. The Chervin CCM is poor at simulating mean daily absorbed solar radiation. the annual cycle of relative humidity at all four locations. 43 Chapter 3 WC I GL GP II III SE Temperature and Precipitation Stations Relative Humidity and Radiation Stations Figure 3-4. NCAR model grid cells and station locations. GREAT PLAINS I, II, III GREAT LAKES 35 35 30 OBSERVED DATA 30 OBSERVED DATA 25 MODEL 25 MODEL 20 20 15 15 TEMPERATURE (°C) 10 TEMPERATURE (°C) 10 5 5 0 0 -5 -5 -10 -10 -15 -15 I a - 20 20 I 60 120 180 240 300 360 60 120 180 240 300 360 DAYS DAYS SOUTHEAST WEST COAST 35 35 30 OBSERVED DATA 30 OBSERVED DATA MODEL 25 25 MODEL 20 20 15 15 TEMPERATURE (°C) 10 5 TEMPERATURE (°C) 10 5 0 0 -5 -5 -10 -10 -15 -15 -20 -20 60 120 180 240 300 360 60 120 180 240 300 360 DAYS DAYS Figure 3-5. Average temperature for a 20-year average year (NCAR model and observations) (Mearns et al., Volume I). 44 Climate Variability Precipitation reproducing mean climate (temperature and precipitation) at the four locations. The Chervin CCM consistently overestimates precipitation, although the seasonal cycle is well The Dickinson model most accurately simulated in the Great Plains region and the West reproduces daily variability of temperature, while Coast grid. The authors do not know why the the other two models overestimate it. This result is model overestimates precipitation, but speculate graphically illustrated in the temperature histograms that it may partly be a result of a precipitation (three models and observed) for two key months for parameterization criterion of 80% relative humidity. the Southeast grid (Figure 3-6). The reasons for these discrepancies have yet to Variability Comparisons of the Chervin CCM be explored in depth, but are likely related to different land surface packages in the models. A Interannual variability of temperature is possible explanation for the lowered daily generally underestimated by the Chervin CCM in temperature variability of the Dickinson model all four regions. Interannual variability of concerns the more sophisticated surface energy precipitation (i.e., relative variability, the standard balance used, which includes consideration of soil deviation relative to the mean) is generally in heat capacity. reasonable agreement with observed data, although it is occasionally overestimated. This is a Control Versus CO2-Perturbed Runs particularly encouraging result for the credibility of predicting climate changes, given how inaccurate the The authors included a preliminary analysis of control precipitation results are in terms of absolute changes in precipitation and temperature, under a values. scenario of doubled CO₂, using the output from Washington's control and doubled CO₂ runs for the In terms of daily variance, the model's relative four regions. Interannual variability could not be humidity tends to be much less variable than analyzed because the time series are too short. observed values at all locations and in most months. However, they examined the daily variability of Results for temperature for January and July temperature and precipitation. indicate that the Chervin model generally overestimates daily temperature variance. An annual temperature increase of about 2 or 3°C (4 to 5°F) occurs at all locations. Annual Intercomparisons of Three CCM Versions and total precipitation increases between 22 and 26% Observed Data at three locations but decreases slightly (2%) in the Southeast. There are also potentially important Comparing different model versions' simulations changes in the seasonal distribution of precipitation. of present-day climate facilitates understanding of For example, at the Southeast grid a smaller the possible ranges of errors and the effect of a percentage of the annual total occurs during the model's structural differences. The present-day summer in the CO2-perturbed case (from 13 to climate runs of models incorporating physics 6%). different from those of the CCM version of Chervin (1986) are compared. Both the Washington and Statistics comparing the daily temperature Dickinson runs consist of 3-year integrations. variance of the control and perturbed runs for January, April, July, and October indicate that the There is considerable variability in how well the temperature variance in general does not models reproduce mean total precipitation for the significantly change (at the 0.05 level of significance) four grids, ranging from the relatively good results at these four grids. Without consideration of of Dickinson's model, to the fair results of statistical significance levels, results are mixed with Washington's model, to the overestimation of both increases and decreases. Chervin's model. On the basis of mean annual and seasonal comparisons, no one model is clearly The percentage of rain days decreases in the superior to the other two in accurately summer under climate change in three of the four 45 Chapter 3 100 100 JANUARY APRIL OBSERVED DATA 80 80 OBSERVED DATA N CHERVIN MODEL V CHERVIN MODEL 60 PERCENTAGE 60 40 PERCENTAGE 40 20 20 0 0 -30 -20 -10 0 10 20 30 40 -30 -20 -10 0 10 20 30 40 (°C) (°C) 100 100 JANUARY APRIL DICKINSON MODEL 80 DICKINSON MODEL 80 1 WASHINGTON MODEL WASHINGTON MODEL PERCENTAGE 60 PERCENTAGE 60 40 40 20 20 0 0 -30 -20 -10 0 10 20 30 40 -30 -20 -10 0 10 20 30 40 (°C) (°C) Figure 3-6. Histograms of daily temperature, observations and three model versions, for two key months of the Southeast grid (Mearns et al., Volume I). grids. Overall, there is a tendency for increased explain discrepancies in variability between model daily precipitation variability at the four locations, control runs and observations. Since the spatial based on analysis of precipitation distribution resolutions of the models differ, the grid boxes of characteristics. the models do not coincide, and so the regions analyzed differ. These are only some of the problems that would affect these comparisons. COMPARISON OF GISS AND Nevertheless, an attempt is made here to compare some of the results that roughly coincide. Some NCAR RESULTS regions, such as the Great Lakes grids, coincide fairly well (see Figures 3-2 and 3-4), and some It is difficult to compare the two studies. The similar analyses were conducted. modeling experiments were conducted partly with different purposes in mind using two different A brief comparison is made of how the models models (which differ not only in how physical reproduce the observed mean climate. In general, processes are modeled but also in their spatial the GISS model is too cool and the NCAR model(s) resolutions). They also use different qualitative and too warm. The GISS model overestimates statistical methods for making comparisons. The precipitation at two grids, and the Chervin version GISS experiment was aimed primarily at examining of the NCAR model overestimates precipitation at the changes in variability with climate change, all grid boxes (although this is not true of two other whereas the immediate purpose of the NCAR versions of the NCAR CCM). experiment was primarily to examine and 46 Climate Variability The following sections compare the observed, West Coast winter. In summer, the GISS model control, and perturbed runs of interannual and daily overestimates, and the NCAR model underestimates variability of temperature and precipitation. Table temperature variability at all locations. 3-4 summarizes the comparisons between the modeled control runs and observations for Regarding the relative variability of precipitation variability. (measured by the coefficient of variation), the results for the two models are rather similar. The Interannual Variability differences between observed and model values are very close (from 1 to 6 percentage points) in each Rind et al. used a 100-year control run for study. The NCAR model slightly underestimates interannual variability calculations. Their the variability at each location, whereas the slight observational data set consists of 30 years (1951-80). errors in the GISS results are mixed. The NCAR study uses a 20-year control run of Chervin (1986) and a 20-year observational data set The reasons for the lack of agreement in the (1949-68). The differences in sample size should be two studies are far from obvious, and speculation noted. can only be rough. Certainly the difference in how the atmosphere-ocean interaction is modeled may Table 3-5 presents the relevant results, winter play a role (i.e., the NCAR model uses fixed sea and summer standard deviations for temperature, surface temperatures, whereas the GISS model and annual coefficients of variation (i.e., a measure computes sea surface temperatures from a simple of relative variability) for precipitation for the four ocean mixed-layer model). regions for both studies. Relative variability values (standard deviation relative to the mean) for the Daily Variability GISS study were provided by its authors (Rind, personal communication). Both models Daily variability of temperature can be overestimate the temperature variability of the compared for two season months (January and July) Great Plains region in winter. (However, the at the four locations using the standard deviations difference in the NCAR study was deemed to be (Table 3-6). Because of certain problems statistically insignificant.) Both models concerning necessary statistical assumptions for underestimate the temperature variability (but the quantitative testing, these comparisons must be NCAR model much more so than the GISS) for the Table 3-4. Variability Results for Control Runs vs. Observationsᵃ Interannual Daily Model Temperature Precipitation Temperature Precipitation (Relative/Absolute)⁵ (Relative/Absolute) GISS High Good/High High Good/High NCARᶜ Low Good/High Highᵈ Good/High ᵃValues in chart refer to how the model estimates compare to the observations. Relative/absolute refers to comparison of coefficients of variation (relative) and standard deviation (absolute). ᶜChervin version of the NCAR model. Values are good or slightly low for the Dickinson version of the NCAR model. 47 Chapter 3 Table 3-5. Interannual Standard Deviations, Temperature and Coefficient of Variation, Precipitation, GISS, and NCAR Control Runs Precipitation Temperature (°C) coefficient of standard deviation variation (%) Model and region Dec.-Feb. June-Aug. (standard deviation/ mean) GISS (n = 100) SGP Model 1.65 1.05 15 Obs. 1.20 0.75 21 SE Model 1.65 1.05 22 Obs. 1.65 0.70 18 WC Model 1.35 1.35 18 Obs. 1.45 0.75 23 GL Model 1.35 1.25 18 Obs. 1.50 0.70 18 NCAR (n = 20) GP III Model 1.3 0.62 17 Obs. 1.1 1.20 22 SE Model 1.0 0.38 10 Obs. 1.8 0.74 12 GL Model 2.2 0.71 10 Obs. 1.6 0.88 11 WC Model 0.8 0.76 17 Obs. 1.6 0.81 17 Abbreviations: SGP = Southern Great Plains; SE = Southeast; WC = West Coast; GL = Great Lakes; GP = Great Plains. Source: Rind, personal communication; Mearns et al. (Volume I). viewed strictly qualitatively. In seven of the eight reduce daily temperature variability). (The relative cases, the studies agree that the models success of the Dickinson version of the CCM in overestimate daily temperature variability. reproducing daily temperature variability partially supports such an explanation, since it has a more In both studies, explanations for the sophisticated surface hydrology scheme compared overestimations are related to the modeling of with the Chervin version.) surface hydrology (i.e., both models fail to completely account for important The models produce, in the majority of cases, surface-atmosphere interactions that would tend to too few light rain days. The GISS model produces 48 Climate Variability Table 3-6. Daily Temperature Standard Deviations (°C) GISS NCAR Month Obs. Model Obs. Model January Great Plains 4.81 8.15 6.18 8.84 Southeast 4.53 6.90 5.41 5.92 Great Lakes 4.97 5.79 5.50 11.20 West Coast 3.63 5.86 4.10 5.00 July Great Plains 1.74 2.56 2.90 2.79 Southeast 1.50 2.34 1.55 1.70 Great Lakes 2.38 3.02 2.67 2.82 West Coast 2.40 3.56 2.18 3.52 Source: Rind et al. (Volume I); Mearns et al. (Volume I). too many extreme rain events in winter at all A slightly clearer picture is gained from locations. The NCAR model tends to produce too comparison of results for daily precipitation. The many high extremes in all four seasons. Neither results of both models point to increased daily study accounts for these discrepancies. precipitation (although not from analysis of the same statistic). This is not true for all locations Comparison of Climate Change during all seasons, however. Comparison of climate change results of the Table 3-7 summarizes the very tentative two models is restricted to changes in daily conclusions that can be drawn given all climate temperature variability and daily precipitation change results regarding changes in climate variability for four months for the four locations, variability from the GISS and NCAR studies. The since the NCAR study includes a quantitative degree of uncertainty in these conclusions should be analysis of only daily variability change. noted, as should the observation that many of the results are from only one model (GISS). The two studies do not agree on the direction of change of daily temperature variability. The Limitations of the Two Studies NCAR results are mixed, showing both increases and decreases, although most of these changes are Both studies underline the importance of statistically insignificant. Rind et al. conclude that viewing the climate change results of the models in in general, there is a decrease in daily temperature the context of how well they reproduce the present variability on the basis of changes in standard climate. Model deficiencies can be expected to limit deviations (but the changes are not statistically the reliability of climate change results, and faith in significant). On the basis of the two research quantitative results is probably misplaced. reports, no clear statement may be made about changes in daily temperature variability under CO₂ A major model deficiency is inability to resolve warming conditions. subgrid-scale atmospheric phenomena that 49 Chapter 3 Table 3-7. Summary of GISS and NCAR Model "Scenarios" for Direction of Variability Changes from Present Climate to Doubled CO₂ Climate for Four U.S. Regionsᵃ Variability Results CO₂-Perturbed Runs Variable Interannual Daily Temperature 1? ??? Precipitation 1? 1?? a Question marks indicate degree of uncertainty: ? = results of only one model; ?? = results of two models, but some conflicting results. contribute to climate variability, such as fronts and The lack of this information resulted in the intense cyclones (hurricanes), and important formation of climate scenarios wherein the temporal variations in atmosphere-ocean coupling, such as El variability of both precipitation and temperature Niño Southern Oscillation (ENSO) events. were not changed (see Chapter 4: Methodology). (However, it appears that more sophisticated GCMs This was considered a limitation or concern in many incorporating complete ocean models do produce studies, some of which are discussed in this section. ENSO-type events (Meehl, 1989).) However, model results do give crude estimates as to the importance In the Johnson et al. study on agricultural runoff of some physical processes responsible for variability and leaching (reviewed in Chapter 6: Agriculture), and what must be done to improve them. Further the results were considered to be limited by the testing is needed to determine how the models' failure to consider changes in storm frequency and deficiencies in reproducing present-day climate duration that would result from climate change. affects "predictions" for a CO2-warmed future The results of this study could be vastly different climate. from those presented, depending upon assumptions concerning precipitation duration, frequency, and intensity, all of which would change if a changed IMPLICATIONS FOR STUDIES daily variability were assumed. OF CLIMATE CHANGE IMPACTS Several studies on hydrology summarized in this report also are highly dependent upon assumptions As indicated in the second section of this about precipitation variability. These include the chapter, virtually all systems affected by climate are Lettenmaier et al. study on the hydrology of affected by climate variability, although some are catchments in the Central Valley and the Sheer and more affected than others. The relative importance Randall study on the impact of climate scenarios on of climate variability and changes in variability, as a water deliveries, both reviewed in Chapter 14: result of climate change, to particular impact areas California. The scenarios assumed that the number is reflected in the results and limitations of some of of days of rainfall remains the same under the the studies summarized in this report. climate change. Model results in terms of predicting runoff amounts would be quite different Of greatest concern is the lack of information if more rainfall events of lower intensity were regarding changes in the variability of temperature and precipitation that would attend climate change. 50 Climate Variability assumed compared with the same number of rainfall daily maximum temperatures and the persistence of events of (generally) higher intensity. such temperatures (i.e., heat waves). The studies for the Southeast (Chapter 16) did It would be impossible to quantitatively or even not consider changes in the frequency of droughts qualitatively estimate how different the results of or severe storms such as hurricanes, which could these studies would be if changes in climate certainly affect the likelihood of flooding for some variability had formed part of the climate scenarios coastal communities. However, these concerns are made available as input for the various climate considered to be secondary to changes in sea level impact models used. Primarily, it is impossible that would dominate in terms of changing the because the variability changes are not known; likelihood of floods. second, it is impossible because most of the studies are so complex that the effect of a change in one Crop yields are very dependent on daily variable (a complex change at that) is not intuitively variability. For example, heat waves occurring obvious in most cases. Analyses of the sensitivity of during the grain filling process lower wheat yields. the impact models involved to changes in variability Whether a drought occurs early or late in the would be required to provide specific answers. growing season has differential effects on yields. What can be said at this point is that the lack of Changes in variability were not considered in the information on climate variability has limited a Rosenzweig, Peart et al., Ritchie, and Dudek studies number of studies in this report and has limited the (see Chapter 6: Agriculture). completeness of the answers they could provide. Changes in the frequencies of extreme events are considered to be of great importance to RESEARCH NEEDS potential forest disturbance, as discussed in Chapter 5: Forests. The possibility of increases in the frequencies of events such as droughts, flooding, The research reported above clearly indicates wind, ice, or snowstorms may be of greater that research of changes in climate variability significance to forest survival than the gradual mean associated with climate change is truly in its infancy. change in climate that has been studied so far. Much needs to be done. Future research needs may be broken into three categories: further The Kalkstein study, which is reviewed in analysis of GCMs; improvements in GCMs; and Chapter 12: Human Health, is strongly dependent sensitivity analysis of impacts. upon the determination of certain maximum temperature threshold values beyond which human Further Investigation of Variability in mortality increases. In applying the death/weather GCMs effects statistical models to scenarios of climate change, Kalkstein held temperature variability Results summarized here represent only an constant, so that temperatures that exceed the initial effort at looking at variability in GCMs. We threshold values are determined unrealistically. need to examine in more models and at many more grid boxes the daily and interannual variability of Changes in the variability of temperature both many climate variables (such as relative humidity, seasonally and daily are important to studies solar radiation, and storm frequency) in addition to concerned with the effect of temperature change on temperature and precipitation. Other time scales electricity demand (discussed in Chapter 10). of variability also should be examined, such as 7- to Although new generating capacity requirements for 10-day scales, which correspond to the lifetime of the nation for 2010 and beyond are calculated many frontal storms. Moreover, the most assuming climate change, the numbers generated sophisticated statistical techniques must be used or, could be considerably different for any particular where needed, developed, such that uniform year, depending mainly on air-conditioning needs, quantitative indicators are available to evaluate both which would be the major use increase for how well the current models reproduce present electricity. Such needs are sensitive to extremes in variability and how they forecast the change 51 Chapter 3 in variability under climate change conditions. The REFERENCES causes for discrepancies in present-day climate variability and control run variability must be better understood to attain a clearer understanding of future climate changes. American Meteorological Society. 1985. Flash floods: A statement of concern by the AMS. Improvements in GCMs Bulletin of the American Meteorological Society 66(7):858-859. The results of Rind et al. and Mearns et al. give Barbecel, O., and M. Eftimescu. 1973. Effects of some indications that oversimplifications in the land Agrometeorological Conditions on Maize Growth surface packages of GCMs contribute to and Development. Bucharest, Romania: Institute overpredictions of daily temperature variability. of Meteorology and Hydrology. pp. 10-31. This possibility is further underlined by the better results obtained with Dickinson's model, which Bates, G.T., and G.A. Meehl. 1986. The Effect of includes a more sophisticated land surface package. CO₂ concentration on the frequency of blocking in More detailed analyses of current GCMs are a general circulation model coupled to a simple necessary to confirm this speculation, as well as to mixed layer ocean model. Monthly Weather Review determine the causes of other errors in variability, 114:687-701. such as for precipitation. Other known causes of error, such as the models' relative inability to Becker, R.J., and R.A. Wood. 1986. Heatwave. simulate subgrid-scale phenomena, must be Weatherwise 39(4):195-6. investigated further. The next step involves altering the GCMs so that variability is properly simulated. Berger, A. 1980. Spectrum of climate variations and Only then can much faith be put in GCM forecasts possible causes. In: Berger, A., ed. Climatic of variability changes with a perturbed climate. Variations and Variability: Facts and Theories. Dordrecht, Holland: D. Reidel. pp. 411-432. Sensitivity Analyses of Impacts Bourke, W. 1974. A multi-level spectral model. I. It also must be determined how important Formulation and hemispheric integrations. Monthly changes in variability will be to different areas of Weather Review 102:687-701. impact. Since the variability of climate variables produced from GCMs cannot be "trusted" or even Bourke, W., B. McAvaney, K. Puri, and R. Thurling. easily analyzed at this point, these sensitivity 1977. Global modeling of atmospheric flow by analyses of impact models should be performed with spectral methods. Methods in Computational statistically simulated time series of climate Physics 17:267-324. variables, as has been performed by Schwarz (1976) and Mearns et al. (1984). By simulating time series, Bridger, C.A., F.P. Ellis, and H.L. Taylor. 1976. different levels of autocorrelation and variance in Mortality in St. Louis, Missouri, during heat waves the time series may be controlled for and in 1936, 1953, 1954, 1955, and 1966. Environmental systematically varied. By this means, important Research 12:38-48. thresholds of variability change for different variables as they affect the output of impact models Brinkmann, W.A.R. 1983. Variability of can be determined. Moreover, ranges of possible temperature in Wisconsin. Monthly Weather impacts of variability change can be determined and Review 111:172-180. can serve as guides until better information is available on how variability will change in a Chervin, R.M. 1986. Interannual variability and CO2-warmed world. seasonal predictability. Journal of the Atmospheric Sciences 43:233-251. 52 Climate Variability Chervin, R.M. 1981. On the comparison of Hansen, J., I. Fung, A. Lacis, S. Lebedeff, D. Rind, observed GCM simulated climate ensembles. R. Ruedy, G. Russell, P. Stone. 1988. Global Journal of the Atmospheric Sciences 38:885-901. climate changes as forecast by the Goddard Institute for Space Studies three-dimensional model. Journal Clark, R.A. 1987. Hydrological design criteria and of Geophysical Research 93(D8):9341-9364. climate variability. In: Soloman, S.I., M. Beran, and W. Hogg, eds. The Influence of Climate Change Heathcote, R.L. 1985. Extreme event analysis. In: and Climatic Variability on the Hydrologic Regime Kates, R.W., J.H. Ausubel, and M. Berberian, eds. and Water Resources. International Association of Climate Impact Assessment, SCOPE 27. Hydrological Sciences Publ. No. 168. Oxfordshire: Chichester, United Kingdom: John Wiley and Sons. IAHS Press. Henz, J.F., and V.R. Scheetz. 1976. The big Coleman, J.E. 1988. Climatic warming and Thompson flood of 1976 in Colorado. Weatherwise increased summer aridity in Florida, U.S.A. Climatic 29(5):278-285. Change 12:164-178. Herbert, P.J., and G. Taylor. 1979. Everything you Dare, P.M. 1981. A study of the severity of the always wanted to know about hurricanes. midwestern winters of 1977 and 1978 using heating Weatherwise 32(2):61-67. degree days determined from measured and wind chill temperatures. Bulletin of the American Jones, T.S., et al. 1982. Morbidity and mortality Meteorological Society 62(7):674-682. associated with the July 1980 heat wave in St. Louis and Kansas City, MO. Journal of the American Diaz, H.F. and R.G. Quayle 1980. The climate of Medical Association 247:3327-3331. the United States since 1895: Spatial and temporal changes. Monthly Weather Review 108:249-266. Kalkstein, L.S., R.E. Davis, J.A. Skindlov, and K.M. Valimont. 1987. The impact of human-induced Dickinson, R.E., A. Henderson-Sellers, P.J. climate warming upon human mortality: A New Kennedy, and M.F. Wilson. 1986. York case study. Washington, DC: U.S. Biosphere-Atmosphere Transfer Scheme (BATS) Environmental Protection Agency. for the NCAR Community Climate Model. NCAR Technical Note 275. Boulder, CO: National Center Katz, R.W. 1984. Procedures for Determining the for Atmospheric Research. Statistical Significance of Changes in Variability Simulated by an Atmospheric General Circulation Ellis, F.P. 1972. Mortality from heat illness and Model. Climate Research Institute Report No. 48. heat-aggravated illness in the U.S. Environmental Corvallis, OR: Oregon State University. Research 5(1):1-58. Linder, K.P., M.J. Gibbs, and M.R. Inglis. 1987. Emmanuel, K.A. 1987. The dependence of Potential Impacts of Climate Change on Electric hurricane intensity on climate. Nature 326:483-485. Utilities. Albany, NY: New York State Energy Research and Development Authority. Report 88.2. Emmanuel, K.A. 1988. Toward a general theory of hurricanes. American Scientist 76(4):370-379. Lough, J.M., T.M.L. Wigley, and J.P. Palutikof. 1983. Climate and climate impact scenarios for Gibbs, W.J. 1984. The great Australian drought: Europe in a warmer world. Journal of Climate and 1982-1983. Disasters 8/2/84:89-104. Applied Meteorology 22:1673-1684. Gibbs, W.J., J.V. Maher, and M.J. Coughlan. 1975. Ludlum, D.M. 1988. Weatherwatch (for Climatic variability and extremes. In: Pittock, A.B. September). Weatherwise 41(6):354-355. L.A. Frakes, D. Jenssen, J.A. Peterson, and J.W. Zillman, eds. Climatic Change and Variability. Manabe, S., and D.G. Hahn. 1981. Simulation of Cambridge, United Kingdom: Cambridge University atmospheric variability. Monthly Weather Review Press. pp. 135-150. 109:2260-2286. 53 Chapter 3 McQuigg, J.D. 1981. Climate variability and crop Oliver, J. 1981. The nature and impact of yield in high and low temperature regions. In: Back Hurricane Allen - August 1980. Journal of W., J. Pankrath, and S.H. Schneider, eds. Climatology 1:221-235. Food-Climate Interactions. Dordrecht, Holland: D. Reidel Publishing Company. pp. 121-138. Parry, M.L. 1978. Climatic Change, Agriculture and Settlement. Folkstone, United Kingdom: Mearns, L.O., S.H. Schneider, S.L. Thompson, and Dawson. L.R. McDaniel. 1989. Analysis of climate variability in general circulation models: comparison Pittock, A.B. 1975. Patterns of variability in with observations and changes in variability in relation to the general circulation. In: Pittock, 2xCO₂ experiments. Journal of Geophysical A.B., L.A. Frakes, D. Jenssen, J.A. Peterson, J.W. Research (accepted). Zillman, eds. Climatic Change and Variability. Cambridge, United Kingdom: Cambridge Mearns, L.O., R.W. Katz, and S.H. Schneider. University Press. pp. 167-178. 1984. Extreme high-temperature events: Changes in their probabilities with changes in mean Policansky, D. 1977. The winter of 1976-77 and the temperature. Journal of Climate and Applied prediction of unlikely weather. Bulletin of the Meteorology 23:1601-1613. American Meteorological Society 58(10):1073-1074. Mederski, H.J. 1983. Effects of water and Quirk, W.J. 1981. Climate and energy temperature stress on soybean plant growth and emergencies. Bulletin of the American yield in humid temperature climates. In: Raper, Meteorological Society 62(5):623-631. C.D., and P.J. Kramer, eds. Crop Reactions to Water and Temperature Stresses in Humid, Reed, D.N. 1986. Simulation of time series of Temperate Climates. Boulder, CO: Westview temperature and precipitation over eastern England Press. pp. 35-48. by an atmospheric general circulation model. Journal of Climatology 6:233-257. Meehl, G. 1989. Seasonal cycle forcing of El Niño phenomena in a coupled ocean-atmosphere model. Reibsame, W.E. 1988. Adjusting water resources Submitted to Science. management to climate change. Climatic Change 13:69-97. Mitchell, J.M. 1976. An overview of climatic variability and its causal mechanisms. Quaternary Rind, D., R. Goldberg, and R. Ruedy. 1989. Research 6:481-493. Change in climate variability in the 21st century. Climatic Change 14:5-37. Mogil, H.M., A. Stern, and R. Hagen. 1984. The great freeze of '83: Analyzing the causes and the Rosenberg, N.J., ed. 1979. Drought in the Great effects. Weatherwise 37(6):304-308. Plains: Research on Impacts and Strategies. Littleton, CO: Water Resources Publications. Neild, R.E. 1982. Temperature and rainfall influences on the phenology and yield of grain Rosenberg, N.J., ed. 1978. North American soybean and maize. Agricultural Meteorology Droughts. Boulder, CO: Westview Press. 27:79-88. Sanders, J.F. 1982. The hurricane dilemma. Newman, J.E. 1978. Drought impacts on American Weatherwise 35(4):174-178. agricultural productivity. In: Rosenberg, N.J., ed. North American Droughts. Boulder, CO: Shugart, H.H., M.Y. Antonovsky, P.G. Jarvis, and Westview Press. pp. 43-63. A.P. Sandford. 1986. CO₂, climatic change, and forest ecosystems. In: Bolin, B., B.R. Doos, J. Oeschsli, F.W., and R.W. Buechley. 1970. Excess Jager, and R.A. Warrick, eds. The Greenhouse mortality associated with three Los Angeles Effect, Climatic Change, and Ecosystems. SCOPE September hot spells. Environmental Research 29. Chichester, United Kingdom: John Wiley and 3:277-284. Sons. pp. 475-521. 54 Climate Variability Schwarz, H.E. 1977. Climatic change and water Warrick, R.A. et al. 1975. Drought Hazard in the supply: how sensitive is the Northeast? In: United States: A Research Assessment. Boulder, Climate, Climate Change and Water Supply. CO: University of Colorado. IBS Monograph No. Washington, DC: National Academy of Sciences. NSF-RA-E-75-004. pp. 111-120. Washington, W.M., and G.A. Meehl. 1984. Shaw, R.H. 1983. Estimates of yield reductions in Seasonal cycle experiment on the climate sensitivity corn caused by water and temperature stress. In: due to a doubling of CO₂ with an atmospheric Raper, C.D., and P.J. Kramer, eds. Crop Reactions general circulation model coupled to a simple to Water and Temperature Stresses in Humid, mixed-layer ocean model. Journal of Geophysical Temperate Climates. Boulder, CO: Westview Research 89(D6):9475-9503. Press. pp. 49-66. White, G.F. et al. 1975. Flood Hazard in the Soloman, A.M., and D.C. West. 1985. Potential U.S.A. Research Assessment. Boulder, CO: responses of forests to CO2-induced climate change. University of Colorado. IBS Monograph No. In: White, M.R., ed. Characterization of NSF-RA-E-75-006. Information Requirements for Studies of CO₂ Effects: Water Resources, Agriculture, Fisheries, Wilford, J.N. 1988. Vast Persistent Pattern Forests, and Human Health. U.S. Department of Spreading Heat Wave. The New York Times, p. 1; Energy. DOE/ER-0236. pp. 145-170. July 19. Thompson, L.M. 1968. Weather and technology in Wilhite, D.A., and M.H. Glantz. 1987. the production of corn. In: Purdue Top Farmer Understanding the drought phenomenon: The role Workshop, Corn Production Proceedings. West of definitions. In: Wilhite, D.A., and W. Easterling, Lafayette, IN: Purdue University. pp. 3-19. eds. Planning for Drought. Boulder, CO: Westview Press. pp. 11-27. Trenberth, K.E., G.W. Branstator, and P.A. Arkin. 1988. Origins of the 1988 North American drought. Wilson, C.A., and J.F.B. Mitchell. 1987. Simulated Science 242:1640-1645. climate and CO2-induced climate change over Western Europe. Climatic Change 10:11-42. Van Loon, H. and J. Williams. 1978. The association between mean temperature and interannual variability. Monthly Weather Review 106:1012-1017. 55 CHAPTER 4 METHODOLOGY NEED FOR CLIMATE CHANGE SCENARIO COMPONENTS SCENARIOS To assess the potential effects of global climate change, regional scenarios of such change should As discussed in Chapter 2: Climate Change, have the following characteristics: there is a scientific consensus that increased atmospheric concentrations of greenhouse gases will 1. The scenarios should be internally likely increase global temperatures, and that such a consistent with global warming caused by global temperature increase will likely increase increases in greenhouse gas emissions. A global precipitation and sea levels. There is no doubling of the CO₂ concentration in the consensus on how regional climates may change. atmosphere is thought to increase global We do not know whether temperatures will rise in temperatures by approximately 1.5 to 4.5°C all regions; we do not know whether precipitation in (3 to 8°F). The regional temperature any particular region will rise or fall or whether we changes and seasonal distributions may be will have seasonal changes, and we are uncertain higher or lower, as long as they are about the rate and magnitude of change. As internally consistent with the global range. discussed in Chapter 3: Climate Variability, scientists do not know how variability - that is, the 2. The scenarios must include a sufficient frequency of droughts, storms, heat waves, and number of meteorological variables to meet similar phenomena -- may change. Without the requirements for using effects models. knowing how regional climate may change, we These effect models include models of crop cannot predict impacts. growth, forest succession, runoff, and other systems. Some models of the relationship Despite these uncertainties, we can get a sense between climate and a system use only of what the future may look like through the use of temperature and precipitation as climate scenarios. Scenarios are plausible combinations of variables, while others also need solar conditions that may be used to illustrate future radiation, humidity, winds, and other events. They may be used to identify possible variables. effects of climate change and to evaluate responses to those effects. To incorporate uncertainties 3. The meteorological variables should be surrounding regional climate change, regional internally consistent. While a scenario is scenarios should include a variety of potential not a prediction, it should at least be climate changes consistent with the state of plausible. The laws of physics limit how knowledge regarding global warming. By analyzing meteorological variables may change in many scenarios, we may be able to identify the relationship to each other. For example, if direction and relative magnitude of impacts. Yet, global temperatures increase, global unless scenarios have probabilities assigned to them, precipitation must also rise. Regional predictions of future impacts cannot be made. In changes should be internally consistent with this report, probabilities are not assigned and results these large-scale changes. do not represent predictions. Only the direction of change and relative magnitude are identified. The 4. The scenarios should provide scenarios used in this report do not represent the meteorological variables on a daily basis. entire range of possible climate change. Thus, the Many of the effects models used in this range of effects identified does not represent the study, such as crop yield and hydrology entire range of potential effects. models, need daily meteorological inputs. 57 Chapter 4 5. Finally, the scenarios should illustrate what arbitrary amount. For example, one could assume climate would look like on a spatial scale that temperatures increase by 2 or 4°C, or that fine enough for effects analysis. Many rainfall rises or falls by 10% and all other variables effects models consider changes in are held constant. Such scenarios are relatively easy individual stands of trees or farm fields. To to use and can help to identify the sensitivities of run them, scenarios must illustrate how systems to changes in different variables. To climate may change locally. determine how sensitive a system is to temperature alone, one could hold other variables at current climate levels and change temperature by arbitrary TYPES OF SCENARIOS amounts. Two questions should be answered in analyzing A major drawback to using scenarios with the potential impacts of the greenhouse effect: arbitrary changes is that they may not be realistic, What would be the effects of a large climate change since evaporation, precipitation, wind, and other in the future? How quickly will the effects become variables will most likely change if global apparent over time? The first question asks what temperatures change. A combination of unrealistic the world will be like in the future; the second is meteorological changes may yield an unrealistic about the speed of change and the sensitivity of the effect. We are not sure how other meteorological system. variables would change on a regional scale if temperature rose a certain amount. Thus, scenarios One way of examining the first question is to with arbitrary changes may be useful for use scenarios of an equilibrium future climate. determining sensitivities to particular variables but Climate equilibrium is defined as climate in which not for determining the possible magnitudes of average conditions are not changing (although year- effects. to-year variations could still occur). Analog Warming A drawback of an equilibrium scenario is that it occurs at an arbitrary point in the future and Many climatologists have advocated the use of assumes that the climate has reached a stable level historic warming periods as an analog of how a corresponding with the higher concentrations of future warming may affect regional climates greenhouse gases. It does not indicate how climate (Vinnikov and Lemeshko, 1987). The instrumental may change between now and the equilibrium weather record can be used by comparing a cool condition or how soon effects may be seen. decade on record, such as the 1880s, with a warm Furthermore, a "stable" climate has never happened, decade, such as the 1930s (Wigley, 1987), or by nor is it likely to occur. comparing a decade such as the 1930s with the present. To help identify sensitivities and give a sense of when effects may occur, this study uses transient Paleoclimatic data may also be incorporated scenarios of climate change. A transient scenario is into an analog warming scenario. For example, a scenario of how climate may change over time. 6,000 years ago the temperatures were about 1°C warmer. Paleoclimatologists have determined how The options for creating regional scenarios of rainfall and temperature patterns on a broad global warming include the following: regional scale differed in the past. The changes associated with past climates that were warmer than 1. arbitrary changes in climate; now may be used as an analog warming scenario. 2. analog warming; and The advantage of using an analog is that it gives a realistic sense of how regional and local 3. use of general circulation models. weather patterns change as global climate warms. For example, climate data from 1880 to 1930 show Arbitrary Changes how daily and local weather changed during a warming period. A simple way of constructing a scenario is to assume that climate variables change by some 58 Methodology However, analogs have several drawbacks. GCMs have several advantages over the other First, they are not consistent with the range of approaches for creating scenarios. First, the models global warming now thought likely under the are used to estimate how global climate may change greenhouse effect: 1.5 to 4.5°C. The warmest in response to increased concentrations of period of the last 125,000 years was 1°C warmer greenhouse gases. Thus, regional outputs are than the present temperature. (Although the internally consistent with a global warming Pliocene Epoch (2 to 5 million years ago) had associated with doubled CO₂. Second, the estimates global temperatures several degrees higher than of climate variables (for example, rainfall, now, there is virtually no information on the temperature, and humidity levels) are physically regional distribution of temperature and rainfall consistent within the bounds of the model physics. during that period.) In addition, the past warmings Third, GCMs estimate outputs for many were not necessarily caused by changes in the meteorological variables (including wind, radiation, concentration of greenhouse gases, but may have cloud cover, and soil moisture) providing enough been due to such factors as shifts in the inclination input for effects models. Fourth, GCMs simulate of the Earth's axis. These factors caused different climate variability on at least a daily basis. regional climate changes than would be associated with increases in radiative forcing. Second, Among the most important limitations are the paleoclimatic and historic records do not provide GCMs' simulations of the oceans. The oceans play enough detail to conduct comprehensive analysis of a critical role in determining the rate of climate the 1°C warming. Paleoclimatic records only change, regional climate differences, and climate indicate broad regional patterns of change for a few variability. The GCMs, however, are coupled to variables, such as temperature, rainfall, and solar relatively simple models of ocean circulation, which radiation. We cannot discern local, daily, or either treat the oceans as a "swamp" or only model interannual climate from these records. Even using the upper layers of oceans. The models' the 1930s data presents some problems. Daily assumptions oversimplify the transfer of heat to and records are available only for temperature and from the oceans. In addition, the GCMs simplify rainfall. Some effects models need more variables, other important factors that affect climate, including such as wind or radiation. Furthermore, the cloud cover and convection, sea ice, surface albedo number of weather stations with 1930s data is (the amount of light reflected, rather than absorbed, limited, which could present problems for creating from the surface) and land surface hydrology (i.e., comprehensive regional scenarios. soil moisture), which may also contribute to uncertainty about the estimates of climate change General Circulation Models (GCMs) (Dickinson, 1986; Schlesinger and Mitchell, 1985; Gates, 1985). For example, some of the GCMs GCMs are dynamic models that simulate the model soil moisture storage in a simple manner, physical process of the atmosphere and oceans to assuming the soils act like a "bucket." (There have estimate global climate. These models have been been recent improvements on this method.) This developed over two decades and require extensive method of modeling raises uncertainties concerning computations to run. They can be run to estimate estimates of runoff from the models. The way current climates and the sensitivity of climate to GCMs simulate such important climate factors as different conditions such as different compositions oceans, clouds, and other features casts some doubt of greenhouse gases. The GCMs are often used to on the validity of the magnitude of global warming simulate climate caused by a doubling of carbon estimated by the models. (For a further discussion dioxide levels, also referred to as doubled CO2. of the role of oceans in climate change, see Chapter Estimates of climate change caused by this effective 2: Climate Change. For a discussion of the GCMs' doubling of co₂¹ are referred to as "doubled CO₂ ability to estimate climate variability, see Chapter 3: scenarios." Output is given in regional grid boxes. Climate Variability.) One of the major disadvantages of using 1 The "effective doubling of CO2" means that the total radiative GCMs for effects analysis is their low spatial forcing of all greenhouse gases (CO2, CH4, N2O, CFCs, etc.) is resolution. GCMs give outputs in grid boxes that the same as the radiative forcing caused by doubling carbon vary in size from 4 by 5 degrees latitude to as much dioxide concentrations, over midcentury levels, alone. In other words, the combination of all greenhouse gases has the same as 8 by 10 degrees longitude. Figure 4-1 shows the radiative forcing as simply doubling CO2. grid boxes from the Goddard Institute for Space 59 Chapter 4 47.0 39.1 31.3 23.5 -135 -125 -115 -105 -95 -85 -75 -65 Figure 4-1. GISS model of the United States. Studies (GISS) model overlaid on a map of the observations on different scales. GCM estimates of United States. Each grid box is 8 by 10 degrees and rainfall are less reliable on a regional scale. As is an area larger than France (Mitchell, 1988). Grotch points out, the disparities between GCM Within each grid box, the actual climate may be estimates of current regional climate and actual quite variable. For example, although both are in conditions calls into question the ability of GCMs to the same grid box, the weather in southern predict climate change on a regional scale. Washington State may be quite different from the weather in northern California. The models, The disparities among GCM estimates on a however, do not account for variations within each regional scale are due to a number of factors. One grid box. For any simulated time, they provide a of the most important is the simplified assumptions single value for temperature, for rainfall, and for concerning the oceans. The assumptions on other other variables for the entire area of the box. factors such as cloud cover, albedo, and land surface hydrology also affect regional estimates. The GCMs A second disadvantage for effects analysis, also simplify topographic features within grid boxes, which may be more critical than the first, is that such as the distribution of mountains or lakes. The GCMs generally do not accurately simulate current large size of the grid boxes means that these regional climate conditions. In general, the features are oversimplified on a geographic scale. accuracy of GCM climate estimates decreases with This contributes to uncertainty regarding estimates increasing resolution. The GCMs do a reasonable of regional climate change. In sum, as Grotch job of estimating observed global and zonal concluded, GCM estimates of regional climate climates, but the estimates of regional climate are, change should not be taken as predictions of in many cases, far from observed conditions. This regional climate change. They should be is shown in Table 2-2 (see Chapter 2: Climate interpreted as no more than illustrations of possible Change), adapted from Grotch (1988), which future regional climate conditions. displays GCM temperature estimates and actual 60 Methodology CHOICE OF DOUBLED CO2 of warm, cold, wet, and dry years. Since the data are from the most recent decades, they are the most SCENARIO complete historic data available. A complete daily record for a number of weather variables only GCM outputs were employed as a basis for began in 1948. constructing the scenarios to be used in our report because they produce the best estimate of climate GCMs Used change due to increased greenhouse gas concentrations and they produce regional climate To obtain a range of scenarios, output from estimates internally consistent with doubled CO₂ three GCMs was used: concentrations. Yet, GCMs are relatively new tools that need a great degree of refinement. Their Goddard Institute for Space Studies (GISS) results must be applied with caution. The regional (Hansen et al., 1988); GCM estimates of climate change are considered to be scenarios, not predictions. Given the Geophysical Fluid Dynamics Laboratory uncertainties about GCM estimates of daily and (GFDL) (Manabe and Wetherald, 1987); interannual variability (see Chapter 3: Variability), and a conservative approach involves using average monthly changes for each grid box. Oregon State University (OSU) (Schlesinger and Zhao, 1988). The scenarios described in this chapter are a hybrid between GCM outputs and historic weather The average seasonal temperature and data. The estimates of average monthly change in precipitation for the U.S. gridpoints for each model temperature, precipitation, and other weather are displayed in Figure 4-2. All three models variables are used from GCM grid boxes. Model estimate that average temperatures over the United simulations of monthly doubled CO₂ conditions are States would rise, but they disagree on the divided by model simulations of average monthly magnitude. OSU gives 3°C, GISS 4.3°C, and GFDL current conditions in each grid. The ratios of 5.1°C. The seasonal patterns are different, with (2xCO₂): (1xCO₂) are multiplied by historic weather GISS having a larger warming in winter and fall, conditions at weather stations in the respective grid GFDL having the highest temperature change in the boxes. Parry et al. (1987) used this approach in an spring, and OSU having little seasonal variability. analysis of impacts of climate change on agriculture. All three models estimate that annual precipitation Thus, if a grid box is estimated to be 2°C warmer over the United States would increase. GISS and under the GCM doubled CO₂ run, all stations in OSU estimate that annual precipitation would rise, that grid are assumed to be 2°C warmer in the respectively, by 73 millimeters (2.92 inches) and 62 doubled CO₂ scenario. The effect of this is to millimeters (2.48 inches), while GFDL estimates a keep geographic variation from station to station rainfall increase of only 33 millimeters (1.31 inches). within a grid the same as in the historic base period. The first two models have precipitation increases in Furthermore, interannual (year to year) and daily all four seasons, while GFDL has a decline in variability remain the same. If rainfall occurs 10 summer rainfall. As can be seen in the regional days in a month, in the scenario it also occurs 10 chapters, the models show greater disagreement on days in the month, and the amount of rainfall is the direction and pattern of regional rainfall adjusted by the GCM output. Since these scenarios changes than on regional temperature. Overall, are hybrids between GCM average monthly OSU appears to be the "mildest" scenario, with the estimates and daily historic weather records, these lowest temperature rise and largest increase in scenarios are not strictly GCM scenarios. Each precipitation. GFDL appears to be the most scenario is referred to by the GCM, whose monthly "extreme," with the highest temperature rise, the output serves as its base (e.g., the "GISS scenario"). smallest increase in precipitation, and a decrease in summer rainfall. Some of the important parameters The years 1951-80 were chosen as the base in the three GCMs are displayed in Table 4-1. period to which average doubled CO₂ changes were applied. Several decades of data give a wide range 61 Chapter 4 Temperature Precipitation 6 0.4 GISS 5 0.3 GFDL 0.2 4 OSU CHANGE (°C) 3 CHANGE (mm/day) 0.1 0 2 -0.1 1 -0.2 O -0.3 Winter Spring Summer Fall Annual Winter Spring Summer Fall Annual Figure 4-2. Average changes in temperature (°C) and precipitation (mm/day) over the grid boxes of the lower 48 states (2xCO₂ less 1xCO₂). The "extreme" values in the GFDL doubled temperatures above a certain level. The studies do CO₂ scenario are due, in part, to assumptions made not identify how these and other systems could be in the model run used in this report. That run did affected by changes in temporal climate variability. not constrain sea surface temperature and sea ice, Holding spatial variability within a grid box constant which yielded seasonal extremes in the northern also affects the results of the analyses performed for hemisphere. A later run, produced too late for use this report. Climate change may also lead to in this study, constrained sea surface temperature changes in wind patterns, which could change storm and sea ice to observed values. Both runs yield the patterns, cloud distribution, deposition of air same average global warming of 4.0°C, while the pollutants, and other systems. In addition, the years later run has greater seasonal extremes in the 1951 to 1980 were a period of relatively low weather southern hemisphere. Both runs show a large variability in the United States. Only adjusting decrease in summer soil moisture (Wetherald, average conditions from the base period in the personal communication, 1988). scenarios may underestimate potential increases invariability. (For further discussion, see Chapter Limitations 3: Variability.) A major limitation of the doubled CO₂ The choice of the three doubled CO2 scenarios scenarios used for this study is the lack of temporal does not necessarily bracket the range of possible and spatial variability. By applying average monthly climate change in the latter half of the next century. changes to the historic data set, it is assumed that Due to the uncertainties about the rate and the daily and interannual patterns of climate remain magnitude of global warming, it is possible that the same. This assumption is probably unrealistic, average global temperatures could be lower or since a change in average conditions will probably higher than indicated by the models. Other climate lead to a change in variability. Furthermore, variables could be different too. Thus, these holding variability constant can have an impact on scenarios should be interpreted as illustrations of effects analysis. possible future conditions, not as predictions. Furthermore, we did not assign probability to these Most climate-sensitive systems are sensitive to scenarios. Currently, there is not enough climate variability. For example, riverflow is very information or a methodology for making such a sensitive to the amount and intensity of rainstorms. determination. Certain crops are sensitive to consecutive days with 62 Methodology Table 4-1. Major Features for the Three GCMsᵃ Model Base Temp for Increase When resolution Model Diurnal 1xCO2 doubled in global GCM calculated (lat. X long.) levelsᵇ cycle (ppm) CO₂ precipitation (°C) (%) GISSC 1982 7.83 X 10d 9 yes 315 4.2 11 GFDLᵈ 1984-85 4.44 X 7.5d 9 no 300 4.0 8.7 OSU 1984-85 4.00 X 5.0ᵈ 2 no 326 2.8 7.8 GISS Transient 1984-85 7.83 X 10d 9 yes 315 -- -- (in 1958) a All models are global in extent and have an annual cycle. All models have a smoothed topography that varies between models. The later GFDL run has been added for information. All models (except the transient) give data for the present climate (1xCO₂) and double CO₂ climate (2xCO₂). b All models make calculations for surface conditions as well as for the listed upper-air levels. CA gridpoint model with stated resolution. This is a spectral model that has 15 waves. Note: Oceans in Models: GISS: This model has a slab ocean not over 65 meters deep; it has some variation of mixed depth over the seasonal cycle (for example, the depth is shallower in summer than winter in mid-latitudes). It has a specified pseudo ocean heat transport designed to reproduce the present day sea surface temperature (SST) in the simulation of the present climate. Ice thickness is predicted. For the GISS transient runs, the ocean depth was not limited in this way. In it, the average annual maximum mixed-layer depth was 127 inches. GFDL: The slab ocean is 68 meters deep. There is no horizontal heat transport that would make the present day SST come out exactly right. Ice thickness is predicted. OSU: This model has a slab ocean that is 60 meters deep (only 5 meters deep during spin-up period). It does not have heat transport that would force the model to reproduce the present day SST (this is being added in 1989). If current emission trends continue, the doubling would not occur at the same time as the effective doubling of CO2 concentrations will occur increase in greenhouse gas concentrations. The around the year 2030. However, that estimate does oceans absorb greenhouse gases and heat from the not account for some recent developments that may atmosphere and serve to delay the warming. The slow the increase in greenhouse gas concentrations. full extent of climate change associated with CO₂ If implemented, the Montreal Protocol would cut doubling could take several decades or more and emissions of chlorofluorocarbons (CFCs) by 50%. may not occur until the latter half of the next If an international agreement is reached on century. reduction of nitrogen oxides (NOx), the concentration of nitrogen dioxide (N₂O) may be In this report, results from doubled CO₂ slightly reduced. Pollution control measures in scenarios are generally not associated with a countries such as the United States may also reduce particular year. When analysis is necessary, we have concentrations of low-level ozone, another generally assumed that the CO₂ warming will occur greenhouse gas. Thus, the effective doubling of in 2060. In some cases, researchers assumed a CO₂ may happen after 2030. different time period for CO2 warming, and those exceptions are noted as appropriate in the text. As discussed in Chapter 2: Climate Change, the change in climate potentially caused by CO₂ 63 Chapter 4 The doubled CO2 scenarios are often analog takes one only as far as a 0.5°C warming or, interpreted as estimates of future static in the case of paleoclimatic records, a 1°C (equilibrium) conditions. The assumption that the warming. It does not indicate what happens in the concentration of greenhouse gases becomes constant decades after the 0.5 to 1.0°C level is reached. In at doubled CO₂ levels is an arbitrary one. In fact, addition, the analog may not represent the regional if emissions are not limited, concentrations could distribution of climate associated with greenhouse become greater and the global climate would forcing. continue to change. In many places in this report, responses are presented as if the climate stabilizes GCM Transient Runs at doubled CO₂ conditions. Natural systems and society, however, may be responding and adapting The Goddard Institute for Space Studies has to continuing and perhaps, accelerating changes in modeled how global climate may change as climate. concentrations of greenhouse gases gradually rise over the next century. This is called the transient run. GISS has modeled climate change under OPTIONS FOR CREATING several assumptions of trace gas growth. The TRANSIENT SCENARIOS transient runs start in 1958 with the atmospheric concentrations of greenhouse gases that existed The options for developing transient scenarios then. The concentrations of the gases and are similar to the options for the doubled CO₂ equivalent radiative forcing were estimated to scenarios: increase from 1958 until an arbitrary point in the future according to several different assumptions 1. arbitrary changes; regarding trace gas growth. The GISS transient run yields daily climate estimates from 1958 until that 2. analog warming; and arbitrary point. 3. GCM transient runs. For example, one of the transient scenarios, which is known as GISS A, assumes that trace gas concentrations continue to increase at historic rates Arbitrary Changes and net greenhouse forcing increases exponentially. The scenario is run from 1958 to 2062. The end of One could examine the manner in which a the transient corresponds with a global warming system responds to an arbitrary 1 or 2°C equivalent to that of the equilibrium climate from temperature warming and to small arbitrary changes the doubled CO2 run. This scenario does not in other variables. The problems of physically account for the potential reduction in CFC inconsistent assumptions about changes among emissions due to the Montreal Protocol or for other variables and regions pertain here also. In addition, activities that may reduce the growth in emissions. the arbitrary warming scenario gives no indication GISS B assumes a decreasing trace gas of when the warming may occur. concentration growth rate such that climate forcing increases linearly (Hansen et al., 1988). It stops in Analog Warming 2029. GISS B includes volcanoes, while GISS A does not. Wigley (1987) has suggested using analogs as scenarios for climates that may occur within the Since the GCMs are used to produce this next several decades. He noted that the warming transient run, the advantages and disadvantages of from the late 19th century to 1940 was about 0.4°C, using this approach are the same as those described which may approximate the transient warming over in the discussion of doubled CO2 scenarios. In the next two decades. The problem is that climate addition, the timing of the changes estimated by the may change faster in the future than in the early GCMs is complicated by the uncertainties regarding 20th century. (The average decadal warming may the growth of greenhouse gas emissions and the be as much as 0.5°C, rather than the 0.1°C roles of the oceans and clouds in delaying climate identified for earlier years.) Furthermore, the changes (Dickinson, 1986). 64 Methodology CHOICE OF TRANSIENT In this study, the historic time series 1951-80 is used, and the transient monthly statistics are applied SCENARIO to the time series. The procedure for creating the transient scenario was to first linearly interpolate This study used transient scenarios based on between decadal means. This smoothes out the the GISS transient run because, of all the different sharp decadal changes from the actual transient approaches, only this one provides internally GISS results and is shown in Figure 4-4(a). The consistent estimates of climate change and allows baseline 1951-80 weather data were repeated for 80 examination of the entire range of climate change years, with the last 20 years consisting of a between current conditions and doubled CO2 repetition of the 1951-70 data. Figure 4-4(b) shows climate. the average U.S. temperatures for 1951-80 repeated for 80 years. The data transformations displayed in In creating the transient scenario, an approach Figures 4-4(a) and (b) were done for data for each similar to that used for the doubled CO₂ scenario month for each grid box, site, and climate variable. was employed. Since relatively little confidence The smoothed month-by-month transient data were exists in the GCM's estimates of changes in added to the repeated 1951-80 data for each site interannual and daily variability, the monthly means and variable. Figure 4-4(c) displays the addition of were calculated for each decade of the transient. the smoothed average U.S. transient temperatures This process gives average decadal temperature, with actual U.S. 1951-80 temperatures, repeated. precipitation, and other changes. The average Although there is a cooling from the 1950s to the decadal temperature changes in GISS A and B for 1960s, followed by a warming in the 1970s, the the United States are shown in Figure 4-3. underlying warming of the transient, which is 3.7°C by the middle of the 2050s in GISS A, is much As in the doubled CO₂ scenario, the average greater than the variability in the base period. meteorological changes from the transient are combined with a historic time series. What is Limitations different from the doubled CO2 scenario is that a gradual change in temperature and other variables Since the transient scenarios were also derived is mixed with a historic time series with its own from GCMs, the same limitations concerning variability. This can produce a regular oscillation. 4 4 3.72 3.5 3.5 2.99 3 3 2.47 TEMPERATURE (°C) 2.5 1.72 1.36 TEMPERATURE (°C) 2.5 2 2 1.5 1.5 1.26 1.02 1 0.88 1 0.70 0.59 0.35 0.5 0.30 0.5 0.18 0 0 1980s 1990s 2000s 2010s 2020s 2030s 2040s 2050s 1980s 1990s 2000s 2010s 2020s TRANSIENT SCENARIO A TRANSIENT SCENARIO B Figure 4-3. GISS transients "A" and "B" average decadal temperature change for lower 48 states gridpoints. 65 Chapter 4 A. SMOOTHED GISS "A" ANNUAL AVERAGE U.S. TEMPERATURE OTHER SCENARIOS 4.0 3.5 In a few cases, researchers used meteorologic 3.0 data from the 1930s as an analog scenario. This 2.5 TEMPERATURE (°C) scenario was used to provide additional information 2.0 on the sensitivity of systems to climate change. In 1.5 a few other cases, researchers only examined 1.0 paleoclimatic records. In these cases, the goal was 0.5 to determine how a system responded to past 0.0 climate change. -0.5 1980 1990 2000 2010 2020 2030 2040 2050 2060 YEAR EPA specified that researchers were to use B. OBSERVED 1951-1980 ANNUAL U.S. TEMPERATURE REPEATED 2.0 three doubled CO2 scenarios, two transient 1.5 scenarios, and an analog scenario in this study. 1.0 Many researchers, however, did not have sufficient TEMPERATURE (°C) time or resources to allow for the use of all 0.5 scenarios. EPA asked the researchers to run the 0.0 -0.5 scenarios in the following order, going as far -1.0 through the list as time and resources allowed: 1980 1990 2000 2010 2020 2030 2040 2050 2060 YEAR C. TRANSIENT SCENARIO: SMOOTHED GISS "A" COMBINED WITH 1. GISS doubled CO2; 1951 AND 1980 REPEATED 4.0 3.5 2. GFDL doubled CO2; 3.0 2.5 3. GISS transient A; TEMPERATURE (Ce) 2.0 1.5 4. OSU doubled CO2; 1.0 0.5 5. Analog (1930 to 1939); and 0.0 -0.5 6. GISS transient B; 1980 1990 2000 2010 2020 2030 2040 2050 2060 YEAR Most researchers were able to use at least the GISS Figure 4-4. Transient scenarios (temperature change). and GFDL doubled CO₂ scenarios. Comparison of results across studies may be limited because of inconsistent use of scenarios. temporal and spatial variability pertain as in the doubled CO2 scenario. An additional limitation in Sea Level Rise Scenarios the transient scenario is the rate of change. The GISS transient runs assume a gradual rate of Unlike the climate scenarios, the alternative change in temperature. The simplistic treatment of sea level rise scenarios were not based solely on the ocean circulation in the GCM affects the rate of differences between various general circulation warming estimated by the model. Broecker (1987) models. Instead, they were based on the range of has shown that past climate changes may have been estimates that previous studies have projected for abrupt. Broecker, however, analyzed a global the year 2100 (Hoffman et al., 1983, 1986; Meier et cooling, and the changes occurred over a much al., 1985; Revelle, 1983; Thomas, 1986), which have longer period than greenhouse warming. A sudden generally considered alternative rates of greenhouse warming could mean that significant effects happen gas emissions, climate sensitivity ranging from 1.5 to sooner and more suddenly than the results of the 4.5°C for a CO₂ doubling, and uncertainties transient analysis used in this study indicate. The regarding ocean expansion and glacial melting. inclusion of the 1951-80 base period in the scenario Estimates for the year 2100 generally range from 50 yields short-term oscillations. to 200 centimeters. 66 Methodology This report uses three scenarios for the year considered. In some cases, researchers ran 2100 -- 50, 100, and 200 centimeters -- and additional scenarios with assumptions about compares them to the current trend of 12 technological and other changes. In addition, centimeters per century. Because most studies have potential responses to climate change were not reported estimates for the intermediate years, considered in some, but not all, cases. For these we followed the convention of a 1987 National and many other reasons, the results should be Research Council report (Dean et al., 1987) and interpreted only as an indication of the sensitivity of interpolated sea level rise using a parabola. The current systems to global warming, not as a rates of sea level rise assumed in this report are prediction of what the effects will be. displayed in Figure 7-8 in Chapter 7: Sea Level Rise. Because various coastal areas are also sinking In some situations, quantitative models of the (and in a few cases rising), relative sea level rise at relationship between climate and a particular system specific locations was estimated by adding current did not exist. In those cases, other approaches were local subsidence trends. Note that sea level rise used to try to identify sensitivities. Some scenarios are presented for the year 2100, while researchers examined how systems responded to doubled CO₂ scenarios are presented for the latter analog warmings. In other cases, expert judgment half of the 21st century. was used. This consisted of literature reviews to assemble information on sensitivities as they appear in the literature, and workshops and interviews to EFFECTS ANALYSES poll experts on how they thought systems would respond to global warming. In this study, the preferred approach for analyzing potential impacts of climate change was to develop quantitative estimates. Most researchers RESEARCH NEEDS estimated impacts by running models that simulate the relationship between weather and the relevant The scenarios used in this report help identify system. The climate scenarios were used as inputs the sensitivities of systems to climate change. into the models. Since the researchers had only Because of the lack of confidence concerning several months to do the analysis, they used either regional estimates of climate change from GCMs, "off-the-shelf" models or analytic techniques. In we cannot predict impacts. In order to predict the many cases, existing models were calibrated to new effects of climate change, major improvements need sites. This lack of time also limited the gathering of to be made in GCMs. These could take many new data to a few studies. years. In the meantime, we will continue to use scenarios to identify sensitivities. As with GCMs, A drawback of using empirical models of scenarios can also be improved. systems to estimate sensitivities is that the models are applied to climates for which they were not GCMs developed. The models estimate relationships with observed climate. This relationship is then To produce better estimates of regional extrapolated to an unprecedented climate. It is climate change, both the resolution of GCMs and possible that in the new climate situation, the the modeling of physical processes need to be statistical relationship may be different owing to the improved. The GCMs used for this report had crossing of a threshold or for some other reason. large grid boxes, in which major geographic With the drawbacks of empirical models, the features, such as the Great Lakes or the Sierra current statistical relationships are the best basis for Nevada Mountains, which have large impacts on quantitatively estimating sensitivities. local climate, were not well represented. Ideally, the higher the resolution, the better the For the most part, researchers analyzed the representation of geographic features. But each potential effects of climate change on systems as increase in resolution means a large increase in they currently exist. Although these changes may be computations and computing power needed to run quite substantial, potential changes in populations, the model. Furthermore, at high resolutions, the the economy, technology, and other factors were not GCMs may require new parameterizations. The 67 Chapter 4 resolution should be increased at least to the point REFERENCES at which major geographic features are well represented in the models. Broecker, W.S. 1987. Unpleasant surprises in the It is also important that the estimates of greenhouse? Nature 328:123-126. physical processes in the models be improved to increase the confidence about estimates of the Dean, R.G., R.A. Darylrumple, R.W. Fairbridge, magnitude and timing of changes. Three areas need S.P. Leatherman, D. Nummendal, M.P. O'Brien, the most attention: oceans, clouds, and hydrology. O.H. Pilkey, W. Sturges, R.L. Wiegel. 1987. The oceans play an important role in delaying Responding to Changes in Sea Level: Engineering climate change and have a large influence on Implications. National Research Council. regional climates. However, the ocean models Washington, DC: National Academy Press. currently used in GCMs are relatively simple. Ocean models that better simulate the absorption Dickinson, R.E. 1986. How will climate change: and transport of heat and gases would give The climate system and modelling of future climate. improved estimates of transient and regional climate In: Bolin, B., B.R. Doos, J. Jager, and R.A. change. Clouds are a major feedback to global Warrick, eds. Scope 29: The Greenhouse Effect, warming and influence regional climate. More Climatic Change and Ecosystems. New York: John realistic modeling of clouds by GCMs would Wiley and Sons. pp. 221-231. improve the estimates of the magnitude of global warming and regional change. Finally, more Gates, W.L. 1985. Modeling as a means of studying sophisticated hydrology in GCMs will yield better the climate system. In: MacCracken, M.C., and estimates of soil moisture and runoff, which will also F.M. Luther, eds. Projecting the Climatic Effects of improve estimates of regional climate changes. Increasing Carbon Dioxide. Washington, DC: U.S. Department of Energy. DOE/ER-0237. Scenarios Grotch, S.L. 1988. Regional Intercomparisons of General Circulation Model Predictions and The scenarios in this report were based on Historical Climate Data. Prepared for U.S. changes in average conditions, either at equilibrium Department of Energy. TR041. (doubled CO₂) or due to a gradual change in average underlying conditions (transient). As Hansen, J., I. Fung, A. Lacis, D. Rind, G. Russell, pointed out in Chapter 3: Variability, many systems S. Lebedeff, R. Ruedy, and P. Stone. 1988. Global are quite sensitive to changes in the frequency and climate changes as forecast by the GISS 3-D model. intensity of extreme events. In the future, scenarios Journal of Geophysical Research 93:9341-9364. should incorporate change in variability to help identify sensitivities to variability. Transient Hansen, J., G. Russell, D. Rind, P. Stone, A. Lacis, scenarios can also be improved. Such scenarios S. Lebedeff, R. Ruedy, and L. Travis. 1983. should be useful for testing sensitivities to changes Efficient three-dimensional global models for in long-term climate trends as well as year-to-year climate studies: Models I and II. Monthly Weather variations. At the same time, it is important to Review 3(4):609-622. keep scenarios simple. More detailed scenarios, involving a lot of data (such as daily data from Hoffman, J.S., D. Keyes, and J.G. Titus. 1983. GCMs) may be difficult to use. The more detailed Projecting future sea level rise. Washington, DC: the scenario, the more likely it will be applied U.S. Government Printing Office. incorrectly, which limits the ability to compare results by different researchers. In addition, Hoffman, J.S., J. Wells, and J.G. Titus. 1986. scenarios should be simple, so the assumptions used Future global warming and sea level rise. In: in creating them can be easily understood. Sigbjarnarson G., ed. Iceland Coastal and River Designers of scenarios will have to wrestle with the Symposium. Reykjavik, Iceland: National Energy competing desires of being more detailed and Authority. maintaining simplicity. 68 Methodology Manabe, S., and R.T. Wetherald. 1987. Large scale Schlesinger, M., and Z. Zhao. 1988. Seasonal changes in soil wetness induced by an increase in Climate Changes Induced by Doubled CO2 or carbon dioxide. Journal of Atmospheric Sciences Simulated by the OSU Atmospheric GCM/Mixed- 44:1211-1235. Layer Ocean Model. Corvallis, OR: Oregon State University, Climate Research Institute. Meier, M.F. et al. 1985. Glaciers, ice sheets, and sea level. Washington, DC: National Academy Thomas, R.H. 1986. In: Titus, J.G., ed. Effects of Press. Changes in Stratospheric Ozone and Global Climate. Washington, DC: U.S. Environmental Mitchell, J.F.B. 1988. Local effects of greenhouse Protection Agency and UNEP. gases. Nature 332:399-400. Vinnikov, K.Y., and N.A. Lemeshko. 1987. Soil Parry, M., T. Carter, N. Konijin, and J. Lockwood. moisture content and runoff in the USSR territory 1987. The Impact of Climatic Variations on with global warming. Journal of Meteorology and Agriculture: Introduction to the IIASA/UNEP Case Hydrology No. 12. Studies in Semi-Arid Regions. Laxenburg, Austria: International Institute for Applied Systems Analysis. Wigley, T.M.L. 1987. Climate Scenarios. Prepared for the European Workshop in Interrelated Revelle, R. 1983. Probable future changes in sea Bioclimate and Land-Use Changes. Boulder, CO: level resulting from increased atmospheric carbon National Center for Atmospheric Research. NCAR dioxide. In: Changing Climate. Washington, DC: 3142-86/3. National Academy Press. Schlesinger, M.E., and J.F.B. Mitchell. 1985. Model projections of the equilibrium climatic response to increased carbon dioxide. In: MacCracken, M.D., and F.M. Luther, eds. Projecting the Climatic Effects of Increasing Carbon Dioxide. Washington, DC: U.S. Department of Energy. DOE/ER-0237. 69 CHAPTER 5 FORESTS FINDINGS Reforestation along northern portions of potential forest ranges could mitigate some of these impacts. Global warming could significantly affect the forests of the United States. Changes could be apparent in 30 to 80 years, depending upon the region, the If elevated CO₂ concentrations substantially increase the water-use efficiency of tree species, quality of a site, and the rate of climate change. the southern declines could be alleviated. There may be northward shifts in species ranges, dieback along the southern reaches of species If climate stabilizes, forests might eventually ranges, and changes in forest productivity. Other regain a generally healthy status (over a period stresses in combination with climate change may of several centuries). In the meantime, exacerbate these impacts. Different migration rates declining forests could be subject to increased and climate sensitivities may result in changes in fires, pest attacks, and replacement with low- forest composition. Without large-scale value trees, grasslands, and shrubs. A reforestation, large reductions in the land area of continually changing climate could result in healthy forests are possible during this century of even greater dislocations among forests. adjustment to climate changes. Although climate fluctuations, timber harvests, disease outbreaks, wildfires, and other factors have affected forests Productivity Changes during the last century, the magnitude of these changes is substantially less than those projected in Dieback along the southern limits of response to climate changes considered in this distribution of many species could result in report. productivity declines of 40 to 100%, depending on how dry soils become. Range Shifts Productivity could increase along the northern The southern ranges of many forest species in limits of some eastern tree species, particularly the eastern United States could die back as a as slow-growing conifers are replaced by more result of higher temperatures and drier soils. rapidly growing hardwoods. The southern boundary could move several hundred to 1,000 kilometers (up to 600 miles) Combined Impacts With Other Stresses in a generally northward direction for the scenarios studied. Large regions of severely stressed forests, combined with possible increases in fires, pests, The potential northern range of forest species disease outbreaks, wind damage, and air in the eastern United States could shift pollution, could produce major regional northward as much as 600 to 700 kilometers disturbances. These factors were not (370 to 430 miles) over the next century. considered for this report. Actual northward migration could be limited to as little as 100 kilometers (60 miles) owing to Additional impacts of changes in forests could the slow rates of migration of forest species. include reductions in biotic diversity, increased Without reforestation, full migration of eastern soil runoff and soil erosion, reduced aquifer forests to potential northern distributions could recharge, changes in recreation, and changes in take centuries. If climate change occurs too wildlife habitat. rapidly, some tree species may not be able to form healthy seeds, thus halting migration. 71 Chapter 5 Policy Implications rate of global warming. This study did not evaluate the effectiveness of reforestation efforts. Institutions such as the U.S. Forest Service, state forest agencies, and private companies EXTENT AND VALUE OF U.S. should begin to consider how to factor climate FORESTS change in their long-term planning. Global climate change may need to be a factor in the Forest Service's 50-year planning horizon. Forests occupy 33% of the U.S. land area and exist on some lands in all 50 states. In total, they Where U.S. forests are clearly reduced by occupy 298 million hectares (738 million acres) and climate change, forest agencies will have to are rich in such resources as water and wildlife. consider intensive strategies to maintain productivity. For example, they could Many biotic and abiotic factors influence the undertake reforestation on a more massive scale condition of forests, but climate is the dominant than now practiced and possibly introduce factor (Spurr and Barnes, 1980). This chapter subtropical species into the Southeast. summarizes the current knowledge and predictions concerning the effects of rapid climate change on A coordinated public and private reforestation U.S. forests. effort, together with development of new and adapted silvicultural practices, would also be Distribution and Ownership required. Forests are major carbon sinks, so a large reforestation program would also reduce Eight major forest regions of the conterminous atmospheric CO₂ concentrations, slowing the 48 states contain 84% of the forested ecosystems of the United States (Figure 5-1). The forested areas WESTERN REGIONS Pacific Northwest Douglas fir/hemlock/fir California Pine/fir/redwood Northern Rockies Pine/fir/birch Southern Rockies Pinyon/juniper/pine EASTERN REGIONS Northeast Spruce/fir Maple/beech/birch Central Maple/beech/birch Oak/hickory Southeast STUDY PLOT SITES Southern pine Botkin et al. 0. Davis ALASKA Lake States Urban and Shugart Spruce/fir Spruce/hardwood Maple/beech/birch Spruce/hemlock Figure 5-1. Major forest regions of the United States and their primary tree groups. 72 Forests Table 5-1. Area of U.S. Forest Lands in 1977 by Federal, State, Private, and Other Ownerships (millions of hectares)ᵃ Commercial Forestsᵇ Private Non- % Region/States Primary Tree Species Federal State Industry Indus Otherᶜ Total Total EAST Northeast spruce-fir 0.3 0.4 3.9 7.8 0.7 13.1 4.4 CT, MA, ME, NH, RI, VT maple-beech-birch Lake States spruce-fir 2.3 2.8 1.7 9.9 4.2 20.9 7.0 MI, MN, WI, ND, SD(E) maple-beech-birch Central maple-beech-birch 1.8 2.0 8.6 22.9 2.6 37.9 12.7 DE, IA, IL, IN, KA, KY, MD, oak-hickory MO, NB, NJ, OH, PA, TN, WV Southeast loblolly, shortleaf 5.8 1.0 14.7 54.3 8.0 83.8 28.1 AL, AR, FL, GA, LA, MS, NC, slash pine OK, SC, TN, TX, VA WEST Northern Rockies - pine-fir-birch 9.1 0.6 0.8 2.7 9.3 22.5 7.6 ID, MT, SD(W), WY Southern Rockies - pinyon-juniper-pine 6.4 0.3 0.0 2.4 24.1 33.2 11.1 AZ, CO, NM, NV, UT Pacific Northwest D. fir-hemlock-fir 7.8 1.2 4.0 3.2 5.3 21.5 7.2 OR, WA California CA pine-fir-redwood 3.4 .03 1.1 2.0 9.8 16.3 5.4 SEPARATE Alaska - AK spruce-hemlock-hardwood 3.3 1.0 0.0 0.1 43.9 48.3 16.2 STATES Hawaii HI ohia .01 0.2 0.0 0.2 0.4 0.8 0.3 TOTAL 40.2 9.5 34.8 105.5 108.3 298.3 100 % TOTAL 13.5 3.1 11.7 35.4 36.3 100 a Hectare X 2.47 = acres. b Commercial forests are those capable of growing at least 1.4 cubic meters per hectare per year (20 cubic feet per acre per year) of industrial wood materials. C Other forests include county and municipal forests and those federal lands withdrawn from industrial and wood production for use as parks, preserves, and wilderness. Source: USDA (1982). of Alaska and Hawaii represent the remaining 16% Superimposed over the natural distribution of (Table 5-1). Each forest region includes one or trees, forests, and ecosystems in the United States more forest types distinguished by the major tree is the human infrastructure. Ownerships include species present. As a general rule, some types in federal, state, and private lands (Table 5-1). Within each region have predominantly coniferous tree the forests classified as "commercial" (64% of 298 species (i.e., evergreen, needle-leaved, and million hectares), the federal government ownership softwoods); other forest types are composed mostly of 40 million hectares (99 million acres) is primarily of deciduous trees (i.e., tree species that are broad- in the national forest system managed by the U.S. leaved, have no winter foliage, and are hardwoods). Department of Agriculture's Forest Service (36 Forest types with a mix of coniferous and deciduous million hectares or 91 million acres); most of the trees, however, are not uncommon. remainder is managed by the Department of 73 Chapter 5 Interior's Park Service, Fish and Wildlife Service, Value of U.S. Forests or Bureaus of Land Management and Indian Affairs. State ownerships total 9 million hectares Most populated regions in the United States are (23 million acres). Private lands are divided located close to or within a forested region. For between those of industrial forest companies (35 instance, the Boston-Washington corridor is within million hectares or 86 million acres) and those of the eastern hardwoods. The populations of Atlanta small, private landowners, who collectively have 106 and the Southeast are interspersed among the million hectares (262 million acres) (USDA, 1982). southern pine forests. Chicago and nearby Great Lakes communities are surrounded by the mixed Another significant segment of American forests conifer-hardwood forests of that region, and the Los consists of those maintained within urban and Angeles to San Francisco populations parallel the suburban areas. Examples are community parks, Sierra Nevadas to the east. In addition, greenbelts, roadside forests, and wooded residential urban/suburban forests exist in or near most of the and industrial zones (USDA, 1981). These forest nation's cities. Forests, therefore, are part of the areas are important sources of outdoor recreation, environmental fabric and general habitability for the wildlife habitat, and real estate values. In total, the majority of U.S. citizens. urban/suburban forests of the United States occupy approximately 28 million hectares (69 million acres) All forests shed water to some degree, and two- (Grey and Deneke, 1978). thirds of the water runoff in the contiguous 48 states comes from forested ecosystems. Precipitation To the degree that all forest lands are owned by passes through forested ecosystems as canopy some individual or organization, all forest lands are throughfall or flows along tree stems, and then flows under some form of management. A continuum of along the ground surface or into the soil; eventually, management policies exists, ranging from lands some of the water flows into streams. Water yields intended to have minimal human intervention except from U.S. forests provide about 750 billion liters for protection from catastrophic wildfire (e.g., some (200 billion gallons) of water each day for major parks and most wilderness areas) to lands where uses such as irrigation, electricity production, silvicultural practices are intensively applied (e.g., manufacturing, and domestic consumption. These the most productive federal, state, and industrial levels of demand are projected to continue to the forest lands dedicated to growing tree crops); year 2030 (USDA, 1981). (Table 5-2). These forests under government and industrial management constitute roughly one-fourth A favorite use of forests is outdoor recreation. of the total and might be the easiest to manage Activities include hiking, camping, hunting, under climatic impacts simply because they are sightseeing, boating, swimming, fishing, skiing, larger blocks of lands already under strong sledding, and snowmobiling. A 1977 survey of U.S. management commitments. Table 5-2. Percentage of Forest Lands by Level of Management within Four U.S. Regions (estimates for 1977) Forest Other Reserved/ U.S. regions plantations commercialᵇ deferredᶜ East North 9 80 11 South 21 69 10 West Rocky Mountains 2 38 60 Pacific Coast 16 44 40 a Intensively managed plantations. b Moderately managed forests. C Recreational and protected forests. Source: USDA (1982). 74 Forests households indicated that a majority of people RELATIONSHIP BETWEEN participated in outdoor recreation four or more times each year (USDA, 1981). FORESTS AND CLIMATE About 190 million hectares (470 million acres), Scientific understanding of forest ecosystems or 64% of the total U.S. forested ecosystems, are has greatly advanced with each decade of this highly productive commercial forest lands. These century. Yet the literature contains little lands represent about 10% of the world's forest information concerning the direct or indirect effects area, but they supplied nearly a quarter of the of climate change on the complex biological and world's industrial forest products in the late 1970s physical processes in forest ecosystems. Some (USDA, 1982). In 1980, 1.7 million people were insights are gained from paleobotanical studies of employed in timber-based occupations across the past rates and magnitudes of ecological change United States. Such employment is basic to the during glacial-interglacial cycles, as well as changes economic well-being of many small towns and in the species composition of forested ecosystems. communities (Schallau, 1988). The total value of Similarly, observations of forest responses to timber products harvested in 1972 was about $6.4 unusual drought or other weather extremes provide billion, and the total value after such processes as some knowledge. Estimates of rate, magnitude, and manufacturing, marketing, transport, and quality of change have also been derived using construction amounted to $48 billion, or 4% of the computer models developed by plant ecologists or nation's gross national product. In 1979, timber forest management scientists for other objectives. product exports and imports were valued at $7 Their validation for understanding how a forest can billion and $9 billion, respectively. Looking ahead, adapt to climate change is only in the initial stages. the consumption of wood products in the United States is projected to increase between current Climate is a primary determinant of existing levels and the year 2030 (USDA, 1982). forests. The ranges of annual average temperature and rainfall variation determine global forest distributions relative to different biotic regions (Figure 5-2). Substantial increases in temperature Desert Grassland 30 Tropical Forest 20 MEAN ANNUAL TEMPERATURE (°C) Deciduous Forest 10 0 Coniferous Forest Artic and -10 Alpine Tundra 100 200 300 400 MEAN ANNUAL PRECIPITATION (cm) Figure 5-2. Approximate distributions of the major groups of world biomass based upon mean annual temperatures and precipitation (Hammond, 1972). 75 Chapter 5 or decreases in rainfall could, for example, produce At the expected rapid rate of climate change, a shift from a forest to a grassland type. Thus, the potential rates of forest migration would accelerated climate change resulting from human become a major concern. Migration rates vary by activities and related effects on U.S. forests is of species. Paleorecords of the Holocene (10,000 years high concern to citizens and policymakers alike. ago to present) show that extension of ranges for tree species of eastern North American (in response Magnitude to glacial retreat) varied from 10 to 20 kilometers (6 to 12 miles) per century for chestnut, beech, maple, Vegetation has been in an almost constant state and balsam fir (Zabinski and Davis, Volume D). of distributional change and adjustment due to an Other species within the oak and pine groups almost constantly changing climate over the past extended at faster rates, i.e., 30 to 40 kilometers (19 10,000 years and even over the past several hundred to 25 miles) per century. It should be noted that years (Spurr and Barnes, 1980). Lines of evidence there is some uncertainty as to whether these come from studies of fossils, tree rings, carbon-14 migration rates were in response to glacial retreat dating, plus peat and pollen analyses (Webb, 1987). plus climate warming or primarily warming alone. Historical climate changes appear to have been Mechanisms associated with such phenomena as fluctuations in solar radiation, earth orbit variations, and volcanic Knowledge of causal links between weather activity. Evidence of repeated continental glacial patterns and forest response is fundamental to advances and contractions in the Northern projecting growth and composition effects resulting Hemisphere dramatically illustrates the large-scale from climate change. Another requirement is to effects of global climate change. understand the climatic influences on processes influencing populations of forest plants and animals. In response to the glaciation, species shifted These include such phenomena as fires, windstorms, south. Evidence from fossil pollen, for example, landslides, pest outbreaks, and other disturbances indicates a southward shift of spruce into Georgia that affect survival and subsequent colonization by and east Texas during the last glacial advance and different species. Furthermore, the processes that treeless tundra in the Great Lakes region (Spurr control the dispersal of seeds through a mosaic of and Barnes, 1980). During the maximum different ecosystem types (such as forest patches interglacial warmth of 6,000 to 9,000 years ago, interspersed with agricultural lands, wetlands, which was 1.5°C (2.7°F) warmer than the present grasslands, and other land-use groups) must be temperature level, plant zones were one to several clearly defined. hundred kilometers (60 to 250 miles) north of present distributions. Among the important factors now known to influence the growth and distribution of forests are Rates the following. Temperature All forested ecosystems experience change on both spatial and temporal scales; each biological and The optimum temperature for growth depends physical forest component may respond to climatic upon the tree species and other conditions. variation on different spatial and temporal scales. Warmer temperatures usually increase the growth of For example, microorganisms, insects, and birds plants. However, high temperatures can decrease come and go with relatively short-term climatic the growth of plants or cause mortality where variation; shrub species' abundances vary within the temperatures greatly exceed optimum ranges for timespan of decades; trees, once established, could growth. Cold temperatures can limit plant persist for centuries. This understanding is distributions by simply limiting growth at critical important from the perspective of climate change, stages or by directly killing plants. since it implies that forested ecosystems do not respond as a unit, but in terms of parts. Different Precipitation parts respond differently; consequently, future forested ecosystems under a rapidly changing Too much or too little precipitation can limit climate could be quite different from those existing forest production and survival. Too much rainfall in today. 76 Forests some areas can cause flooding or raise the water disturbances, and by the time required for forests to table, thus drowning roots by reducing soil air that shift into new ranges. The length of day exerts contains oxygen required for respiration or by considerable control on physiological processes such promoting fungal attack. Too little rainfall can as release from and onset of dormancy. Significant reduce growth, cause susceptibility to fire or northward shifts of forests would alter their day- pestilence, and possibly kill plants. The seasonal length regime, producing uncertain results. timing of rainfall is more important than total annual rainfall, although forests also require some Nutrient Status minimum total annual rainfall (see Figure 5-2). In addition to climate, most forest growth is CO₂ Concentration strongly influenced by availability of soil nutrients. Disturbances over vast regions, such as drought High CO2 concentrations could increase tree followed by fire, can release large quantities of growth through increases in photosynthesis rates essential nutrients into the atmosphere or into and water-use efficiency (primarily hardwood surface waters. This leaves soils nutrient deficient. species) when water and other nutrients are not Lengthy periods of soil development are usually limited (Strain and Cure, 1985). Plant responses to required to replenish the soil nutrients before a CO2 have been investigated largely in growth large, mature forest stand can be supported. In chambers and are difficult to extrapolate to the real turn, soils reflect properties of the forests that they world. Responses are varied and do indicate some support. This results from decades of nutrient measure of adaptive capability most likely imparted uptake, litterfall, decomposition, and other from ancestral exposure to much higher and lower processes. levels in the geologic past. However, in natural situations, water nutrients or temperature usually Atmospheric Chemistry are limiting factors in forest growth, thus making the impacts of CO2 enrichment uncertain. If water- Much of the nutrient budget of forests involves use efficiency increases, then tolerance to drought deposition of chemicals from the atmosphere as might increase, ameliorating declines in southern gases, aerosols, or particles, or in solution with parts of ranges. Unfortunately, the current state of water as precipitation. Although most of these act knowledge does not allow generalizations on this as nutrients, some produce acid deposition that can subject. leach important soil nutrients (e.g., SO₄⁻) produce a fertilizing effect (e.g., NO₃), or damage leaf tissue Another important relationship between forests (e.g., O3). Climate change will alter transport paths and CO2 is the role forests play as carbon sinks. of air pollutants, and increased temperature could Globally, forest vegetation and supporting soil increase the rates at which gases convert to contain about 60% of the organic carbon stored on deleterious forms. world land surfaces. This organic carbon is largely cycled between forest ecosystems and the Disturbances atmosphere by photosynthesis (uptake of CO₂) and respiration (CO₂ release) in the plants (Waring and Almost continually, forests experience natural Schlesinger, 1985). Anthropogenically caused disturbances or stresses from biotic or abiotic agents reductions of forests either directly (e.g., alone or in combination. Examples are insect and urbanization, mismanagement) or indirectly (as a disease outbreaks, plant competition, wildfire, response to CO2-induced global warming) would drought, cold extremes, and windstorms. tend to increase the "greenhouse effect." These disturbances, which are among the Light primary factors controlling the successional processes in forests (Pickett and White, 1985), may The amount of sunlight bathing an ecosystem range from an opening of small gaps in the canopy sets the upper limit on net primary productivity. as the result of single tree death or of windthrow Thus, the tropics exhibit higher productivity than do occurring when trees are blown down by heavy the boreal regions. This potential productivity winds (predominant successional mechanisms in would, of course, be limited by other climatic effects eastern hardwoods) to large clearings from fire, such as drought, cold, heat, and natural 77 Chapter 5 windthrow, or pestilence (predominant successional The methods used in the previous studies are mechanisms in western forests). quite similar to those used in this report. They include computer modeling of forest processes, Landscape Processes literature surveys, studies of fossil evidence, and empirical relationships constructed by experts. The The horizontal movements of materials such as estimates of future change produced from these soil and biological organisms, together with human studies are generally based on the output of one or disturbances across the landscape, are critical to more of the general circulation models (GCMs) processes controlling tree migration, species used for this report. Thus, the results of the diversity in forests, and the spread of fire, previous studies are consistent with those reported windthrow, and pestilence effects. These processes here. are very poorly understood; quantification in the emerging field of landscape ecology is just beginning. STUDIES IN THIS REPORT Multiple Stresses Six studies on forest effects contributed to the regional case studies reported in this volume. The In general, trees or forests stressed by one purpose was to use existing data bases analyzed in factor, e.g., accelerated climate change, are more new ways to estimate effects on U.S. forests from susceptible to natural stresses (secondary climate change scenarios. The selection of the six disturbances) such as insects, disease, or invading studies was based upon three criteria: use of weed species. The concept of multiple stresses established statistical methods; hypotheses testing leading to forest declines is becoming more widely concerning causal mechanisms; and selection of a recognized (Manion, 1981). Regional climate mix of studies that complemented each other, such changes, even if temporary, frequently predispose that the strengths in one approach might overcome forests to damage by other natural or anthropogenic the weaknesses of another. stresses. This report focuses primarily on forests within the contiguous 48 states. It is worth noting, PREVIOUS STUDIES ON THE however, that the largest magnitude of warming is NATIONAL EFFECTS OF expected in the northern latitudes encompassing the boreal forest and other forest types in Alaska and CLIMATE CHANGE ON FORESTS Canada. Thus, these large forests could also be under significant risk from climate warming. Concern regarding effects of climate change on U.S. forests has prompted several excellent reviews. One of the most comprehensive (Shands and RESULTS OF FOREST STUDIES Hoffman, 1987) was the result of a conference sponsored by EPA, the National Forest Products Association, and the Society of American Foresters. Design of the Studies While pointing out the high uncertainty associated with current predictions of climate change, several Characteristics of the six studies are briefly authors suggested that if predictions are true, listed in Table 5-3. With the exceptions of the distributions of key forest species in the United Overpeck and Bartlein study and the Woodman States will change significantly. study, the methods are discussed in the regional case study chapters and will be mentioned only Other recently produced compilations broadly briefly here. All of the forest studies are in Volume consider forest effects along with other impacts D. (e.g., those on agriculture, prairie land, and the Great Lakes) (White, 1985; Titus, 1986; Meo, 1987; Two studies used correlations between tree Tirpak, 1987). These reviews are largely pioneering distributions and climate (Overpeck and Bartlein; efforts and some overlap occurs, but each presents Zabinski and Davis). Overpeck and Bartlein's some key points. approach consisted of correlating the modern pollen distributions of important tree species with January and July mean temperature and annual rainfall. 78 Forests Table 5-3. Principal Investigators, Regional Focus, and Method of Approach for the Regional Forested Ecosystem Studies Principal investigator Region Method Overpeck and Bartlein Eastern North America Correlation and fossil studies Urban and Shugart Southeast Uplands Forest dynamics model Botkin et al. Great Lakes Forest dynamics model Zabinski and M. Davis Great Lakes Correlation O. Davis California Fossil studies Woodman et al. Southeast, California, Literature review and National The correlation was then tested by reconstructing ranges in tolerance to stresses of temperature, past pollen distributions from general circulation moisture, and shade. Both studies explored forest model simulations of past climates (during the most response starting with bare ground on a range of recent glacial-interglacial cycle) for each species and soil types from well drained to poorly drained. comparing them to observed pollen distributions Forest growth simulations from bare ground from those periods. Future pollen distributions represent conditions after a fire, logging, or similar were then constructed from the expected doubled disturbance. Mature stand simulations are useful CO2 climate projected from the different model for investigating the potential response of present climate scenarios. The correlations were forests to gradual climate change in the immediate constructed on modern pollen distributions, rather future. than tree distributions, to allow the direct comparison to fossil pollen data. Modern pollen For the California case study, Davis distributions are similar to, but not exactly the same reconstructed vegetation patterns in the Sierra as, modern tree distributions. The verification Nevadas from fossil pollen studies for the studies indicated that the approach works interglacial warm periods that occurred between reasonably well at a coarse spatial resolution. That about 6,000 and 9,000 years ago. These is, northern trees are in the north and southern reconstructions represent possible analogs of a trees are in the south, with the regional patterns future warm period at the lower magnitude of the being reasonably well represented. predicted future warming. The approach of Zabinski and Davis was Woodman conducted a literature review for the essentially the same as that of Overpeck and Southeast and California forested regions and Bartlein, except that the correlations were peripherally for the entire nation. The purpose was constructed from the actual modern tree to ascertain the attributes of the forest resource in distributions rather than from the modern pollen terms of extent, ownership, economic and distributions (see Chapter 15: Great Lakes). recreational value, and policy considerations. Two of the studies used computer models of Limitations forest dynamics (Botkin et al.; Urban and Shugart). Growth characteristics of each tree species Although predicted effects vary, these six occurring in the study region are used by the models analytical studies have results that are collectively to determine the growth and development consistent enough to advance our knowledge and ofindividual trees on a site. These growth justify concern regarding the future of U.S. forests characteristics include such attributes as maximum under rapid climate change. The range of age, maximum height, maximum diameter, and 79 Chapter 5 predicted effects is large; however, uncertainties indices of environmental stress, such as July exist regarding (1) the climate scenarios; (2) the temperature or annual rainfall, are usually related to kind and rates of responses of individual tree factors that more directly affect plant growth, such species; and (3) changes in forested ecosystems as as accumulated warmth or summer drought. a whole resulting from environmental change. All However, large uncertainties exist in some instances. of these factors significantly influence the precision This is particularly true with regard to the climatic and accuracy of the results. controls of the southern limits of southeastern forests, simply because of their association with the A major uncertainty in the simulation model continental margin. Does the climate at that approach involves the rates of species dispersal into latitude represent the actual climatic limitation to a region. The current generation of models has no the distributions, or are the species simply stopped dispersal mechanisms. A species is simply present by a geographic barrier? No one really knows. or it is not present. For example, Botkin et al. These uncertainties were partially addressed by excluded most southern tree species so that their Overpeck and Bartlein, who compared their fossil dispersal was unrealistically nonexistent, and these pollen approach to the modeling approach. The southern species could never enter the Great Lakes two approaches use similar relations to climate, and region. But if they had been included in the both can be used reasonably well to simulate forest simulations, these species would have entered the distributions in the geologic past. northern forests at the same rate as the climate change. This would have assumed dispersal rates Several uncertainties with the pollen-climate far in excess of reality. This limitation can, in part, correlation approach limit its precision and be overcome by studies, such as those of Zabinski accuracy. First, many of the plant taxa used in the and Davis, that provide some insight into actual study are plant genera (e.g., pine, oak) rather than dispersal rates and species migration. The species, and thus the simulated results are not simulations did not consider the impact of taxonomically precise. Second, the results are transplanting southern species in these areas. applicable only on a regional scale; local scale predictions are not made. Third, and very The timing of forest declines as estimated by significant, the simulated results assume that all the the models should be interpreted with caution. plants are in equilibrium with the new climate. Declines are triggered by periods of high Rates of dispersal vary between species, and several environmental stress. Forest models are usually not hundred years may pass before plant communities operated far beyond current conditions, such as for are again in equilibrium with climate. How this lag extremely dry soils. Therefore, the extreme climate would affect plant community dynamics is not simulated by these models may not estimate the addressed in this study and is an important research timing and behavior of forest declines as accurately question. as desired. It should also be remembered that there is much uncertainty concerning the rate and timing The paleoecological analysis of the past of the climate change itself. vegetation in the Sierra Nevadas (O. Davis) presents several uncertainties. First, differences with respect A further cautionary point is that although the to weather variations (i.e., season to season and year models considered temperature limitations, nutrient to year) could produce strikingly different types of deficiencies, and soil moisture stress, other vegetation. Also, there is much uncertainty about important factors might affect the timing and what the most appropriate analog period might be magnitude of tree responses. Examples of factors -- or if one even exists. Furthermore, the rate of in need of consideration include disturbance effects climate change in the future is predicted to be much (e.g., impacts from wildfires, pests, and pathogens), faster than the rate of climate change during the age-dependent differences in tree sensitivities to past 20,000 years. Lags in the response of species stress (e.g., older trees are often more susceptible), to the future climate could strongly affect the type and potential CO2-induced increases in water-use of forest at any one location, whereas in the past, efficiency. with a more slowly varying climate, lags in species response were not as important in determining The models also carry assumptions about the forest composition. environmental controls of species limits. In most cases these assumptions are reasonable, given that All of the studies are deficient in some very important processes controlling forest responses to 80 Forests climate, particularly disturbance regimes such as migration coupled with a fairly rapid decline in the fires, windstorms, hurricanes, landslides, and pest southern and western parts of species ranges. Drier outbreaks. Over some forest areas, periods of cloud forest conditions in the United States, induced as cover could change. This is an important much by increased temperature as by changes in uncertainty, for if the annual total is significantly rainfall, would mean less tree growth and therefore increased, reductions in solar radiation could mean reduced forest productivity in general. reduced photosynthesis and thus less forest growth. The forest simulation models provide an In addition, the responses of mature trees to indication of the importance of uncertainties elevated CO2 under conditions of moisture, imparted by the climate scenarios. The climate temperature, or other nutrient limitations remain scenarios differ primarily in their representation of largely unexplored. Most research on elevated CO2 regional rainfall patterns. The model results on trees has been performed in controlled chambers indicate that temperature has a large effect on using seedlings, and results show an increase in forest health, either directly through cold and heat photosynthesis and improved water-use efficiency in stress or indirectly through exaggeratedg drought some cases (Strain and Cure, 1985). However, the effects. Thus, the overall characteristics of forest seedlings were not previously grown in or responses are remarkably similar among the three acclimatized to high CO₂ environments. Evidence climate scenarios. However, this generalization is has shown that plants grown under high CO2 will uncertain because models usually do not incorporate respond differently to changes in temperature, light, all possible mechanisms of impact. and moisture conditions (Strain and Cure, 1985). Magnitude Another shortcoming is that methods to extrapolate CO₂ fertilization results from laboratory Eastern Forests Northern Limits experiments to the natural world are limited, and an understanding of regional changes in water-use All of the study results suggest a northward efficiency is even more limited. Furthermore, expansion of most eastern tree species (Figure 5-3 complex interactions between fertilization effects displays results from Overpeck and Bartlein). That and changes in water-use efficiency can produce is, spruce, northern pine, and northern hardwood unexpected problems such as increased heat loads species would move north by about 600-700 due to effects on evaporation cooling. These kilometers (375-440 miles) into the Hudson Bay interactions are not well understood but could region of the Canadian boreal forest (Overpeck and produce major regional changes in forest responses. Bartlein; Zabinski and Davis). New England Therefore, it is not yet possible to quantitatively coniferous forests would be replaced by more incorporate the direct effects of CO₂ on forests into hardwood forests and especially by the oak species studies such as these. Further, if water or nutrients from the eastern mid-United States (Botkin et al.; are limiting to forest growth, they would likely Overpeck and Bartlein; Zabinski and Davis). As exceed the fertilization effects of elevated CO2. the northern mixed forests shift from spruce-fir to Also, forest canopies at optimum development have sugar maple, some sites could actually triple their multilayered leaf areas so that light limitations exist present productivity (Botkin et al.). for the lower portion of the foliage in addition to frequent water and nutrient limitations. This adds Additionally, southern pine species could shift further weight to the belief that CO₂ enrichment about 500 kilometers (310 miles) into the present may not significantly affect forest productivity. hardwood forest lands of eastern Pennsylvania and New Jersey (Overpeck and Bartlein; Urban and Results Shugart; Solomon and West, 1986; Miller et al., 1987). Depending upon the severity of climate The six studies conducted for EPA consistently change, Urban and Shugart estimated that near the indicate that climate changes would significantly northern limits of slash pine in East Tennessee, affect the natural forests of the United States. The aboveground woody biomass in 100 years could distribution of healthy forests in the eastern United range from little change to an extremely low States appears to become greatly reduced from their biomass with almost no trees (i.e., a grassland, present areas during the next century (Figures 5-3 savanna, or scrub). However, even with little and 5-4). This results from a very slow northward decrease in productivity, species shifts would alter 81 Chapter 5 Current Climate A Spruce Birch N. Pines Oak S. Pines Prairie Forbs GISS Model Output B GFDL Model Output C OSU Model Output D Spruce Birch N. Pines Oak S. Pines Prairie Forbs Figure 5-3. Maps of eastern North America depicting present distributions of major forest genera and herbacious vegetation compared with potential future distributions after reaching equilibrium with the climate predicted for doubled CO₂. The comparison is based upon (A) simulations using modern pollen data and simulated future pollen abundances for each of the three doubled CO₂ scenarios: (B) GISS; (C) GFDL; and (D) OSU. The three levels of shading in each scenario map indicate estimated future pollen abundances ranging from 20% (darkest or strongest chance of future distributions) to 5% and 1% (lightest or least chance of future distributions) (Overpeck and Bartlein, Volume D). 82 Forests Sugar Maple Present Range Range After 2050: GISS Range After 2050: GFDL Scale 0 400Km Potential Range Inhabited Range Figure 5-4. Present and future geographical range for sugar maple (Zabinski and Davis, Volume D). the forest composition from shortleaf to loblolly conditions. This estimation results from scenario pine, a more commercially valuable tree species. conditions of heat that would exceed the tolerance limits for most tree species. Under the mildest Eastern Forests - Southern Limits scenario (OSU), even forest areas in South Carolina and southward would be marginal, supporting about Ultimately, forest decline and mortality could half their current biomass. truncate southern distributions of tree species by as much as 1,000 kilometers (625 miles) in many Biomass accumulations in 100 years for mature northern hardwood species (Zabinski and Davis; natural forests in productive sites in the Great Overpeck and Bartlein) or by as little as a few Lakes region could be reduced to 23-54% of their hundred kilometers (about 120 miles) in southern present values (Botkin et al.; Solomon and West, pines and hardwoods (Urban and Shugart; Solomon 1986). On poor sites, forests could be converted to and West, 1986). Under the driest scenario grassland or savanna with very low productivity, (GFDL), Zabinski and Davis estimate local ranging from 0.4 to 28% of their present values. extinction in the Great Lakes region of many eastern tree species such as eastern hemlock and Western Forests sugar maple (Figures 5-3 and 5-4). These estimates bear considerable uncertainty for all species. Similar projections were made for six western coniferous species: ponderosa and lodgepole pine, These uncertainties are particularly true for the Douglas-fir, western hemlock, western larch, and southern limits of southeastern species that border Englemann spruce (Leverenz and Lev, 1987). the continental margin. The actual southern Estimations are mixed for the West. Because of climatic limitations of these species are not well the mountainous conditions in the West, upslope known (Urban and Shugart). Nevertheless, under shifts are possible for Douglas-fir, ponderosa pine, the most severe climate scenario in the Southeast and western hemlock in the northern Rocky with increased temperatures and decreased growing- Mountains. In the coastal mountains of California season precipitation, Urban and Shugart's results and Oregon, Douglas-fir could shrink in the suggest that the 18 tree species they considered lowlands and be replaced by western pine species would no longer grow in the southern half of the (O. Davis; Leverenz and Lev, 1987). Overall, the region. Present forest lands in the region would be western forest lands are estimated to favor more replaced by scrub, savanna, or sparse forest drought-tolerant tree species, such as the hard pine 83 Chapter 5 A. MISSISSIPPI FORESTS B. SOUTH CAROLINA FORESTS 180 180 160 160 140 140 WOODY BIOMASS (T/ha) 120 100 WOODY BIOMASS (T/ha) 120 100 80 80 60 60 40 No Climate Change 40 No Climate Change GISS A GISS A 20 20 0 0 1980 2000 2020 2040 2060 1980 2000 2020 2040 2060 YEAR YEAR Figure 5-5. Estimated changes in biomass of mature forests in Mississippi (A) and South Carolina (B) under the GISS transient climate change scenario (Urban and Shugart, Volume B). group, at the expense of fir, hemlock, larch, and These rapid declines, coupled with the expected spruce species. magnitude of climate change, raise the question of how fast forests can migrate. Based upon fossil If regional drought persisted, the frequency of records, Zabinski and Davis have estimated that the fires could increase, significantly reducing total maximum dispersal rate of several tree species in forested area. Also, with massive upslope response to the last glacial retreat was roughly 50 movement, some species could be pushed off the kilometers (30 miles) per century. Under the tops of mountains into local extinction. expected rapid warming, they estimated that a dispersal rate of about 1,000 kilometers (600 miles) No quantitative estimates have been derived for per century would be required to maintain species productivity for the western forests under potential distributions near their current extent. Such warming conditions. However, using the analog migration rates are doubtful, suggesting greater approach of Davis, under the most severe conditions reductions in species ranges under rapid climate projected for California, the species composition of change, with declines in the drier western portions. the west-side Sierra Nevada forests would become more similar to that of the east-side forests. This Mechanisms of Migration could reduce the standing biomass to about 60% of current levels. Distribution changes (i.e., migrations) suggested by these studies must be considered carefully. Rates of Decline and Migration Reproductive processes are essential for the migration of tree species across the landscape. For In the Great Lakes region, significant forest many tree species, climate change could reduce decline and forest compositional change could natural regeneration in an existing location and become evident within 30 to 60 years (Figure 5-5A; introduce the species at different latitudes or Botkin et al.). In the Southeast region, forest altitudes. Reproductive processes in trees, such as declines could become most evident in 60 to 80 flowering, pollination, seed set, seed germination, years with declines in the drier western portions and seedling competitive success, are particularly occurring even earlier, perhaps in about 30 years sensitive to climate. (Figure 5-5B and C); Urban and Shugart). As previously discussed in this chapter (see Limitations) Specific regional climate scenarios vary as a there is considerable uncertainty about these function of the GCM. All scenarios estimate numbers. increases in temperature; however, some include 84 Forests increases in rainfall, and others have decreases. The region, for example, beech could decrease in northward shifts of species appear to result from a abundance (Zabinski and Davis), and birch and release from cold temperature stress, which maple could increase (Botkin et al.). On some normally freezes flowers, seedlings, and even adult lands, forest productivity could remain about the trees. However, the western and southern limits of same as today, but changes to less economically eastern tree species appear to result from important species could be significant. insufficient moisture and excessive heat stress, which primarily affect sensitive life history stages but can Not considered quantitatively in any of the also affect mortality rates of adult trees. Though studies are changes in forest disturbance regimes. difficult to detect in the early phases of rapid These changes should not be considered lightly. climate change, tree mortality is sensitive to chronic Extreme and more frequent climatic variations (see moisture stress and mortality rates would likely Chapter 3: Variability) could cause much higher increase among the major forest regions of the mortality in U.S. forests than the current United States. experience. Although little is known as yet, some locations may experience an increase in the Two points are important about regional frequency of extreme weather events, for example, uncertainties of future rainfall distribution. First, wind, ice, or snow storms, droughts, and flooding. changes in the seasonal distribution of rainfall are as Besides the direct damage these events can cause, or more important than relatively small changes in they can predispose forests to damage from the annual total. If summer rainfall decreases while secondary stresses such as insects, disease, and winter rainfall increases, the trees may still wildfires. experience summer drought stress. Second, evapotranspiration is a log function of temperature. Therefore, as temperature goes up, water loss from ECOLOGICAL AND SOCIO- trees and soils can increase tremendously. If minor increases in rainfall are not sufficient to override the ECONOMIC IMPLICATIONS evapotranspirational losses of water, drought impacts will pervade. Both of these mechanisms The effects of doubled CO₂ climate changes appear to dominate the forest impacts in this study. may be considered from two perspectives: ecological and socioeconomic. Evidence for All of the study approaches used under all of significant national implications is strong from both the climatic scenarios estimate major forest declines viewpoints. in the southern parts of species ranges and expansions to the north. These declines, resulting Ecological Implications primarily from drought stress, would occur despite the differing rainfall predictions among the climate Ecological implications for forests commonly scenarios used in this study. Global precipitation is start with tree response. But strong implications generally projected to increase slightly with global also exist for other ecosystem components, e.g., warming (see Chapter 4: Methodology), but it is not animals, soils, water, secondary impacts, and as known whether this increase would be sufficient to noted, the atmosphere through which climate compensate for potential increases in plant moisture change is mediated. Forest effects are described in stress caused by higher temperatures. Precipitation terms of tree distribution changes and biomass in some regions may decline. Droughts would production changes, but many other processes become more common. The western limits of interact among the other major components. Thus, eastern forests could similarly retract as the climate significant changes in tree response would be warms. accompanied by ecological reverberations throughout all the forested areas of major U.S. Existing forests probably would not shift intact, regions. but would change in composition. Variations in migration rates and sensitivities to weather variables Tree Distributions and Biomass Productivity produce individual responses to climate change. These changes are consistent with the well-known As discussed, migrations of forest tree species dynamic nature of ecosystems and were projected to the North in response to rapid warming in North for the forests of all regions. In the Great Lakes America during the next century will be likely. 85 Chapter 5 However, significant lag is possible. Even under the effects. All of these factors illustrate that climate maximum rates of species dispersal estimated by change could influence the regional patterns of Zabinski and Davis, healthy forest areas may not biotic diversity in both plants and animals (see redevelop for several centuries. Furthermore, if Chapter 8: Biodiversity). climate continues to change beyond the next century, then healthy forests may never redevelop. Soils Meanwhile, distribution ranges may not be under such constraints, so the extent of healthy forested Soils under warmer climates also would change, regions in the United States probably would be although at a much slower rate than shifts in species greatly reduced. Though some locations may have distribution. Increased soil temperatures, however, increased productive potential from a biomass per would affect the entire range of physical, chemical, hectare standpoint, the large reductions in areas and biological soil processes and interactions. For with healthy forests would likely create a net example, populations of bacteria, fungi, and animals reduction in forest productivity for the United States could increase in a way that would accelerate for several centuries or longer. decomposition of litter and thereby reduce the availability of nutrients essential for forest growth Even if a massive reforestation effort were (Spurr and Barnes, 1980). undertaken, the new forests resulting from species shifts might or might not be as productive as Considerable time may be required to develop existing forests. More northern latitudes or higher optimum soil conditions for high forest productivity elevations raise other considerations. Farther north, supporting species at more northern latitudes or days are longer in the summer and shorter in the higher elevations. Furthermore, it is not at all clear winter. At higher elevations, damaging ultraviolet how well some northern soils could support more light intensity is greater. All of these conditions southern species. The soils of the boreal forest could lower forest productivity below present levels. differ from those under the deciduous forests to the Furthermore, it is not clear that reforestation would south. be successful. A major intent of reforestation would be to artificially speed up northward migration of Water tree species. However, seedlings that would appear to be favored on some northern sites several Where forests give way to drier conditions (e.g., decades in the future may not survive there now in the Great Lakes region and California), many because of constraints imposed by temperature, day lands now serving as watersheds might be used for length, or soil conditions. Similarly, seedlings that different purposes. Furthermore, regional-scale could not survive on those sites now might not be disturbances (such as fire) and applications of the best adapted species for those same sites several chemicals (such as fertilizers and pesticides) could decades in the future. degrade regional water quality and increase airborne toxic chemicals (see Chapter 9: Water Resources). Animals Sea level rise may impact some coastal forests. A change in the size and relative homogeneity Many forest lands of high value for timber of forests could influence whether some animals can production (e.g., in the Southeast) or recreation (in continue to live in their present locations. Often, the Northeast, Northwest, and California) are close animals are finely adapted to habitats specific to a to ocean coasts. Inundations, decreases in depth to certain location. For some animals, migration can the water tables, and saltwater intrusions could be hindered by boundaries between forests and trigger rapid forest declines near these areas. other land types or facilitated as animals move along edges. Furthermore, some animals (e.g., Secondary Impacts many game species) prefer young forests, and others (e.g., many rare and endangered species) prefer old As the southern bounds of forests tend to shift forests. In turn, animals can exert a profound north, forest decline (sick and dying trees) could influence on forest structure and composition become extreme over large areas that would through selective browsing of seedlings, insect attack become highly susceptible to weed competition, pest of different tree species, seed dispersal, and other 86 Forests outbreaks, or wildfire. As forests decline, species of elicit a strong concern. In the Atlanta-Southeast lower economic value, as well as weedy shrubs and region, the southern pine forests, while undergoing herbs, could invade via wind dispersion. Under a gradual expansion of their northern boundaries, stressful environments, such species are severe would have less vigor in the remaining stands. This competitors with most commercial tree species. could raise their vulnerability to damage from insects and disease, reducing esthetic values -- Trees experiencing less favorable growth atleast an intermediate impact for most of the local conditions are more stressed and will be vulnerable citizens. In contrast, within some portions of the to insect and disease attack. These secondary pest Southeast, the Great Lake region, and California, impacts could last "until the most vulnerable forest drier climates may cause the loss of some forest stands or tree species are eliminated" (Hedden, lands to prairie or desert conditions -- a severe 1987). In addition, it is estimated that the incidence change for the people there, not only in their living of catastrophic wildfires will increase in U.S. forests environment but also in the whole spectrum of with higher temperatures. Simand and Main (1987) forest land use. estimated that fire occurrence and fire-suppression costs would increase 8 and 20%, respectively. Recreation Socioeconomic Implications Forests must be in a relatively healthy condition to support quality recreational use (Clawson, 1975). The United States enjoys substantial economic Forests undergoing gradual composition changes and cultural benefits from its forests. Until recently, might remain healthy, but rapid changes would most the nation's forest managers assumed that these likely cause stressed or declining forests. Such benefits could be sustained by maintaining forests in forest conditions would have less recreational appeal a healthy condition (Fosberg, 1988). This was because of such factors as less pleasing appearance, achieved, for example, by preventing fires or pest greater threat of wildfire, and reduced hunting invasions, avoiding careless use, and enhancing quality when game populations change or are productivity through good silviculture. diminished. Furthermore, drier conditions in U.S. forests would harm recreational opportunities that Beginning with the possibility of regional air depend on abundant water or snow. pollution damage to forests, suspected in the 1980s, alterations of the environment external to forests Wood Products presented a new concern. Research and policy discussions to deal with this issue are ongoing. Altered U.S. forest productivity resulting from climate change would have obvious major economic If climate changes as rapidly as predicted, this impacts. Significant yield reductions could lead to additional external influence with its more global unemployment, community instability, industrial dimensions looms as a possible hazard to forests dislocation, and increased net imports of wood and their use. As can be imagined, a list of products. potential socioeconomic concerns would be large. To provide a brief perspective, three issues are Reforestation projects could make up for some considered. losses in forest productivity and artificially advance migrations forced by climate change. Reforestation Quality of the Human Environment technology has greatly improved in recent decades so that success rates also have increased greatly. The forest amenities enjoyed by most U.S. Examples are high-vigor seedlings developed citizens will be affected according to different forest through improved nursery practices, genetic responses. In the Boston-Washington corridor, a selection, and vegetative propagation. composition change from predominantly hardwood Improvements in the field include machine planting, to predominantly pine forests, though ecologically fertilization, and weed control on selected sites. significant, may not be noticed by most people if it Results are evident from the large acreages of occurs gradually. However, a delay of years or plantations established in the United States in decades between the decline of existing forests and recent decades, particularly with loblolly pine in the replacement by migrating tree species would likely Southeast and Douglas-fir in the Pacific Northwest 87 Chapter 5 (Table 5-2). Large-scale reforestation in the United What constraints (e.g., mandatory forest States and elsewhere could significantly add to the practices) should be placed on forest total carbon sink provided by world forests, thereby managers to ensure national environmental offsetting some of the buildup of atmospheric CO2. goals? Although this was not studied, attempts to reforest some very dry sites may be unsuccessful. Who should pay the additional costs incurred in implementing new policies? Innovative manufacturing trends should prove to be timely during times of rapid forest change. The large array of forest ownerships in the High-strength and durable products from United States, public and private, makes reconstituted wood (e.g., new particle board development and implementation of forest policy concepts, warp-proof hardwood lumber, paper more complicated than in most countries. Around products of fiber from multispecies) are now in use the world, about 77% of all forests are in some or well along in development. These new methods form of public ownership (Hummel, 1984). The will lessen the present overdependency on a few diversity of owners and managers results in widely commercial conifer species from stands above divergent goals and objectives. minimum size and quantity (Ince, 1987). The result will be an ability to use the timber resources of the How Much Land Should Be Forested? future, however they change in composition. Changes in forest composition or regional boundaries induced by rapid climate change would FOREST POLICY AND CLIMATE magnify the complexity of national forest policy even CHANGE further. Lands in forests now would require review relative to such competing needs as agriculture and Historically, U.S. forest policies have undergone residential use, which would also be adjusting to climate change. continued development to meet national change (Young, 1982). The earliest policies were adapted How Much Should Be Withdrawn From by the New England colonies in the 1600s to regulate overcutting near settlements. Wood was Timber Production? needed for fuel and buildings, but existing methods were not capable of long distance log transportation. Where the productivity of wood is significantly Development of U.S. forest policies has continued reduced, increased, or shifted, a policy question that and has been particularly intense this century, as the would surely arise concerns whether forest lands national forests, national parks, and wilderness areas should be reallocated to maintain timber have been established. production. If so, how should competing forest uses, such as watersheds, parks, and wilderness, be At present, forest managers are dealing with treated? How much of each can the United States many additional policy issues. Five of these afford under changed climatic conditions? Should (Clawson, 1975) are important to climate the federal government purchase more forest lands change/forest response: to support all public needs? How much U.S. land should be devoted to In the short term, forest managers could forests? compensate for some loss of productivity by improved technology, although at increased costs. How much forest land should be withdrawn An example would be establishment of more from timber production and harvest? drought-tolerant plantations through genetic selections, improved nursery stock, and more How should the federal forest lands be intensive silvicultural practices (e.g., weed control managed? (That is, the lands under the and thinning). Introducing new species adapted to USDA Forest Service, USDI Park Service, warmer climates might be possible in some Bureaus of Land Management and Indian locations, but this would call for development of Affairs, and other federal agencies that new silvicultural regimes and utilization methods manage forest lands.) possible, but time consuming and costly. In the long 88 Forests term, if growing conditions become extremely control, fertilization, and reforestation be employed difficult on some U.S. forest lands because of in an attempt to preserve them? This question and climate changes, establishing trees for wood others will challenge the fundamental concepts of production on such sites may not be economically the benefits of multiple use and sustained yield of justified. U.S. forests. How Should We Manage Federal Forests? How Can We Ensure National Goals? The national forests under the USDA Forest At the minimum, federal agencies must plan Service are managed according to a series of and act in concert with the state and private forest complex legal directives and administrative organizations. In the first half of this century, the procedures, beginning with the Organic Act of 1897 federal government attempted to regulate forest (Woodman and Furiness, Volume D). Ultimately, harvests on all federal, state, and private lands. the objective became to manage the national forests Development of this policy did not survive strong for multiple uses, with timber and other forest public concern and intense political debate against resources on a sustained-yield basis and certain such policy (Worrell, 1970); the same sentiment lands set aside as wilderness areas. The National would likely exist today. However, under the Forest Management Act of the mid-1970s requires influence of climate change, the nation may once management plans for each national forest subject again have to face the touchy issue of what to public review. The plans look ahead 50 years restraints or forest practices must be regulated for and are to be updated every 10 years. all public and private lands. Lands managed by the Department of Interior Solomon and West (1985) point out that while are under similar mandates. For example, a climate change might disrupt forest ecosystems in congressional act passed in 1976 charged the Bureau the future, it is uncertain whether forest managers of Land Management to manage its 2.3 million could or would be able to apply silvicultural hectares (5.1 million acres) of forest and range land practices on a scale large enough to maintain the according to multiple-use and sustained-yield net productivity of commercial forest lands in the principles. Similarly, the National Park Service is United States. Some states (e.g., Washington, mandated to manage national parks, monuments, Oregon, and California) have laws specifying fire historic sites, and so forth, for the recreational protection requirements, control burn practices, and enjoyment of people. Such activities as timber reforestation minimums following timber harvests. harvesting, hunting, mining, and grazing are not Zoning, permits, licenses, and various taxation permitted. In addition to the federal government, measures also have been attempted with mixed most states, many counties, and some municipalities results. It is much easier to prevent owners from own forest lands. destroying forests than to compel them to implement silvicultural practices. The Forest and Rangeland Renewable Resources Planning Act of 1974 requires the Reforestation Secretary of Agriculture to make periodic reviews of the nation's forest and rangeland resources. In the To keep pace with the global climate changes future, these assessments and planning efforts estimated, the U.S. reforestation effort conceivably should include consideration of the possible effects would need to be doubled or tripled in size. In of predicted climate changes. recent years, about 800,000 hectares (2 million acres) per year (approximately 700+ million A key issue is the level of priority given to seedlings) have been reforested in the United States maintaining forest health under changed climate (USDA, 1982). Costs range from $200 to $700 per conditions. For instance, under more adverse hectare ($80 to $280 per acre) depending upon environments, should national forests be left to species, site preparation, plantation density, and decline as a natural process, thereby losing esthetic planting method. Using $500 per hectare ($200 per values in parks, water yields from watersheds, and acre) as a mode, the total annual expenditure is highly productive timber crops? Or should near $400 million. About 0.4% of the commercial silvicultural forest techniques such as thinning, weed land base is reforested annually. At this rate, it 89 Chapter 5 would take 100 years to reforest 40% of the U.S. the middle of the 21st century? While subsets of forest lands, assuming no repeat hectares to cover this question must include extent, magnitude, and failures or harvests of the first plantations. risk considerations, additional knowledge is needed concerning the following: An expansion on the scale suggested above would require large investments in seed 1. Forest migration processes and rates, procurement, tissue culture capability, nursery including the landscape processes that capacity, and research to improve knowledge about control the horizontal movements of forests, the establishment and silviculture of drought- animals, and disturbances; resistant plantations. Even if the dollar commitments were made, reforestation at this scale 2. Interactions among the different landscape might be possible only if all forest lands were components and land-use practices that managed by one organization. The complex forest affect biodiversity, and water quantity and ownership pattern in the United States, therefore, quality; would be an issue to overcome in a national reforestation program. 3. The impact of climate change alone and in combination with other natural or Who Should Pay? anthropogenic influences, such as insects, pathogens, CO₂ enhancement, air Adjusting forest policies to address the issues pollutants, UV-B radiation, and acid arising from climate change will most likely raise deposition on U.S. forests; and the costs of using the nation's forests whether for water, recreation, esthetics, or timber. Additional 4. The processes and mechanisms that play research to answer many new questions will also key roles in forest ecosystem effects both require more funds. A major question will be who biologically as in photosynthesis and should pay for these costs. Land owners? Forest respiration, and physically as in flows of users? Consumers? All taxpayers? The answers energy, carbon, water, and nutrients will come when better information is available on through ecosystems. resulting forest effects, followed by public debate establishing new priorities for forest use in a Methods changed climate. How can forest ecosystems be measured to reliably detect the effects of rapid climate change? RESEARCH NEEDS Today, the response of ecosystems to environmental change is largely based upon extrapolating from The forest effects resulting from rapid climate field observations, from knowledge about seedlings or individual trees of a small number of change are at present hypothetical. The change has not yet occurred, and many uncertainties are commercially valuable species, and from computer associated with the predictions. Effective policies to models. The following must be accomplished: deal with new forest effects will require more 1. A determination of the most useful information and fewer uncertainties that must come through forest ecosystem research. Four broad integrating variables for forest ecosystems questions concerning U.S. forests frame the that indicate the effects of climate change research needs for the 1990s: What will the effects -- particularly variables that are early- be? How can they be measured reliably? How warning indicators of ecosystem response; should they be managed? How can we ensure that research will be conducted in a timely fashion? 2. Effective sampling designs developed for experiments and long-term monitoring at the forest ecosystem scale; and Effects of Climate Change 3. Improved models capable of projecting What will be the effect on the nation's forest regional effects on forests across multiple ecosystems if climate changes occur as predicted by spatial and temporal scales. 90 Forests Forest Management Policy Issues. New Orleans, LA: U.S. Environmental Protection Agency. What options are available to the public and private forest managers and owners in the United Davis, M.D., and D.B. Botkin. 1985. Sensitivity of States to address the changes in the nation's forests the cool-temperate forests and their pollen to rapid that might occur in the next century? Research is climatic change. Quaternary Research 23:327-340. needed to accomplish the following: Fosberg, M.A. 1988. Forest productivity and health 1. Understand the socioeconomic impacts of in a changing atmospheric environment. In: all forest ecosystem effects to clarify Berger, A., et al., eds. Climate and Geosciences: A economic risks and alternatives; and Challenge for Science and Society in the 21st Century. NATO ASI Series. Series C; 2. Develop technology to mitigate the adverse Mathematical and Physical Sciences, Vol. 285. effects or to exploit the benefits of forest Dordrecht, The Netherlands: Kluwer Academic change, such as breeding, bioengineering, Publishers. pp. 681-688. transplanting, fertilization, irrigation, and other management approaches. Grey, G.W., and F.J. Deneke. 1978. Urban Forestry. New York: John Wiley and Sons. Timing of Research Hammond, A.L. 1972. Ecosystem analysis: biome approach to environmental science. Science 175:46- The timing of the research is critical. The 48. effects of climate change may be some decades away, but this should not lessen the urgency to Hedden, R. 1987. Impact of climate change on begin research toward better information and forest insect pests in the southern U.S. In: Meo, methods. The complexities of the science are very M., ed. Proceedings of the Symposium on Climate large. Developing a base of knowledge to identify Change in the Southern U.S.: Future Impacts and potential forest changes before they are upon the Present Policy Issues. New Orleans, LA: U.S. nation will require significant time and resources. Environmental Protection Agency. Hummel, F.C., ed. 1984. Forest Policy, a REFERENCES Contribution to Resource Development. The Hague: Martinus Nijhoff/Dr. W. Junk, Publishers. Barbour, M.G., J.H. Burk, and W.D. Pitts. 1987. Terrestrial Plant Ecology, 2nd Ed. Menlo Park, Ince, P.J. 1987. Technology, timber demand and CA: Benjamin/Cummings Publishers. timberland investment. In: A Clear Look at Timberland Investment, Milwaukee, WI; April 27- Botkin, D.B. 1979. A grandfather clock down the 29. Conference proceedings. Forest Products staircase: stability and disturbance in natural Research Society. ecosystems. In: Waring, R.H., ed. Proceedings of the 40th Annual Biological Colloquium. Forests: Lavdas, L.G. 1987. The impact of climate change Fresh Perspectives From Ecosystem Analysis. on forest productivity. In: Meo, M., ed. Corvallis, OR: Oregon State University Press, pp. 1- Proceedings of the Symposium on Climate Change 10. in the Southern U.S.: Future Impacts and Present Policy Issues. New Orleans, LA: U.S. Clawson, M. 1975. Forests for Whom and for Environmental Protection Agency. What? Resources for the Future. Baltimore, MD: Johns Hopkins University Press. Leverenz, J.W., and D.J. Lev. 1987. Effects of carbon dioxide-induced climate changes on the Cubbage, F.W., D.G. Hodges, and J.L. Regens. natural ranges of six major commercial tree species 1987. Economic implications of climate change in the western United States. In: Shands, W.E., and impacts on forestry in the South. In: Meo, M., ed. J.S. Hoffman, eds. The Greenhouse Effect, Climate Proceedings of the Symposium on Climate Change Change, and U.S. Forests. Washington, DC: in the Southern U.S.: Future Impacts and Present Conservation Foundation, pp. 123-156. 91 Chapter 5 Manion, P.D. 1981. Tree Disease Concepts. New Spurr, S.H., and B.V. Barnes. 1980. Forest Jersey: Prentice Hall. Ecology, 3rd Ed. New York: John Wiley and Sons. Meo, M., ed. 1987. Proceedings of the Symposium Strain, B.R., and J.D. Cure, eds. 1985. Direct on Climate Change in the Southern United States: Effect of Increasing Carbon Dioxide on Vegetation. Future Impacts and Present Policy Issues; May 28- Washington, DC: U.S. Department of Energy. 29. New Orleans, LA: University of Oklahoma and DOE/ER-0238. U.S. Environmental Protection Agency. Tirpak, D.A., ed. 1987. Potential Effects of Future Miller, F.W., P.M. Dougherty, and G.L. Switzer. Climate Changes on Forest and Vegetation, 1987. Effect of rising carbon dioxide and potential Agriculture, Water Resources and Human Health, climate change on loblolly pine distribution, growth, Vol. V. Assessing the Risks of Trace Gases That survival and productivity. In: Shands, W.E., and Can Modify the Stratosphere. Washington, DC: J.S. Hoffman, eds. The Greenhouse Effect, Climate U.S. Environmental Protection Agency. EPA 400/1 Change, and U.S. Forests. Washington, DC: - 87/001E. Conservation Foundation, pp. 157-188. Titus, J.G., ed. 1986. Climate Change, Vol. 3. Pickett, S.T.A., and P.S. White. 1985. The Ecology Effects of Changes in Stratospheric Ozone and of Natural Disturbance and Patch Dynamics. Global Climate. Washington, DC: U.N. Academic Press, Inc. Harcourt Brace Jovanovich. Environmental Program and U.S. Environmental Protection Agency. Schallau, C.H. 1988. The forest products industry and community stability: the evolution of the issue. USDA. 1981. U.S. Department of Agriculture, Montana Business Quarterly Summer: 1-8. Forest Service. An Assessment of the Forest and Range Land Situation in the U.S. Forest Resource Shands, W.E., and J.S. Hoffman, eds. 1987. The Report No. 22. Greenhouse Effect, Climate Change, and U.S. Forests. Washington DC: Conservation USDA. 1982. U.S. Department of Agriculture, Foundation. Forest Service. An Analysis of the Timber Situation in the U.S. 1952-2030. Forest Resource Report No. Simand, A.J. and W.A. Main. 1987. Global climate 23. change: the potential for changes in wildland fire activity in the Southeast. In: Meo, M., ed. Waring, R.H., and W.H. Schlesinger. 1985. Forest Proceedings of the Symposium on Climate Change Ecosystems, Concepts and Management. Orlando, in the Southern U.S.: Future Impacts and Present FL: Academic Press, Inc. Policy Issues. New Orleans, LA: U.S. Environmental Protection Agency. Webb, T. 1987. The appearance and disappearance of major vegetational assemblages: long-term Solomon, A.M., and D.C. West. 1985. Potential vegetational dynamics in eastern North America. responses of forests to CO₂ induced climate change. Vegetation 69:177-187. In: White, M.R., ed. Characterization of Information Requirements for Studies of CO₂ White, M.R., ed. 1985. Characterization of Effects: Water Resources, Agriculture, Fisheries, Information Requirements for Studies of CO₂ Forests and Human Health. Washington, DC: U.S. Effects: Water Resources, Agriculture, Fisheries, Department of Energy. DOE/ER-0236. pp. 145- Forests and Human Health. Washington, DC: U.S. 1709. Department of Energy. DOE/ER-0236. Solomon, A.M., and D.C. West. 1986. Atmospheric Worrell, A.C. 1970. Principles of Forest Policy. carbon dioxide change: agent of future forest New York: McGraw-Hill. growth or decline? In: Titus J.G., ed. Effects of Changes in Stratospheric Ozone and Global Young, R.A., ed. 1982. Introduction to Forest Climate. Vol. 3: Climate Change. Washington, DC: Science. New York: John Wiley and Sons. U.S. Environmental Protection Agency. 92 CHAPTER 6 AGRICULTURE FINDINGS above temperature thresholds for particular crops in some locations. The exact magnitude of change will Climate change would affect crop yields and result be sensitive to changes in climatic variability, in northward shifts in cultivated land, causing particularly the frequency of droughts. significant regional dislocations in agriculture with associated impacts on regional economies. It would Economic Impacts expand crop irrigation requirements, stress livestock production, and increase infestations of agricultural Under three out of four scenarios, a small to pests and diseases. Preliminary results suggest that moderate aggregate reduction in the nation's although U.S. crop production could decline, agricultural output was estimated. The supplies would be adequate to meet domestic needs. estimated production levels appeared to be The potential for reduction of the national adequate to meet domestic consumption agricultural capacity and the many uncertainties needs. If droughts occur more frequently surrounding the interactive effects on the under changing climate, effects on agriculture agricultural system create the necessity to respond may be more severe. to the climate change issue. Assuming no change in export demand, Crop Yields reduced outputs would decrease exports, which could negatively affect global food supplies and The effects of climate change alone may the U.S. trade balance. This report did not reduce average yields of corn, soybeans, and analyze global changes in agriculture, which wheat, both rainfed and irrigated, except in the could have a major effect on demand for U.S. northernmost latitudes where warmer products. conditions provide a longer frost-free growing season. Decreases in modeled yields result Under the most severe climate change primarily from higher temperatures, which scenarios, continued technological shorten a crop's life cycle. improvements, similar to those in recent years, would have to be sustained to offset losses. When the direct effects of CO2 on crop Increasing food demand from higher U.S. and photosynthesis and transpiration are world population would aggravate the approximated along with the effects of climate economic losses due to climate change. change, average rainfed and irrigated corn, soybean, and wheat yields could overcome the The economic response of agriculture to negative effects of climate change in some changes in regional productivity may be to locations. If climate changes are severe, yields shift crop production and associated could still decline. The extent to which the infrastructure in a northward direction. This beneficial direct effects of CO2 will be seen is because yields in northern areas generally under field conditions with changed climate is increase relative to yields in southern areas. uncertain. Although availability of agricultural soils was included in the economic analysis, neither the Even if the patterns of climate variability are sustainability of crop production in northern unchanged, yield stability may decrease, areas nor the introduction of new crops into particularly under rainfed conditions. This southern areas was studied. may occur because there would be more days 93 Chapter 6 Irrigation Demand into southern regions of the United States. Cold stress conditions may be reduced in the winter, but The demand for irrigated acreage is likely to heat stress is likely to increase in the summer. increase in all regions. This is due to the Reproductive capabilities may also decrease. reliability of irrigated yields relative to dryland yields and to higher commodity prices that Policy Implications make expansion of irrigated production more economically feasible. Actual increases in Global climate change has important irrigated acreage would depend on the implications for all parts of the agricultural adequacy of water supply and on whether the system. The agricultural research structure, cost of water to farmers increases. which is dedicated to maintaining U.S. farm productivity, should expand climate change Demand for more irrigation would increase research in activities ranging from the field stress on and competition for regional water level to the national policy level. supplies. If irrigation does increase, it could increase surface and groundwater pollution Current U.S. Department of Agriculture and other forms of environmental degradation. (USDA) research on heat- and drought- tolerant crops and practices and maintenance Agricultural Pests of crop germ plasm should be sustained and enhanced to limit vulnerability to future Climate warming could change the ranges and climate change. populations of agricultural pests. Temperature increases may enhance the survival of insect The USDA should evaluate current legislation pests in the winter, extend their northward in regard to its ability to allow adaptation to ranges, increase pest species with more than global warming. Flexibility in shifting crop one generation per year, and allow pest types and farm practices will speed adjustment. establishment earlier in the growing season. Such adaptation strategies should consider the These effects could result in a substantial rise impacts on soil erosion and water quality. in pesticide use, with accompanying environmental hazards. Changes in pests will The USDA, the Department of Commerce, also depend on regional shifts in crop the U.S. Trade Representative, and the State production. Department should consider the implications of potential long-term changes in the level of Farm-Level Adjustments U.S. crop exports for the U.S. balance of trade and strategic interests. Farmers may adjust to climate change by using A national drought policy is strongly needed to full-season and heat-resistant crop species or coordinate federal response to the possibility varieties, by altering planting dates, by planting of increased droughts due to climate change. two crops during one growing season, by Even without climate change, such a policy is increasing or altering their scheduling of necessary not only for the agricultural sector irrigation, by using more pesticides, and by but also for other sectors. harvesting earlier. If climate change is not severe, these adjustments may mitigate losses in crop yields; more severe climate change is likely to make major adaptation necessary. SENSITIVITY OF AGRICULTURE TO CHANGES IN CLIMATE Livestock Effects Agriculture is a critical American industry, Higher temperatures may increase disease and providing food for the nation's population and as heat stress on livestock in some regions. much as $42.6 billion in exports for the nation's Existing livestock diseases may shift north, trade balance (Figure 6-1). Agriculture employs 21 while tropical diseases may extend their ranges million people more than any other industry, 94 Agriculture 50 Others Fruits, nuts, and vegetables Cotton 40 Livestock and by-products Oilseeds and by-products Grains and preparations 30 Billions of $ 20 10 0 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 YEAR Figure 6-1. Value of U.S. agricultural exports by commodity, 1972-86 (not adjusted for inflation). Livestock excludes poultry and dairy products (The World Food Institute, 1987; U.S. Department of Agriculture, Economic Research Service, Foreign Agricultural Trade of the United States, Washington, DC, January-February 1987, and various other issues). when taking into account workers on farms and in For example, failure of the monsoon season meat, poultry, dairy, baking, and food-processing caused shortfalls in crop production in India, activities (Council for Agricultural Science and Bangladesh, and Pakistan in 1987. The 1980s have Technology, 1988). The U.S. agricultural also seen the continued deterioration of food production system includes farm equipment production in Africa, despite adequate world food manufacture, fertilizer and seed supplies, rural supplied elsewhere, because of persistent drought, banking, and shipping. Total farm assets were $771 internal wars, poor distribution, weak infrastructure, billion in 1985; food and fiber were 17.5% of the and a deteriorating environment. Climate extremes total gross national product in the same year. have had large effects on U.S. agriculture. During Wheat, corn, soybeans, cotton, fruits and vegetables, the Dust Bowl years of the 1930s, U.S. wheat and and livestock are among the most important U.S. corn yields dropped by up to 50%. Midsummer agricultural commodities. 1983 saw an unpredicted drought in the U.S. Corn Belt and in the southeastern United States, causing Worldwide, agricultural products must provide U.S. corn yields to fall by about a third, from over sustenance for the world's growing population, now 7,000 kilograms per hectare to about 5,000 estimated at about 5 billion and projected to rise to kilograms per hectare (from about 110 to 80 bushels 8.2 billion by 2025 (Zachariah and Vu, 1988). per acre). Global production and consumption of grain have grown steadily since 1960, although regional food The 1988 drought recently demonstrated the shortages continue to occur owing to climate impact that climate variability can have on variability and socioeconomic factors. Technological agricultural productivity. This drought decreased advances, such as improved hybrids and irrigation U.S. corn yields by almost 40%, and the cost of the systems, have reduced the dependence of crop yields 1988 Drought Relief Bill is estimated to be $3.9 on local environmental conditions, but weather is billion (Schneider, 1988). The 1988 drought still an important factor in agricultural productivity. emphasizes anew the close link between agriculture and climate. 95 Chapter 6 Light from the sun, frost-free growing seasons, Thus, climate plays a major role in and the hydrologic cycle largely govern the determining crop and livestock productivity. suitability of geographic areas for crop production Agricultural productivity determines profitability and and affect crop productivity. Livestock production decisionmaking at the farm level, which in turn is responsive to climate through differing levels of define farming systems at the regional level and heat and cold stress and altered ranges of disease- import-export supply and demand at the national carrying vectors such as mosquitoes and ticks. and international levels. These complex interrelationships necessitate a broad consideration Higher levels of CO₂ in the air would also of the impacts of potential climate change on U.S. affect crops. Increased CO2 has enhanced crop agriculture. photosynthesis and has improved crops' use of water in experimental settings. Because experimental research has rarely simultaneously investigated both PREVIOUS STUDIES OF the climatic and the direct effects of CO2 on plants, it is difficult to assess the relative contributions of CLIMATE CHANGE AND CO2 and increased temperature to plant responses. AGRICULTURE This remains one of the most crucial questions in the analysis of impacts of climate change and increased CO2 on agriculture. Relationships between climate and agriculture have been studied intensively for many years. However, relatively few studies have specifically The presence and abundance of pests affecting addressed both the climatic and the direct effects both crops and livestock are highly dependent on that the growth in trace gases will have on climate. The severity of the winter season, wind agriculture. Even fewer studies have addressed patterns, and moisture conditions determine in large these potential effects in an integrated approach part where pests will be prevalent. The that links both biophysical and economic spheres of geographical distribution of pests also depends on analysis. locations of crop types. Most research attention in the United States, Much of U.S. agricultural production takes supported primarily by the U.S. Department of place under technologically advanced cropping systems that are primarily monocultural. Likewise, Energy, has focused on the direct effects of CO2 on crops. These studies are reviewed by Acock and livestock production is highly specialized, both Allen (1985) and Cure (1985), who found an technically and geographically, and a high degree of average increase in yields of about 30% and integration exists between grain and livestock production. Any significant level of economic increases in water-use efficiency for crops growing in air with doubled CO2 (660 ppm) and favorable, robustness associated with general, multiple- current climate conditions. Kimball (1985) and enterprise farms has long since passed from the Decker et al. (1985) suggested that the potential scene. The ability of our agricultural system to adapt to climate change may be more limited now effects of CO2 and/or climate change on agricultural production systems may include shifts in in some ways than it was in the past. production areas and changes in levels of livestock stresses, water availability, and pest control Agriculture strongly affects the natural management. environment. It often increases soil erosion, intensifies demand for water, degrades water Integrated approaches to the impacts of quality, reduces forested land, and destroys wildlife climate change on agriculture involving both habitats. Many agricultural practices contribute to biophysical and economic processes have been soil degradation, groundwater overdraft, loss of considered in studies by Callaway et al. (1982), the plant and aquatic communities, and generally Carbon Dioxide Assessment Committee (1983), reduced resilience in environmental and genetic Warrick et al. (1986), and the Land Evaluation resources. Therefore, climate-driven shifts in Group (1987). A benchmark international study on agricultural regions have implications for both the agronomic and economic effects of climate environmental quality. change on agriculture was conducted by the International Institute for Applied Systems Analysis (Parry et al., 1988). No study has as yet 96 Agriculture comprehensively examined the combined effects of three crops were selected for the modeling studies climate change and the direct effects of CO2 on on the effects of climate change on yields. U.S. agriculture. The results from the regional studies of crop production (not including California), hydrological CLIMATE CHANGE STUDIES IN predictions from the climate models, and an agricultural economics model were linked in an THIS REPORT integrated approach to enable investigators to translate the estimated yield changes from the crop Structure of and Rationale for the Studies modeling studies and predicted changes in water availability into economic consequences (see Figure The regions studied for this report are 6-2). Such a coordinated analytical framework is important agricultural production areas (see Table necessary to account for the effects of market forces 6-1). The Great Lakes and Southeastern States are on the total agricultural sector, including livestock, major corn and soybean producers, and the Great and to evaluate the adequacy of the nation's Plains States grow mainly wheat and corn. resource base for agricultural production under California annually produces about 10% of U.S. climate change. Economic forces may lead farmers cash farm receipts from cotton, grapes, tomatoes, to grow more crops in areas with relatively high lettuce, and many other crops. productivity and fewer crops in areas with relatively low productivity. The agricultural studies involve the following research topics (see Table 6-2): (1) crop growth and The studies of demand for irrigation water, yield, (2) regional and national agricultural water quality, and farm-level adjustment were also economics, (3) demand for water for irrigation, (4) linked with the integrated modeling studies by water quality, (5) pest-plant interactions, (6) direct common assumptions, sites, or outputs. Because effects of CO2 on crop growth and yield, (7) California grows a large and diverse number of crop impacts of extreme events, (8) potential farm-level commodities, a simple approach was used to adjustments, (9) livestock diseases, and (10) estimate crop yield changes for the California case agricultural policy. study based on heat, sunlight, and photosynthetic response to CO2. These yield changes were then Production of corn, wheat, and soybeans is used in a model of agricultural land and water use critical to the economic well-being of the nation's in California. Adjustment experiments were farmers and the national trade balance. These included in several studies to test possible crops make up about two-thirds of the total U.S. adaptation mechanisms, such as changes in planting agricultural acreage, and their economic value is dates and crop varieties. equal to that of all other crops combined. These Table 6-1. Crop Production by Region Harvested Wheat acres EPA study areas Corn (thousands of bushels) Soybeans (thousands) Southeast 311 272 306 29 Great Lakes 4,644 297 822 92 Great Plains 921 755 136 71 California 38 63 -- 6 Total (48 states) 8,209 2,507 1,990 337 Source: U.S. Department of Commerce (1983). 97 Chapter 6 Table 6-2. Agriculture Projects for EPA Report to Congress on the Effects of Climate Change Regional Studies Effects of Projected CO2-Induced Climate Changes on Irrigation Water Requirements in the Great Plains States - Allen and Gichuki, Utah State University (Volume C) Climate Change Impacts upon Agriculture and Resources: A Case Study of California - Dudek, Environmental Defense Fund (Volume C) Farm-Level Adjustments by Illinois Corn Producers to Climate Change - Easterling, Illinois State Water Survey (Volume C) Impacts of Climate Change on the Fate of Agricultural Chemicals Across the USA Great Plains and Central Prairie - Johnson, Cooter, and Sladewski, Oklahoma Climatological Survey (Volume C) Impact of Climate Change on Crop Yield in the Southeastern U.S.A.: A Simulation Study - Peart, Jones, Curry, Boote, and Allen, University of Florida (Volume C) Effects of Global Climate Change on Agriculture: Great Lakes Region - Ritchie, Baer, and Chou, Michigan State University (Volume C) Potential Effects of Climate Change on Agricultural Production in the Great Plains: A Simulation Study - Rosenzweig, Columbia University/NASA Goddard Institute for Space Studies (Volume C) National Studies The Economic Effects of Climate Change on U.S. Agriculture: A Preliminary Assessment - Adams, Glyer, and McCarl, Oregon State University and Texas A&M University (Volume C) Analysis of Climate Variability in General Circulation Models - Mearns, Schneider, Thompson, and McDaniel, National Center for Atmospheric Research (Volume I) Direct Effects of Increasing CO₂ on Plants and Their Interactions with Indirect (Climatic) Effects - Rose, Consultant (Volume C) Potential Effects of Climatic Change on Plant-Pest Interactions - Stinner, Rodenhouse, Taylor, Hammond, Purrington, McCartney, and Barrett, Ohio Agricultural Research and Development Center and Miami University (Volume C) Agricultural Policies for Climate Changes Induced by Greenhouse Gases - Schuh, University of Minnesota (Volume C) Changing Animal Disease Patterns Induced by the Greenhouse Effect Stem, Mertz, Stryker, and Huppi, Tufts University (Volume C) Effect of Climatic Warming on Populations of the Horn Fly Schmidtmann and Miller, USDA, Agricultural Research Service (Volume C) 98 Agriculture Variability All of the modeling studies used the doubled Trace Gases GCMs CO2 climate change scenarios developed for the report (see Chapter 4: Methodology). These scenarios were developed from estimated changes in monthly mean climate variables from general circulation models (GCMs), without alterations in Climate Change Scenarios climate variability. For example, the number of days of precipitation remains the same in the baseline and climate change scenarios, and the amount of precipitation on each of those days is adjusted by the GCM ratio for climate change. Crop-Response Extreme events, such as maximum temperature, Models vary in the climate change scenarios according to the ratios, but the daily and interannual patterns of warm episodes are determined by the observed baseline climate. Yield Predictions by Crop The lack of changes in the daily and interannual patterns of extreme events may result in underestimation of impacts of climate change. This is because runs of extreme climate variables (for Soil and Water Trade example, prolonged heat spells during grain filling Resource Availability Assumptions and drought) can decrease crop productivity. For rainfed crops, yields may change considerably, depending on whether a change in precipitation is Agro-Economic caused by more or fewer events or by higher or Model lower precipitation per event. The frequency, intensity, and/or duration of extreme climatic events can be much more consequential to crop yields than Economic Consequences are simple changes in means. Land Use and Irrigated Acreage Changes Timing of Effects The timing of climate change is uncertain -- Figure 6-2. Flow chart of model interactions in rates of future emissions of trace gases, as well as EPA studies of the effects of global climate change when the full magnitude of their effects will be on U.S. agriculture (Dudek, 1987). realized, are unknown. CO₂ concentrations are estimated to be about 450 ppm in 2030 and 555 ppm in 2060 if current emission trends continue (Hansen et al., 1988). Other greenhouse gases The agricultural studies performed for this besides CO₂ (e.g., methane (CH₄), nitrous oxide EPA report explore the sensitivities of the different (N₂O), and chlorofluorocarbons (CFCs)) are also parts of the agricultural system (shown in Table 6- increasing. The effective doubling of CO2 means 2) to climate change scenarios. They are not meant that the combined radiative forcing of all to be predictions of what will happen; rather, they greenhouse gases has the same radiative forcing as aim to define ranges and magnitudes of the doubled CO₂ (usually defined as 600 ppm). The potential responses as the system is currently effective doubling of CO₂ concentrations will occur understood. Regional results were extrapolated to around the year 2030, if current emission trends other areas to give estimates of changes in national continue. The climate change caused by an effective production. doubling of CO₂ may be delayed by 30 to 40 years or longer. 99 Chapter 6 RESULTS OF AGRICULTURAL Table 6-3. Increase in Daily Canopy Photosynthesis STUDIES Rates Used in Crop Modeling Studies (%) Regional Crop Modeling Studies Soybean Wheat Corn Design of the Studies Increase in Widely validated crop growth models -- photosynthesis 35 25 10 CERES-Wheat and CERES-Maize (Ritchie and (%) Otter, 1985; Jones and Kiniry, 1986) and SOYGRO (Jones et al., 1988) -- were used to simulate wheat, Source: Peart et al. (Volume C); Ritchie et al. corn, and soybean yields at selected geographically (Volume C); Rosenzweig (Volume C). distributed locations within the Great Lakes, the Southeast, and the Great Plains. Representative agricultural soils were modeled at each site. California crop yield changes were predicted yields. Technology and cultivars were assumed not separately by using an agroclimatic index. (See the to change from present conditions. regional chapters, Chapters 14 through 17 of this report, for descriptions of individual studies.) The CERES and SOYGRO models describe Changes in temperature, precipitation, and solar relationships between plant processes and current radiation were included in the crop modeling climate. These relationships may or may not hold studies. The crop models simulated both rainfed under differing climatic conditions, particularly the and irrigated production systems. The crop high temperatures estimated for the greenhouse modeling approach allowed for analysis of warming. Lack of analysis of the nature and extent latitudinal gradients in changes in crop yields and of agricultural soils at each modeling site adds provided compatible results for each climate change uncertainty to the results. scenario to be used as inputs in the agricultural economics study. (See Ritchie et al., Peart et al., The direct effects of CO₂ in the crop modeling and Rosenzweig, Volume C.) results may be overestimated for two reasons. First, experimental results from controlled environments The direct effects of CO2 -- i.e., increased may show more positive effects of CO₂ than would photosynthesis and improved water-use efficiency actually occur in variable, windy, and pest-infested -- were also included with the climate change (weeds, insects, and diseases) field conditions. scenarios in some model runs to evaluate the Second, since the study assumed higher CO2 levels combined effects. The direct effects were (660 ppm) in 2060 than will occur if current approximated by computing ratios of elevated CO₂ emission trends continue (555 ppm), the simulated (660 ppm) to ambient CO2 (330 ppm) values for beneficial effects of CO2 may be greater than what daily photosynthesis (Table 6-3) and will actually occur. evapotranspiration rates (see Peart et al., Volume C, for detailed description of method). Results Limitations Under climate change scenarios alone, without the direct effects of CO2, yields of corn, soybeans, Uncertainties in the crop modeling studies and wheat were generally estimated to decrease in reside in climate model predictions, locations of the the Great Lakes, Southeast, and Great Plains climate stations (not always in production centers), regions, except in the northernmost latitudes, where crop growth models, and estimates of the direct warmer conditions provided a longer frost-free effects of CO₂. In particular, the climate change growing season. Figures 6-3 and 6-4 show change scenarios did not include changes in climate in modeled rainfed corn and soybean yields for the variability, even though changes in the frequencies GISS and GFDL scenarios. The northern locations of extreme events may considerably affect crop where yields increased included sites in Minnesota. 100 Agriculture 145 100 80 60 40 20 0 -20 -40 -60 -80 40 Duluth, MN 20 0 60 -20 40 -40 20 -60 0 -80 Des Moines, IA 20 -40 09 -60 40 -80 20 Columbus, OH 0 -20 40 40 -40 20 20 - 0 c 08- -20 Wichita, KS -20 -40 -40 -60 -80 -80 Memphis, TN -80 Charleston, SC 40 60 20 0 40 -20 20 100 -40 0 80 - -20 60 -80 -40 40 -100 -60 20 Macon, GA -80 0 San Antonio, TX 20 -40 -60 80 Meridan, MS GISS GFDL GISS + Direct Effects of CO₂ GFDL + Direct Effects of CO₂ Figure 6-3. Percent change in rainfed corn yields simulated by the CERES-Maize model for baseline (1951- 80) and GISS and GFDL climate change scenarios with and without the direct effects of CO2 for selected locations (Peart et al., Volume C; Ritchie et al., Volume C; Rosenzweig, Volume C). Decreases in modeled yields resulted primarily from Southeast, the direct effects of CO₂ would not fully higher temperatures, which would shorten the crop compensate for changes in climate variables -- net life cycle thus curtailing the production of usable yields were estimated to decrease significantly from biomass. In the Southeast, rainfall reductions were the base case. Elsewhere, yields were generally a major factor in the GFDL results. Modeled estimated to increase, with relatively greater rainfed yields were estimated to decrease more than increases at the northern locations. irrigated yields. The crop models were also used to test several When increased photosynthesis and improved possible adaptations by farmers to the predicted water-use efficiency were included in the crop climate changes. For example, a corn variety that is models along with the climate change scenarios, better adapted to longer growing seasons was tested yields increased over the baseline in some locations in Indiana. Use of this later maturing variety would but not in others (see Figures 6-3 and 6-4). not compensate entirely for the yield decreases Particularly when combined with the hotter and caused by the warmer climate change scenarios. drier GFDL climate change scenario in the 101 Chapter 6 100 80 60 100 40 120 80 20 60 80 40 0 Duluth, MN 40 20 0 0 -20 40 -40 60 -80 Columbus, OH Des Moines, IA 0 -10 -20 -30 -40 40 -50 -60 20 -70 0 -80 -20 Memphis, TN 40 20 -60 -80 0 0 -10 -100 -20 -20 Charleston, SC -30 -40 -40 -50 -60 -60 -80 -70 -80 -100 Macon, GA Meridan, MS GISS GFDL GISS + Direct Effects of CO₂ GFDL + Direct Effects of CO₂ Figure 6-4. Percent change in rainfed soybean yields simulated by the SOYGRO model for baseline (1951-80) and GISS and GFDL climate change scenarios with and without the direct effects of CO2 for selected locations (Peart et al., Volume C; Ritchie et al., Volume C). Implications This could further tighten water supply problems in some areas and increase pollution from nonpoint The potential for climate change-induced sources (i.e., pollution that is not traceable to any decreases in crop yields exists in many agricultural one distinct source, such as agricultural chemicals regions of the United States. In some northern from farmers' fields). Considerable uncertainty areas, crop yields may increase. Farmers would exists regarding the future availability of surface need varieties of corn, soybeans, and wheat that are water and groundwater supplies with climate better acclimated to hotter and possibly drier change, and concerning the competing demands for conditions to substitute for present varieties. and costs of using or extracting the water (see Chapter 9: Water Resources). If the major agricultural areas are to continue to provide a stable supply of food under the Regional and National Economics Study predicted changes in climate, supplemental irrigation may be required for many soils. Pressure The estimated yield changes from the crop for increased irrigation may grow in these regions. modeling studies (not including California) and 102 Agriculture projected changes in irrigation water demand and Acreage available for production is based on current availability were introduced into an agricultural definition of agricultural land classes. Both irrigated economic model to translate the physical effects of and nonirrigated crop production and water supply climate change into economic consequences. relationships are included for most regions. The Adams et al. (see Volume C) estimated the regional model simulates a long-run, perfectly competitive and national economic implications of changes in equilibrium and was developed using 1980-83 yields of wheat, corn, soybeans, and other crops and economic and environmental parameters. in the demand for and availability of water associated with alternative global climate change A set of model runs was conducted, using the scenarios. GISS and GFDL climate change scenarios, with and without the direct effects on crop yields. Potential Study Design changes in technology and in future U.S. and world food demand due to population growth were also A spatial equilibrium agricultural model introduced into the climate change analysis. developed by Adams et al. (1984) was used to represent production and consumption of numerous Limitations agricultural commodities for the U.S. farm production regions as designated by the USDA The economic approach used in this study has (Figure 6-5). The model has been used to estimate several limitations. The economic model is static in agricultural losses due to increased ultraviolet-B the sense that it simulates an equilibrium response (UV-B) radiation caused by stratospheric ozone to climate change, rather than a path of future depletion (Adams et al., 1984). It consists of farm- changes. Substitution of crop varieties, new crops, level models for production regions, integrated with and adjustments in farm management techniques a national-level model of the agricultural sector. were not included; thus, the negative effects of PACIFIC LAKE STATES NORTHEAST MOUNTAIN PLAINS CORN BELT MOUNTAIN APPALACHIAN DELTA SOUTHEAST SOUTHERN STATES PLAINS Figure 6-5. Farm production regions in the United States (USDA, 1976). 103 Chapter 6 climate change were possibly overestimated. Since are not explicitly included in the model. Such CO₂ levels were assumed to be high in the crop changes could have major impacts on U.S. modeling study, estimates of the beneficial direct agriculture. For example, warming may enhance effects of CO2 on crop yields may have biased the the agricultural capabilities of high-latitude countries economic results in the positive direction in some such as Canada and the U.S.S.R. While the net scenarios. effect of climate change on the rest of the world is uncertain, global changes could overwhelm U.S. Furthermore, changes in yields used as inputs national impacts. A net negative effect on to the economic model were modeled for only agriculture abroad would improve the position of wheat, corn, and soybeans for a limited number of U.S. agricultural producers through enhanced sites and regions. The regional crop yield analyses exports, but could increase the negative impacts on cover 72% of current U.S. corn production, 33% of U.S. consumers through increases in global wheat production, and 57% of the soybean output. commodity prices. National estimates were extrapolated from these for all other crop commodities in the model. Changes Results in risk, where risk is defined as increases in variance of crop yields, were not explicitly included in the It is important to note that the results of the economic analysis. The accuracy of the estimates of economic study are not predictions. Rather, they changes in water supply and crop water are initial estimates of how the current agricultural requirements derived from the GCMs cannot be system would respond to the projected climate ascertained. Potential increases in the demand for change scenarios. water by nonagricultural users, which would reduce water available for irrigation, were not included. All The economic model showed a small to of these assumptions introduce uncertainties into moderate aggregate loss in economic welfare the results. associated with the estimated crop yield and hydrologic changes derived from the climate change Potential changes in international agricultural scenarios (see Table 6-4). For the moderate GISS supply, demand, and prices due to climate change climate change scenario, net losses were small; for Table 6-4. Aggregate Economic Effects of GISS and GFDL Doubled CO₂ Climate Change on U.S. Agriculture with and without the Direct Effects of CO2 on Crop Yields Economic effects (billions of 1982 dollars) Run Consumer Producer Total GISS Analysis 4ª: -7.3 1.5 -5.9 without CO2 GISS Analysis 4: 9.4 1.3 10.6 with CO₂ GFDL Analysis 4: -37.5 3.9 -33.6 without CO2 GFDL Analysis 4: -10.3 0.6 -9.7 with CO2 a Analysis 4 includes the crop yield and irrigation water supply and demand consequences of climate change throughout the United States. Source: Adams et al. (Volume C). 104 Agriculture the more extreme GFDL scenario, they were Production of most crops was reduced because greater. The magnitudes of these changes, which of yield declines and limited availability of land and are annual, may be compared with the estimated resources. With climate change alone, corn $2.5 billion (in 1982 dollars) in agricultural losses production decreased 12 and 47% in the GISS and due to increased UV-B radiation caused by GFDL scenarios, respectively, while soybean stratospheric ozone depletion of 15% (Adams et al., production was estimated to be reduced by 12 and 1984). In general, consumers lose and producers 53% for the same scenarios. In all scenarios, land gain because of the increased prices of agricultural under production in Appalachia, the Southeast, the commodities and inelastic demand (i.e., insensitivity Mississippi Delta, and the Southern Plains could to price changes) for agricultural crops. decrease on average by 11 to 37%, while in the Lake States, the Northern Plains, and the Pacific it Higher CO2 levels could reduce negative could increase by small amounts (see Figure 6-6). economic impacts (Table 6-4). Under the less While availability of agricultural soils was included severe GISS climate scenario, the CO₂ direct effects in the economic analysis, the sustainability of crop were estimated to sufficiently counter the climatic production in northern areas was not studied. effects in most regions, so that both producers and consumers gain. With the more severe GFDL Irrigated acreage was estimated to increase in climate change scenario combined with the direct all areas, primarily because irrigation becomes effects of CO2, lower yields led to higher prices, but economically feasible as agricultural prices rise (see not by as much as occurred with the climate change Figure 6-7). These changes reflect both increased scenarios alone. However, significant changes in demand by farmers for irrigation water and changes regional agricultural land use occurred even when in water availability as estimated by the GCM the beneficial direct effects of CO2 were taken into scenarios, but do not take into account changes in account. competition with industrial or municipal users. 20+ 20t 10 10 0 0 10 20- 10 20+ Lake States 20 10 VI 30 0 140 40 10 -50- Northeast 20- 20+ Northern Plains 20 10 10 0 88 888 0 10 / 20th 10 20- 10 20 Corn Belt 20 0 Mountain 10 10 0 20 10 30 20- 20t 40 Pacific 10- 20t 20t 50 0 10 10 60 10 0 0 70 20 10 10 80L 30 20 20 Appalachia 40 30 30 50- 40 GISS 40 Southern Plains 50 50 GFDL 60 60 -70 70 GISS + Direct Effects of CO₂ 80L 80L GFDL Delta States Southeast + Direct Effects of CO₂ Figure 6-6. Percent change in regional agricultural acreage simulated by an economic model of the U.S. agricultural sector for the GISS and GFDL climate change scenarios with and without the direct effects of CO2 on crop yields (Adams et al., Volume C). 105 Chapter 6 80t 70 60 50 40 40 30 20 30 10 0 20 10 20- 10 Northern Plains 0 40t 10 30 20 20 Mountain 40 10 2.5 80t 30 2.0 70 0 1.5 60 20 1.0 50 10 0.5 40 10 0 30 20 0.5 20 0 1.0 10 Pacific GISS Delta States 0 10 GFDL 10 20- GISS + Direct Effects of CO2 20 Southeast Southern Plains GFDL + Direct Effects of CO2 Figure 6-7. Change (100,000s of acres) in regional irrigation acreage simulated by an economic model of the U.S. agricultural sector for the GISS and GFDL climate change scenarios with and without the direct effects of CO₂ on crop yields. Changes are not shown in the Great Lakes, Corn Belt, Appalachia, and Northeast because currently irrigated acreage is small (2% of total U.S. irrigated acreage) in these regions (Adams et al., Volume C). Technological changes, such as higher yielding Implications crop varieties, chemicals, fertilizers, and mechanical power, have historically enabled agriculture to Food Supply and Exports increase production with the same amount of, or less, land, labor, and other resources. When the The economic analysis implies that although effect of future technological change (based on yield climate change could reduce the productive capacity increases from 1955 to 1987) was modeled along of U.S. agriculture, major disruption in the supply of with the less severe GISS climate change (without basic commodities for American consumers would the direct effects of CO2), most of the adverse not occur. Domestic consumers would face slightly climate effects were estimated to be offset. Under to moderately higher prices under some analyses, the severe GFDL climate change scenario, but supplies could be adequate to meet current and continued and substantial improvements in yields projected domestic demand. However, if droughts would be required to overcome the climate change occur more frequently under changed climate, effects. Stated another way, the adverse effects of effects on agriculture may be more severe. climate change could negate most of the higher output attributable to improved technology over the Exported commodities in some scenarios next 50 years. It is important to note, however, that decline by up to 70%, assuming the demand for the rate of future technological advances is very exports remains constant. Thus, climate change difficult to predict. Increasing food demand from could affect the United States in its role as a higher U.S. and world population aggravated the reliable supplier of agricultural export commodities. estimated economic losses from the climate change It is likely that supply of and demand for scenarios. 106 Agriculture agricultural commodities could shift among prairie potholes for ducks and flyways for bird international regions, and responses of U.S. migrations. agriculture will take place in this global context. There is a great need to determine the nature of In addition, many of the glacial till soils in the these changes in global agriculture by analyzing the northern latitudes are not as productive as Corn potential impacts of climate change on both major Belt soils. Thus, large increases in production of world agricultural production regions and potentially crops would most likely require greater applications vulnerable food deficit regions. of chemical fertilizers. The use of these fertilizers in humid regions on glacial till and sandy soils is Regional Economics and Land Use now creating an environmental hazard to the underlying groundwater, receiving waters, and Regional shifts in U.S. agricultural production aquatic habitats in many areas. With climate patterns (not only grain crops but also vegetables change, water and fertilizer use would have to be and fruits) are highly likely, as all climate change carefully managed to minimize still more leaching of scenarios tested show that the southern areas of the water-soluble nutrients such as nitrogen and potash. United States become less productive relative to the northern areas. This is primarily because the high Demand for Water for Irrigation temperatures estimated for climate change would stress crop production more in southern areas than Water is the single most critical factor in in northern areas where crops are currently limited determining the development, survival, and by lower temperatures and shorter growing seasons. productivity of crops. The amount of water that However, increased agricultural production may be crops use and thus the demand for irritation water difficult to sustain in the North, because some soils are governed largely by the evaporation process. may be less fertile and may have lower water- Higher air temperatures due to increasing trace holding capacity. Crops grown in soils with lower gases in the atmosphere could heighten evaporative water-holding capacity require more evenly demands. Increased irrigation to satisfy these distributed rainfall to produce comparable yields. higher demands could accelerate depletion of groundwater and surface water resources. Also, the Regional changes in agriculture would have rate of evaporation might outstrip precipitation, thus important implications for rural communities. As decreasing crop yields. production areas shift, climate change effects would reverberate through these communities and are Studies reported in the California and the likely to result in structural changes in local Great Plains case studies (see Chapters 14 and 15) economies, such as relocation of markets and explicitly examined the potential changes in demand transportation networks. At its most extreme, for water for irrigation. The studies did not climate change could cause dislocation of rural consider changes in competing demands for water communities through farm abandonment. such as industrial and residential use, which also may change in a warmer climate. The California Environmental Concerns study, however, considered changes in supply due to earlier snowmelt and sea level rise. In these Regional agricultural adjustments could place regions, water is a critical resource for agriculture; environmental resources at risk. Where agricultural California and the parts of the Great Plains fed by acreage would increase, demands for natural the Ogallala Aquifer, in particular, depend very resources, such as soil and water, might intensify heavily on irrigation for crop production. current pressures on environmental elements, such as rivers, lakes, aquifers, wetlands, and wildlife Irrigation Requirements in the Great Plains habitats. Northern States, such as Minnesota and North Dakota, could become more productive for Allen and Gichuki (see Volume C) computed annual crops like corn and soybeans because of irrigation water requirements for sites in the Great warmer temperatures and a longer frost-free Plains for the baseline climate and the GISS and growing season. Given the presence of forests and GFDL climate change scenarios. The direct effect wetlands in these regions, increased agricultural of CO2 on water use was also included. (For study production in the area might threaten natural design and limitations, see Chapter 17: Great ecosystems, including wildlife habitats such as Plains.) Major changes in irrigation water 107 Chapter 6 requirements were estimated for all locations in the Great Plains and for all crops (see Figure 6-8). The CORN most significant would be the persistent increases in 80 seasonal net irrigation water requirements for 70 alfalfa, which would be driven by the climate 60 changes in temperature, wind, humidity, and solar season. Decreases in irrigation requirements were Percent Change From Baseline Value 50 radiation, and by the lengthening of the growing 40 30 estimated for winter wheat in most regions. These 20 decreases would be the result of earlier planting 10 dates and shorter crop life cycle due to high 0 temperatures. When crop varieties appropriate to 10 the longer growing season were modeled, irrigation Nebraska Kansas Oklahoma Texas water requirements for winter wheat were estimated GISS to increase. Simulated irrigation water GFDL requirements during peak periods increased in almost all areas (see Figure 6-9). Figure 6-9. Percent change in peak irrigation requirements of corn for GISS and GFDL climate change scenarios with direct effect of CO₂ on crop ALFALFA water use included (Allen and Gichuki, Volume C). 120 100 Percent Change From Baseline Value 80 While farmers in the Great Plains would 60 probably shift to longer season crops, climate 40 change conditions (warmer temperatures and drying 20 in some areas) during the later summer months 0 could increase irrigation requirements and elevate Nebraska Kansas Oklahoma Texas leaf temperatures to a point that exceeds optimum CORN temperatures required for high productivity. This 30 might make it uneconomical to take full advantage Percent Change From Baseline Value 20 of the longer growing season, especially if the higher CO₂ levels increase photosynthesis and offset the 10 effects of a shorter season to some degree. 0 Water Resources for Agriculture in California -10 20 Nebraska Kansas Oklahoma Texas In the California regional case study, Dudek (see Volume C) characterized the potential shifts in WHEAT 2 demand for water for agricultural production that 0 would accompany shifts in cropping patterns driven Percent Change From Baseline Value -2 by changing climate. Changes in competing 4 -6 demands for water from industrial or municipal -8 users were not considered. (For description of -10 study design and limitations, see Chapter 14: -12 GISS California.) When climate change was considered -14 GFDL -16 alone, groundwater extraction and surface water use Nebraska Kansas Oklahoma Texas were estimated to decline in California as a result of changes in both supply of (derived from GCM Figure 6-8. Percent change in net seasonal climate change scenarios) and agricultural demand irrigation requirements for GISS and GFDL climate for water. When the direct effects of CO2 on crop change scenarios with direct effect of CO₂ on crop yields were included, groundwater extraction would water use included (Allen and Gichuki, Volume C). increase because of improved yields of all crops 108 Agriculture except corn and because of enhanced economic and sprinklers currently requires about $1,500 to welfare. Institutional responses to changes in $5,000 per hectare in capital investment (Postel, surface and groundwater use could include water 1986). transfers, which could improve irrigation efficiency. When water markets were included in the Direct Effects of CO2 on Crops simulations, economic welfare was improved by 6 to 15% over the base, while crop acreage increased Global increases in CO₂ are likely to influence and groundwater extraction decreased. crop metabolism, growth, and development directly through physiological processes and indirectly Implications for Demand for Irrigation Water through climate. Rose (see Volume C) reviewed recent experimental work performed on the direct Expanded use of irrigation is implied from the effects of CO2 on crops, with emphasis on wheat, regional crop modeling studies for the Great Lakes, corn, soybeans, and cotton. the Southeast, and the Great Plains (see Chapters 15, 16, and 17, respectively). Increases in irrigated Elevated concentrations of CO2 directly affect acreage are also estimated for most regions when plant processes such as photosynthesis and the economics of crop production are factored in transpiration. Higher CO2 concentrations are also (see Adams et al., Volume C). When these results expected to influence these processes indirectly are considered along with the irrigation studies, it through predicted increases in temperature and appears that climate change is likely to increase the other changes in climate variables such as demand for water from the agricultural sector in precipitation. Because experimental research has many regions. rarely simultaneously studied both the direct and indirect effects of plant responses, it is difficult to In the Great Plains, heightened evaporative assess the relative contributions of elevated CO2 demand and variability of rainfall may increase the and climate changes to predictions of crop need for irrigation in dryland farming regions. The responses. simulated changes in irrigation water requirements are varied, and specific crops and locations probably Research on the physiological effects has would be affected differently. Higher peak focused primarily on responses of rates of irrigation water requirements for some crops may photosynthesis and transpiration to increasing require larger capacity irrigation systems and may concentrations of atmospheric CO2. Photosynthesis enlarge energy demands. rates have increased in these crops in relatively ideal experimental environments. At moderate Intensified extraction of water poses serious temperatures, most crops will probably show environmental and economic problems, especially in increases in size and possibly yield as CO₂ areas where groundwater is being overdrawn. concentrations rise. However, plants also have Streamflows also may slacken if more surface water internal regulation mechanisms that may lessen is used for irrigation, thereby aggravating water these effects under field conditions. quality problems. This in turn would harm fish, wildlife, and recreational activities. Transpiration rates per unit leaf area decrease, while total transpiration from the entire plant Regional changes in cropping locations and sometimes increases because of greater leaf area. patterns of water use also could exacerbate Drought-stressed plants exposed to high partial agricultural, nonpoint source pollution, and could pressures of CO₂ should be better able to cope with further deplete groundwater resources. Institutional water deficits. Leaf temperatures in all species are responses, such as markets for water transfers, expected to rise even more than air temperatures; could help improve irrigation water management this may inhibit plant processes that are sensitive to and alleviate some of these negative effects. high temperature. The economic and social costs of shifting the Few studies have examined the interactive location of irrigated agriculture could be effects of CO2, water, nutrients, light, temperature, considerable. The construction of irrigation systems pollutants, and sensitivity to daylength on consisting of reservoirs, wells, ditches, pipes, pumps, photosynthesis and transpiration. Even fewer 109 Chapter 6 studies have examined the effects of these For example, the potato leafhopper, a serious pest interactions on the growth and development of the on soybeans and other crops, at present overwinters whole plant. Therefore, considerable uncertainty only in a narrow band along the coast of the Gulf of exists concerning the extent to which the beneficial Mexico (Figure 6-10). Warmer winter temperatures effects of increasing CO2 will be seen in crops in the GFDL and GISS scenarios could cause a growing in the field under normal farming doubling or tripling of the overwintering range in conditions with climate change. the United States, respectively. This would increase the invasion populations in the northern states by Climate Impacts on Pest-Plant similar factors. The invasions also would be earlier Interactions in the growing season, assuming planting dates do not change. Both features are likely to lead to Compared with the existing information on the greater insect density and damage. This pattern is potential effects of climate change on crop repeated with the other three pests studied and production, relatively little effort has been directed indicates that these pests, and possibly others, may toward assessing the influence of climate change on move northward and invade cropping systems earlier plant-pest interactions. Atmospheric increases in in the growing season under climate change conditions. temperature and CO2, and changes in moisture regimes, all can directly or indirectly affect interactions between pests and crops. Changes in pests will also depend on regional shifts in crop production. Although crop pests may be defined as weeds, insects, or disease pathogens, the EPA work Potato leafhopper on this subject focused on insects. Study Design and Results Stinner et al. (see Volume C) conducted a literature survey and modeling experiments on the GISS GFDL major mechanisms through which climate change Present may affect pest-plant interactions. This study emphasized the major insect pest and pathogen species of corn and soybeans. The survey indicates that temperature and precipitation patterns are the key variables that affect crop-pest interactions. The temperature increases associated with the climate change scenarios would bring about the following trends: (1) increased survival for migratory and Figure 6-10. Present and potential (GISS and nonmigratory insect pest species in the winter; (2) GFDL climate change scenarios) overwintering northern range extensions of current pests in the range of the potato leafhopper, Empoasca fabae, a higher latitudes and migration of southern species major pest of soybeans (Stinner et al., Volume C). into the northern Grain Belt regions; (3) an increase in pest species with more than one generation per year in the northern Grain Belt; (4) earlier The Soybean Integrated Crop Management establishment of pest populations in the growing (SICM) model (Jones et al., 1986) was run with the season; and (5) increased abundance of pests during GISS and GFDL climate change scenarios to more susceptible crop growth stages. estimate changes in damages caused by corn earworm. Modeling results show that earworm The potential changes in the overwintering damage to soybeans would increase in severity in the Grain Belt under a warmer climate. Such ranges of four major pests were mapped for the GISS and GFDL climate change scenarios and were damage could cause grain farmers in the Midwest to compared to present ranges. The overwintering suffer significant economic losses. These results capability of the four major pests may extend were particularly marked with the warmer and drier GFDL scenario. northward with both climate change scenarios. 110 Agriculture Limitations wheat, corn, and cotton production regions in the Great Plains and the Corn Belt. (For details of the Lack of knowledge about the physiological study, see Chapter 17: Great Plains.) They used effects of CO2 on crop plants and lack of the Pesticide Root Zone Model (PRZM) (Carsel et experimental evidence of direct CO₂ effects on al., 1984), which simulates the vertical movement of insect-plant interactions make the study of pest- pesticides in the soil. The model consists of plant interactions particularly difficult. Only one hydrological and chemical transport components cultivar was used in the modeling study under both that simulate runoff, erosion, plant uptake, leaching, the baseline and the climate change scenarios, and decay, foliar washoff, and volatilization of a planting dates remained the same. In reality, pesticide. The interactions among soil, tillage, farmers would probably switch to a more management systems, pesticide transport, and climatically adapted cultivar as climate changed, and climate change were studied. they would advance planting dates in response to longer growing seasons. Limitations Implications The frequency and duration of precipitation remain the same in the climate change scenarios, Increased pest-related crop damage could even though these storm characteristics are critical intensify pesticide use. The economic and factors in determining the transport of agricultural environmental ramifications of such an increase chemicals and may change. The scenarios assume could be substantial, not only in current farming that the number of days with rainfall does not regions but also in new areas if agriculture shifts to change, but the intensity of rainfall increases or the more northern regions such as the northern decreases. Runoff and leaching estimates would Plains, the Great Lakes States, and the Pacific most likely be different if the number of days of Northwest (see Figure 6-6). rainfall changed and daily rainfall amounts were held constant. Increased use of pesticides would create additional threats to the integrity of ecosystems The PRZM is a one-dimensional, point model through soil and water contamination and could that does not simulate the transport of water below increase risks to public health. If agricultural the root zone. Thus, results on a regional basis production is not to rely increasingly on chemicals must be extrapolated with care. The direct effects that are potentially harmful to the environment, an of CO₂ on crop growth, which may increase the size increased need will exist for alternative pest of the plants and the extent to which crops cover management strategies such as biological control, the soil, are not included. genetic resistance, and innovative cropping systems. Results Effects of Climate Change on Water Quality Regional changes in chemical loadings of water and sediment are likely due to climate change Agricultural pesticides are ranked as a high- but probably will not be uniform. There appears to priority pollution problem in many rural regions. be some consensus between the GCM scenarios Potentially toxic agricultural chemicals can be concerning the estimated regional changes (Table transported away from fields via runoff of surface 6-5). Modeled pesticides in runoff increase in the soils and via downward leaching and percolation cotton production area, and pesticides carried by through the soil. An understanding of these sediments decrease in the spring wheat and corn processes is needed to evaluate potential threats to regions. Leaching of pesticides tends to be less drinking water quality caused by climate change. everywhere owing to changes in seasonal precipitation and increased evaporation. Study Design Implications Johnson et al. (see Volume C) modeled the partitioning of agricultural pesticides among uptake, When the changes in water quality from the degradation, surface runoff, and soil leaching for predicted climate change scenarios are considered 111 Chapter 6 Table 6-5. Summary of GISS and GFDL GCM Model Consensus of PRZM Pesticide Transport by Cropping Region and Pesticideᵃ Surface Surface Crop and pesticide pesticide Pesticide pesticide type runoff losses erosion losses leaching Spring wheat Highly soluble/short-livedᵇ + + - Highly soluble/long-lived + - Slightly soluble/long-lived - Winter wheat Highly soluble/short-lived + + Highly soluble/long-lived + - Slightly soluble/long-lived - Cotton Highly soluble/short-lived + + Highly soluble/long-lived + + - Slightly soluble/long-lived + + - Corn Highly soluble/short-lived - - Highly soluble/long-lived - - Slightly soluble/long-lived - - - a + indicates that median values increase under climate change; - indicates that median values decrease under climate change; blank indicates no consensus among median values. b Example: median value of all tillage, soil, weather site scenarios for highly soluble/short-lived pesticides in the spring wheat crop area. Source: Johnson et al. (Volume C). in conjunction with the estimated increases in pests potential changes in climate variability, a review of and implied higher applications of pesticides literature on agriculture and extreme events that described in the study on pest-plant interactions, the focuses on the nature and magnitudes of significant potential for changes in the nation's water quality impacts is included in Chapter 3: Climate becomes apparent. Any deterioration in water Variability. quality could adversely affect public drinking water supplies and human health. Corn, soybeans, wheat, and sorghum are sensitive to high maximum temperatures during Climate Variability blooming. Lower yields of corn, wheat, and soybeans have been correlated with high The impacts of climate change result not only temperatures. The damaging effect of runs of hot from a slow change in the mean of a climate days on corn yields was particularly evident in the variable but often from shifts in the frequency of U.S. Corn Belt in 1983. extreme events. Droughts, freezes, and prolonged periods of hot weather have strong effects on Although the problems associated with low agricultural production. Although the agricultural temperatures may diminish with climate change, modeling studies did not include the effects of risks of frost damage to crops may change in the 112 Agriculture growing areas of certain crops. Citrus trees are very a longer growing season (see Chapter 15: Great vulnerable to low minimum temperatures. Winter Lakes). Rosenzweig (see Chapter 17: Great Plains) wheat is often damaged by low temperatures known showed that adjusting the planting date of winter as winter kill, especially in the absence of snow. wheat to later in the fall would not ameliorate the Even with warmer winters and fewer frosts, more effects of climate change, but that changing to damage may occur at less extreme temperatures. varieties more suited to the predicted climate could For example, the effect of freezing temperatures is overcome yield decreases at some locations. exacerbated if crops have not yet been hardened by cold temperatures or if the crops are no longer Dudek's California study found that flexible dormant and a cold snap occurs. institutional responses to climate change would help to compensate partly for negative climate change Drought is a major cause of year-to-year effects (see Chapter 14: California). By allowing variability in crop production. In the Dust Bowl movement of water around the state by transferral years of the 1930s, yields of wheat and corn in the of water rights, California's water resource Great Plains dropped to as much as 50% below managers could alleviate some groundwater normal. In 1988, agricultural disaster in areas of the extraction and compensate for surface water northern Great Plains demonstrated a high reductions. vulnerability to drought, and nationwide corn yields decreased by nearly 40%. Reduction in vegetative Easterling (see Chapter 15: Great Lakes) cover associated with drought also brings about found that potential farmer adjustments to climate severe wind erosion of soils, which will affect future change include changes in tillage practices, crop productivity. Low yields of forage crops during increased application of fertilizers, selection of more droughts result in food shortages for livestock and full-season and heat-resistant varieties, changes in premature selling of livestock. If frequency of planting densities, higher use of pesticides, earlier drought increases with climate change, impacts on harvest, and reduced artificial drying. Different agriculture can be severe. adjustments could occur at different times in the cropping season. With the hotter and drier GFDL Farm-Level Management and Adjustments scenario, farmers may have to adopt production to Climate Change practices different from those in use today. Climate changes that leave soils drier during summer than they are at present will most likely lead to an Adjustments to existing production practices increased use of irrigation in the Corn Belt. This would be the first course of action in the face of increased irrigation is also supported by the climate change. The net effect of climate change projected price increases for all crops grown in with adjustment by farmers may be significantly Illinois. different from the estimated effects of climate change alone. Implications Study Design Although detrimental climate change effects on agriculture may be partly offset naturally by Several studies addressed possible adjustments increased photosynthesis and water-use efficiency that could modify the effects of climate change. caused by higher levels of atmospheric carbon These adjustments include changes in planting and dioxide, farmers themselves would use a variety of harvesting dates, tillage practices, crop varieties, adjustments to adapt to climate change. Market application of agricultural chemicals, irrigation forces also would aid adaptation to climate change technology, and institutional responses for water because they help to allocate resources efficiently. resource management. Each crop and region would respond differently to climate change, and adjustment strategies would Results need to be tailored to each situation. Ritchie et al. demonstrated that the yield Costs of adjustments are likely to vary reduction in corn in the Great Lakes could be partly considerably from region to region. Costs would be overcome with selection of new varieties that have 113 Chapter 6 relatively small in regions where farmers can switch vectors responsible for the transmission of infectious from one variety to another or from one grain crop diseases in livestock. The activity and reproduction to another, thus enabling continued use of existing of disease-carrying vectors infecting livestock, farm machinery and marketing outlets. However, at humans, and crops are driven primarily by locations near the present limit of major agricultural temperature, humidity, and precipitation. These regions (e.g., the boundary between wheat farming impacts are likely to be similar to those on mortality and ranching), relatively small changes in climate and morbidity of disease in humans (see Chapter may require a substantial switch in type of farming. 12: Human Health), and they also are similar to This may require substantial costs in new equipment changes predicted for crop pests. and other changes in agricultural infrastructure. Severe climate change may necessitate farm Design of Studies abandonment in some regions. Stem et al. (see Volume C) studied the Improvements in agricultural technology also available literature on four livestock diseases to may be expected to ease adjustment through evaluate the range of potential changes in disease development of appropriate farming practices, crop distribution and occurrence under climate change varieties, and livestock species. Adjustment and conditions. Schmidtmann and Miller (see Volume adaptation to climate change should be included in C) used a population dynamics simulation model to agricultural research programs to enable this estimate the effects of the GFDL climate change process to occur. scenario on the life cycle of the horn fly, a ubiquitous pest of pastured cattle throughout the Livestock United States. Animal products are a critical source of Limitations protein, energy, vitamins, and minerals. U.S. livestock production, mainly from cattle, swine, The horn fly model is based on population sheep, and poultry, was estimated to be worth over counts taken at various times under different $31 billion in 1986 (USDA, 1987). weather and management conditions. However, the prediction of current horn fly populations appears to Climate is known to significantly affect many be well correlated with observations. The model is aspects of animal health and production. The direct not validated for the high temperatures predicted effects of climate warming on animal health include for the climate change. Schmidtmann and Miller differences in incidence of heat and cold stress, used only the hottest climate change scenario, changes in weight gain, and decline in reproductive GFDL; the other scenarios may have resulted in a capabilities. Indirect effects may involve trends in smaller geographic shift in the range of the horn fly. the availability and prices of animal feeds and the It should also be noted that the horn fly analysis is expanded geographic distribution and activity of based on current livestock management, breeds, and disease-carrying vectors. distribution. Possible changes in these factors are beyond the scope of this study. For example, Higher winter temperatures may lower the changes in location and extent of grassland regions incidence of respiratory diseases in livestock and forage production caused by climate warming (Webster, 1981). Conversely, warmer summers may would affect livestock production and horn fly necessitate more hours of indoor cooling during distributions. which pathogens are confined to housing structures. Climate warming may significantly increase the costs Results of air-conditioning in poultry housing. Changes in reproductive capabilities such as decreased ovulation Stem et al. found that under warmer rates, shortened intensity and duration of estrus, conditions, livestock diseases currently causing decreased fertility of males, and increased serious economic losses in tropical countries could embryonic mortality also have been shown to occur spread into the United States. Rift Valley fever is with high temperatures (Ames, 1981). transmitted principally by mosquitoes, and the disease may spread as rising winter temperatures Climate change may also affect the become able to support an increase in the mosquito survivability, activity, and geographic distribution of 114 Agriculture population (see Figure 6-11). African swine fever United States could be extended by 8 to 10 weeks. also may become a greater threat. The increase in horn fly populations could substantially reduce the average daily gain of The ranges and activities of disease-carrying growing beef cattle. Also under the GFDL agents of blue tongue and anaplasmosis, diseases simulation, increased pest activity was estimated in currently causing severe losses in cattle and sheep dairy cattle in the North and Northwest -- a result production in the United States, may expand. If that could significantly decrease milk production. disease-carrying insects increase their winter survival Conversely, under the same scenario, the and reproduce year-round in more states, the summertime activity of the horn fly could decrease geographical distribution of blue tongue, which is in the South because the warmer climate would caused by a virus, may expand northward and exceed the horn fly's tolerance to high temperatures. eastward. Anaplasmosis, a rickettsial infection of ruminants, is the second most important disease of Implications cattle in the United States. Distribution of the insect carrier's habitat could expand to northern With climate change, patterns of livestock states with climate change, and the insects' day-to- diseases and pests may also change. Tropical day activity may increase; this process may also livestock diseases may become an increased threat, cause an increase in disease transmission. because more geographical areas are potential ranges for the insect carriers of the diseases. The horn fly causes annual losses of $730.3 Temperature conditions may improve in the winter million in the beef and dairy cattle industries but may be exacerbated in the summer. (Drummond, 1987). Schmidtmann and Miller found Reproductive capabilities may be lower. Livestock that with the very warm GFDL climate change production would also be affected if rangeland areas scenario, the horn fly season throughout most of the shift and forage production levels change. CURRENT & DOUBLED CO₂ DOUBLED CO₂ Figure 6-11. States where significant Culex spp. activity permits establishment of Rift Valley fever for current and doubled CO₂ levels (Stem et al., Volume C). 115 Chapter 6 ECONOMIC AND ECOLOGICAL the agricultural system strives to adapt to a changing IMPLICATIONS OF climate, there may be no chance of optimizing for static conditions. Rather, the system may be caught AGRICULTURAL STUDIES in forever playing catch-up. The U.S. agricultural system has historically Effects of CO2 been able to adopt new technologies rapidly and may be less vulnerable to climate change than It is also important to note that the crop natural ecosystems. In fact, global warming may modeling studies showed that the direct CO₂ effects cause a number of benefits. Potential benefits of on crop photosynthesis and water-use efficiency induced climate change include increases in ameliorate the negative effects of climate change in length of growing season and in air temperatures, some locations under certain climate conditions; which would benefit regions where crop growth is however, such effects do not occur uniformly, and constrained by short summers and low they do not occur everywhere. Regional changes in temperatures. Longer growing seasons would likely U.S. agriculture occurred with the GISS and GFDL lead to increased yields of hay and other perennial climate change scenarios both with and without the crops. Energy costs for grain drying may be direct effects of CO2. While much work must be reduced, since annual crops would reach maturity done to improve both climate and crop models, earlier and would have more opportunity to dry in policy analysis should consider that the beneficial the fields. Furthermore, in places where direct effects of CO₂ may not offset the negative precipitation increases during the growing season, effects of climate change. irrigation requirements could be reduced. If irrigation requirements are lessened, demand on Environmental Quality regional water resources and associated costs to farmers may fall. Changes in the agricultural production system are likely to have significant impacts on resource However, many reasons to avoid complacency use and the environment. Many of the agricultural about the predicted climate change remain. studies suggest that climate warming could result in Concern for our major resources (especially land accelerated rates of demand for water for irrigation and water), rural communities, and the environment (see Chapter 9: Water Resources), increases in is justified. While many critical uncertainties exist pesticide usage to control changes in pest vectors, regarding the magnitude and timing of impacts, it and changes in water quality from agricultural appears that climate change is likely to affect U.S. chemicals. Decreases in biological diversity may agriculture significantly in the coming century. limit the adaptive capacity of agriculture, which requires a broad base of germ plasm for modifying Costs and Timing of Adjustment current crops and developing new ones (see Chapter 8: Biodiversity). Since our agricultural production system primarily consists of specialized farms producing A northward migration of agriculture would commodities in geographically specialized increase the use of irrigation and fertilizers on sandy production patterns, the costs of adjusting to soils, thus endangering underlying groundwater changed comparative advantage among agricultural quality. From South Dakota to southern Canada, regions, with ensuing changed resource use and critical prairie wetlands may be lost to drainage and changed agricultural infrastructure, may be quite conversion to cropland. Many of these areas are high in some regions. These shifts would also entail important wildlife habitats. Shifts in agricultural involvement of and costs to the federal government. activities may increase the susceptibility of soils to wind and water erosion. Climate change could thus If warming occurs rapidly, U.S. agriculture will exacerbate many of the current trends in have less time to adjust and costs may be greater. environmental pollution and resource use associated As climate continues to warm, costs may rise at an with agriculture as well as initiate new ones. increasing rate. Finally, unless CO₂ and other trace gas emissions are limited, we may be facing a Sea level rise, an associated impact of climate continual and possibly accelerating rate of change, will threaten low-lying coastal agricultural atmospheric accumulations and climate change. As regions with seasonal and in some instances 116 Agriculture permanent flooding, saltwater intrusion of forces as well as government programs would play freshwater aquifers and rivers, and salt a crucial role in creating the flexibility to respond to contamination of soils. Agricultural lands in coastal climate changes by sending signals on the efficient regions may be lost. (See Chapter 9: Water use of resources, and in mitigating their ultimate Resources, and Chapter 7: Sea Level Rise, for impact as they have done in the past. Agricultural linkages with agriculture.) policies should be evaluated to ensure that they are appropriate to both current and possible future Furthermore, climate change will act on conditions in regard to their ability to facilitate agriculture simultaneously with other environmental adaptation to climate change. For example, stresses. Levels of UV-B radiation caused by flexibility in shifting crop types and farm practices depletion of stratospheric ozone are likely to will speed adjustments. increase in the future, as are levels of tropospheric ozone and acid precipitation. The interactions Land-Use Programs among these multiple stresses and climate change need to be studied in agricultural settings. Federal legislation aimed at reducing the use of newly plowed grasslands, e.g., the "Sod-Buster Global Agriculture Bill," and the related "Swamp-Buster Bill," which restricts agricultural encroachment into wetlands Finally, U.S. agriculture is an integral part of subject to flooding and water-logging, are examples the global, international agricultural system. of new policies meant to protect marginal lands. Consequently, the adjustment of U.S. agriculture to The basic goals of these new laws, which are part of climate change cannot be considered in isolation the 1985 Farm Bill, are to protect the most erodible from the rest of the world. The optimal farmland by removing it from crop production and configuration of U.S. adjustments will depend very to use conservation as a tool for reducing much on how simultaneous changes in regional overproduction. Nearly 80 million acres of U.S. climates affect global agriculture and how other cropland were retired under these and other farm countries, in turn, respond to those changes. programs in 1988. Policy research should address how these programs may fare under changing climate conditions. POLICY IMPLICATIONS Another program established in the 1985 Farm Since climate change appears likely to Bill that may help alleviate the negative effects of reconfigure the agricultural activities and climate change is the Conservation Reserve demographics of rural America, policies should be Program. This program is aimed at removing from examined in light of these potential effects. crop production the cropland classified as "highly Agricultural policies should be designed to ease erodible" by the Soil Conservation Service. The bill adjustments to climate change and to ensure the created a new form of long-term contract of up to sustainability of our natural and human resources 10 years and provides payments to farmers who (see Schuh, Volume C, and Dudek, Volume C). apply conservation practices, such as maintaining a Following are specific policy areas that policymakers grass cover, on those acres. If successful, the could investigate to respond appropriately to the Conservation Reserve Program may reduce the projected climate change. impact of climate fluctuations on total grain production by taking the most sensitive lands out of Commodity Policies use. The 1988 drought, however, demonstrated that Agricultural pricing and production policies the Conservation Reserve Program may be difficult should promote efficient adjustment to the changing to maintain in the face of climate stress. As the conditions of global supply and demand induced by the greenhouse effect, which may include shifts in drought worsened during the summer, use of the set-aside lands was requested so that badly hit comparative advantage among regions and increased farmers could salvage some economic benefits from likelihood of droughts in some regions. Although these acres. Such conflicts may be more common these shifts may be slow, the cumulative effects may in the future, and land retirement strategies must be be large and they deserve close monitoring. Market 117 Chapter 6 weighed against possible needed increases in The frequency and magnitude of climate production. extremes may be altered with climate change. Responding to the changes may be costly for the Awareness of potential changes in agricultural government if crops fail frequently. The Drought land use due to regional climate change should be Relief Bill for the drought of 1988 is scheduled to built into land-use planning programs, especially in cost $3.9 billion to cover just 1 year of a climatic regions where agricultural activities may expand into extreme. On the other hand, some areas that natural, unmanaged ecosystems. Large-scale currently suffer from climate extremes may benefit drainage and water projects would need from climate change. Risk policy mechanisms for environmental impact studies to carefully assess this relief, recovery, and mitigation of climate change potential expansion of agricultural land (see should be examined so that they will be ready to Baldwin, Volume J). help farmers adjust. Water-Resource Management Programs A national drought policy is strongly needed to coordinate federal response to the possibility of Current water supply policies do not generally increased frequency and duration of future droughts encourage optimum water-use efficiency. A greater due to climate change. Even without climate degree of water efficiency should promote flexibility change, such a policy is needed not only for the in light of the potential for increased irrigation agricultural sector but also for other sectors. demands with climate change. Policies such as water transfers and markets should be considered International Trade Agreements for irrigated areas. Policies designed to ease the adjustment to Water Quality Policy greenhouse effects must be global in scope because the effects, although varied, are global in nature. The increased use of agricultural chemicals, Comparative advantage will likely shift significantly along with changes in the hydrological cycle, both within the United States and in other potentially threaten both soil and water supplies, countries. Population and economic activities also and eventually, public health. Negative would change geographically with climate change, consequences could be avoided or lessened by thus affecting the location of demand for including potential climate change effects in water agricultural products. It is already a goal of U.S. quality planning and by supporting alternative pest agricultural policy to incorporate global conditions management strategies that use such techniques as of supply and demand into the agricultural sector. biological control, genetic resistance, and innovative The potential seriousness of the impacts on the cropping systems. agricultural production system of the greenhouse effect may provide added incentive to establish such Risk Management and Drought Policy policies both nationally and internationally. The vulnerability of current and potential food-deficit Changes in the frequency, intensity, and regions to climate change should also be considered. location of extreme events are important for agriculture and the regional income that it produces. Agricultural Contributions to the The adequacy of the private crop insurance and Greenhouse Effect federal disaster payment programs should be assessed in the face of climatic uncertainty. For Agriculture itself is an active contributor to the example, only about 20 to 25% of potentially greenhouse effect. Clearing of forested land for insurable acreage is currently covered by crop agriculture often involves burning of trees and insurance. Farmers tend to rely on federal disaster shrubs that release CO₂. The biomass that is not relief programs to bail them out of such disasters as burned tends to decay gradually, also emitting CO2. droughts, floods, hail, and windstorms. Financial Agricultural activities release other radiatively active risk is also part of the credit structure that covers trace gases. Flooded rice fields emit methane land, equipment, and production in modern farming. (CH₄) as a product of the anaerobic decomposition of organic matter. Ruminants also release methane 118 Agriculture as a consequence of their digestive processes. In exports and the role of the United States as addition, soils may volatilize some of the a reliable supplier of agricultural export nitrogenous fertilizer applied to them in the form of commodities. nitrous oxide (N₂O). Finding effective ways to reduce these emissions presents a major challenge 2. Crop and livestock productivity -- Study the to the agricultural research community. In this interactive effects of climate variability and regard, the Conservation Reserve Program and change, CO2, tropospheric ozone, UV-B forestation efforts could provide a partial solution, from stratospheric ozone depletion, and since vegetation fixes CO₂ from the air. (See other environmental and societal variables Lashof and Tirpak, 1989, for further discussion of on agricultural productivity. Determine agriculture's contribution to the greenhouse effect.) how changed climatic variability may amplify or lessen the preliminary EPA Agricultural Research results. The agricultural research community should Because of the significant production enhance climate change research from the field level changes indicated by these studies, the need to the national policy level. It should continue to for better simulation of the direct effects of breed heat- and drought-resistant crop varieties and CO2 in the crop models, and the limited new crop species in preparation for global warming. adjustment studies performed, further crop Research in biotechnology may also be directed research should be conducted on a longer toward alleviating the negative effects of climate term basis. Necessary work includes change. Improved water-use and irrigation resolving the differences in forecasts of the efficiency also take on renewed importance in the GCMs, and designing more appropriate light of potential climate change. Energy scenarios including transient climate change requirements of the agricultural system under and changes in climatic variability. climate change should be defined, given the Physiologically based submodels are needed potential for increases in energy-intensive activities for the effects of increased CO2 on various such as irrigation and application of agricultural crops. The effects on other major crops chemicals. Research attention also should be such as cotton also should be studied. Crop directed toward reducing agricultural emissions of models should be improved in their trace gases. simulation of the effects of increasing temperatures. RESEARCH NEEDS Research on the direct CO2 effects on crops to this point has provided windows of knowledge concerning certain crops at 1. International agriculture -- Study the specific stages of their life cycles. Both the potential shifts in international comparative direct and the climate change effects of advantage and the vulnerability of food- deficit regions, and evaluate the high CO2 are probably quite different at different stages of development. Research implications of such shifts for the United should evaluate the interactive effects of States. CO2 and temperature over the whole life One of the most crucial areas for further cycle of the plant, with varying conditions of water and nutrition, rather than with plants research is the projection of potential under optimal conditions. Then crop climate change effects at the international response to the combined climatic and level. Potential changes in agricultural physiological effects of CO2 may be yields and production of major crops, and predicted more realistically. Much more impacts on regions that are food-deficient research on climate change and livestock now or that may become food-deficient in production is needed. Important research the future, all need to be studied. areas include crop-livestock interactions, Economics and policy research should reproduction, and diseases. consider the implications of shifts in global agriculture for the levels of U.S. crop 119 Chapter 6 3. Adaptation strategies -- Study the dynamic Adams, R.M., S.A. Hamilton, and B.A. McCarl. nature of climate change: What is the rate 1984. The Economic Effects of Ozone on of adaptation of regional agricultural Agriculture. Corvallis, OR: U.S. Environmental systems compared with the rate of climate Protection Agency. EPA-600/3-84-090. change? Evaluate the thresholds of sensitivity of U.S. agriculture. Studies Ames, David R. 1981. Effect of climate on should analyze the ability of various aspects livestock production data. In: Knapp, F.W., ed. of the agricultural production systems to Systems Approach to Animal Health and adapt to various rates and degrees of Production. Lexington, KY: University of climate change to determine these Kentucky. pp. 148-148. thresholds of sensitivity. It would also be useful to identify the costs of different types Callaway, J.M., F.J. Cronin, J.W. Currie, and J. of adjustments and the regions most likely Tawil. 1982. An analysis of methods and models to experience greater costs. for assessing the direct and indirect impacts of CO2- induced environmental changes in the agricultural 4. Agricultural economics -- Expand the sector of the U.S. economy. Richland, WA: Pacific national analysis to include crops and Northwest Laboratory, Battelle Memorial Institute. regions not now included (for example, PNL-4384. cotton and grasslands, and the western regions of the United States). Conduct Carbon Dioxide Assessment Committee. 1983. further analyses of regional shifts in Changing Climate. Washington, DC: National agriculture. Studies that link water Academy of Sciences. resource and agriculture models are needed to estimate changes in water demand Carsel, R.F., C.N. Smith, L.A. Mulkey, J.D. Dean, among agriculture and competing users. and P. Jowise. 1984. Users Manual for the Thus, estimates of actual changes in Pesticide Root Zone Model (PRZM). Athens, GA: irrigated acreage could be made. U.S. Environmental Protection Agency. EPA- 600/3-84-109. 5. Environmental impacts -- Elucidate the impacts of climate change on water Council for Agricultural Science and Technology. quantity, water quality, and other 1988. Long-Term Viability of U.S. Agriculture. components of the environment caused by Ames, IA: Council for Agricultural Science and shifts in crop and livestock production and Technology. Report No. 114. related industries. Cure, J.D. 1985. Carbon dioxide doubling 6. Agricultural emissions of trace gases -- responses: A crop survey. In: Strain, B.R., and J.D. Discover effective ways to reduce emissions Cure, eds. Direct Effects of Increasing Carbon of methane from livestock, nitrous oxide Dioxide on Vegetation. Washington, DC: U.S. from fertilizer application, and other Department of Energy. DOE/ER-0238. pp. 99- agricultural sources of trace gases. 116. Decker, W.L., V. Jones, and R. Achutuni. 1985. REFERENCES The impact of CO2-induced climate change on U.S. agriculture. In: White, M.R., ed. Characterization of Information Requirements for Studies of CO2 Acock, B., and L.H. Allen, Jr. 1985. Crop Effects: Water Resources, Agriculture, Fisheries, responses to elevated carbon dioxide concentrations. Forests and Human Health. Washington, DC: U.S. In: Strain, B.R., and J.D. Cure, eds. Direct Effects Department of Energy. DOE/ER-0236. pp. 69-93. of Increasing Carbon Dioxide on Vegetation. Washington, DC: U.S. Department of Energy. DOE/ER-0238. pp. 33-97. 120 Agriculture Dudek, D.J. 1987. Economic implications of Lashof, D., and D. Tirpak, eds. 1989. Policy climate change impacts on southern agriculture. In: Options for Stabilizing Global Climate. Draft Meo, M., ed. Proceedings of the Symposium on report. Washington, DC: U.S. Environmental Climate Change in the Southern United States: Protection Agency. Future Impacts and Present Policy Issues. Norman, OK: University of Oklahoma, Science and Public Parry, M.L., T.R. Carter, N.T. Konijn, eds. 1988. Policy Program. pp. 44-72. The Impact of Climatic Variations on Agriculture. Vol. 1. Assessments in Cool Temperate and Cold Drummond, R.O. 1987. Economic aspects of Regions. Dordrecht: Kluwer. ectoparasites of cattle in North America. In: Leaning, W.H.D., and J. Guerrero, eds. Proceedings Postel, S. 1986. Altering the Earth's Chemistry: of the MSD AGVET Symposium, The Economic Assessing the Risks. Worldwatch Paper 71. Impact of Parasitism in Cattle. XXIII. World Washington, DC: Worldwatch Institute. Veterinary Congress. Montreal, Quebec. pp. 9-24. Ritchie, J.T., and S. Otter. 1985. Description and Hansen, J., I. Fung, A. Lacis, D. Rind, G. Russell, performance of CERES-Wheat: A user-oriented S. Lebedeff, R. Ruedy, and P. Stone. 1988. Global wheat yield model. In: Willis, W.O., ed. ARS climate changes as forecast by the GISS 3-D model. Wheat Yield Project. Washington, DC: U.S. Journal of Geophysical Research 93(D8):9341-9364. Department of Agriculture, Agricultural Research Service. ARS-38. pp. 159-175. Jones, C.A., and J.R. Kiniry, eds. 1986. CERES- Maize: A Simulation Model of Maize Growth and Schneider, K. 1988. Drought cutting U.S. grain Development. College Station, TX: Texas A&M crop 31% this year. The New York Times August University Press. 12:A1. Jones, J.W., K.J. Boote, S.S. Jagtap, G. USDA. 1976. U.S. Department of Agriculture. Hoogenboom, and G.G. Wilkerson. 1988. Handbook of Agricultural Charts. Agricultural SOYGRO V5.41: Soybean Crop Growth Simulation Handbook 504. Washington, DC: U.S. Government Model. User's Guide. Florida Agr. Exp. Sta. Printing Office. Journal No. 8304, IFAS. Gainesville, FL: University of Florida. USDA. 1987. U.S. Department of Agriculture. Agricultural Statistics. Washington, DC: U.S. Jones, J.W., J.W. Mishoe, G. Wilkerson, J.L. Stimac, Government Printing Office. and W.G. Boggess. 1986. Integration of soybean crop and pest models. In: Frisbie, P.E., and P. U.S. Department of Commerce. 1983. U.S. Adisson, eds. Integrated Pest Management on Department of Commerce, Bureau of the Census. Major Agriculture Systems. Texas Agriculture 1982 Census of Agriculture, Vol. 1. Geographical Experiment Station. Publication No. MP-1616. Area Series, Part 51, United States Summary and College Station, TX: Texas A&M University. State Data. Washington, DC: U.S. Government Printing Office. Kimball, B.A. 1985. Adaptation of vegetation and management practices to a higher carbon dioxide Warrick, R.A., R.M. Gifford, M.L. Parry. 1986. world. In: Strain, B.R., and J.D. Cure, eds. Direct CO2, climatic change and agriculture. Assessing the Effects of Increasing Carbon Dioxide on Vegetation. response of food crops to the direct effects of Washington, DC: U.S. Department of Energy. increased CO₂ and climatic change. In: Bolin, B., DOE/ER-0238. pp. 185-204. B.R. Doos, J. Jager, and R.A. Warrick, eds. The Greenhouse Effect, Climatic Change and Land Evaluation Group. 1987. Implications of Ecosystems. A Synthesis of the Present Knowledge, climatic warming for Canada's comparative position SCOPE 29. New York: John Wiley and Sons. pp. in agricultural production and trade. Guelph, 393-473. Ontario: University of Guelph, University School of Rural Planning and Development. Webster, A.J.F. 1981. Weather and infectious disease in cattle. The Veterinary Record 108:183- 187. 121 Chapter 6 World Food Institute. 1987. World Food Trade Zachariah, K.C., and Vu, M.T. World Bank. 1988. and U.S. Agriculture, 1960-1986. Ames, IA: Iowa World Population Projections. 1987-1988 Ed. Short- State University. and Long-Term Estimates. Baltimore, MD: Johns Hopkins University Press. 122 CHAPTER 7 SEA LEVEL RISE FINDINGS The Southeast would bear approximately 90% of the land loss and 66% of the shore protection Global warming could cause sea level to rise 0.5 to costs. 2 meters by 2100. Such a rise would inundate wetlands and lowlands, erode beaches, exacerbate Policy Implications coastal flooding, and increase the salinity of estuaries and aquifers. Many of the necessary responses to sea level rise, such as rebuilding ports, constructing A 1-meter rise could drown approximately 25 to levees, and pumping sand onto beaches, need 80% of the U.S. coastal wetlands; ability to not be implemented until the rise is imminent. survive would depend largely on whether they On the other hand, the cost of incorporating sea could migrate inland or whether levees and level rise into a wide variety of engineering and bulkheads blocked their migration. Even land use decisions would be negligible compared current sea level trends threaten the wetlands of with the costs of not responding until sea level Louisiana. rises. A 1-meter rise could inundate 5,000 to 10,000 Many wetland ecosystems are likely to survive square miles of dryland if shores were not sea level rise only if appropriate measures are protected and 4,000 to 9,000 square miles of implemented in the near future. At the state dryland if only developed areas were protected. and local levels, these measures include land use planning, regulation, and redefinitions of Most coastal barrier island communities would property rights. The State of Maine has already probably respond to sea level rise by raising issued regulations to enable wetlands to migrate land with sand pumped from offshore. Wide landward by requiring that structures be and heavily urbanized islands may use levees, removed as sea level rises. while communities on lightly developed islands may adjust to a gradual landward migration of The coastal wetlands protected under Section the islands. 404 of the Clean Water Act will gradually be inundated. The act does not authorize measures Protecting developed areas against such to ensure survival of wetland ecosystems as sea inundation and erosion by building bulkheads level rises. and levees, pumping sand, and raising barrier islands could cost $73 to $111 billion The National Flood Insurance Program may (cumulative capital costs in 1985 dollars) for a wish to consider the implications of sea level 1-meter rise by the year 2100 (compared with rise on its future liabilities. A recent HUD $6 to $11 billion under current sea level trends). authorization act requires this program to Of this total, $50 to $75 billion would be spent purchase property threatened with erosion. The (cumulative capital costs in 1985 dollars) to act may imply a commitment by the federal elevate beaches, houses, land, and roadways by government to compensate property owners for the year 2100 to protect barrier islands losses due to sea level rise. (compared with $4 billion under current trends). Developed barrier islands would likely be The need to take action is particularly urgent in protected from sea level rise because of their coastal Louisiana, which is already losing 100 high property values. square kilometers per year. 123 Chapter 7 CAUSES, EFFECTS, AND gulf coasts, to a slight drop in much of the Pacific Northwest (Figure 7-1). Areas such as Louisiana RESPONSES provide natural laboratories for assessing the possible effects of future sea level rise (Lyle et al., Global warming from the greenhouse effect 1987). could raise sea level approximately 1 meter by expanding ocean water, melting mountain glaciers, and causing ice sheets in Greenland to melt or slide TIME, YEARS into the oceans. Such a rise would inundate coastal 1850 1865 1880 1895 1910 1925 1940 1955 1970 1985 wetlands and lowlands, erode beaches, increase the risk of flooding, and increase the salinity of SITKA, AK estuaries, aquifers, and wetlands. In the last 5 years, many coastal communities throughout the world have started to prepare for the 7.0 possibility of such a rise. In the United States, NEW YORK, NY Maine has enacted a policy declaring that shorefront 6.6 buildings will have to be moved to enable beaches and wetlands to migrate inland to higher ground. SCALE, FEET 6.2 Maryland has shifted its shore-protection strategy CHARLESTON, SC from a technology that can not accommodate sea MIAMI BEACH, when FL 5.8 level rise to one that can. Seven coastal states have held large public meetings on how to prepare for a 5.4 rising sea. Australia, the Netherlands, and the Republic of Maldives are beginning to undergo a 5.0 similar process. Causes GALVESTON, TX Ocean levels have always fluctuated with changes in global temperatures. During the ice ages when the earth was 5°C (9°F) colder than today, Figure 7-1. Time series graph of sea level trends much of the ocean's water was frozen in glaciers for New York, Charleston, Miami, Galveston, and and sea level often was more than 100 meters (300 Sitka (Lyle et al., 1987). feet) below the present level (Donn et al., 1962; Kennett, 1982; Oldale, 1985). Conversely, during the last interglacial period (100,000 years ago) when Global sea level trends have generally been the average temperature was about 1°C (2°F) estimated by combining the trends at tidal stations warmer than today, sea level was approximately 20 around the world. Studies combining these feet higher than the current sea level (Mercer, measurements suggest that during the last century, 1968). worldwide sea level has risen 10 to 15 centimeters (4 to 6 inches) (Barnett, 1984; Fairbridge and When considering shorter periods of time, Krebs, 1962). Much of this rise has been attributed worldwide sea level rise must be distinguished from to the global warming that has occurred during the relative sea level rise. Although climate change last century (Meier, 1984; Gornitz et al., 1982). alters worldwide sea level, the rate of sea level rise Hughes (1983) and Bentley (1983) estimated that a relative to a particular coast has greater practical complete disintegration of West Antarctica in importance and is all that monitoring stations can response to global warming would require a 200- to measure. Because most coasts are sinking (and a 500-year period, and that such a disintegration few are rising), the range of relative sea level rise would raise sea level 20 feet. Most recent varies from more than 3 feet per century in assessments, however, have focused on the likely Louisiana and parts of California and Texas to 1 rise by the year 2100. Figure 7-2 illustrates recent foot per century along most of the Atlantic and estimates of sea level rise, which generally fall into the range of 50 to 200 centimeters. 124 Sea Level Rise purposes. In many cases, the responses to sea level 4.0 rise are sufficiently well established and the probability of no response is sufficiently low that it would be misleading to discuss the potential effects Hoffman (1983) High without also discussing responses. For example, much of Manhattan Island is less than 2 meters 3.0 above high tide; the effect of sea level rise would almost certainly be the increased use of coastal SEA LEVEL RISE RELATIVE TO 1986 (Meters) engineering structures and not the inundation of downtown New York. Glacier Volume Estimate of Polar Hoffman (1983) Mid-High A rise in sea level would inundate wetlands and 2.0 Board Augmented With Thermal Expansion Estimates by NRC Meier (1985b) High lowlands, accelerate coastal erosion, exacerbate (1983) coastal flooding, threaten coastal structures, raise WMO (1986) High water tables, and increase the salinity of rivers, Hoffman (1983) Mid-Low III bays, and aquifers (Barth and Titus, 1984). Most of the wetlands and lowlands are found along the gulf 1.0 II coast and along the Atlantic coast south of central Revelle (1983) New Jersey, although a large area also exists around Hoffman (1983) Low San Francisco Bay. Similarly, the areas vulnerable I Meier (1985b) Low* to erosion and flooding are also predominately in Past Century Estimated WMO (1986) Low the Southeast; potential salinity problems are spread 0.12 m Rise 0.0 more evenly along the U.S. Atlantic coast. We now 2000 2050 2100 YEAR discuss some of the impacts that would result if no responses were initiated to address sea level rise. Figure 7-2. Estimates of future sea level rise (derived from Hoffman, 1983, 1986; Meier, 1985; Destruction of Coastal Wetlands Revelle, 1983). Coastal wetlands are generally found between the highest tide of the year and mean sea level. Although most studies have focused on the Wetlands have kept pace with the past rate of sea impact of global warming on global sea level, the level rise because they collect sediment and produce greenhouse effect would not necessarily raise sea peat upon which they can build; meanwhile, they level by the same amount everywhere. Removal of expanded inland as lowlands were inundated (Figure water from the world's ice sheets would move the 7-3). Wetlands accrete vertically and expand inland. earth's center of gravity away from Greenland and Thus, as Figure 7-3 illustrates, the present area of Antarctica and would thus redistribute the oceans' wetlands is generally far greater than the area that water toward the new center of gravity. Along the would be available for new wetlands as sea level U.S. coast, this effect would generally increase sea rises (Titus et al., 1984b; Titus, 1986). The level rise by less than 10%. Sea level could actually potential loss would be the greatest in Louisiana drop, however, at Cape Horn and along the coast of (see Chapter 16: Southeast). Iceland. Climate change could also affect local sea level by changing ocean currents, winds, and In many areas, people have built bulkheads just atmospheric pressure; no one has estimated these above the marsh. If sea level rises, the wetlands will impacts. be squeezed between the sea and the bulkheads (see Figure 7-3). Previous studies have estimated that if Effects the development in coastal areas were removed to allow new wetlands to form inland, a 1.5- to 2-meter In this section and in the following sections, the rise would destroy 30 to 70% of the U.S. coastal effects of and responses to sea level rise are wetlands. If levees and bulkheads were erected to presented separately. However, the distinction is protect today's dryland, the loss could be 50 to 80% largely academic and is solely for presentation (Titus, 1988; Armentano et al., 1988). 125 Chapter 7 5000 YEARS AGO TODAY D CURRENT X SEA LEVEL SEA LEVEL SEDIMENTATION AND PAST PEAT FORMATION SEA LEVEL FUTURE COMPLETE WETLAND LOSS WHERE HOUSE IS PROTECTED SUBSTANTIAL WETLAND LOSS WHERE THERE IS VACANT UPLAND IN RESPONSE TO RISE IN SEA LEVEL FUTURE FUTURE SEA LEVEL X SEA LEVEL CURRENT CURRENT SEA LEVEL SEA LEVEL PEAT ACCUMULATION Figure 7-3. Evolution of marsh as sea rises. Coastal marshes have kept pace with the slow rate of sea level rise that has characterized the last several thousand years. Thus, the area of marsh has expanded over time as new lands have been inundated. If in the future, sea level rises faster than the ability of the marsh to keep pace, the marsh area will contract. Construction of bulkheads to protect economic development may prevent new marsh from forming and result in a total loss of marsh in some areas. Such a loss would reduce the available habitat for barriers are generally long narrow islands and spits birds and juvenile fish and would reduce the with the ocean on one side and a bay on the other. production of organic materials on which estuarine Typically, the oceanfront block of an island ranges fish rely. from 5 to 10 feet above high tide, and the bay side is 2 to 3 feet above high water. Thus, even a 1- The dryland within 2 meters of high tide meter sea level rise would threaten much of this includes forests, farms, low parts of some port cities, valuable land with inundation. cities that sank after they were built and are now protected with levees, and the bay sides of barrier Erosion threatens the high part of these islands islands. The low forests and farms are generally in and is generally viewed as a more immediate the mid-Atlantic and Southeast regions; these would problem than the inundation of the bay sides. As provide potential areas for new wetland formation. Figure 7-4 shows, a rise in sea level can cause an Major port cities with low areas include Boston, ocean beach to retreat considerably more than it New York, Charleston, and Miami. New Orleans is would from the effects of inundation alone. The generally 8 feet below sea level, and parts of visible part of the beach is much steeper than the Galveston, Texas City, and areas around the San underwater portion, which comprises most of the Francisco Bay are also well below sea level. active "surf zone." While inundation alone is Because they are already protected by levees, these determined by the slope of the land just above the cities are more concerned with flooding than with water, Bruun (1962) and others have shown that the inundation. total shoreline retreat from a sea level rise depends on the average slope of the entire beach profile. Inundation and Erosion of Beaches and Barrier Islands Previous studies suggest that a 1-foot rise in sea level would generally cause beaches to erode 50 to Some of the most important vulnerable areas 100 feet from the Northeast to Maryland (e.g., are the recreational barrier islands and spits Kyper and Sorensen, 1985; Everts, 1985); 200 feet (peninsulas) of the Atlantic and gulf coasts. Coastal along the Carolinas (Kana et al., 1984); 100 to 1,000 126 Sea Level Rise upon. A 1-meter sea level rise would enable a 15-year storm to flood many areas that today are flooded only by a 100-year storm (e.g., Kana et al., 1984; Leatherman, 1984). (2) Beach erosion also would leave oceanfront properties more vulnerable A to storm waves. (3) Higher water levels would reduce coastal drainage and thus would increase flooding attributable to rainstorms. In artificially drained areas such as New Orleans, the increased need for pumping could exceed current capacities. (4) Finally, a rise in sea level would raise water PREVIOUS SEA LEVEL tables and would flood basements, and in cases B where the groundwater is just below the surface, perhaps raise it above the surface. Saltwater Intrusion A rise in sea level would enable saltwater to penetrate farther inland and upstream into rivers, bays, wetlands, and aquifers. Salinity increases C would be harmful to some aquatic plants and animals, and would threaten human uses of water. Figure 7-4. The Bruun Rule: (A) initial condition; For example, increased salinity already has been (B) immediate inundation when sea level rises; (C) cited as a factor contributing to reduced oyster subsequent erosion due to sea level rise. A rise in harvests in the Delaware and Chesapeake Bays, and sea level immediately results in shoreline retreat to conversion of cypress swamps to open lakes in due to inundation, shown in the first two examples. Louisiana. Moreover, New York, Philadelphia, and However, a 1-meter rise in sea level implies that much of California's Central Valley obtain their the offshore bottom must also rise 1 meter. The water from areas located just upstream from areas sand required to raise the bottom (X') can be where the water is salty during droughts. Farmers supplied by beach nourishment. Otherwise, waves in central New Jersey and the city of Camden rely will erode the necessary sand (X) from upper part on the Potomac-Raritan-Magothy Aquifer, which of the beach as shown in (C). could become salty if sea level rises (Hull and Titus, 1986). The South Florida Water Management District already spends millions of dollars every year feet along the Florida coast (Bruun, 1962); 200 to to prevent Miami's Biscayne Aquifer from becoming 400 feet along the California coast (Wilcoxen, 1986); contaminated with seawater. and perhaps several miles in Louisiana. Because most U.S. recreational beaches are less than 100 Responses feet wide at high tide, even a 1-foot rise in sea level would require a response. In many areas, The possible responses to inundation, erosion, undeveloped barrier islands could keep up with and flooding fall broadly into three categories: rising sea level by "overwashing" landward. In erecting walls to hold back the sea, allowing the sea Louisiana, however, barrier islands are breaking up to advance and adapting to the advance, and raising and exposing the wetlands behind them to gulf the land. Both the slow rise in sea level over the waves; consequently, the Louisiana barrier islands last thousand years and the areas where land has have rapidly eroded. been sinking more rapidly offer numerous historical examples of all three responses. Flooding For over five centuries, the Dutch and others If sea level rises, flooding would increase along have used dikes and windmills to prevent inundation the coast for four reasons: (1) A higher sea level from the North Sea. By contrast, many cities have provides a higher base for storm surges to build been rebuilt landward as structures have eroded; the 127 Chapter 7 town of Dunwich, England, has rebuilt its church Most of the measures for counteracting seven times in the last seven centuries. More saltwater intrusion attributable to sea level rise have recently, rapidly subsiding communities (e.g., also been employed to address current problems. Galveston, Texas) have used fill to raise land For example, the Delaware River Basin Commission elevations; the U.S. Army Corps of Engineers and protects Philadelphia's freshwater intake on the coastal states regularly pump sand from offshore river and New Jersey aquifers recharged by the river locations to counteract beach erosion. Venice, a by storing water in reservoirs during the wet season hybrid of all three responses, has allowed the sea to and releasing it during droughts, thereby forcing the advance into the canals, has raised some lowlands, saltwater back toward the sea. Other communities and has erected storm protection barriers. have protected coastal aquifers by erecting underground barriers and by maintaining freshwater Most assessments in the United States have pressure through the use of impoundments and concluded that low-lying coastal cities would be injection wells. protected with bulkheads, levees, and pumping systems, and that sparsely developed areas would adapt to a naturally retreating shoreline (e.g., Dean et al., 1987; Gibbs, 1984; Schelling, 1983). This HOLDING BACK THE SEA: A conclusion has generally been based on estimates NATIONAL ASSESSMENT that the cost of structural protection would be far less than the value of the urban areas being The studies referenced in the previous section protected but would be greater than the value of have illustrated a wide variety of possible effects undeveloped land. from and responses to a rise in sea level from the greenhouse effect. Although they have identified Studies on the possible responses of barrier the implications of the risk of sea level rise for islands and moderately developed mainland specific locations and decisions, these studies have communities show less agreement but generally not estimated the nationwide magnitude of the suggest that environmental factors would be as impacts. This report seeks to fill that void. important as economics. Some have suggested that barrier islands should use seawalls and other "hard" It was not possible to estimate the nationwide engineering approaches (e.g., Kyper and Sorensen, value of every impact of sea level rise. The studies 1985; Sorensen et al., 1984). Others have pointed to thus far conducted suggest that the majority of the the esthetic problems associated with losing beaches environmental and economic costs would be and have advocated a gradual retreat from the shore associated with shoreline retreat and measures to (Howard et al., 1985). Noting that new houses on hold back the sea, which can be more easily barrier islands are generally elevated on pilings, assessed on a nationwide basis. Because the Titus (1986) suggested that communities could hold eventual impact will depend on what people actually back the sea but keep a natural beach by extending do, a number of important questions can be the current practice of pumping sand onto beaches addressed within this context: to raising entire islands in place. Would a gradual abandonment of Responses to erosion are more likely to have moderately developed mainland areas adverse environmental impacts along sheltered significantly increase the amount of water than on the open coast (Titus, 1986). wetlands that survived a rise in sea level? Because the beach generally is a barrier island's most important asset, economics would tend to Would the concave profiles of coastal areas encourage these communities to preserve their ensure that more wetlands would be lost natural shorelines; actions that would prevent the than gained, regardless of land-use island from breaking up also would protect the decisions? adjacent wetlands. However, along most mainland shorelines, economic self-interest would encourage Should barrier islands be raised in place by property owners to erect bulkheads; these would pumping sand and elevating structures and prevent new wetland formation from offsetting the utilities? loss of wetlands that were inundated. 128 Sea Level Rise Park Weggel Leatherman Yohe Cost of Cost to Value Loss of Cost Levees Rebuild of Wet/Dry Elevations of Erosion Island Infra- Threatened Land Sand Retreat structure Property Decision to Use Island Raising Scenario CASE STUDY NATIONAL ANALYSIS Park Weggel Leatherman Titus Cost Area of Elevations Assumptions Barrier Islands Cost of Nourishing Cost of Loss of Beaches and Raising Non-Sand Cost Protecting Wet/Dry Barrier Elevations of Raising Sheltered Coasts Land (assuming fixed costs Coastal Barriers 2-m Scenario of sand) Titus Titus Titus Confidence Cost of Protecting Intervals Sheltered Shores Increasing Sand 50 and 100 cm Cost Scenario for Wetland Loss Scenarios Figure 7-5. Overview of sea level rise studies and authors. Would a landward migration of developed STRUCTURE OF STUDIES FOR barrier islands or encircling them with dikes and levees be feasible alternatives? THIS REPORT How much property would be lost if barrier A central theme underlying these questions is islands were abandoned? that the implications of sea level rise for a community depend greatly on whether people adjust to the natural impact of shoreline retreat or 129 Chapter 7 undertake efforts to hold back the sea. Because no noted that the cost of this option would be one knows the extent to which each of these considerably less than the resources that would be approaches would be applied, this study was lost if the islands were not protected as shown in designed to estimate the impacts of sea level rise for Figure 7-6. (1) holding back the sea, and (2) natural shoreline retreat. Once the case study was complete, Park, Leatherman, and Weggel proceeded independently The tasks were split into five discrete projects: with their studies (although Park provided Weggel with elevation data). When those studies 1. Park et al. estimated the loss of coastal were complete, Titus synthesized their results, wetlands and dryland. developing a nationwide estimate of the cost of holding back the sea and interpolating Weggel's 2. Leatherman estimated the cost of pumping 200-centimeter results for the 50- and 100- sand onto open coastal beaches and barrier centimeter scenarios. islands. In presenting results from the Park and Weggel 3. Weggel et al. estimated the cost of studies, the sites were grouped into seven coastal protecting sheltered shores with levees and regions, four of which are in the Southeast: New bulkheads. England, mid-Atlantic, south Atlantic, south Florida/gulf coast peninsula, Louisiana, other gulf 4. Yohe began a national economic (Texas, Mississippi, Alabama, Florida Panhandle), assessment by estimating the value of and the Pacific coast. Figure 7-7 illustrates these threatened property. regions. 5. Titus and Greene synthesized the results of other studies to estimate ranges of the 200 175 nationwide impacts. Lost Rent 150 From Not 125 Raising the Island 100 Figure 7-5 illustrates the relationships between 90 the various reports. (All of the sea level rise studies 80 70 are in Volume B of the Appendices to this report.) 60 50 As the top portion shows, the assessment began 40 with a case study of Long Beach Island, New Jersey, 30 which was necessary for evaluating methods and Millions 25 providing data for purposes of extrapolation. The Costs of Raising Island Park and Leatherman studies performed the same 20 calculations for the case study site that they would 15 subsequently perform for the other sites in the nationwide analysis. However, Weggel and Yohe 10 conducted more detailed assessments of the case study whose results were used in the Leatherman 5 and Titus studies. 0 1980 2000 2020 2040 2060 2080 2100 Year Because it would not be feasible for Leatherman to examine more than one option for Figure 7-6. Annual cost of raising island versus the cost of protecting the open coast, Weggel annual costs (lost rent) from not protecting the estimated the cost of protecting Long Beach Island island (in 1986 dollars) (Titus and Greene, Volume by three approaches: (1) raising the island in place; B). (2) gradually rebuilding the island landward; and (3) encircling the island with dikes and levees. Yohe estimated the value of threatened structures. Titus SCENARIOS OF SEA LEVEL RISE analyzed Weggel's and Yohe's results and concluded that raising barrier islands would be the most reasonable option for the Leatherman study and Although the researchers considered a variety of scenarios of future sea level rise, this report focuses 130 Sea Level Rise NORTHEAST MID ATLANTIC SOUTH ATLANTIC WEST COAST OTHER GULF LOUISIANA OTHER GULF SOUTHERN WEST FLORIDA Figure 7-7. Coastal regions used in this study. on the impacts of three scenarios: rises of 50, 100, A rise in sea level greater than the rate of and 200 centimeters by the year 2100. All three of vertical wetland accretion would result in these scenarios are based on quantitative estimates a net loss of coastal wetlands. of sea level rise. No probabilities were associated with these scenarios. Following the convention of a The loss of wetlands would be greatest if all recent National Research Council report (Dean et developed areas were protected, less if al., 1987), the rise was interpolated throughout the shorelines retreated naturally, and least if 21st century using a quadratic (parabola). For each barrier islands were protected while site, local subsidence was added to determine mainland shores retreated naturally. relative sea level rise. Figure 7-8 shows the scenarios for the coast of Florida where relative sea The loss of coastal wetlands would be level rise will be typical of most of the U.S. coast. greatest in the Southeast, particularly Sea level would rise 1 foot by 2025, 2040, and 2060 Louisiana. for the three scenarios and 2 feet by 2045, 2065, and 2100. Study Design Park's study was based on a sample of 46 RESULTS OF SEA LEVEL coastal sites that were selected at regular intervals. This guaranteed that particular regions would be STUDIES IN THIS REPORT represented in proportion to their total area in the coastal zone. The sites chosen accounted for 10% Loss of Coastal Wetlands and Dryland of the U.S. coastal zone excluding Alaska and Hawaii. To estimate the potential loss of wet and dry land, Park first had to characterize their Park (Volume B) sought to test a number of elevations. For wetlands, he used satellite imagery hypotheses presented in previous publications: to determine plant species for 60- by 80-meter parcels. Using estimates from the literature on the 131 Chapter 7 and assumed that mangroves would begin to replace 2.5 marsh after that year. 200cm Limitations 2.0 The greatest uncertainty in Park's analysis is a poor understanding of the potential rates of vertical accretion. Although this could substantially affect 1.5 SEA LEVEL RISE (Meters) the results for low sea level rise scenarios, the practical significance is small for a rise of 1 meter 100cm because it is generally recognized that wetlands 1.0 could not keep pace with the rise of 1 to 2 centimeters per year that such a scenario implies for 0.70 2 Ft. 50cm the second half of the 21st century. 0.5 1 Ft. 0.35 Errors can be made when determining Baseline vegetation type based on the use of infrared "signatures" that satellites receive. Park noted, for 0.0 1986 2000 2050 2100 example, that in California the redwoods have a YEAR signature similar to that of marsh grass. For only a Figure 7-8. Sea level scenarios (Miami Beach). few sites, Park was able to corroborate his estimates of vegetation type. frequency of flooding that can be tolerated by Park's study did not consider the potential various wetland plants, Park determined the implications of alternative methods of managing percentage of time that particular parcels are riverflow. This limitation is particularly serious currently under water. From this, Park inferred regarding application to Louisiana, where widely wetland elevation based on the known tidal range. varying measures have been proposed to increase For dryland, he used spot elevation measurements the amount of water and sediment delivered to the to interpolate between contours on U.S. Geological wetlands. Finally, the study makes no attempt to Survey topographic maps. predict which undeveloped areas might be developed in the next century. Park estimated the net loss of wetlands and dryland for no protection, protection of developed At the coarse (500-meter) scale Park used, the areas, and protection of all shores. For the assumption of protecting only developed areas no-protection scenario, estimating the loss of amounts to not protecting a number of mainland dryland is straightforward. However, for calculating areas where the shoreline is developed but areas net wetland loss, Park had to estimate the loss of behind the shoreline are not. Therefore, Park's existing wetlands as well as the creation of new estimates for protecting developed areas should be wetlands. For calculating losses, Park used interpreted as applying to the case where only published vertical accretion rates (see Armentano et densely developed areas are protected. Finally, al., 1988), although he allowed for some Park's assumption that dryland would convert to acceleration of vertical accretion in areas with vegetated wetlands within 5 years of being ample supplies of sediment, such as tidal deltas. inundated probably led him to underestimate the Park assumed that dryland would convert to net loss of wetlands due to sea level rise. wetlands within 5 years of being inundated. Results For sites in the Southeast, Park also allowed for the gradual replacement of salt marshes by Park's results supported the hypotheses mangrove swamps. The upper limit for mangroves suggested by previous studies. Figure 7-9 shows is around Fort Lauderdale. Park used the GISS nationwide wetlands loss for various (0- to 3-meter) transient scenario to determine the year particular sea level rises for the three policy options sites would be as warm as Fort Lauderdale is today investigated. For a 1-meter rise, 66% of all coastal wetlands would be lost if all shorelines were protected, 49% would be lost if only developed 132 Sea Level Rise areas were protected, and 46% would be lost if Figure 7-10 illustrates Park's estimates of the shorelines retreated naturally. inundation of dryland for the seven coastal regions. If shorelines retreated naturally, a 1-meter rise As expected, the greatest losses of wetlands would inundate 7,700 square miles of dryland, an would be in the Southeast, which currently contains area the size of Massachusetts. Rises of 50 and 200 85% of U.S. coastal wetlands (Figure 7-9). For a 1- centimeters would result in losses of 5,000 and meter sea level rise, 6,000 to 8,600 square miles 12,000 square miles, respectively. Approximately (depending on which policy is implemented) of U.S. 70% of the dryland losses would occur in the wetlands would be lost; 90 to 95% of this area Southeast, particularly Florida, Louisiana, and North would be in the Southeast, and 40 to 50% would be Carolina. The eastern shores of the Chesapeake and in Louisiana alone. By contrast, neither the Delaware Bays also would lose considerable Northeast nor the West would lose more than 10% acreage. of its wetlands if only currently developed areas are protected. Costs of Defending Sheltered Shorelines Study Design COMPOSITE FOR UNITED STATES ALL DRYLAND PROTECTED This study began by examining Long Beach 100 Island in depth. This site and five other sites were SALT MARSH used to develop engineering rules of thumb for the 80 BEACH/FLAT PERCENT OF 1986 WETLAND AREAS @@@@@@@@@@@ cost of protecting coastal lowlands from inundation. 60 MANGROVE Examining the costs of raising barrier islands 40 FRESH MARSH required an assessment of two alternatives: (1) SWAMP 20 building a levee around the island; and (2) allowing the island to migrate landward. 0 0.0 0.1 0.3 0.6 1.0 1.5 2.2 3.0 SEA LEVEL RISE (Meters) After visiting Long Beach Island and the adjacent mainland, Weggel (Volume B) designed DEVELOPED AREAS PROTECTED and estimated costs for an encirclement scheme consisting of a levee around the island and a 100 drainage system that included pumping and 80 PERCENT OF 1986 WETLAND AREAS underground retention of stormwater. For island 60 migration, he used the Bruun Rule to estimate 40 oceanside erosion and navigation charts to calculate the amount of sand necessary to fill the bay an 20 equivalent distance landward. For island raising and 0 0.0 0.1 0.3 0.6 1.0 1.5 2.2 3.0 island migration, Weggel used the literature to SEA LEVEL RISE (Meters) estimate the costs of elevating and moving houses and of rebuilding roads and utilities. NO PROTECTION Weggel's approach for estimating the nationwide 100 costs was to examine a number of index sites in 80 PERCENT OF 1986 WETLAND AREAS depth and thereby develop generalized cost 60 estimates for protecting different types of shorelines. 40 He used the topographic information collected by 20 Park for a sample of 95 sites to determine the area 0 and shoreline length that had to be protected. He 0.0 0.1 0.3 0.6 1.0 1.5 2.2 3.0 then applied the cost estimation factors to each site SEA LEVEL RISE (Meters) and extrapolated the sample to the entire coast. Figure 7-9. Nationwide wetlands loss for three After assessing Long Beach Island, Weggel shoreline-protection options. Note: These wetlands conducted less detailed studies of the following include beaches and flats that are not vegetated areas: metropolitan New York; Dividing Creek, wetlands; however, results cited in the text refer to vegetated wetlands (Park, Volume B). 133 Chapter 7 A. DRYLAND LOSS BY 2100 WITHOUT SHORE PROTECTION 3.0 2.8 2.6 2.4 2.2 2.0 LOSS OF DRYLAND (THOUSANDS OF SQ. MILES) 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 IIIIIIII 0.0 Northeast Mid- South South Louisiana Other Gulf West Atlantic Atlantic & West Florida B. DRYLAND LOSS BY 2100 WITH PROTECTION OF DEVELOPED AREAS 3.0 2.8 2.6 2.4 2.2 2.0 LOSS OF DRYLAND (THOUSANDS OF SQ. MILES) 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Northeast Mid- South South Louisiana Other Gulf West Atlantic Atlantic & West Florida SEA LEVEL RISE BASELINE 50 CM 100 CM 200 CM SCENARIO: Figure 7-10. Loss of dryland by 2100: (A) if no areas are protected, and (B) if developed areas are protected with levees (derived from Park, Volume B; see also Titus and Greene, Volume B). New Jersey; Miami and Miami Beach; the area Finally, Weggel was able to examine only one around Corpus Christi, Texas; and parts of San scenario: a 2-meter rise by 2100. This scenario was Francisco Bay. chosen over the more likely 1-meter scenario because an interpolation from 2 meters to 1 meter Limitations would be more reliable than an extrapolation from 1 meter to 2 meters. (See the discussion of Titus The most serious limitation of the Weggel study and Greene for results of the interpolation.) is that cruder methods are used for the national assessment than for the index sites. Even for the Results index sites, the cost estimates are based on the literature, not on site-specific designs that take into Case Study of Long Beach Island consideration wave data for bulkheads and potential savings from tolerating substandard roads. Weggel Weggel's cumulative cost estimates clearly did not estimate the cost of pumping rainwater out indicate that raising Long Beach Island would be of areas protected by levees. much less expensive ($1.7 billion) than allowing it to 134 Sea Level Rise migrate landward ($7.7 billion). Although the cost Nationwide Costs of building a levee around the island ($800 million) would be less, the "present value" would be greater. Table 7-2 shows Weggel's estimates for the Weggel concluded that the levee would have to be index sites and his nationwide estimate. The index built in the 2020s, whereas the island could be sites represent two distinct patterns. Because urban raised gradually between 2020 and 2100. Thus, the areas such as New York and Miami would be (discounted) present value of the levee cost would entirely protected by levees, the cost of moving be greater, and raising the necessary capital for a buildings and rebuilding roads and utilities would be levee at any one time could be more difficult than relatively small. On the other hand, Weggel gradually rebuilding the roads and elevating houses concluded that in more rural areas such as Dividing as the island was raised. Moreover, a levee would Creek, New Jersey, only the pockets of development eliminate the waterfront view. A final disadvantage would be protected. The roads that connected them of building a levee is that one must design for a would have to be elevated or replaced with bridges, specific magnitude of sea level rise; by contrast, an and the small number of isolated buildings would island could be raised incrementally. have to be moved. The Weggel analysis shows that landward Weggel estimates that the nationwide cost of migration is more expensive than island raising, protecting developed shorelines would be $25 primarily because of the increased costs of billion, assuming bulkheads are built, and $80 billion rebuilding infrastructure. Thus, migration might be assuming levees are built. Unlike wetlands loss, the less expensive in the case of a very lightly developed cost of protecting developed areas from the sea island. Levees might be more practical for wide would be concentrated more in the Northeast than barrier islands where most people do not have a in the Southeast because a much greater portion of waterfront view. the southeastern coast is undeveloped. Table 7-1. Total Cost of Protecting Long Beach Island from a 2-Meter Rise in Sea Level (millions of 1986 dollars) Island Island Protective measure Encirclement raising migration Sand costs: Beach 290 290 0 Land creation/maintenance NA 270 321 Moving/elevating houses NA 74 37 Roads/utilities 0 1072 7352 Levee and drainage 542 0 0 Total 832 1706 7710 NA = Not applicable. Source: Leatherman (Volume B); Weggel (Volume B) 135 Chapter 7 Table 7-2. Cumulative Cost of Protecting Sheltered Waters for a 2-Meter Rise in Sea Level (millions of 1986 dollars) New Raise old Move Roads/ bulkhead bulkhead building utilities Total Index sites New York 57 205 0.5 9.5 272.3 Long Beach Island 3 4 2.7 3.8 13.7 Dividing Creek 4 6 4.8 18.2 33.0 Miami area 11 111 0.3 8.3 130.7 Corpus Christi 11 29 2.8 40.9 83.4 San Francisco Bayᵃ 3 19 2.0 20.0 44.0 Nationwide estimate Low High Northeast 6,932 23,607 Mid-Atlantic 4,354 14,603 Southeast 9,249 29,883 West 4,097 12,802 Nation 24,633 80,176 Site names refer to the name of U.S. Geological Survey quadrant, not to the geographical area of the same name. Source: Weggel et al. (Volume B). Case Study of the Value of Threatened "lost" when the house was within 40 feet of the Coastal Property spring high tide mark. (See Titus and Greene, Volume B, for discussion.) Study Design Limitations Yohe's (Volume B) objective was to estimate Yohe's results for a sea level rise of less than 18 the loss of property that would result from not holding back the sea. Using estimates of erosion inches are sensitive to the assumption regarding and inundation for Long Beach Island from when a property would be lost. On the bay side, Leatherman and Park et al., Yohe determined people might learn to tolerate tidal inundation. which land would be lost from sea level rise for a Unless a major storm occurred, people could sample of strips spanning the island from the ocean probably occupy oceanfront houses until they were to the bay. He then used the Ocean County, New flooded at high tide. However, the resulting loss of Jersey, tax assessor's estimates of the value of the recreational use of the beach probably would have land and structures that would be lost, assuming a greater impact than abandoning the structure. that the premium associated with a view of the bay Tax maps do not always provide up-to-date or ocean would be transferred to another property estimates of property values. However, the distinction between the tax assessor's most recent owner and not lost to the community. He estimated estimate of market value and the current market the annual stream of rents that would be lost by assuming that the required return on real estate is value is small compared with the possible changes 10% after tax. Yohe assumed that a property on in property values that will occur over the next the bay side was "lost" whenever it was flooded at century; hence, Titus and Greene used tax assessors estimates of market values. high tide, and that property on the ocean side was 136 Sea Level Rise Results also examined one representative site in each of the remaining states. Yohe's results suggest that the cost of gradually raising Long Beach Island would be far less than Limitations the value of the resources that would be protected. Figure 7-6 compares Yohe's estimates of the annual Although the samples of sites in the Northeast loss in rents resulting from not holding back the sea and Northwest are representative, complete with Weggel's estimates of the annual cost of raising coverage would have been more accurate. the island for the 2-meter scenario. With the Furthermore, Leatherman used conservative exception of the 2020s, the annual loss in rents assumptions in estimating the unit costs of sand. resulting from not holding back the sea would be Generally, a fraction of the sand placed on a beach far less than the annual costs of pumping sand and washes away because the sand's grain is too small. elevating structures. Titus and Greene point out Moreover, as dredges have to move farther offshore that the cost would be approximately $1,000 per to find sand, costs will increase. year per house, equivalent to 1 week's rent (peak season). For Florida, Leatherman used published estimates of the percentage of fine-grain sand and Nationwide Cost of Pumping Sand Onto assumed that the dredging cost would rise $1 per Recreational Beaches cubic yard for every additional mile offshore the dredge had to move. For the other states, however, Leatherman's goal (Volume B) was to estimate he assumed that the deposits mined would have no the cost of defending the U.S. ocean coast from a fine-grain sand and that dredging costs would not rise in sea level. increase. (To test the sensitivity of this assumption, Titus and Greene developed an increasing-cost Study Design scenario.) Leatherman assumed no storm worse than the 1-year storm, which underestimates the Owing to time constraints, it was possible to sand volumes required. consider only one technology. Based on the Long Beach Island results, Leatherman assumed that the A final limitation of the Leatherman study is cost of elevating recreational beaches and coastal that it represents the cost of applying a single barrier islands by pumping in offshore sand would technology throughout the ocean coasts of the provide a more representative cost estimate than United States. Undoubtedly, some communities assuming that barrier islands would be abandoned, (particularly Galveston and other wide barrier would migrate landward, or would be encircled with islands in Texas) would find it less expensive to dikes and levees. erect levees and seawalls or to accept a natural shoreline retreat. The first step in Leatherman's analysis was to estimate the area of (1) the beach system, (2) the Results low bayside, and (3) the slightly elevated oceanside of the island. Given the areas, the volume of sand Table 7-3 illustrates Leatherman's estimates. A was estimated by assuming that the beach system total of 1,900 miles of shoreline would be nourished. would be raised by the amount of sea level rise. Of 746 square miles of coastal barrier islands that The bay and ocean sides of the island would not be would be raised for a 4-foot sea level rise, 208 raised until after a sea level rise of 1 and 3 feet, square miles would be for a 2-foot rise. As the respectively. Cost estimates for the sand were table shows, two-thirds of the nationwide costs derived from inventories conducted by the U.S. would be borne by four southeastern states: Texas, Army Corps of Engineers. Louisiana, Florida, and South Carolina. Leatherman applied this method to all Figure 7-11 illustrates the cumulative nationwide recreational beaches from Delaware Bay to the costs over time. For the 50- and 200-centimeter mouth of the Rio Grande, as well as California, scenarios, the cumulative cost would be $2.3 to $4.4 which accounts for 80% of the nation's beaches. He billion through 2020, $11 to $20 billion through 2060, and $14 to $58 billion through 2100. By 137 Chapter 7 Table 7-3. Cost of Placing Sand on U.S. Recreational Beaches and Coastal Barrier Islands and Spits (millions of 1986 dollars) Sea level rise by 2100 State Baseline 50 cm 100 cm 200 cm Maineᵃ 22.8 119.4 216.8 412.2 New Hampshireᵃ 8.1 38.9 73.4 142.0 Massachusettsᵃ 168.4 489.5 841.6 1,545.8 Rhode Islandᵃ 16.3 92.0 160.6 298.2 Connecticut 101.7 516.4 944.1 1,799.5 New Yorkᵃ 143.6 769.6 1,373.6 2,581.4 New Jerseyᵃ 157.6 902.1 1,733.3 3,492.5 Delaware 4.8 33.6 71.1 161.8 Maryland 5.7 34.5 83.3 212.8 Virginia 30.4 200.8 386.5 798.0 North Carolina 137.4 655.7 1,271.2 3,240.4 South Carolina 183.5 1,157.9 2,147.7 4,347.7 Georgia 25.9 153.6 262.6 640.3 Florida 120.1 786.6 1,791.0ᵇ 7,745.5ᵇ (Atlantic coast) Florida 149.4 904.3 1,688.4 4,091.6 (gulf coast) Alabama 11.0 59.0 105.3 259.6 Mississippi 13.4 71.9 128.3 369.5 Louisiana 1,955.8 2,623.1 3,492.7 5,231.7 Texas 349.6 4,188.3 8,489.7 17,608.3 California 35.7 174.1 324.3 625.7 Oregonᵃ 21.9 60.5 152.5 336.3 Washington Stateᵃ 51.6 143.0 360.1 794.4 Hawaiiᵃ 73.5 337.6 646.9 1,267.5 Nation 3,788.0 14,512.0 26,745.0 58,002.0 ᵃIndicates states where estimate was based on extrapolating a representative site to the entire state. All other states have 100% coverage. b Florida estimates account for the percentage of fine-grain sediment, which generally washes away, and for cost escalation as least expensive sand deposits are exhausted. All other estimates conservatively ignore this issue. Source: Leatherman (Volume B) (baseline derived from Leatherman). contrast, if current trends continue, the total cost of proceeded independently. Titus and Greene's sea level rise for beach nourishment would be about primary objectives (Volume B) were to combine $35 million per year. various results to estimate the nationwide cost of holding back the sea for various sea level rise Synthesis of the Three National Studies scenarios and to derive ranges for the specific impacts. Their objectives were as follows: Study Design 1. Use Park's results to weight Weggel's high and Although Weggel used Park's topographic data, low scenarios according to whether levees or the analysis in the three nationwide studies bulkheads would be necessary, and interpolate 138 Sea Level Rise on Weggel's assumption that the cost of building bulkheads and levees rises as a function of the structure's height. 60 200 Cm Cost of Raising Barrier Islands Other Than Dredging 50 40 Weggel's case study of Long Beach Island provided cost estimates for elevating structures and 30 rebuilding roads, while Leatherman estimated the 100 Cm 20 area that would have to be raised. Many barrier $ Billions 50 Cm islands have development densities different from 10 those of Long Beach Island because they have large tracts of undeveloped land or larger lot sizes. Therefore, Titus and Greene used census data to 5 estimate a confidence interval for the average building density of barrier islands, and they applied Weggel's cost factors. 0 Sensitivity of Sand Costs to Increasing Scarcity of 1980 2000 2020 2040 2060 2080 2100 Year Sand Titus and Greene used Leatherman's escalating cost assumptions for Florida to estimate sand Figure 7-11. Nationwide cost of sand for protecting pumping costs for the rest of the nation. ocean coast (in 1986 dollars) (Leatherman, Volume B). Confidence Intervals Weggel's cost estimate for the 2-meter rise to The Park and Weggel studies involved sampling, rises of 50 and 100 centimeters; but the researchers did not calculate statistical confidence intervals. Therefore, Titus and Greene 2. Use results from Leatherman and Weggel, developed 95% confidence intervals for the cost of along with census data, to estimate the protecting sheltered coasts, the area of wetlands loss nationwide cost (other than pumping sand) of for various scenarios. raising barrier islands; Limitations 3. Develop an increasing-cost scenario for the cost of protecting the open ocean coast; and Besides all of the limitations that apply to the Park, Leatherman, and Weggel studies, a number of 4. Develop statistical confidence intervals for others apply to Titus and Greene. wetland loss, impacts of the various policy options, and costs of protecting developed Cost of Protecting Sheltered Shores shores. Titus and Greene assumed that the portion of Cost of Protecting Sheltered Shores the coast requiring levees (instead of bulkheads) would be equal to the portion of lowlands that Titus and Greene developed a single estimate otherwise would be inundated. This assumption for protecting each site with bulkheads and levees tends to understate the need for levees. For by assuming that the portion of developed areas example, a community that is 75% high ground protected with levees would be equal to the portion often would still have very low land along all of its of the lowlands that Park estimated would be shoreline and hence would require a levee along inundated. They interpolated the resulting 2-meter 100% of the shore. But Titus and Greene assume estimate to 50- and 100-centimeter estimates, based that only 25% would be protected by levees. 139 Chapter 7 sand costs would increase by the same pattern Cost of Raising Barrier Islands nationwide as they would in Florida. The data provided by Weggel focused only on Results elevating roads, buildings, and bulkheads. Thus, Titus and Greene do not consider the cost of Loss of Wetlands and Dryland replacing sewers, water mains, or buried cables. On the other hand, Weggel's cost factors assume that Table 7-4 illustrates 95% confidence intervals rebuilt roads would be up to engineering standards; for the nationwide losses of wetlands and dryland. it is possible that communities would tolerate If all shorelines were protected, a 1-meter rise substandard roads. In addition, the census data would result in a loss of 50 to 82% of U.S. coastal Titus and Greene used were only available for wetlands, and a 2-meter rise would result in a loss incorporated communities, many of which are part of 66 to 90%. If only the densely developed areas barrier island and part mainland; thus, the data were protected, the losses would be 29 to 69% and provide only a rough measure of typical road 61 to 80% for the 1- and 2-meter scenarios, density. respectively. Except for the Northeast, no protection results in only slightly lower wetland loss Sensitivity of Sand Costs to Increased Scarcity of than protecting only densely developed areas. Sand Although the estimates for the Northeast, mid- Atlantic, the gulf regions outside Louisiana, and the Finally, Titus and Greene made no attempt to Florida peninsula are not statistically significant (at determine how realistic their assumption was that the 95% confidence levels), results suggest that wetlands loss would be least in the Northeast and Northwest. Table 7-4. Nationwide Loss of Wetlands and Drylandᵃ (95% confidence intervals) Square milesᵇ Baseline 50-cm rise 100-cm rise 200-cm rise Wetlands Total protection N.C. 4944-8077 6503-10843 8653-11843 (38-61) (50-82) (66-90) Standard 1168-3341 2591-5934 3813-9068 4350-10995 protection (9-25) (20-45) (29-69) (33-80) No protection N.C. 2216-5592 3388-8703 3758-10025 (17-43) (26-66) (29-76) Dryland Total protection 0 0 0 0 Standard 1906-3510 2180-6147 4136-9186 6438-13496 protection No protection N.C. 3315-7311 5123-10330 8791-15394 ᵃWetlands loss refers to vegetative wetlands only. ᵇNumbers in parentheses are percentages. NC = Not calculated. Source: Titus and Greene (Volume B). 140 Sea Level Rise Table 7-5. Cumulative Nationwide Cost of Protecting Barrier Islands and Developed Mainland Through the Year 2100 (billions of 1986 dollars)ᵃ Sea level scenario Baseline 50-cm rise 100-cm rise 200-cm rise Open coast Sand 3.8 15-20 27-41 58-100 Raise houses, roads, utilities 0 9-13 21-57 75-115 Sheltered shores 1.0-2.4 5-13 11-33 30-101 Totalᵇ 4.8-6.2 32-43 73-111 119-309 a Costs due to sea level rise only. ᵇRanges for totals are based on the square root of the sum of squared ranges. Source: Titus and Greene (Volume B). Costs of Holding Back the Sea coastal wetlands acreage (outside Louisiana) by 17 to 76% if no mainland areas were protected, by 20 Table 7-5 illustrates the Titus and Greene to 80% if only currently developed areas were estimates of the costs of holding back the sea. The protected, and by 38 to 90% if all mainland areas low range for the sand costs is based on were protected. These estimates of the areal losses Leatherman's study, and the high range is based on understate the differences in impacts for the various the increasing cost scenario Titus and Greene land-use options. Although a substantial loss would developed. The uncertainty range for the costs of occur even with no protection, most of today's elevating structures reflects the uncertainty in census wetland shorelines would still have wetlands; the data regarding the current density of development. strip simply would be narrower. By contrast, High and low estimates for the cost of protecting protecting all mainland areas would generally sheltered shorelines are based on the sampling replace natural shorelines with bulkheads and errors of the estimates for the 46 sites that both levees. This distinction is important because for Park et al. and Weggel et al. examined. many species of fish, the length of a wetland shoreline is more critical than the total area. Titus and Greene estimated that the cumulative nationwide cost of protecting currently developed Options for State and Local Governments areas in the face of a 1-meter rise would be from $73 to 111 billion, with costs for the 50- and 200- Titus (1986) examined three approaches for centimeter scenarios ranging from $32 to 309 maintaining wetland shorelines in the face of a billion. These costs would imply a severalfold rising sea: (1) no further development in lowlands; increase in annual expenditures for coastal defense. (2) no action now but a gradual abandonment of Nevertheless, compared with the value of coastal lowlands as sea level rises; and (3) allowing future property, the costs are small. development only with a binding agreement to allow such development to revert to nature if it is threatened by inundation. POLICY IMPLICATIONS The first option would encounter legal or financial hurdles. The extent to which the Wetlands Protection "due-process" clause of the Constitution would allow governments to prevent development in anticipation The nationwide analysis showed that a 50- to of sea level rise has not been specifically addressed 200-centimeter rise in sea level could reduce the 141 Chapter 7 by the courts. Although purchases of land would be The Federal Role feasible for parks and refuges, the cost of buying the majority of lowlands would be prohibitive. Section 404 of the Clean Water Act discourages Moreover, this approach requires preparation for a development of existing wetlands, but it does not rise in sea level of a given magnitude; if and when address development of areas that might one day be the sea rises beyond that point, the wetlands would necessary for wetland migration. This program will be lost. Finally, preventing future development provide lasting benefits, even if most coastal would not solve the loss of wetlands resulting from wetlands are inundated. Although marshes and areas that have already been developed. swamps would be inundated, the shallow waters that formed could provide habitat for fish and Enacting no policy today and addressing the submerged aquatic vegetation. No one has assessed issue as sea level rises would avoid the costs of the need for a federal program to protect wetlands planning for the wrong amount of sea level rise but in the face of rising sea level. would probably result in less wetlands protection. People are developing coastal property on the Coastal Protection assumption that they can use the land indefinitely. It would be difficult for any level of government to State and Local Efforts tell property owners that they must abandon their land with only a few years' notice, and the cost of State and local governments currently decide purchasing developed areas would be even greater which areas would be protected and which would than the cost of buying undeveloped areas. be allowed to erode. Currently, few localities contribute more than 10% of the cost of beach Economic theory suggests that under the third nourishment, with the states taking on an increasing alternative, people would develop a property only if share from the federal government. However, many the temporary use provided benefits greater than coastal officials doubt that their states could raise the costs of writing it off early. This approach the necessary funds if global warming increased the would result in the greatest degree of flexibility, costs of coastal protection over the next century by because it would allow real estate markets to $50 to $300 billion. If state funds could not be incorporate sea level rise and to determine the most found, the communities themselves would have to efficient use of land as long as it remains dry. take on the necessary expenditures or adapt to erosion. This approach could be implemented by regulations that prohibit construction of bulkheads Long Beach Island, New Jersey, illustrates the as sea level rises or by the use of conditional long- potential difficulties. The annual cost of raising the term leases that expire when high tide falls above a island would average $200 to $1,000 per house over property's elevation. the next century (Titus and Greene, Volume B). Although this amount is less than one week's rent The State of Maine (1987) has implemented during the summer, it would more than double this third approach through its coastal dune property taxes, an action that is difficult for local regulations, which state that people building houses governments to contemplate. Moreover, the island along the shore should assume that they will have to is divided into six jurisdictions, all of which would move their houses if their presence prevents the have to participate. natural migration of coastal wetlands, dunes, or other natural shorelines. A number of states also More lightly developed communities may decide have regulations that discourage bulkheads, although that the benefits of holding back the sea are not they do not specifically address sea level rise. The worthwhile. Sand costs would be much less for an option can be implemented through cooperative island that migrated. Although Weggel estimated arrangements between developers, conservancy that higher costs would be associated with allowing groups, and local governments. (See Titus and Long Beach Island to migrate landward than with Greene, Volume B, for additional details.) raising the island in place, this conclusion resulted largely from the cost of rebuilding sewers and other utilities that would still be useful if the island were raised. 142 Sea Level Rise Regardless of how a barrier island community protective beach would have constituted a intends to respond to sea level rise, the eventual considerably higher fraction of the total benefits costs can be reduced by deciding on a response well (Titus, 1985). in advance. The cost of raising an island can be reduced if roads and utilities are routinely elevated Wetlands Loss in Louisiana or if they have to be rebuilt for other reasons (e.g., Titus et al., 1987). The cost of a landward By preventing freshwater and sediment from migration also can be reduced by discouraging reaching the coastal wetlands, federal management reconstruction of oceanfront houses destroyed by of the Mississippi River is increasing the storms (Titus et al., 1984a). The ability to fund the vulnerability of coastal Louisiana to a sea level rise required measures also would be increased by (e.g., Houck, 1983). For example, current fostering the necessary public debate well before navigation routes require the U.S. Army Corps of the funds are needed. Engineers to limit the amount of water flowing through the Atchafalaya River and to close natural Federal Efforts breaches in the main channel of the Mississippi; these actions limit the amount of freshwater and While state governments generally are sediment reaching the wetlands. Alternative routes responsible for protecting recreational beaches, the have been proposed that would enable water and U.S. Army Corps of Engineers is responsible for sediment to reach the wetlands (Louisiana Wetland several major federal projects to rebuild beaches Protection Panel, 1987). These include dredging and for efforts to curtail land loss in Louisiana. The additional canals parallel to the existing Mississippi long-term success of these efforts would be River gulf outlet or constructing a deepwater port improved if the corps were authorized to develop east of the city. comprehensive long-term plans to address the impacts of sea level rise. Either of these options would cost a few billion dollars. By contrast, annual resources for correcting Beach Erosion land loss in Louisiana have been in the tens of millions of dollars. As a result, mitigation activities In its erosion-control efforts, the corps has have focused on freshwater diversion structures and recently shifted its focus from hard structures (e.g., on other strategies that can reduce current wetland seawalls, bulkheads, and groins) to soft approaches, loss attributable to high salinities but that would not such as pumping sand onto beaches. This shift is substantially reduce wetlands loss if sea level rises consistent with the implications of sea level rise: 50 to 200 centimeters (Louisiana Wetland groins and seawalls will not prevent loss of beaches Protection Panel, 1987). due to sea level rise. Although more sand will have to be pumped than current analyses suggest, this The prospect of even a 50-centimeter rise in sea approach could ensure the survival of the nation's level suggests that solving the Louisiana wetlands beaches. loss problem is much more urgent than is commonly assumed. Because federal activities are now a Nevertheless, consideration of accelerated sea major cause of land loss, and would have to be level rise would change the cost-benefit ratios of modified to enable wetlands to survive a rising sea, many corps erosion control projects. As with the the problem is unlikely to be solved without a operations of reservoirs (discussed in Chapter 16: congressional mandate. A recent interagency report Southeast), the corps is authorized to consider flood concluded that "no one has systematically protection but not recreation. When they evaluated determined what must be done to save 10, 25, or 50 the benefits of erosion control at Ocean City, percent of Louisiana's coastal ecosystem" (Louisiana Maryland, the corps concluded that less than 10% Wetland Protection Panel, 1987). Until someone of the benefits would be for flood control (most estimates the costs and likely results of strategies were related to recreation). Had they considered with a chance of protecting a significant fraction of accelerated sea level rise, however, the estimated the wetlands in face of rising sea level, it will be flood protection benefits from having a difficult for Congress to devise a long-term solution. 143 Chapter 7 Flood Insurance was great enough to outweigh the likely damages from floods. However, statutes limiting the rate at In 1968, Congress created the National Flood which flood insurance rates can increase could keep Insurance Program with the objective of reducing rates from rising as rapidly as the risk of flooding, federal disaster relief resulting from floods. The thereby increasing the federal subsidy. Federal Emergency Management Agency (FEMA), which already had responsibility for administering No assessment of the impacts of sea level rise disaster relief, was placed in charge of this program on the federal flood insurance program has been as well. undertaken. The National Flood Insurance Program sought Sewers and Drains to offer localities an incentive to prevent flood-prone construction. In return for requiring Sea level rise also would have important impacts that any construction in a floodplain be designed to on coastal sewage and drainage systems. Wilcoxen withstand a 100-year flood, the federal government (1986) examined the implications of the failure to would provide subsidized insurance to existing consider accelerated sea level rise in the design of homes and a fair-market rate for any new San Francisco's West Side (sewerage) Transport, construction (which was itself a benefit, since private which is a large, steel-reinforced concrete box insurers generally did not offer flood insurance). buried under the city's ocean beach. He found that Moreover, as long as a community joined the beach erosion will gradually expose the transport to program, it would continue to be eligible for federal the ocean, leaving the system vulnerable to disaster relief; if it did not join, it would no longer undermining and eventual collapse. Protection costs be eligible. As a result of this program, new coastal for the $100 million project would likely amount to houses are generally elevated on pilings. an additional $70 million. Wilcoxen concludes that had sea level rise been considered, the project Although Congress intended to prevent coastal probably would have been sited elsewhere. disasters, the National Flood Insurance Act does not require strategic assessments of long-term issues The impacts of sea level rise on the construction (see Riebsame, Volume J). Thus, FEMA has not grants program probably would be less in most conducted strategic assessments of how the program other cases. As sea level rises, larger pumps will could be managed to minimize flood damage from be necessary to transport effluents from settling shoreline retreat caused by both present and future ponds to the adjacent body of water. However, sea rates of sea level rise. level rise would not necessarily require alternative siting. The projects serving barrier islands often are Congress recently enacted the Upton-Jones located on the mainland, and projects located on Amendment (Public Housing Act of 1988), which barrier islands are generally elevated well above commits the federal government to pay for flood levels. If barrier islands are raised in rebuilding or relocating houses that are about to response to sea level as the nationwide analysis erode into the sea. Although the cost of this suggests, sewerage treatment plants will be a small provision is modest today, a sea level rise could part of the infrastructure that has to be modified. commit the federal government to purchase the houses on all barrier islands that did not choose to Engineering assessments have concluded that it hold back the sea. Furthermore, this commitment is already cost-beneficial to consider sea level rise could increase the number of communities that in the construction of coastal drainage systems in decided not to hold back the sea. urban areas. For example, the extra cost of installing the larger pipes necessary to accommodate The planned implementation of actuarially a 1-foot rise in sea level would add less than 10% to sound insurance rates would ensure that as sea level the cost of rebuilding a drainage system in rise increased property risk, insurance rates would Charleston, South Carolina; however, failure to rise to reflect the risk. This would discourage consider sea level rise would require premature construction of vulnerable houses, unless their value rebuilding of the $4 million system (Titus et al., 1987). 144 Sea Level Rise RESEARCH NEEDS Barth, M.C., and J.G. Titus, eds. 1984. Greenhouse Effect and Sea Level Rise: A A much better understanding of erosion Challenge for This Generation. New York: Van processes is needed to (1) understand how much Nostrand Reinhold Company. erosion will take place if no action is taken; and (2) help identify the most cost-effective means for Bentley, C.R. 1983. West Antarctic ice sheets: protecting sandy shores. An improved diagnosis and prognosis. In: Proceedings of the Carbon Dioxide Research Conference: Carbon understanding of how wetland accretion responds to different temperatures, higher CO₂ Dioxide, Science, and Consensus. Conference concentrations, changing mineral content, and the 820970. Washington, DC: Department of Energy. drowning of adjacent wetlands is needed. This will refine our ability to project future wetlands loss and, Bruun, P. 1962. Sea level rise as a cause of shore perhaps, devise measures for artificially enhancing erosion. Journal of Waterways and Harbors their vertical growth. Division (ASCE) 1:116-130. This report did not examine the impacts of Dean, R.G. et al. 1987. Responding to Changes in increased flooding because flood models have not Sea Level. Washington, DC: National Academy Press. been applied to the large numbers of coastal sites that would be necessary to conduct a nationwide assessment. Time-dependent estuarine salinity Dony, W.L., W.R. Farrand, and M. Ewing. 1962. models, such as that of the Delaware River Basin Pleistocene ice volumes and sea level lowering. Commission, should be applied to major estuaries Journal of Geology 70:206-214. to examine impacts on ecosystems and drinking water supplies. Everts, C.H. 1985. Effect of sea level rise and net sand volume change on shoreline position at Ocean Assessments of the impacts of global warming City, Maryland. In: Titus, J.G., ed. Potential on coastal environments would be greatly improved Impact of Sea Level Rise on the Beach at Ocean by better estimates of future sea level rise. In City, Maryland. Washington, DC: U.S. addition to the improved ocean modeling that will Environmental Protection Agency. be necessary for better projections of surface air temperatures (see Chapter 2: Climate Change), this Fairbridge, R.W., and W.S. Krebs, Jr. 1962. Sea will also require a substantial increase in the level and the southern oscillation. Geophysical Journal 6:532-545. resources allocated for monitoring and modeling glacial processes. Finally, this report assumed that winds, waves, and storms remained constant; future Gibbs, M. 1984. Economic analysis of sea level studies will need estimates of the changes in these rise: methods and results. In: Barth, M.C., and climatic variables. J.G. Titus, eds. Greenhouse Effect and Sea Level Rise: A Challenge for This Generation. New York: Van Nostrand Reinhold Company. REFERENCES Gornitz, V.S., S. Lebedeff, and J. Hansen. 1982. Global sea level trend in the past century. Science Armentano, T.V., R.A. Park, and C.L. Cloonan. 215:1611-1614. 1988. Impacts on coastal wetlands throughout the United States. In: Titus, J.G. ed. Greenhouse Hoffman, J.S., D. Keyes, and J.G. Titus. 1983. Effect, Sea Level Rise, and Coastal Wetlands. Projecting Future Sea Level Rise. Washington, DC: Washington, DC: U.S. Environmental Protection U.S. Environmental Protection Agency. Agency. Hoffman, J.S., J. Wells, and J.G. Titus. 1986. Barnett, T.P. 1984. The estimation of "global" sea Future global warming and sea level rise. In: level change: a problem of uniqueness. Journal of Sigbjarnarson, G., ed. Iceland Coastal and River Geophysical Research 89(C5):7980-7988. Symposium. Reykjavik, Iceland: National Energy Authority. 145 Chapter 7 Houck, O.A. 1983. Land loss in coastal Louisiana: Louisiana Wetland Protection Panel. 1987. Saving causes, consequences, and remedies. Tulane Law Louisiana's Coastal Wetlands and the Need for a Review 58(1):3-168. Long-Term Plan of Action. Washington, DC: U.S. Environmental Protection Agency/Louisiana Howard, J.D., O.H. Pilkey, and A. Kaufman. 1985. Geological Survey. EPA-230-02-87-026. Strategy for beach preservation proposed. Geotimes 30(12):15-19. Lyle, S.D., L.E. Hickman, and H.A. Debaugh. 1987. Sea Level Variations in the United States. Rockville, Hughes, T. 1983. The stability of the West MD: National Ocean Service. Antarctic ice sheet: what has happened and what will happen. In: Proceedings of the Carbon Meier, M.F. et al. 1985. Glaciers, Ice Sheets, and Dioxide Research Conference: Carbon Dioxide, Sea Level. Washington, DC: National Academy Science, and Consensus. Conference 820970. Press. Washington, DC: Department of Energy. Meier, M.F. 1984. Contribution of small glaciers to Hull, C.H.J., and J.G. Titus. 1986. Responses to global sea level. Science 226(4681):1418-1421. salinity increases. In: Hull, C.H.J., and J.G. Titus. 1986. Greenhouse Effect, Sea Level Rise, and Mercer, J.H. 1968. Antarctic ice and Sangamon Salinity in the Delaware Estuary. Washington, DC: sea level. Geological Society of America Bulletin U.S. Environmental Protection Agency and 79:471. Delaware River Basin Commission. Oldale, R. 1985. Late quaternary sea level history Hull, C.H.J., and J.G. Titus. 1986. Greenhouse of New England: a review of published sea level Effect, Sea Level Rise, and Salinity in the Delaware data. Northeastern Geology 7:192-200. Estuary. Washington, DC: U.S. Environmental Protection Agency and Delaware River Basin Revelle, R. 1983. Probable future changes in sea Commission. level resulting from increased atmospheric carbon dioxide. In: Changing Climate. Washington, DC: Kana, T.W., J. Michel, M.O. Hayes, and J.R. National Academy Press. Jensen. 1984. The physical impact of sea level rise in the area of Charleston, South Carolina. In: Schelling, T. 1983. Climatic change: implications Barth, M.C., and J.G. Titus, eds. Greenhouse for welfare and policy. In: Changing Climate. Effect and Sea Level Rise: A Challenge for This Washington, DC: National Academy Press. Generation. New York: Van Nostrand Reinhold Company. Sorensen, R.M., R.N. Weisman, and G.P. Lennon. 1984. Control of erosion, inundation, and salinity Kennett, J. 1982. Marine Geology. Englewood intrusion. In: Barth, M.C., and J.G. Titus, eds. Cliffs, NJ: Prentice-Hall. Greenhouse Effect and Sea Level Rise: A Challenge for This Generation. New York: Van Kyper, T., and R. Sorenson. 1985. Potential Nostrand Reinhold Company. impacts of sea level rise on the beach and coastal structures at Sea Bright, New Jersey. In: O.T. State of Maine. 1987. Dune Rule 355. Augusta, Magson, ed. Coastal Zone '85. New York: ME: Maine Department of Environmental American Society of Civil Engineers. Protection. Leatherman, S.P. 1984. Coastal geomorphic Thomas, R.H. 1986. Future sea level rise and its responses to sea level rise: Galveston Bay, Texas. early detection by satelite remote sensing. In: Titus, In: Barth, M.C., and J.G. Titus, eds. Greenhouse J.G., ed. Effects of Changes in Stratosphere Ozone Effect and Sea Level Rise: A Challenge for This and Global Climate. Washington, DC: U.S. EPA Generation. New York: Van Nostrand Reinhold and United Nations Environment Programme. Company. 146 Sea Level Rise Thomas, R.H. 1985. Responses of the polar ice Titus, J.G., M.C. Barth, J.S. Hoffman, M. Gibbs, sheets to climatic warming. In: Meier, M.F., D.G. and M. Kenney. 1984b. An overview of the causes Aubrey, C.R. Bentley, W.S. Broecker, J.E. Hansen, and effects of sea level rise. In: Barth, M.C., and W.R. Peltier, and R.J.C. Somerville, eds. Glaciers, J.G. Titus, eds. Greenhouse Effect and Sea Level Ice Sheets, and Sea Level. Washington, DC: Rise: A Challenge for This Generation. New York: National Academy Press. Van Nostrand Reinhold Company. Titus, J.G. 1988. Sea level rise and wetlands loss: Titus, J.G., T. Henderson, and J.M. Teal. 1984. an overview. In: Titus, J.G., ed. Greenhouse Sea level rise and wetlands loss in the United States. Effect, Sea Level Rise, and Coastal Wetlands. National Wetlands Newsletter 6:4. Washington, DC: U.S. Environmental Protection Agency. Titus, J.G., C.Y. Kuo, M.J. Gibbs, T.B. LaRoche, M.K. Webb, and J.O. Waddell. 1987. Greenhouse Titus, J.G., ed. 1988. Greenhouse Effect, Sea Level effect, sea level rise, and coastal drainage systems. Rise, and Coastal Wetlands. Washington, DC: U.S. Journal of Water Resources Planning and Environmental Protection Agency. Management. American Society of Civil Engineers 113(2):216-227. Titus, J.G. 1987. The greenhouse effects, rising sea level, and society's response. In: Devoy, R.J.N. Wilcoxen, P.J. 1986. Coastal erosion and sea level Sea Surface Studies. New York: Croom Helm. rise: implications for ocean beach and San Francisco's West Side Transport Project. Coastal Titus, J.G. 1986. Greenhouse effect, sea level rise, Zone Management Journal 14:3. and coastal zone management. Coastal Zone Management Journal 14(3):147-171. WMO. 1986. World Meteorological Organization. Atmospheric ozone 1985. Assessment of our Titus, J.G. 1985. Sea level rise and the Maryland understanding of the processes controlling its coast. In: Potential Impacts of Sea Level Rise on present distribution and change. Global Ozone the Beach at Ocean City, Maryland. Washington, Research and Monitoring Project, Report No. 16. DC: U.S. Environmental Protection Agency. Geneva, Switzerland: World Meteorological Organization. Titus, J.G. 1984a. Planning for sea level rise before and after a coastal disaster. In: Barth, M.C., and J.G. Titus, eds. Greenhouse Effect and Sea Level Rise: A Challenge for This Generation. New York: Van Nostrand Reinhold Company. 147 CHAPTER 8 BIOLOGICAL DIVERSITY FINDINGS Rapid climate change would add to the already existing threats biodiversity faces from Unlike most other impacts, loss of species and anthropogenic activities, such as deforestation reduced biological diversity are irreversible. The and habitat fragmentation. ability of a natural community to adapt to changing climate conditions will depend on the rate of climate Marine Ecosystems change, the size of species ranges, the dispersal rates of the individual species, and whether or not The loss of coastal wetlands and coastal barriers to species migration are present. If climate habitat resulting from sea level rise and changes rapidly, many species will be lost. saltwater intrusion may profoundly affect the populations of all inhabitants of these Species Diversity ecosystems, including mollusks, shellfish, finfish, and waterfowl. However, there is no The effect of climate change on species and evidence to indicate these species would ecosystems will most likely vary, with some become extinct. species benefiting and others facing extinction. The uncertainties surrounding the rate of Freshwater Ecosystems warming, individual species response, and interspecies dynamics make impacts difficult to Freshwater fish in large bodies of water, such assess. However, climate change would alter as the Great Lakes, may increase in competitive outcomes and destabilize natural productivity, but some significant species could ecosystems in unpredictable ways. decline. Fish in smaller bodies of water may be more constrained in their ability to respond In many cases, the indirect effects of climate to climate change. They also may be harmed change on a population, such as changes in by reductions in water quality. habitat, in food availability, and in predator/prey relationships, may have a Migratory Birds greater impact than the direct physiological effects of climate change. Migratory birds are likely to experience mixed effects from climate change, with some arctic Natural and manmade barriers, including nesting herbivores benefiting and continental roads, cities, mountains, bodies of water, nesters and shorebirds suffering. The loss of agricultural land, unsuitable soil types, and wintering grounds due to sea level rise and habitat fragmentation, may block migration of changing climate could harm many species, as species in response to climate change and would the loss of inland prairie potholes due to exacerbate losses. potentially increased continental dryness. The areas within the United States that appear to be most sensitive to changes in climate are Policy Implications those that have a number of threatened and endangered species, species especially sensitive Existing refuges, sited to protect a species or to heat or drought stress, and species ecosystem under current climate, may not be inhabiting coastal areas. properly located for this purpose if climate changes or as species migrate. 149 Chapter 8 Wildlife agencies such as the Department of may face a more rapid change in climate that may the Interior, state government agencies, and have important consequences for biological diversity. conservation organizations may wish to assess the feasibility of establishing migratory The National Resource corridors to facilitate species migration. Public and private lands in the United States Areas that may become suitable future habitat provide sanctuary for an abundant diversity of plants for threatened and endangered species, such as and animals. About 650 species of birds reside in lowland areas adjacent to current wetlands, or pass through the United States annually. Over need to be identified and protected. 400 species of mammals, 460 reptiles, 660 freshwater fishes, and tens of thousands of The practice of restoration ecology may need invertebrates can be found in this country, in to be broadened to rebuild parts of ecosystems addition to some 22,000 plants (U.S. Fish and in new areas as climates shift. Wildlife Service, 1981). These species compose a wide variety of ecosystem types within the United The increase in the number of species at risk States, including coniferous and broad-leaf forest, as a result of climate change may require new grassland, desert, freshwater, marine, estuarine, strategies for balancing ecosystem level inland wetland, and agricultural ecosystems. Figure concerns with single species concerns. Agency 8-1 shows the major ranges of natural vegetation in programs such as the Fish and Wildlife the United States. Service's Endangered Species Program, may wish to assess the relative risk of climate The U.S. national parks, forests, wilderness change and more current stresses on ecological areas, and fish and wildlife refuges are among the systems. public lands that provide sanctuary for wildlife resources, including many endangered species. U.S. public lands, which encompass over 700 million VALUE OF BIOLOGICAL acres (about 32% of the land area of the United DIVERSITY States), support about 700 rare species and communities (Roush, 1986). Over 45% of the lands held by the Forest Service, Fish and Wildlife Maintaining the biological diversity of our Service, National Park Service, and Bureau of Land natural resources is an important goal for the Management are in Alaska, and over 48% are nation. The preamble to the Endangered Species located in the 11 most western states (U.S. Act of 1973 emphasizes the value of individual Department of the Interior, 1987). However, much species, stating that endangered and threatened of the nation's biological diversity lies outside these species of fish, wildlife, and plants "are of aesthetic, areas. ecological, educational, historical, recreational and scientific value to the Nation and its people." We Private land holdings also account for a great depend upon our nation's biological resources for deal of this nation's biological endowment. Private food, medicine, energy, shelter, and other important groups, such as the Nature Conservancy and the products. In addition to species diversity, the Audubon Society, manage 500,000 acres and 86,000 genetic variability within a species and the wide acres, respectively, for biological diversity. variety of ecosystems add to biological diversity. Reduced biological diversity could have serious implications for mankind as untapped resources for GENERAL COMPONENTS OF research in agriculture, medicine, and industry are irretrievably lost. BIOLOGICAL DIVERSITY The evolving biological diversity of this planet Biological diversity can be broadly defined as is inevitably affected by climate change. Historic the full range of variety and variability within and climate changes have resulted in major changes in among living organisms. It includes species species diversity. This has been true for the millions diversity, genetic diversity, and ecosystemic or of years life has existed on Earth. Now our planet community diversity. This report concentrates on 150 Biological Diversity Sf Sg Mt Ne Tg S Oh SW Op Op Se C Rb Mg M Sf Spruce-fir forest SW Southwest broadleaf woodland S Sagebrush shrubland Tf Transition pine-aspen forest Sg Short grassland Mg Mesquite and desert grassland Ne Northeast hardwood forest Tg Tall grassland C Creosote bush shrubland Oh Pacific coast forest Oak-hickory forest north: spruce, hemlock Rb Riverbottom cypress-tupelo-sweetgum south: douglas fir Op Oak-pine forest Mt Coast Range-Rocky Mountain conifer forest Mangrove swampland lower: pine, douglas fir higher: spruce-fir Se Southeast pine forest summits: alpine meadow Figure 8-1. Natural vegetation in the United States (Hunt, 1972). species diversity, but only because it is better The stresses brought about by development, understood. Genetic and ecosystemic diversity are overuse, and alteration of habitat have fragmented equally important. much of the world's natural habitat and have created many new barriers. Consequently, for many Species Diversity species, dispersal has become much more difficult than it was in the past. For other species, humans Each species occurs in a characteristic range or have inadvertently aided dispersal and have caused geographical area. The factors controlling species rapid spread in recent years. Such practices as ranges are critical constraints on biological diversity. clearcut logging prevent the dispersal of species The presence of a species in an area suggests that adapted to dense forest conditions (e.g., flying the species must have successfully achieved the squirrels) and promote the dispersal of species following: (1) dispersal into an area (no barriers to suited to open areas (e.g., deer). dispersal, such as the presense of bodies of water or unsuitable soil types); (2) survival in that area (the Currently, 495 species are listed as endangered physical characteristics of the area were suited to within the United States, and over 2,500 species the species' physiology, and food was available); and await consideration for that status by the Fish and (3) establishment in the area (the organism found Wildlife Service. The list of endangered species is an appropriate place in the food web in the absence dominated by plants, birds, fishes, and mammals but of excessive competition and predation, and was also includes insects, amphibians, reptiles, mollusks, able to reproduce). and crustaceans (U.S. Fish and Wildlife Service, 1988). 151 Chapter 8 New species are created through the contribute to the extinction of a species by reducing evolutionary process of speciation, whereas existing its ability to adapt to changing environmental species are lost through extinction. Speciation conditions. generally requires at least hundreds of thousands of years. However, extinction as a result of human Generally, species with larger populations have activities, even without climate change, is occurring greater genetic diversity. Species near extinction rapidly and at an increasing rate. Owing to its represented by few individuals in few populations slowness, the process of speciation does little to have lower genetic diversity, a situation exacerbated offset species' loss to extinction. by inbreeding. Additionally, extreme climatic events may cause bouts of natural selection that reduce Stressed Biological Diversity genetic variability (Mayr, 1963). Biological diversity continues to erode steadily Community and Ecosystemic Diversity around the globe as a result of human activities. Habitat destruction, degradation, and fragmentation Ecosystemic diversity is the number of have resulted in the loss of many species and have distinctive assemblages of species and biotic reduced the ranges and populations of others. processes that occur in different physical settings. These impacts affect all three levels of biological A long-leaf pine forest, a sand dune, and a small diversity. Through providing an additional pressure pond are all part of our diversity at this level. on ecological systems, climate change will further Ecosystems come into existence through complex reduce the biological diversity in this nation and physical and biological processes not now well around the globe. understood. They may be lost by outright replacement of one by another (as in the It is difficult to determine the exact rate of desertification of a grassland) or by the gradual species extinction because the number of species on merging of two formerly separate ecosystems (as in the Earth is known only to an order of magnitude. the loss of some estuarine systems when they A recent estimate by Wilson (1988) places the total become saltier and take on more of the number of species between 5 and 30 million. characteristics of a purely saltwater ecosystem). Assuming 10 million species, Wilson made the Ecosystems can also be eliminated because of rough calculation that one in every 1,000 species is human activities (as in the filling in of a wetland). lost each year. Wilson then compared this to estimates of extinction rates over geologic time, which ranged between 1 in every 1 million and 1 in FACTORS AFFECTING THE every 10 million each year. Thus, human activities RESPONSE OF BIOLOGICAL may be eliminating species at least 1,000 times faster than natural forces. DIVERSITY TO CLIMATE CHANGE The significance of rare species should not be underestimated. A narrowly or sparsely distributed species may be a keystone in an ecosystem, Species respond to environmental change on a controlling the structure and functioning of the hierarchy of time scales. For relatively small community, or it may be a species of great and yet changes occurring within the lifetime of an unknown value to humans. individual, each member of the species can respond through a variety of physiological adjustments. Individual species differ in their ability to adjust to Genetic Diversity change. Some can withstand a great deal of climate change, whereas others are restricted to a narrow Each species that persists has a characteristic range. Over several generations, natural selection genetic diversity. The pool of genetic diversity can cause genetic adaptation and evolution in within a species constitutes an adaptation to its response to the change. Alternatively, a species can present environment as well as a store of adaptive respond to climate change by moving into a new options for some possible changes in the area through migration and dispersal. This can environment. The loss of genetic diversity can occur over a relatively short period of time if the 152 Biological Diversity species has the biological ability to move quickly. negative response, such as local extinction in an The discussion of response to climate change area, is usually quicker than the positive response of centers on migration as the response that could new species' colonization of a region (see Chapter occur over a relatively short period of time. 5: Forests). In the Arctics, the lag period between climate change and species response by migration The distributions of species are significant and colonization may be several hundred years indicators of climate change. Local climate appears (Edlund, 1986). This lag period will leave areas to be the primary factor defining an environmental open for weedy, opportunistic species that can setting and determining the species composition and quickly migrate and propagate in a region. spatial patterns of communities in terrestrial zones (Bolin et al., 1986). Temperature means, The rate of climate change will be crucial to temperature extremes, and precipitation are the the survival of the species in an ecosystem. A 3°C factors most often affecting the potential natural (5°F) increase in temperature, for example, would distributional limits of a species (Ford, 1982), while effect a several hundred kilometer poleward shift in the actual distribution of a species is also affected the temperate vegetation belts (Frye, 1983). If this by soil type, soil moisture, ecological dynamics, and change took place within a century, species would regional isolation. need to migrate several kilometers each year to adapt to this warming. Plants have a wide range of Rate of Climate Change migration rates, and only some may be able to achieve this rate. Failure of a species to "keep up" Predicting how a species or ecosystem might with suitable environmental conditions would respond to a given environmental change is difficult. eventually result in extinction. Adaptation to climate change will inextricably depend on the rate of climate change. For some Many factors make evaluating the impact of species, migration rates may be inadequate to keep climate change on ecosystems difficult. The great up. interdependencies among species in an ecosystem add considerable uncertainty to the effect that the The large number of combinations of dispersal various responses of individual species will have on range and age to reproduction make the potential the system. An impact upon a single species could rate of migration different for every species. profoundly affect the entire ecosystem. Certain Paleorecords suggest migration rates between 10 species are vital to the workings of their ecosystems. and 20 kilometers per century for chestnut, maple, Among them are large carnivores that regulate and balsam fir, and between 30 and 40 kilometers predator-prey relationships, large herbivores that per century for some oak and pine species (see significantly change vegetation, and organisms that Chapter 5: Forests). On the other hand, cattle pollinate plants (WRI, 1988). Plants can also be key egrets have shown a much quicker migration rate by species within an ecosystem. For example, colonizing all of the North American tropics within elimination of a tree species in a region could have approximately 40 years. a significant effect on the whole forest ecosystem, including birds, insects, and mammals. As species shift at different rates in response to climate change, communities may disassociate Animal populations are generally much more into new arrangements of species. Local extinction mobile than plants. But animal distributions heavily can result either directly from physiological depend on vegetation for food, protection, and pressures or indirectly from changes in interspecies nesting habitat. Species not directly dependent on dynamics. Hence, the effect of climate change on vegetation ultimately depend on some other species an area will be to cause sorting and separation of that is. The ranges of the fig wasp and the fig species as a result of the differential rate of depend entirely upon one another. In this case, the migration and species retreat (Ford, 1982). plant species depends on a single pollinator, and Ecosystems, therefore, will not migrate as a unit. the insect species relies upon a single species of plant for food (Kiester et al., 1984). Species do not immediately respond to changed and changing environmental conditions. A 153 Chapter 8 Effect on Genetic Diversity change frequently result from insufficient habitat area or isolation from other populations. The With regard to genetic diversity, rapid climate problem of isolation is similar to that of island change would select for those genotypes species and has become known as the island (combinations of genes) that were best suited to the dilemma. Species on reserves are often remnants of new climate regime and would tend to eliminate larger populations and are more susceptible to others. This process of natural selection would environmental stress and extinction. usually decrease the genetic variability within a population. In the long term (evolutionary time), it Species on reserves are likely to be pressured is possible that greater climatic variability could from two directions as a result of climate change. select for greater genetic variability. A population isolated on a reserve surrounded by altered or unsuitable habitat receives little Barriers to Response immigration from populations outside the reserve. Also, that population may not be able to colonize The rate of species migration is also affected areas outside the reserve as these areas become by natural and manmade barriers and by suitable because of development or other alterations competition. Peters and Darling (1985) examined of habitat. the potential responses of species to climate change, ecological interactions, and barriers to adaptation. Even without the added pressure of climate Physical barriers include mountains, bodies of water, change, reserve populations are vulnerable because roads, cities, agricultural land, inappropriate soil many reserves are not large enough to support a type, and habitat heterogeneity (landscape self-sustaining population (Lovejoy, 1979). The patchiness). A species whose migration rate is predictive theory of island biogeography showed sufficient to keep up with changing conditions could that, other factors being equal, small islands become constrained by a physical barrier. Inability accommodate smaller numbers of species than do to cross the barrier could result in a reduction of large islands (MacArthur and Wilson, 1967). This the range of the species and its eventual extinction. held true for other ecological "islands," such as mountaintops, woodlots, and lakes. Also, when Reserve and Island Species large ecosystems become smaller through fragmentation, the number of species always Additional constraints on the ability of declines. Figure 8-2 shows how mammalian populations living on reserves to respond to climate extinctions have been inversely related to refuge area in North American parks. REFUGE AREA VS, SPECIES LOSS 45 2 40 35 1 3 5 PERCENT OF ORIGINAL SPECIES LOSS (AS PERCENT OF ORIGINAL) PARK AREA (km²) SPECIES LOST 30 4 6 1) Bryce Canyon 144 36 25 2) Lassen Volcano 426 43 7 3) Zion 588 36 8 4) Crater Lake 641 31 20 5) Mount Rainier 976 32 6) Rocky Mountain 1,049 31 7) Yosemite 2,083 25 15 8) Sequoia-Kings Canyon 3,389 23 9) Glacier-Waterton 4,627 7 10 10) Grand Teton-Yellowstone 20,736 4 5 9 10 0 4 6 8 10 REFUGE AREA (LOG SQ. Km) Figure 8-2. Habitat area and loss of large animal species in North American parks (1986) (Newmark, 1987). 154 Biological Diversity Reserves that originally may have been well often respond more easily to changing conditions on sited to protect a vulnerable population and its a slope because a shorter distance is required to habitat may, after climate change and population migrate to achieve the same temperature change. response, exist outside the now suitable range. Figure 8-3 illustrates this problem. Large reserves Among the problems associated with altitudinal and buffer zones around reserves help to lessen migration are displacement of the species at the top these problems. Corridors between reserves lessen (Peters and Darling, 1985). Also, with the increase the problem of spatial isolation by allowing for in altitude, the area available for colonization some migration between reserves. usually becomes smaller, communities become isolated, and these smaller populations are more DISTRIBUTION BEFORE MAN prone to extinction. A. N CLIMATE EFFECTS RESEARCH FUTURE RESERVE This section reviews some previous studies of ecological response to past changes in climate, SL recent studies of potential response to climate change, and studies done for this report, which use climate change scenarios from general circulation CURRENT DISTRIBUTION models for a doubled CO2 environment (see Table B. 8-1). Forest Ecosystems RESERVE The tree species that make up any forest are SL major factors in determining the biological diversity found there. Trees provide a multitude of habitats and are the basis of much of the food web in a forest. DISTRIBUTION WITH CLIMATE CHANGE C. Changes in forest composition resulting from New SL climate change (see Chapter 5: Forests) would have FORMER significant implications for biological diversity. RESERVE Potential northerly range shifts of several hundred to a thousand kilometers may be limited by the tree Old SL species' ability to disperse. One possibility is that southern pine forests will move farther north into the regions currently occupied by mixed hardwood Figure 8-3. Effect of climate change on biological species. Some of these hardwood forests contain reserves. Hatching indicates the following: the highest tree species diversity found anywhere in (A) species distribution before human habitation the United States (Braun, 1950). If they migrated (SL indicates southern limit of species range); north, species would inevitably be lost, and overall (B) fragmented species distribution after human biological diversity would substantially decrease. habitation; (C) species distribution after warming (Peters and Darling, 1985). If forests were disrupted by the extinction of the dominant tree species, the land would be Mountain Species invaded by weedy, opportunistic species. This would create a system with very low diversity, similar to Just as species can migrate latitudinally, they that following logging. Ultimately, these new can respond altitudinally to climate change by systems would not persist as succession took place, moving up or down a mountain slope. Species can but the pattern of succession following the removal of a forest by rapid climate change is unknown. 155 Chapter 8 Table 8-1. Studies Conducted for This Report and Freshwater Ecosystems Cited in This Chapter A study conducted by Magnuson et al. (Volume E) concludes that in most areas of the Potential Responses of Great Lakes Fishes and Great Lakes, climate warming would increase the Their Habitat to Global Climate Warming - amount of optimal thermal habitat for warm-, cool- Magnuson, Regier, Shuter, Hill, Holmer, and and coldwater fishes (see Chapter 15: Great Lakes). Meisner, University of Wisconsin (Volume E) Although overall productivity would increase, overall biological diversity could decrease through The Effects of Global Climate Change on the intensified species interactions. Water Quality of Mountain Lakes and Streams - Byron, Jassby, and Goldman, University of A study by Byron et al. (Volume E) on California at Davis (Volume E) mountain lakes suggests that climate change would cause a range of impacts, including higher The Effects of Climate Warming on Lake Erie productivity, changes in species composition, and Water Quality - Blumberg and DiToro, decreased water quality resulting from an increase HydroQual, Inc. (Volume A) in algal growth (see Chapter 14: California). Blumberg (Volume A) found that thermal Ecological Effects of Global Climate Change: stratification in Lake Erie could decrease dissolved Wetland Resources of San Francisco Bay - oxygen levels. Josselyn and Callaway, San Francisco State University (Volume E) The combined pressures of warmer waters, saltwater intrusion, and a rising sea level would Projected Changes in Estuarine Conditions significantly affect estuaries. The regional studies Based on Models of Long-Term Atmospheric suggest that coastal estuaries would see a growth in Alteration - Livingston, Florida State University marine species and a loss of some estuarine species. (Volume E) A study by Josselyn (Volume E) on the San Francisco Bay estuary suggests a decline in species that use the delta for spawning (see Chapter 14: California). Livingston (Volume E) concluded that Tropical Forest Ecosystems crabs, shrimp, oysters, and flounder in the Apalachicola estuary could not survive the warming The greatest concentration of biological in the GISS and GFDL scenarios (see Chapter 16: diversity in the world is in the rain forests of the Southeast). Tropics (Wilson, 1988). Besides reducing diversity, deforestation contributes to disruption of the global Saltwater Ecosystems carbon cycle by releasing CO₂ into the atmosphere and will directly affect the rate of climate change In general, a warmer global climate would (Prance, 1986). Indeed, on a global scale, the increase productivity in ocean fisheries, but the problems of tropical deforestation, rapid climate location and relative abundance of species are likely change through (among other factors) increased to change (Sibley and Strickland, 1985). Up to CO₂ production, and the loss of biological diversity some threshold temperatures, such as 2°C (4°F), can be seen as aspects of the same problem. warmer ocean temperatures would increase ocean productivity in many species, but beyond that Tropical forests are also important as wintering threshold, productivity could decline (Glantz, grounds for migratory birds coming from the United Volume J). It is likely that as productivity decreases, States and as sources of new knowledge, because biological diversity would decrease as well. Warmer the patterns of interactions between species and temperatures would most likely cause fish to climate are at their most sensitive and complex migrate poleward, although many other factors, such there (Robinson, 1978; Janzen, 1986). The Tropics as shifts in upwelling, may affect this. may provide important leading indicators of the ecological effects of climate change. 156 Biological Diversity Coral Reef Ecosystems Migratory Birds Coral reefs provide the structural base for the Migratory waterfowl are likely to experience very biologically diverse reef ecosystems. Coral very mixed effects as a result of warmer reefs in the Caribbean and the Pacific may be temperatures (Boyd, 1988). Herbivorous, arctic severely stressed as a result of warmer water nesting species, such as geese, could benefit from temperatures and the rising sea level associated with the shortened winter season and from the increases climate change. Extensive bleaching of coral (the in vegetation, in nesting habitat, and in ecosystem expelling of symbiotic algae in response to productivity (Harington, 1987). Smaller arctic environmental stress) occurred in the Pacific after nesting shorebirds, on the other hand, would be the 1982-83 El Niño (Glynn, 1984) and in the harmed by the encroachment of taller vegetation, Caribbean following a summer of elevated water potentially eliminating the preferred low-lying temperatures in 1987 (Roberts, 1987). Loss of the tundra breeding ground. Other effects on algae, the primary food source of the coral, is shorebirds could result from changes in ecosystem thought to kill coral, making the reef ecosystem predator-competitor relationships and changes in vulnerable to erosion and physical devastation. the seasonal timing of such events as larval blooms, upon which these birds depend for nourishment Coral reefs also will very likely be affected by while they are in a flightless stage and during sea level rise. Studies by Buddemeier and Smith migration (Myers, 1988). (1988) and Cubit (1985) suggest that vertical accretion of reef flats eventually may be unable to Waterfowl that breed in the continental interior keep up with an accelerating rise in sea level. Reef may suffer more than arctic nesters. Over half of flats also may be subject to the stress of increasingly all waterfowl in North America originate in the large waves, erosion, and sedimentation, which can prairie pothole region, a large agricultural area inhibit coral growth (Buddemeier, 1988). riddled with ecologically productive permanent and semipermanent wetlands. Increased temperature Arctic Ecosystems and changes in seasonal precipitation could reduce the highly variable number of potholes (wetlands) Within the North American Arctics, plant size, in the area and could significantly impair the vigor, and reproduction could be expected to productivity of breeding ducks. increase with higher temperatures in the near term (years to decades). Some low-lying plants would Because of the drought of 1988, over 35% of most likely become upright, and there would be a the seasonal wetlands within the prairie pothole northerly movement of the tree line and all region were dry during the breeding season (U.S. vegetative zones (Edlund, 1986). Fish and Wildlife Service, 1988). The Fish and Wildlife Service forecast that only 66 million ducks Over the longer term, however, rising would migrate during the fall of 1988, a total of 8 temperatures may be a mixed blessing. Overall million fewer than in 1987 and the second-lowest biological productivity is likely to increase, and some migration on record (Irion, 1988). The productivity species may be able to increase their range. index for mallards (number of young per adult) was However, some arctic plant species are likely to be 0.8, which was down by over 20% from the out-competed by invading species, and many others historical average (U.S. Fish and Wildlife Service, would face the same type of problem that 1988). mountaintop species face: they would have nowhere to go once they reach the Arctic Ocean. Thus, Waterfowl and other migratory birds are likely native arctic species may be especially at risk. to be affected on both ends of their migratory Other arctic species may face their own problems. journey and at staging areas along the way. The For example, caribou would be severely harmed if loss of coastal wetlands, already an area of great rivers do not freeze for periods long enough to concern in the United States, reduces the amount of allow for migration. habitat available to waterfowl, creating population pressures on a limited resource. Of the 215 million 157 Chapter 8 acres of wetlands in the coterminous United States NATIONAL POLICY at the time of settlement, fewer than 99 million IMPLICATIONS acres (46%) remain (U.S. Fish and Wildlife Service, 1988). Loss of an additional 26 to 82% of existing coastal wetlands could occur over the next century Climate change presents new challenges for as a result of a 1-meter rise in sea level, saltwater policymakers, regulators, and resource managers. intrusion, and human development (see Chapter 7: Planning for climate change may help to minimize Sea Level Rise). Loss of wintering habitat along the the disruption to natural systems and facilitate Gulf of Mexico would affect many waterfowl, adaptation under changing conditions. Decisions including mallards, pintails, and snow geese. will need to be made in an environment of increased pressure on many other resources. The Tropics, the winter home for many species of migratory birds, may be significantly altered by Policies regarding rare and endangered species rapid climate change. The need to protect a species are likely to change as the number of species at risk in all parts of its range underscores the truly global greatly increases. As more species become stressed nature of the effects of rapid climate change on and potentially threatened by climate change, biological diversity (Terborgh, 1974). reevaluation of protection policies may be required. The tradeoffs between protection of individual Endangered Species species and species' habitats and the broader protection of biodiversity at the level of ecosystems Hundreds of species are currently listed as may need to be reexamined. As a part of this endangered in the United States, and several question, decisions concerning whether to protect thousand await consideration for that status. These existing communities or to foster establishment of species are likely to be stressed further as a result new communities may need to be made. of climate change. Management Options to Maintain Threatened and endangered species of the Biological Diversity Southeast would be very susceptible to the impacts of sea level rise. Some species potentially at risk in Only a limited number of techniques are that region include the Key deer, manatee, Florida available for maintaining biological diversity. panther, and Everglades kite (Breckenridge, 1988). However, these techniques can be adapted and Climate change could also greatly increase the intensified to meet the potentially great impacts of number of rare, threatened, and endangered species rapid climate change. in the United States. Maintenance of Native Habitats Other Direct and Indirect Stresses The most direct way to maintain biological As plant and animal species experience diversity is to manage land to retain ecosystems, increasing pressures from changes in temperature, communities, and habitats. This already has been precipitation, and soil moisture, so too will successfully undertaken on a broad scale by federal agriculture and urban water supplies. The changes and state governments and by private organizations. that result from the human response to climate Ecosystem conservation, especially as represented by change may have the greatest impact on biological the national parks and other large reserves, diversity. If the continental interior of North maintains much of our national biological diversity. America dries, for example, wetlands that dry out These ongoing efforts will be the crucial first step may be cultivated, and our current uses of water for maintaining biological diversity in the face of resources may change. These secondary effects may climate change. significantly compound the loss of biological diversity. 158 Biological Diversity Land acquisition and management policies species is propagated in captivity. Indeed, some should take climate change into account. Climate rare species, such as the Pere David deer and the change and the future requirements of whole California condor, now exist only in captivity. This ecosystems should be considered in siting and technique can be made to work for a variety of managing reserves. To preserve functioning species, depending on their biology and the degree ecosystems, large areas of land will be required. to which they successfully adapt to captive Preserves would need to be at least large enough to conditions. As more species become threatened support self-sustaining populations. Lands that with extinction due to climate change, the effort could be more important as future plant and animal applied in this area may have to increase habitats need to be identified and evaluated. Land dramatically. However, only a tiny fraction of the managers should consider whether these lands nation's species can be maintained in this way. should be set aside. Although identification of Existing seed bank programs also provide an appropriate future habitats is difficult and highly important method for conserving plant genetic dependent on the future rate and extent of climate diversity. change, some areas, such as lowland areas adjacent to current wetlands, hold good potential for habitat Restoration of Habitat protection. Restoration ecology is a new discipline whose To protect a species, alternative sites should goal is to develop methods to restore damaged be considered with regard to the ecological needs of ecological communities to their prior unaltered target species under changing conditions. Siting state. Except in forestry, where reforestation has a reserves in mountainous areas is beneficial because longer tradition, restoration ecology has been in it allows for the shorter-distance altitudinal shifts of existence for only a few years. Nonetheless, it offers adjustment to changing climate. Stream corridors, some real promise for ameliorating the effects of which can be effective avenues of dispersal for rapid climate change. terrestrial as well as aquatic organisms, should be protected wherever possible. Providing corridors Normally, restoration is done at the site where for migration between reserves also should enhance the community previously existed and was altered or the ability of wildlife to adapt to climate change. damaged. Historical and baseline information is Ideally, these corridors should be wide enough to used to manage the species in such a way as to maintain the ecosystem characteristics of the reserve eliminate unwanted new species and to encourage in their center. Some species do not find the and possibly reintroduce native species. habitat conditions of narrow corridors suitable for migration. Perhaps the theory and practice of restoration ecology could be expanded to include rebuilding The pressures caused by changing climate are natural communities on sites where they have not likely to exacerbate competing land-use demands. previously existed. This activity has not yet been Acquisition of land for preserving biological attempted but may be necessary to save diversity will often be difficult, especially in areas communities displaced by climate change. If the where agriculture or forestry may be expanding. climate changes so that many of the key species of Flexible management strategies that reserve the a community can no longer survive in their original possibility of land management for biological range, and if the species are incapable of dispersing diversity in the future, while allowing for other use and establishing themselves elsewhere, then the in the interim, hold potential for reducing resource artificial transplantation of components of entire conflicts and maintaining biological diversity. communities may become necessary. This Creative approaches such as encouraging transplantation of communities would be a hedgerows, which may serve as migratory corridors, monumental task and could help to save much should also be considered. biological diversity, but it cannot possibly be undertaken on the scale necessary to preserve all Maintenance of Species in Artificial Conditions species threatened by climate change. Restoration ecology can be useful for extending reserve When individual species are threatened with boundaries and for providing migratory corridors. extinction, a possible option is to ensure that the 159 Chapter 8 Planning Options The federal government manages an enormous amount of land and should consider management While there are only a few management options to preserve biological diversity on much of techniques to maintain biological diversity, many that land. The major management techniques of different groups in our society can implement them. habitat maintenance and restoration ecology could These groups can be divided into the private and be applied by the agencies actively responsible for public sectors. managing the nation's public lands. Many different groups in the private sector, ranging from private individuals to large RESEARCH NEEDS conservation organizations, will have an interest in maintaining biological diversity. However, all would The ability to protect biological diversity is need information about the current and probable severely restricted by a lack of knowledge regarding future state of biological diversity. The federal the rate of climate change, the precise nature of the government may be able to play a role here by change, how individual species will respond, and providing information on the state of biological how ecological balances will shift. Research should diversity, including the systematics and distribution be expanded in two areas: identification of of species; on the genetic variability of species; and biological diversity, and species interactions and on the distribution of communities and ecosystems. biological diversity. New management options for biological diversity should be derived from these The four major federal land management studies. agencies develop plans intended to lay out a comprehensive framework and direction for Identification of Biological Diversity managing federal land. Land Resource Management Plans, required for each national forest, define the direction of management in the First and most important, an intensified, better forest for the next 10 to 15 years. In addition, the coordinated research effort, involving both Forest Service prepares 50-year plans, as required systematics (organism classification) and ecology, is by the Resource Conservation Act. The National required to identify the biologically diverse Park Service prepares a General Management Plan resources of our country. There should be more for each unit in the system that defines a strategy coordination to identify U.S. plants and animals, for achieving management objectives within a 10- range maps, and habitat requirement information for those species. year time frame. A Statement for Management is also prepared for each national park and is evaluated every 2 years; this includes a The apparently simple task of identifying the determination of information needs. The Bureau species of plants and animals that exist in a given of Land Management's (BLM) Resource area is actually a major barrier to further Management Plans and the Fish and Wildlife understanding. Although common species are Service's Refuge Master Plans are prepared and usually easy to identify, serious problems are often revised as needed for BLM resource areas and encountered in attempts to determine whether a wildlife refuges (U.S. Department of the Interior, widespread group is, for example, one or two 1987). These periodic reviews of the management species. For example, there is currently no federally plans for public lands should include consideration sponsored Flora (listing of all known plants) of the United States. of the possible effect of climate change on biodiversity. Although it is necessary to describe the genetic diversity of our nation's species, it is difficult to do Some federal land management agencies are so in a direct fashion. What may be feasible is the beginning to devote resources to the climate change issue. The Forest Service, for example, has begun further development of population genetic theory planning the Forest Atmosphere Interaction (FAI), and of data that would predict the genetic diversity which will be concerned with the relationship of a species based on species' properties, such as between the atmosphere and our national forests. population size and habitat range variability. The FAI has been designated a priority research program for the Forest Service. 160 Biological Diversity The challenge in describing ecosystem diversity REFERENCES is to find the system of classification that best helps make decisions intended to minimize the loss of Bolin, B., B. Doos, J. Jager, and R. Warrick, eds. biological diversity. Such a system will most likely 1986. The Greenhouse Effect, Climatic Change, only be found through experience. For now, we and Ecosystems (SCOPE 29). Chichester, England: should continue with the many different approaches John Wiley and Sons. of ecosystem classification, and we should look for the strengths of each. Boyd, H. 1988. Impact of Climate Change on Arctic-Nesting Geese. Paper presented at Climate Species Interactions and Biological Institute Wildlife Symposium, Washington, DC; Diversity January 21. The second area to which research should be Braun, E.L. 1950. Deciduous Forests of Eastern devoted is the direct effects of climate on species North America. New York: Hafner Publishing and the indirect interactions of species with other Company. species dependent on climate. Comprehensive mapping of species' ranges along temperature and Breckenridge, R. 1988. Challenge of Assessing moisture gradients would provide valuable Climate Change Effects on Fish and Wildlife information. The direct effect of climate change on Resources in the Western United States. Paper vegetation needs to be better assessed, and more presented at the Western Association of Fish and estimates of species' dispersal rates would Wildlife Agencies, Western Division of the significantly improve our ability to identify species American Fisheries Society, Albuquerque, NM; July at greatest risk. 10-13. A variety of ecosystems within a diversity of Brown, L., W. Chandler, A. During, C. Flavin, L. climatic regions and terrains should be intensively Heise, J. Jacobson, S. Postel, C. Shea, L. Starke, studied using analog climate regions under changed and E. Wolf. 1988. State of the World 1988. climate conditions. Although an ecosystem's Worldwatch Institute. New York: W.W. Norton response under changing climate conditions will not and Co., Inc. be wholly predictable, modeling individual ecosystemic responses would enhance knowledge of Buddemeier, R. 1988. Impacts of Climate Change the likely effects. Further research on how species on Coral Reefs, Islands and Tropical Coasts. Paper interact and how trophic structures might change presented at the Climate Institute Workshop on the with climate would help predictive capabilities. Impact of Climate Change on the Third World, Washington, DC; March 24-25. There should be further study on the question of the relationship between ecosystem function and Buddemeier, R., and D. Hopley. 1988. Turn-Ons species diversity to resolve the uncertainty in this and Turn-Offs: Causes and Mechanisms of the area. Modeling the effect of climate change on Initiation and Termination of Coral Reef Growth. ecosystem function and its relationship to diversity Paper presented for submittal to the Proceedings would help with predictive capabilities. of the Sixth International Coral Reef Symposium, James Cook University, Townsville, Australia; It will be impossible to study in detail even a August 8-12. fraction of the nation's species. The groups chosen for study either must be representative of many Buddemeier, R., and S. Smith. 1988. Coral Reef species or must possess some special properties Growth in an Era of Rapidly Rising Sea Level: (such as extreme sensitivity to climate change). The Predictions and Suggestions for Long-Term method of deciding which group to study is itself a Research. Coral Reefs 7:51-56. major outstanding research question. 161 Chapter 8 CEQ. 1980. Council on Environmental Quality. Kiester, A.R., R. Lande, D.W. Schemske. 1984. Environmental Quality: Eleventh Annual Report. Models of Coevolution and Speciation in Plants and Washington, DC: U.S. Government Printing Office. Their Pollinators. American Naturalist 124(2):220- 243. Cubit, J. 1985. Possible Effects of Recent Changes in Sea Level on the Biota of a Caribbean Reef Flat Leatherman, S. 1987. Effects of Sea Level Rise on and Predicted Effects of Rising Sea Levels. Beaches and Coastal Wetlands. Climate Institute, Preprint from the Proceedings of the Fifth Proceedings of the First North American International Coral Reef Congress, Tahiti. Conference on Climate Change, Washington, DC, pp. 140-146. Edlund, S. 1986. Modern Arctic Vegetation Distribution and Its Congruence With Summer Lovejoy, T. 1979. Refugia, refuges and minimum Climate Patterns. Environment Canada, critical size: problems in the conservation of the Proceedings: Impact of Climatic Change on the neotropical herpetofauna. In: Duellman, W., ed. Canadian Arctic. The South American Herpetofauna: Its Origin, Evolution and Dispersal. University of Kansas Ford, Michael. 1982. The Changing Climate. Museum Natural Hist. Monograph 7:1-485. London: George Allen and Unwin. MacArthur, R., and E. Wilson. 1967. The Theory Frye, R. 1983. Climatic Change and Fisheries of Island Biogeography. Princeton, NJ: Princeton Management. Natural Resources Journal 23:77-96. University Press. Glynn, P. 1984. Widespread Coral Mortality and the Mayr, E. 1963. Animal Species and Evolution. 1982-83 El Niño Warming Event. Environmental Cambridge, MA: The Becknap Press of Harvard Conservation 11(2):133-140. University Press. Harington, C. 1987. The Impact of Changing Myers, J.P. 1988. Remarks to Climate Institute Climate on Some Vertebrates in the Canadian Wildlife Symposium. Washington, DC; January 21. Arctic. Climate Institute, Proceedings of the First North American Conference on Preparing for Myers, N. 1979. The Sinking Ark. New York: Climate Change. Pergamon Press. Hunt, C.B. 1972. Geology of Soils. San Francisco, Naveh, Z., and A. Lieberman. 1984. Landscape CA: W.H. Freeman. Ecology. New York: Springer-Verlag. Hunt, C.B. 1974. Natural Regions of the United Newmark, W.D. 1987. A land-bridge island States and Canada. San Francisco, CA: W.H. perspective on mammalian extinctions in western Freeman. North American parks. Nature 325:430-432. Irion, R. 1988. Drought Helps Two Endangered Peters, R., and J. Darling. 1985. The greenhouse Species Rebound, but It's a Dismal Year for Ducks. effect and nature reserves. Bioscience 35(11):707- The Washington Post, August 1. 717. Janzen, D. 1986. The Future of Tropical Ecology. Prance, G.T. 1986. Tropical Rain Forests and the Annual Review of Ecology and Systematics 17:305- World Atmosphere. AAAS Selected Symposium 324. No. 101. Boulder, CO: Westview Press, pp. xxi-105. Jordan, W., R. Peters, and E. Allen. 1988. Regier, H., J. Holmes, J. Meisner. 1987. Likely Ecological Restoration as a Strategy for Conserving Effects of Climate Change on Fisheries and Biological Diversity. Environmental Management Wetlands, With Emphasis on the Great Lakes. 12(1):55-72. 162 Biological Diversity Climate Institute, Report for the First North Tiner, R. Jr. 1984. Wetlands of the United States: American Conference on Preparing for Climate Current Status and Recent Trends. U.S. Fish and Change. Wildlife Service. Newton Corner, MA: Habitat Resources. Roberts, L. 1987. Coral bleaching threatens Atlantic reefs. Science 238:1228-9. Topping, J.C., and J. Bond. 1988. The Potential Impact of Climate Change on Fisheries and Wildlife Robinson, M.H. 1978. Is tropical biology real? in North America. Report of the Climate Institute Tropical Ecology 19(1):30-50. to the U.S. EPA. Roush, J. 1986. Private action for public lands. U.S. Department of the Interior. 1987. National The Nature Conservancy News 36(4):4-7. Park Service and National Recreation and Park Association. Trends 24(2):2-43. Sibley, T., and R. Strickland. 1985. Fisheries: some relationships to climate change and marine U.S. Fish and Wildlife Service. 1981. Endangered environmental factors. Chapter 5. In: White, M.R., Means There's Still Time. Washington, DC: ed. Characterization of Information Requirements Department of the Interior, U.S. Fish and Wildlife for Studies of CO₂ Effects. Washington, DC: U.S. Service. Department of Energy. U.S. Fish and Wildlife Service. 1988. Endangered Strain, B. 1986. The biosphere and links to Species Technical Bulletin; May. climate. In: Rosenzweig, C., and R. Dickinson, eds. Climate-Vegetation Interactions. Proceedings of a Wilcox, B. 1982. Biosphere Reserves and the NASA/Goddard Space Flight Center Workshop; Preservation of Biological Diversity. Towards the January 27-29. Boulder, CO: Office for Biosphere Reserve: Exploring Relationships Interdisciplinary Earth Studies, University Between Parks and Adjacent Lands. Kalispell, MT: Corporation for Atmospheric Research. Parks Canada and U.S. National Park Service; June. Strain, B. 1987. Direct effects of increasing Wilson, E.O. 1988. Biodiversity. Washington, DC: atmospheric CO2 on plants and ecosystems. Tree National Academy Press. 2(1):18-21. WRI. 1988. World Resources Institute. World Terborgh, J. 1974. Preservation of natural Resources Report. New York: Basic Books. diversity: the problem of extinction prone species. BioScience 24(12):715-722. 163 CHAPTER 9 WATER RESOURCES FINDINGS to evaporative cooling. New plants may also locate in coastal areas to obtain a water source that is reliable and that may be used without violation of Higher temperatures will most likely result in thermal restrictions, although sea level rise could be greater evaporation and precipitation; earlier a problem. This would have important implications snowmelt and reduced water availability in summer; for land use, transmission lines, and the costs of and, during dry periods, more rapid declines in soil power. moisture and water levels, volumes, and flows. Although a general warming and global increase in Where water availability is reduced, conflicts precipitation are likely, the distribution of among users could increase. These include precipitation is highly uncertain and may change in conflicts over the use of reservoir systems for unexpected ways. As a result, the frequency, flood control storage, water supply, or flow seasonality, variability, and spatial distribution of droughts, water availability constraints, floods, and regulation; and conflicts over water rights among agricultural, municipal, and industrial water quality problems will very likely change. users of water supply. Some regions could benefit from changing precipitation patterns, while others could experience Should extreme flood events become more great losses. frequent in a river basin as a result of earlier snowmelt and increased precipitation, activities Although great uncertainty is associated with the located in the floodplain would endure more projection of future hydrologic conditions and their water-use implications, we must be most concerned damages or could require more storage about current vulnerabilities to climate extremes capacity (whether by construction, reallocation, that could become exacerbated under climate or changes in operating procedures), often at the expense of other water uses. change. For instance, certain dry regions could become more vulnerable to drought as a result of higher temperatures, earlier snowmelt, and/or shifts Policy Implications in precipitation. Water management responses to current vulnerabilities are available and in use, and can Impacts on Water Uses appropriately be brought into play to respond to changing hydrologic conditions. These responses If climate in a given region were to become include the following: warmer and drier, water availability would decrease and water demand would increase, Build new storage capacity, provided that the especially demand for irrigation and electric structures show positive net benefits under a power production. variety of possible climatic conditions; Lower riverflows resulting from drier Modify water system operations to improve conditions could adversely affect instream uses performance under extreme conditions, to such as hydropower production, navigation, enhance recovery from extreme conditions, aquatic ecosystems, wildlife habitat, and and to accept greater risk to low-valued uses to recreation. protect high-valued uses; and Lower streamflow and lower lake levels could Encourage a reduction in water demand and cause powerplants to shift from once-through an increase in water-use efficiency through 165 Chapter 9 conservation, water markets, water quality owing to weather variability, 675 bgd of the 1,435 control, drought contingency planning, and bgd of runoff water in the coterminous United coordinated uses of regional and interstate States is considered to be available for use in 95 water resources, provided that such measures years out of 100 (Figure 9-1). do not reduce the performance and recovery capabilities of supply systems. Surface and groundwaters are managed by controlling and diverting flows through impoundments and aqueducts; by withdrawing water IMPACTS OF CLIMATE CHANGE for such "offstream" applications as irrigation and municipal use; by regulating flows to maintain ON THE WATER RESOURCES IN "instream" water quality and such uses as navigation, THE UNITED STATES hydropower, and recreation; and by controlling flows under flood conditions to avoid loss of life, damage to property, or inconvenience to the public. Current Status of Water Resources Water may be "withdrawn" and returned to the source more than once, or "consumed" and not The potential effects of climate change on returned to the source. water resources must be examined within the context of the existing and projected supply of, and In 1985, freshwater withdrawals for offstream demands for, water. uses totaled 338 bgd. Of the withdrawals, 92 bgd were consumed, mostly for irrigation. Withdrawals The United States is endowed with a bountiful and consumption of freshwater by major offstream supply of water, but the water is not always in the uses in 1985 are summarized in Figure 9-1. right place at the right time, or of the right quality. On the average, 4,200 billion gallons per day (bgd) Our investment in water infrastructure is of precipitation fall on the lower 48 states. substantial. Water supply for municipal and However, a large portion of this water (66%) industrial use represented a $108 billion national evaporates, leaving 1,435 bgd (34%) for surface investment in infrastructure in 1984 (National water runoff and groundwater recharge. Largely Council on Public Works Improvement, 1988). Government agencies and industries spent $336 WITHDRAWALS CONSUMPTION RETURN FLOWS 76 EVAPORATION 2765 IRRIGATION/LIVESTOCK 140 64 4 THERMOELECTRIC POWER 131 127 PRECIPITATION 4200 DOMESTIC/COMMERCIAL 36 29 5 INDUSTRIAL/MINING 31 26 WITHDRAWALS 338 RETURN FLOW SURFACE/ 246 GROUND-WATER FLOWS SURFACE/ 1435 INSTREAM/SUBSURFACE USE GROUND- 1097 WATER FLOW TO OCEANS 1343 NUMBERS INDICATE BILLIONS OF GALLONS PER DAY Figure 9-1. Water withdrawals and consumption by offstream uses, coterminous United States, 1985 (Solley et al., 1988). 166 Water Resources billion (in constant 1982 dollars) from 1972 to 1985 gap between demand for water and available supply (Farber and Rutledge, 1987) on water pollution is narrow, or the variability in water supply is high, abatement and control activities. In other areas, or both. For example, average surface water excess water periodically floods agricultural and withdrawal exceeds average streamflow in the Great urban areas, causing annual average damages valued Basin, Rio Grande, and Colorado River Basins. In at $3 billion (in constant 1984 dollars) during the these water-short basins, offstream water uses often past decade (National Council on Public Works conflict with instream uses, such as recreation and Improvement, 1988). maintenance of environmental quality. Degraded water quality further limits water availability in On a national scale, water supplies are many regions. Table 9-1 summarizes the current adequate, and water availability exceeds withdrawals status of water supply by major river basin. The and consumption. However, in some regions, the regions are delineated in Figure 9-2. Table 9-1. Current Status of Water Supply Average Stream- Ground- renewable With- Consump- flow water supply drawalsᵇ tionᵇ Reservoir exceeded overdraft River basin (bgd)ᵃ (1985) (1985) storageᵇ 95% of timeᶜ (%) New England 78.4 11.7 0.9 15 62.4 0 Mid-Atlantic 80.7 29.5 2.1 11 62.2 1.2 South Atlantic-Gulf 233.5 13.5 2.1 15 55.5 6.2 Great Lakes 74.3 42.9 2.9 8 62.6 2.2 Ohio (exclusive of 139.5 22.3 1.5 13 59.0 0 Tennessee region) Tennessee 41.2 22.3 0.8 24 77.0 0 Upper Mississippi (exclusive of Missouri region) 77.2 21.9 2.3 14 54.0 0 Mississippi (entire basin) 464.3 3.7 1.2 32 46.7 0.5 Souris-Red Rainy 6.5 4.3 1.9 110 32.1 0 Missouri 62.5 55.2 20.3 120 40.7 24.6 Arkansas-White-Red 68.6 22.3 11.7 41 36.5 61.7 Texas-Gulf 33.1 41.4 18.2 67 27.5 77.2 Rio Grande 5.1 109.8 43.7 182 33.3 28.1 Upper Colorado 14.7 51.4 16.3 229 39.0 0 Colorado (entire basin) 15.6 47.4 27.2 403 75.0 48.2 Great Basin 9.9 81.8 36.4 30 50.0 41.5 Pacific Northwest 276.2 12.9 4.5 20 70.7 8.5 California 70.2 53.6 29.9 49 44.0 11.5 Alaska 975.5 0.4 0.0 0.1 78.5 0 Hawaii 7.4 17.2 1.8 0.1 60.3 0 Caribbean 5.1 11.9 3.2 5 35.6 5.1 ᵃAverage renewable supply is defined as the average flow potentially or theoretically available for use in the region; units = billion gallons per day. ᵇWithdrawals, consumption, and reservoir storage are expressed as a percentage of the average renewable supply. ᶜAs a percentage of average streamflow. Source: U.S. Water Resources Council (1978); U.S. Geological Survey (1984); Solley et al. (1988). 167 Chapter 9 SOURIS-RED RAINY REGION 9 PACIFIC NORTHWEST NEW ENGLAN MISSOURI REGION 10 GREAT LAKES REGION 4 UPPER MISSISSIPPI MID-ATLANTIC GREAT BASIN REGION 7 REGION 16 OHIO REGION 5 UPPER COLORADO REGION 14 CALIFORNIA REGION 18 REGION ARKANSAS WHITE RED TENNESSEE LOWER COLORADO REGION 11 RIO GRANDE REGION 13 LOWER MISSISSIPPI REGION SOUTH REGION 3 ATLANTIC GULF 8) TEXAS-GULF REGION 12 HAWAII ALASKA REGION 20 REGION 19 CARIBBEAN REGION 21 Puerto Rico Figure 9-2. Water resources regions (U.S. Geological Survey, 1985). Water supply and use have changed on instream uses have made diversion of water for significantly during the past decade. For the first such applications as agriculture in the West and for time since 1950, when the United States Geological powerplant cooling in the East more difficult. Survey began recording water withdrawals, national total fresh and saline water withdrawals dropped Climate Change, Hydrologic Conditions, 10% from 1980 to 1985 (from 443 billion gallons to and Water Resources 399 billion gallons, of which 338 billion gallons were freshwater) (Solley et al., 1988). Increased As shown in Figure 9-3, weather controls conservation and water recycling in agriculture, hydrologic conditions through precipitation (mean industry, and energy production, slower growth in and frequency), runoff, snowmelt, transpiration and energy demand, and decline in availability of new evaporation, soil moisture, and the variability of water supply reduced or tempered water use in all storms and drought. In turn, the ability to use sectors (Solley et al., 1988). Withdrawals declined water resources is greatly influenced by variability in by 7% in irrigation, by 33% in industry, and by 13% hydrologic conditions. in thermal power during the same period. Of the major users, only municipal/domestic water supply Climate change will affect both the supply of increased (by 7%). and demand for water. Figure 9-4 outlines the major potential impacts of global warming and The value of instream uses has risen relative changes in precipitation on water resources. to that of offstream uses. Navigation and hydropower have retained their importance as If climate warms in the United States, there society has begun to place greater value on will likely be greater evaporation and, in turn, wastewater dilution, ambient water quality, fish and greater precipitation; earlier snowmelt and, in turn, wildlife habitats, and recreation. Higher values reduced water availability in summer; and, during 168 Water Resources ATMOSPHERIC MOISTURE 40,000 bgd PRECIPITATION 4,200 bgd EVAPORATION AND TRANSPIRATION FROM SURFACE-WATER BODIES, LAND SURFACE AND VEGETATION 2,800 bgd CONSUMPTIVE USE 100 bgd EVAPORATION FROM OCEANS WELL RECHARGE STREAMFLOW TO OCEANS WATER TABLE 1,230 bgd TOTAL SURFACE AND GROUND-WATER FLOW TO OCEANS FRESH GROUND WATER 1,300 bgd INTERFACE OCEAN, SALINE GROUND WATER bgd=billion gallons per day Figure 9-3. Hydrologic cycle showing the gross water budget of the coterminous United States (Langbein et al., 1949; Solley et al., 1983). dry periods, more rapid declines in soil moisture than in humid areas. In addition, each degree of and water levels, volumes, and flows. Over the very temperature increase causes a relatively greater long term, groundwater availability may be affected decline in runoff and water availability in arid by altered recharge rates. Transpiration may not regions as compared with humid regions. If increase as much because increased levels of carbon regional climate becomes warmer and drier, more dioxide may shrink the stoma or pores of plants vulnerability to interruptions in water availability (Rosenberg, 1988). Although general warming is may be observed. likely to occur, the distribution of precipitation is highly uncertain and may change in unexpected As a result, the frequency, seasonality, ways. variability, and spatial distribution of droughts, water availability constraints, floods, and water Earlier studies have shown that small changes quality problems would probably change. In many in regional temperature, precipitation, and locations, extreme events of dryness and flooding evaporation patterns can cause significant changes could become more frequent. Some regions may in water availability, especially in arid areas (see experience more drought conditions, others more Nemec and Shaake, 1982; Klemes and Nemec, 1985; flooding, others degraded water quality, and others Beran, 1986). Precipitation is more variable in arid a combination. 169 Chapter 9 Climate Change Temperature Increase in Regional Weather Variability All Regions Increased demand Greater evapotranspiration Less precipitation More precipitation Soil moisture loss for air conditioning Less runoff and More runoff and Earlier snowmelt streamflow streamflow Reduced water supply Increased flooding in hotter, drier in hotter, wetter regions regions Increased demand for Increased demand cooling water for Increased demand for flood electric power for irrigation control production Increased water Increased surface consumption and water withdrawals groundwater mining Adverse effects Conflicts between Conflicts between Conflicts between on water off-stream and irrigation and flood control and quality in-stream uses municipal/ industrial all other uses uses Storage/supply Nonstructural/demand policy alternatives policy alternatives NOTE: Figure does not trace the impacts of reduced flows and increased evapotranspiration on navigation, hydropower, and the Great Lakes. Figure 9-4. National impacts of climate change on water supply and demand. 170 Water Resources Global warming may have a significant impact instream uses that depend on levels and on the demand for water in some regions. Warmer flows; and temperatures may raise the demand for air- conditioning in the South without a proportionate domestic supplies that are vulnerable to decrease in demand for electric heat. Increased hazardous and toxic substances in ground demand for cooling water for electricity powerplants and surface water. would result (see Chapter 10: Electricity Demand). Warmer temperatures may also prompt more Table 9-2 highlights the vulnerability of major farmers to irrigate crops (see Chapter 6: water uses in each region to climate change. Agriculture). Irrigation Impacts of Climate Change on Water Uses Irrigation accounts for 42% of freshwater withdrawals and 82% of freshwater consumption in Models of global climate change do not yet the United States. Although irrigated land provide reliable data to predict regional changes in comprises about 10% of harvested cropland acreage the water supplies; however, we can indicate nationwide, it contributes 30% of the value of possible directions of impacts and the water uses cropland production. Many of these crops are and sectors affected. fruits, vegetables, and specialty crops (U.S. Water Resources Council, 1978; Bajwa et al., 1987). The The following sections outline the potential 17 western states account for 85% of the irrigated impacts of climate change on offstream and lands in the country (Bajwa et al., 1987). instream water uses. The uses most likely to be affected are those currently vulnerable to water Water-short western states are exploring quantity and quality constraints: numerous options for minimizing water requirements. Because of depleted groundwater irrigation, the major source of withdrawals supplies, the rising cost of obtaining groundwater, and consumption in the West; and the high cost and limited availability of sites for new surface water developments, irrigated acreage thermal power production, a major source has stabilized or is declining in some areas of the of heat effluent and evaporative West (Solley et al., 1988). Groundwater pumping consumption, especially in the East; for irrigation has already started to decline in the Table 9-2. Potential Regional Impacts of Climate Change on Water Uses: Areas of Vulnerability Arid Western Use Pacific River Northwest California Basins Great Plains Great Lakes Mississippi Southeast Northeast Irrigation X X X X X Thermal X X power Industrial X X X Municipal/ X X X domestic Water quality X X X X X X X Navigation X X X Flood control X X X X X Hydropower X X X X Recreation X X 171 Chapter 9 southern Great Plains States and in Arizona, temperatures raise air-conditioning use (see Chapter although the impacts on production have been 10: Electricity Demand). If streamflows are mitigated by the adoption of more efficient reduced as a result of climate change, powerplants irrigation systems and by a switch to crops offering using once-through cooling could be adversely higher returns to water (Frederick and Kneese, affected. Increased demand for power may 1989). In contrast, supplemental irrigation is rising reinforce existing trends in powerplant design in the Southeast, largely because of expansion in toward evaporative cooling, and in powerplant siting Georgia (Bajwa et al., 1987). toward coastal locations. With less water available, low-flow conditions may interrupt power production Climate change may significantly affect and may increase power production costs and agriculture. Summer drought and earlier runoff are consumer electricity prices. likely to change agricultural practices and increase demands for irrigation in most areas east of the Industrial Uses Rocky Mountains. Since 1954, self-supplied industry steadily used Thermal Power Generation less and less water per unit of production (Solley et al., 1988). This decline was partly due to Steam electric powerplants withdraw almost as efficiencies achieved to comply with federal and much freshwater as irrigation but consume much state water pollution legislation that restricts the less than irrigation. Although the freshwater discharge of untreated water. The trend toward withdrawn to produce the nation's electricity totals more efficient industrial water uses is likely to 131 bgd, only 4.35 bgd are actually consumed continue. (Solley et al., 1988). In regions where flows are reduced, there Future demand for water for power could be a reduction in both the quantity and the production will depend on energy demand, quality of water available for industrial production. technology, and on federal and state regulations In addition, if the climate becomes drier, the governing instream water quality, instream flow, and potential for interruption of industrial supply will be thermal pollution. Although a large amount of increased. installed capacity exists along eastern rivers, freshwater withdrawals by powerplants in the East Domestic Water Uses have decreased, and siting of plants in coastal areas has increased, so that by 1987, 30% of installed Domestic uses account for 10% of total water capacity in coastal areas used saline surface water withdrawn and 11% of consumption. Over the past (Solley et al., 1988). In addition, the thermal 20 years, domestic water use has increased from 16 regulations have caused a shift in the design of new to 25 bgd owing to growth in the number of cooling systems from once-through cooling, which households, with little change in usage per discharges heat back into the water sources, to household (Solley et al., 1988). evaporative cooling with towers and ponds (Breitstein and Tucker, 1986). Although Most municipal water supply systems are evaporative cooling alleviates thermal pollution, it designed to provide reliable water at all times (safe increases water consumption. yield). However, urban growth depends upon developed water supply, which is approaching During droughts, federal and state regulations exhaustion in some areas. For instance, in the protecting instream uses and limiting thermal Southeast and parts of the West, a large percentage discharges may constrain withdrawals for of municipal water supply comes from groundwater powerplant cooling. In addition, powerplant water (U.S. Water Resources Council, 1978; Solley et al., needs on some eastern rivers are so large that 1988). These regions withdraw more groundwater insufficient water may be available to dissipate heat than can be recharged; consequently, any increased during low-flow conditions (Hobbs and Meier, drought caused by climate change could accelerate 1979). groundwater mining (see Chapter 14: California and Chapter 16: Southeast). Demand for electric power and construction of new generating capacity may increase as warmer 172 Water Resources Municipalities in the West are purchasing and streams depends in part on water quantity. irrigators' water rights to ensure adequate water Water supply is needed for dilution of wastewaters supplies for urban growth. If climate change results that flow into surface and groundwater sources. in reduced municipal supply, this trend will continue Freshwater inflows are needed to repel saline waters or accelerate, leading to the loss of irrigated in estuaries and to regulate water temperatures in acreage. order to forestall changes in the thermal stratification, aquatic biota, and ecosystems of lakes, In the East, Midwest, and Southeast, streams, and rivers. municipalities may be able to increase safe yield by repairing and replacing existing leaking water The Federal Clean Water Act of 1972 and delivery systems and by consolidating fragmented subsequent amendments ushered in a new era of water supply districts. These actions could provide water pollution control. Massive expenditures for the margins of safety necessary to accommodate treatment facilities and changes in water-use climate change. practices by government and industry have decreased the amount of "conventional" water Navigation pollutants, such as organic waste, sediment, oil, grease, and heat, that enters water supplies. Total If riverflow and lake levels became lower, public and private, point and nonpoint, and capital navigation would be impeded. Systems that are and operating water pollution abatement and particularly vulnerable are those with unregulated control expenditures from 1972 to 1985 totaled $336 flows or levels and high traffic, such as the billion in 1982 dollars (Farber and Rutledge, 1987). Mississippi River and the Great Lakes. The effects of dry conditions and reduced water levels on barge Nevertheless, serious surface water quality traffic on the Mississippi in 1988 illustrate the problems remain. Groundwater pollution problems, potential impacts of climate change. especially toxic contamination and nonpoint source pollution, are receiving increased recognition (U.S. Hydropower EPA, 1987b). Because of the decline in water availability One-third of municipal sewage treatment that could result from climate change, hydropower plants have yet to complete actions to be in full output and reliability, which depend on flows, could compliance with the provisions of the Clean Water decline in the West and the Great Lakes. If the Act (U.S. EPA, 1987a). Federal and state Southeast became drier, it could face the same regulation of previously unregulated toxic and problems unless it sacrificed water supply reliability hazardous water pollutants has just begun. In the to maintain hydropower production. West, irrigation has increased the salinity levels in the return water and soils of several river basins Recreation (the lower Colorado, the Rio Grande, and the San Joaquin) to an extent that threatens the viability of If the Southeast becomes drier, there may be irrigation (Frederick and Kneese, 1989). an increase in the conflict among water uses, especially over reservoir releases and levels in the Should climate change involve reduced flows, Tennessee Valley and the Lake Lanier, Georgia, less freshwater may be available in some regions for system. The conflicts are among flood control, diluting wastewater salt and heat, especially in low- which relies on storage; recreation, which depends flow periods (Jacoby, 1989). Dissolved oxygen levels on stable reservoir pool elevations; and downstream in the water would decline while temperature and uses and water supply, which depend on flows. salinity levels would increase, affecting the viability of existing fish and wildlife. Increased thermal Climate Change and Water Quality stratification and enhanced algal production due to higher temperatures may degrade the water quality Water quality directly affects the availability of of many lakes (see Chapter 15: Great Lakes; water for human and environmental uses, since Blumberg and DiToro, Volume A). Finally, the water of unsuitable quality is not really "available." combination of declining freshwater availability Likewise, water quality in the nation's rivers, lakes, 173 Chapter 9 and rising sea level would move salt wedges up many large dams are designed to pass a "probable estuaries, changing estuarine ecology and maximum flood" (an extreme flood event much threatening municipal and industrial water supplies. greater than the 100-year flood). On the other hand, should climate change involve increased flows, greater dilution of pollutants would Smaller structures, such as urban drainage be possible in some regions. culverts and sewers and local flood protection projects, are currently more susceptible to failure Groundwater is the source for over 63% of and are in poorer condition than large structures domestic and commercial use (Solley et al., 1988). (National Council on Public Works Improvement, Although only a small portion of the nation's 1988). One-third of the non-federal flood control groundwater is thought to be contaminated, the dams inspected under the national non-federal dam potential consequences may be significant and may program were found to be unsafe, mostly owing to include cancer, damage to human organs, and other inadequate spillways (National Council on Public health effects (U.S. Congress, 1984). Works Improvement, 1988). The capacity of these non-federal, smaller, mostly urban flood control and Adequate recharge of aquifers is needed not stormwater structures is more likely to be exceeded. only to perpetuate supplies but also to flush Urbanization upstream from many dams and water contaminants. Should climate change result in control structures is already resulting in increased reduced flows and reduced recharge, the quality as impervious surfaces (such as pavement) and well as the available quantity of groundwater could increased peak runoff, making some structures be adversely affected. increasingly vulnerable to failure. Climate Change and Flood Hazards Climate Change and Conflicts Among Water Uses Because of the buffering and redundancy designed into large structures, major federal flood There is no doubt that climate change has the control projects may be able to contain or mitigate potential to exacerbate water availability and quality the impacts of more frequent severe floods. problems and to increase conflicts between regional However, continued performance for flood control water uses as a result. The foregoing discussion has may come at the expense of other uses. For highlighted a number of such conflicts: example, drawing down the levels of reservoirs to contain floodwaters from anticipated increases in conflicts between instream and offstream precipitation or earlier snowmelt may curtail water uses; availability for water supply. (This aggravated conflict is a distinct possibility in California, for conflicts among offstream uses, such as example; see Chapter 14: California.) agriculture, domestic use, and thermal power production; The major concern with existing dams and levees is the consequence of failure under extreme conflicts between water supply and flood conditions. For instance, an increased probability of control in the West; great floods, whether due to urbanization of upstream watersheds or to climate change, would conflicts between all uses and recreation in cause dams with inadequate spillways to fail. the Southeast; and (Spillways are designed to prevent dam failure through overtopping.) conflicts between thermal power production and instream uses, especially in the East. The majority of large dams that provide substantial flood storage are in good condition. The In some areas, increased precipitation due to National Dam Safety Inventory shows that the climate change could alleviate water quality/quantity overall condition of the U.S. Army Corps of problems and conflicts, but only after water Engineers' more than 300 flood control reservoirs is infrastructure is modified to accommodate the sound (National Council on Public Works increased probability of extreme events. Improvement, 1988). In addition, the spillways of 174 Water Resources REGIONAL IMPACTS OF climate change models and forecasts need to CLIMATE CHANGE address regional impacts. The regional studies conducted by the U.S. Water resources supply and management occurs Environmental Protection Agency for this document at the regional, river basin, state, and local levels. (see Table 9-3) examine the potential regional To be of use to water resources decisionmakers, impacts of climate change. (With the exception of Table 9-3. Regional Water Resource Studies California Interpretation of Hydrologic Effects of Climate Change in the Sacramento-San Joaquin River Basin, California - Lettenmaier, University of Washington (Volume A) Methods for Evaluating the Potential Impact of Global Climate Change - Sheer and Randall, Water Resources Management, Inc. (Volume A) The Impacts of Climate Change on the Salinity of San Francisco Bay - Williams, Philip Williams & Associates (Volume A) Great Lakes Effects of Climate Changes on the Laurentian Great Lakes Levels - Croley, Great Lakes Environment Research Laboratory (Volume A) Impact of Global Warming on Great Lakes Ice Cycles - Assel, Great Lakes Environment Research Laboratory (Volume A) The Effects of Climate Warming on Lake Erie Water Quality - Blumberg and DiToro, HydroQual, Inc. (Volume A) Potential Climatic Changes to the Lake Michigan Thermal Structure McCormick, Great Lakes Environment Research Laboratory (Volume A) Great Plains Effects of Projected CO2-Induced Climate Changes on Irrigation Water Requirements in the Great Plains States - Allen and Gichuki, Utah State University (Volume C) Southeast Potential Impacts of Climatic Change on the Tennessee Valley Authority Reservoir System - Miller and Brock, Tennessee Valley Authority (Volume A) Impacts on Runoff in the Upper Chattahoochee River Basin - Hains, C.F. Hydrologist, Inc. (Volume A) Methods for Evaluating the Potential Impact of Global Climate Change - Sheer and Randall, Water Resources Management, Inc. (Volume A) 175 Chapter 9 Allan and Gichuki (Volume C), all studies listed in Climate change may exacerbate water shortage Table 9-3 are found in Volume A.) The studies use and quality problems in the West. Higher scenarios generated from up to four global temperatures could cause earlier snowmelt and circulation models (GCMs) as their starting points runoff, resulting in lower water availability in the (see Chapter 4: Methodology) and match them with summer. Some GCM scenarios predict midsummer regional or subregional water resource models. drought and heat, less groundwater recharge, and This section reviews the findings from the studies on less groundwater and surface water availability for California, the Great Plains, the Great Lakes, and irrigation in the middle latitudes of the country. the Southeast; from previous studies of the impacts The sensitivity analyses conducted by Stockton and of climate change on these and other regions; and Boggess (1979) indicated that a warmer and drier from previous hydrologic studies and models of climate would severely reduce the quantity and individual river basins. quality of water in arid western river basins (Rio Grande, Colorado, Missouri, California) by The GCMs do not yet provide definitive increasing water shortages. Water shortages and forecasts concerning the frequency, amount, and associated conflicts between instream and offstream seasonality of precipitation and the regional uses, between agricultural and urban/industrial distribution of these hydrologic effects (see Chapter water uses, and between flood control and other 2: Climate Change; Chapter 3: Variability; Chapter water uses of reservoirs may be expected under 4: Methodology; Rind and Lebedeff, 1984; Hansen these scenarios. Hydropower output also would et al., 1986; Gleick, 1987; Rosenberg, 1988). The decline as a result of lower riverflow. uncertainty of the forecasts is partially due to the limitations and simplifications inherent in modeling Pacific Northwest complex natural and manmade phenomena. Modeling efforts are made more difficult by the The competition for water for irrigation, feedbacks and interconnections between changes in hydropower, and fisheries habitat is increasing in temperature; and the amount and frequency of the Pacific Northwest (Butcher and Whittlesey, precipitation, runoff, carbon dioxide, growth and 1986). Climate change may alter the seasonality transpiration of foliage, cloud cover, ocean and volume of precipitation and snowmelt, circulation, and windspeed. increasing the risk of flooding, changing reservoir management practices, and affecting the output and However, the regional studies commissioned reliability of hydroelectric power production and the for this report are a significant step in the effort to availability of water for irrigation. bring GCM and regional water resources models together to examine the regional impacts of climate California change. The diversion of water from water-rich northern The West California and from the Colorado River to southern California via federal and state systems of dams, The arid and semiarid river basins west of the aqueducts, and pumping stations has transformed Mississippi River have significant surface and California into the nation's leading agricultural state groundwater quantity and quality problems and are and has made possible the urbanization of southern vulnerable to restricted water availability. Total California. Irrigation accounted for 83% of the water use exceeds average streamflow in 24 of 53 total value of California's agricultural output in 1982 western water resource regions (U.S. Water (Bajwa et al., 1987). Because of this high economic Resources Council, 1978), with the majority of the dependence on water in an arid area, southern West's water withdrawals going to irrigation. California is vulnerable to droughts and any altered Surface and groundwater quality in the West have temporal pattern of runoff that may be caused by deteriorated as a result of low flow, salts atmospheric warming. concentrated by irrigation, and pesticide use. The West also depends upon nonrenewable groundwater Total annual runoff from the mountains supplies for irrigation (Solley et al., 1988). surrounding the Central Valley is estimated to increase slightly under GCM scenarios, but runoff in the late spring and summer may be much less 176 Water Resources than today because higher temperatures cause Plains). The region heavily depends on earlier snowmelt (Lettenmaier, Volume A). The groundwater mining (when pumping exceeds aquifer volume of water from the State Water Project may recharge) for irrigation. The region was severely decrease by 7 to 16% (see Chapter 14: California; affected during the "Dust Bowl" years of the 1930s Sheer, Volume A). Existing reservoirs do not have and suffered from severe drought in 1988. the capacity to increase storage of winter runoff and at the same time to retain flood control capabilities. Because of the greater reliability in irrigated In addition, flows required to repel saline water yields relative to dryland yields, the demand for near the major freshwater pumping facilities in the irrigation could rise (Allen and Gichuki, Volume C; upper Sacramento-San Joaquin River Delta may Adams et al., Volume C). Thus, while total have to be doubled as a result of sea level rise, agricultural acreage could decrease, irrigated further reducing water available to southern acreage and groundwater mining may increase in California (Williams, Volume A). the southern Great Plains. Greater demand may be placed on the Ogallala Aquifer, which underlies Decreases in water availability may also reduce much of the region, causing further mining of the hydroelectric power produced in California. In the aquifer. 1976-77 drought, hydroelectric production in northern California dropped to less than 50% of Great Lakes normal, a deficiency relieved by importing surplus power from the Pacific Northwest and by burning Based on analyses for this report (Croley and additional fossil fuels at an approximate cost of $500 Hartmann, Volume A), higher temperatures may million (Gleick, 1989). overwhelm any increase in precipitation and may evaporate lakes to below the lowest levels on Colorado, Rio Grande, and Great Basins record. However, changes in Great Lakes evaporation under climate change are highly Total consumption is more than 40% of uncertain and depend on such variables as renewable supply in these river basins. The basinwide precipitation, humidity, cloud cover, and Colorado River Basin has huge reservoir storage, windspeed. Under a possible set of conditions, lake but demand exceeds supply in the lower half of the levels could rise. The winter ice cover would be basin. Ordinarily all of the Colorado River's water reduced but would still be present, especially in is consumed before it reaches the Gulf of California shallow areas and northern lakes (Assel, Volume in Mexico. The Colorado River Compact of 1922, A). Navigation depths, hydropower output, and the 1963 Supreme Court decision in Arizona V. water quality all would be adversely affected, but California, the treaties with Mexico of 1944 and losses of existing shorelands from erosion would be 1973, and other agreements allocate Colorado River reduced as a result of lower lake levels (see Chapter water to seven states and Mexico (Dracup, 1977). 15: Great Lakes). Some studies show that the Upper Colorado region will use all of its allocation by the year 2000, Mississippi River reducing water hitherto available to lower Colorado and California (Kneese and Bonem, 1986). The Mississippi River historically has been affected by both spring floods and drought. In 1988, Climate change may further reduce the low flows due to drought received national availability of water in these basins. A model by attention. Low flows disrupt navigation, permit Stockton and Boggess (1979) of a 2°C temperature saltwater intrusion into the drinking water of increase and a 10% precipitation decrease shows southern Louisiana cities, reduce the dilution of decreases in the water supply in the upper Colorado contaminants transported from upstream locations, and the Rio Grande of 40 and 76%, respectively. and reduce the inflow of water to the vast Mississippi Delta wetlands (see Glantz, Volume J). Great Plains Northeast The southern Great Plains States of Kansas, Nebraska, Oklahoma, and Texas produce almost Although the Northeast is humid, cities and 40% of the nation's wheat, 15% of its corn, and powerplants demand large amounts of water at 50% of its fattened cattle (see Chapter 17: Great 177 Chapter 9 localized points in a watershed, necessitating storage of a global warming trend on the nation's water and interbasin transfers. Because of the small resources. How will we manage water resources amount of storage in the Northeast, the region is given the possibility of change and uncertainties vulnerable to prolonged drought. No new major about its nature and timing? storage has been built in the Northeast during the past 20 years, except the Bloomington Dam on the Policy approaches to water resources may be Potomac River. Water supply in lower New grouped under supply (or structural) approaches England, New York, and Pennsylvania, and power and demand (or nonstructural) approaches. Supply production in the Northeast, remain vulnerable to approaches mitigate hydrologic variability and drought, which may occur more frequently climate change; demand approaches modify (Schwartz, 1977; Kaplan et al., 1981). During behaviors that create vulnerability to such change. periodic droughts in the Northeast, such as those in For example, water shortages may be addressed 1962-65 and 1980-81, instream flow regulations either by developing surface water storage capacity ration water and threaten shutdowns of electrical and improving the quality of water from available powerplants (U.S. Army Corps of Engineers, 1977; sources (supply approaches), or by decreasing water Schwartz, 1977; Kaplan et al., 1981). use and consumption (a demand approach). Southeast Many of the policy approaches discussed below have been recommended by water resource experts In the Southeast, the experience with drought for 20 years and are in use to address existing water in recent years is increasing the use of groundwater problems and vulnerabilities. The potential of and surface water for irrigation and is prompting climate change provides another reason for farmers to consider shifting crops. In Georgia, for expanded use of these approaches. instance, the use of groundwater for irrigation has grown quite rapidly. However, the GCMs disagree Supply and Structural Policy Approaches on whether the Southeast may become wetter or drier (see Haines, Volume A; Miller and Brock, The supply-related policy approaches to water Volume A). Most reservoirs in the area have resources include design for uncertainty, surface sufficient capacity to retain flood surges and to water development, and optimization of water maintain navigation, hydropower, water supply, and resource systems. instream uses (e.g., dilution, wildlife) under both wetter and drier conditions (see Chapter 16: Design for Uncertainty Southeast; Sheer and Randall, Volume A). However, drier conditions would pose conflicts Most water resource decisions in the past have between recreational uses (which would be hurt by been based on the assumption that the climate of a changes in reservoir levels) and all other instream region varies predictably around a stationary mean. and offstream uses. Water managers develop water resources plans based on statistical analyses of historical Should the Southeast become drier, a decline climatological and hydrologic data. However, the in the inflow of freshwater could alter the estuarine frequency of extreme events, which has been ecology of the gulf coast, which may be most assumed to be fixed or to be modified only by the vulnerable to sea level rise (see Chapter 16: urbanization of watersheds, may be changed Southeast). significantly by altered climatic conditions. In addition to being uncertain about hydrologic POLICY IMPLICATIONS conditions, we are uncertain about future demographic, economic, and institutional factors Decreases in water availability and quality, that affect offstream water uses and social and increased risk of flood damages, and the economic values attached to instream uses. As an exacerbation of conflicts between water users example, water withdrawals in 1985 declined overall competing for an increasingly scarce or difficult to from 1980, falling far short of projections made manage resource are the major potential impacts starting in 1960 and as recently as 1978 (Solley et al., 1988). 178 Water Resources Finally, we are uncertain about how our decisionmaking, conditions beyond 10 or 20 years economic, regulatory, and institutional systems will may have little impact on design and investment respond to climate change in the absence of decisions (see Chapter 19: Preparing for a Global concerted governmental action. It would be a Warming; Hanchey et al., 1988). mistake to attempt to project the impacts of climate change simply by superimposing projected future Surface Water Development hydrologic conditions on today's social systems. Surface water structures increase developed or The planners and designers of water resources available water supply, provide for the regulation of must address such uncertainties. Three types of flows for instream uses, prevent flooding, or response are often used to address conditions of perform some combination of these functions. These great uncertainty: structures include dams, reservoirs, levees, and aqueducts. Because of high costs of construction, Avoid inflexible, large-scale, irreversible, adverse impacts on the environment, the limited and high-cost measures; opt for shorter number of sites available for new structures, and term, less capital-intensive, smaller scale, opposition by citizen groups, the trend during the and incremental measures. past decade has been away from large excess-capacity, capital-intensive projects. Only the Conduct sensitivity analysis and risk-cost Central Utah Project and the Central Arizona exercises in the design of structural and Project have gone forward in recent years. Only management systems to address the one major project in the Northeast has been potential range of climate change impacts. completed in past 20 years: the Bloomington Dam Sensitivity analysis describes the sensitivity on the Potomac River. In 1982, California citizens of projections to variables affecting their voted down funds for the proposed Peripheral Canal accuracy; risk-cost analysis identifies the that would have permitted increased diversion of costs, for various conditions other than water from north to south in the state. In addition, those projected, associated with the national trend toward increased local/state underdesign or overdesign of a structure. financing and reduced federal financing for projects The consideration of hydrologic extremes has reduced funds available for large projects and the use of risk analysis in the design of (National Council on Public Works Improvement, specific projects to mitigate the adverse 1988). consequences of hydrologic variability may incidentally mitigate many of the physical These current trends in water resources impacts of climate change (Hanchey et al., management may be reevaluated in light of possible 1988). new demands for developed water caused by climate changes. Pressure to build proposed projects such Design structures and systems for rare as the Narrows Project in Colorado, the Garrison events. Matalas and Fiering (1977) found Diversion in North Dakota, the Peripheral Canal in that many large systems have substantial California, and structures to divert water from redundancy (margins of safety) and northern New England to southeastern robustness (ability to perform under a Massachusetts may be renewed if droughts reoccur variety of conditions) that enable them to or demand increases. The pace at which existing adapt technologically and institutionally to projects are upgraded, modified, or expanded may large stresses and uncertain future events. also accelerate. Although the principle of design for rare Optimization of Water Resource Systems extremes may provide robustness, it has a cost and may conflict with the principle of maximizing the Water resources can be managed to maximize economic return from a project. Most public and the water availability from a given resource base private water developers subject projects to "net such as a dam, watershed, or aquifer. Adoption of present value" or "internal rate or return" analyses. systemwide strategies for a large-scale water system These analyses discount future benefits relative to may allow for substantial operating flexibility related present benefits. If a high discount rate is used in to releases of stored water. This flexibility can have 179 Chapter 9 an enormous influence on the overall performance Demand Management and Nonstructural and resilience (recovery abilities) of the system, and Policy Approaches may provide additional yields that mitigate the impacts of climate change. For example, the U.S. Demand-related adaptations encourage a Department of the Interior's Bureau of Reclamation reduction in water demand and an increase in water (1987) is adopting operational, management, or use efficiency through pricing, market exchange of physical changes to gain more output from the same water rights, conservation, protection of water resources. Water management agencies nationwide quality, education and extension service assistance, are implementing methods to protect groundwater technological innovation, and drought management recharge areas and to use ground and surface planning. Policies that discourage activities in flood- waters conjunctively (U.S. EPA, 1987b). Watershed prone areas are the nonstructural counterparts for management practices also affect water supply; for reducing flood damage. example, water yields can be significantly affected by timber harvest practices. Water Pricing, Water Markets, and Water Conservation In the East, consolidation of or coordination among fragmented urban water supply authorities In the past, many people considered that water can achieve economies of scale in water delivery, was too essential a resource or too insensitive to decrease the risk of shortage in any one subsystem price to allow market forces to allocate its use, within a region, increase yields, and provide especially during shortages. Policy took the form of effective drought management procedures. Sheer direct controls and appeals to conserve (Hrezo et (1985) estimated that coordinated water authority al., 1986). In recent years, greater attention has activities in the Potomac River basin eliminated the been given to market-based policies and need for new reservoirs, saving from $200 million to mechanisms that allocate limited water supplies $1 billion. among competing uses and promote water conservation. River basin and aquifer boundaries in many cases traverse or underlie portions of several states. Water prices that reflect real or replacement Regional and interstate cooperation to manage costs and the exchange of water rights by market water resources has a long tradition in some U.S. mechanisms can promote conservation and efficient river basins. Although numerous opportunities exist use. Since water use is sensitive to price (Gibbons, for additional coordination of water management 1986) water users faced with higher prices will between states, within basins, or between basins, the conserve water and modify their technologies and agreements required for regional compacts and crop selection to use less without substantial operating procedures and sharing of water supplies reduction in output. If there is a market for water may require substantial and lengthy negotiations. rights, those willing to pay more may purchase rights from those less willing to pay. As a Several interstate water authorities have consequence, water will be transferred out of significant water allocation authority. For example, marginal uses and will be conserved. the Delaware River Basin Commission allocates water to users in the Delaware Basin and transfers Three related pricing and conservation it to New York City under authority of a 1954 approaches are irrigation conservation, municipal Supreme Court ruling (347 U.S. 995) and federal and industrial water use, and water markets and legislation, which established the Commission in transfers. 1961 and granted it regulatory, licensing, and project construction powers. Similarly, water authorities in Irrigation Conservation the Washington, D.C., metropolitan area operate Potomac River water supply projects as integrated Relatively small reductions in irrigation demand systems under a 1982 agreement. Both the can make large amounts of water available for Delaware and Potomac regional compacts include urban and industrial uses. For instance, nearly 83% provisions for drought allocations. (See Harkness et of the withdrawals and 90% of the consumptive use al., 1985, for management actions taken by the of western water is for irrigation. A 10% reduction Delaware River Basin Commission during a 1984-85 in irrigation use would save 20 million acre-feet drought.) (maf) in water withdrawn and 10 maf in water 180 Water Resources consumed annually, effectively doubling the water Water Markets and Transfers available for municipal and industrial uses in the West (Frederick, 1986). (For comparison, the The "first in time, first in right" appropriation average annual flow of the Upper Colorado River doctrine, which favors the longest standing water Basin is 15 maf.) rights, governs much of the West's surface water and some groundwater. The appropriation doctrine Inexpensive water was a key factor in the has the potential to establish clear, transferable settlement of the West and the expansion of property rights to water -- a precondition for agriculture (Frederick, 1986). The Bureau of effective operation of water markets. The potential Reclamation was established early in this century to for water transfers to the highest value users has not promote the development of irrigation in the West. yet been fully realized because the nature and The Bureau provides irrigation for about 11 million transferability of the rights are obscured by legal acres, more than one-fifth of the total irrigated and administrative factors (Trelease, 1977; acreage. Since the Bureau accounts for nearly Frederick, 1986; Saliba et al., 1987). Following are one-third of all surface water deliveries and about some examples: one-fifth of total water deliveries in the 17 western states, actions by the Bureau to use this water more Rather than grant absolute ownership, efficiently have an impact throughout the West states with prior appropriation rules grant (Frederick, 1986). rights to use water for beneficial purposes. Water rights not put to beneficial use may In the past, demand for Bureau water was not be forfeited. This encourages a use-it-or- based on the real cost of the water, because more lose-it attitude. than 90% of the Bureau's irrigation projects have been subsidized, and payments on some projects no Federal and Native American water rights longer even pay for operation and maintenance remain unquantified in some areas such as (Frederick and Hansen, 1982). Irrigators fortunate the Colorado River Basin. enough to receive such inexpensive water may have little or no incentive to conserve. However, the The emergence in law of the "public trust Bureau's more recently stated objectives include doctrine," which states that all uses are revising their water marketing policy, promoting subject to the public interest, has cast a conservation, and pricing water to reflect its real cloud over some water rights. This has cost (U.S. Department of the Interior, 1987). been true in California, where the public interest has driven a reexamination of Municipal and Industrial Water Use withdrawals from Mono Lake, and where existing permits have been modified to Municipalities throughout the country are protect the Sacramento-San Joaquin Delta finding it difficult and expensive to augment their from saltwater intrusion. Montana is supplies to meet the demands of population and increasingly basing water management plans economic growth and are finding that users would on its instream flow requirements and is rather use less than pay more (Gibbons, 1986). exploring ways to have these requirements Traditional average-cost pricing provides adequate for all future beneficial instream uses count service to customers and adequate returns to water as a bona fide use of the Missouri River to companies, but is being reevaluated because it tends slow the growth rate of water diversion for to cause overinvestment in system capacity (U.S. offstream uses (Tarlock, 1987). Congress, 1987). Marginal-cost pricing (charging for the cost of the last-added and most expensive- In resolving interstate water disputes, a increment of supply) or progressive-rate pricing federal common law of "equitable (charging more per unit to users of large amounts) apportionment" has developed under which can reduce domestic and industrial water an informed judgment, based on consumption because water use is sensitive to price consideration of many factors, secures a (Gibbons, 1986). "just and equitable" water allocation (see Strock, 1987). The Supreme Court 181 Chapter 9 decided in Colorado V. New Mexico (456 U.S. 176, 1982) that equitable Frederick and Kneese (1989) caution that water apportionment may be used to override transfers occur gradually and are not likely to affect prior appropriation priorities in cases of more than a small percentage of agricultural water major flow reductions. The Supreme Court rights for the foreseeable future. However, legal specifically mentioned climatic conditions in and institutional changes facilitating water markets ruling that prior appropriation systems and demand for water by high-value users may be would otherwise protect arguably wasteful accelerated under the stress of climate change and inefficient uses of water at the expense (Trelease, 1977). of other uses (see Strock, 1987). Drought Management Policies Because of imperfect competition, third-party effects, uncertainty over Integrating drought planning into water resource administrative rules, and equity management may assume greater priority if climate considerations, water market prices may change aggravates water shortages. The Model not appropriately measure water values Water Use Act (Hrezo et al., 1986) advocates that according to economic efficiency criteria states or water supply authorities integrate drought (Gibbons, 1986; Saliba et al., 1987). management and advance planning into their policies by designating a governmental authority for It is possible to control groundwater drought response and by adopting mechanisms for withdrawals, but for a number of reasons it automatically implementing and enforcing water-use is difficult to establish market mechanisms restrictions. In 1986, only seven states had for groundwater allocation. Because all comprehensive management plans for water groundwater users essentially draw from a shortages (Hrezo et al., 1986). Most states rely on shared pool, groundwater resources are water rights appropriations, emergency conservation treated as "common property." As a result, programs, and litigation to allocate water during property rights are difficult to define, shortages. Improved capabilities in surface third-party impacts of transfers of hydrology and in water system modeling and groundwater rights are significant, and monitoring would be required to support broadened interstate agreements concerning allocation drought contingency planning. of interstate aquifer water are difficult to attain (Emel, 1987). Water Quality Despite the obstacles, transfer of water rights Federal and state legislation and regulations for among users -- especially from irrigators to control of instream water quality have had a municipalities and power companies seeking water dramatic effect on reducing conventional water for urban expansion and electricity production is pollutants since the enactment of the 1972 Clean becoming common in many western states (Wahl Water Act. The reduced riverflows and lake levels and Osterhoudt, 1986; Frederick, 1986). Methods that are possible under altered climate conditions include negotiated purchases, short-term exchanges could necessitate more stringent controls on point during droughts, and water banks and markets and nonpoint sources to meet water quality (Wahl and Osterhoudt, 1985; Saliba et al., 1987; standards. Promotion of nonpolluting products, Wahl and Davis, 1986). waste minimization, and agricultural practices that reduce the application of chemicals will also Legislation in many western states has enhance water quality, making more water of facilitated water transfers (Frederick, 1986; suitable quality available for use. Frederick and Kneese, 1989). For instance, Arizona's new water law facilitates the purchase of Many states have adopted measures to protect agricultural land for water rights, and the use of that instream water uses. These include reserving flows water for urban development. Strict technical or granting rights for particular instream uses and standards imposing conservation on municipal and directing agencies to review impacts before granting industrial water uses, such as watering golf courses new rights (U.S. Water Resources Council, 1980; with wastewater, are also part of Arizona's laws Frederick and Kneese, 1989). Regulations limiting (Saliba et al., 1987). 182 Water Resources water use may have to be modified where climate Research activities should include the following: change has resulted in reduced flows during droughts. Monitor atmospheric, oceanic, and hydrologic conditions to detect evidence of Policies for Floodplains water resources impacts of climate change. The National Flood Insurance Program was Continue to develop and refine regional enacted in 1968, with major amendments in 1973. hydrologic models that are capable of The program provides subsidized flood insurance modeling the changes in runoff, water for existing structures in flood-prone areas, provided availability, water use, and evapotran- that the community with jurisdiction regulates the spiration induced by changes in location and construction of new buildings to temperature and atmospheric conditions. minimize future flood losses. New structures that This research should focus on vulnerable comply with the restrictions are eligible for river basins where demand approaches or insurance at full actuarial rates. exceeds safe yield or where hydrologic variability is high. In 1979, the program took in $140 million in premiums and paid $480 million in claims. Refine global climate change models and Recently, the program was authorized to relocate link them to regional hydrologic models so structures exposed to repeated flood or erosion that regional water resource planners, damage rather than pay claims for such structures. engineers, and managers can use their projections more confidently. Where rainfall and flooding increase, the 100-year floodplain would expand, and rate maps Study the sensitivity of existing water would need revision. Premium payments and claims systems to possible changes in climate would rise. conditions. At the same time, the following research is RESEARCH NEEDS needed to identify opportunities for adopting measures to adjust and adapt to climate change. Water is the principal medium by which changes in atmospheric conditions are transmitted Quantify federal and Native American to the environment, the economy, and society. water rights in the West. Hydrology is the key discipline that enables us to understand and project these effects. Improvements Examine how present institutions and in both the GCMs and regional hydrologic models markets can better allocate water among are needed so that we may understand the impacts users and provide incentives to conserve of climate change and devise appropriate water water. resources management strategies. Specifically, GCMs do not yet provide regional forecasts at the Assess the extent to which laws and level of certainty and temporal and spatial regulations may exacerbate the effects of resolution required for decisionmakers. To be more climate change. (Examples include thermal helpful, the GCMs should provide forecasts specific controls for rivers and federal pricing and to individual river basins or demand centers, and reallocation policies for irrigation water.) should describe hydrologic conditions over the typical design-life of water resource structures. Identify, project, and quantify the demographic and institutional adjustments that may occur in the absence of public action in response to climate-induced impacts on water resources. This research will reduce uncertainty for policymakers regarding where concerted public action may be or not be needed. 183 Chapter 9 REFERENCES Gibbons, D.C. 1986. The Economic Value of Water. Washington, DC: Resources for the Future. Bajwa, S., M. Grosswhite, and J.E. Hostetler. 1987. Agricultural Irrigation and Water Supply. Gleick, P.H. 1987. Regional Hydrologic Agriculture Information Bulletin 532. U.S. Consequences of Increases in Atmospheric CO₂ and Department of Agriculture. Washington, DC: U.S. Other Trace Gases. Climatic Change 10:137-161. Government Printing Office. Gleick, P.H. 1989. The sensitivities and Beran, M. 1986. The water resource impact of vulnerabilities of water supply systems to climatic future climate change and variability. In: Titus, J., changes. In: Waggoner, P.E., ed. Climatic ed. Effects of Changes in Stratospheric Ozone and Variability, Climate Change, and U.S. Water Global Climate. Vol. 1: Overview. Washington, DC: Resources. New York: John Wiley and Sons. In U.S. Environmental Protection Agency, pp. 299-330. press. Breitstein, L., and R.C. Tucker. 1986. Water Hanchey, J.R., K.E. Schilling, and E.Z. Stakhiv. Reuse and Recycle in the U.S. Steam Electric 1988. Water resources planning under climate Generating Industry An Assessment of Current uncertainty. In: Preparing for Climate Change. Practice and Potential for Future Applications. Proceedings of the First North American Report to U.S. Geological Survey, Grant No. 14-08- Conference on Preparing for Climate Change. 0001-G1074. Bethesda, MD: Dames and Moore. Rockville, MD: Government Institutes, Inc. pp. 394-405. Butcher, W.R., and N.K. Whittlesey. 1986. Competition between irrigation and hydropower in Hansen, J., A. Lacis, D. Rind, G. Russell, I. Fung, the Pacific Northwest. In: Frederick, K.D., ed. P. Ashcraft, S. Lebedeff, R. Ruedy, and P. Stone. Scarce Water and Institutional Change. 1986. The greenhouse effect: projections of global Washington, DC: Resources for the Future climate change. In: Titus, J., ed. Effects of Changes in Stratospheric Ozone and Global Dracup, J.A. 1977. Impact on the Colorado River Climate. Vol. 1: Overview. Washington, DC: U.S. Basin and southwest water supply. In: The Environmental Protection Agency, pp. 199-218. National Research Council, ed. Climate, Climatic Harkness, W.E., H.F. Lins, and W.M. Alley. 1985. Change, and Water Supply. Washington, DC: National Academy of Sciences, pp. 121-132. Drought in the Delaware River Basin. In: National Water Summary. 1985: Hydrologic Events and Emel, J.L. 1987. Groundwater rights: definition Surface-Water Resources. Washington, DC: U.S. and transfer. Natural Resources Journal 27:653-674. Geological Survey, U.S. Government Printing Office, pp. 29-40. Farber, K.D., and G.L. Rutledge. 1987. Pollution abatement and control expenditure, 1982-85. Survey Hobbs, B.F., and P.M. Meier. 1979. An analysis of water resources constraints on power plant siting in of Current Business May:21-26. the Mid-Atlantic States. Water Resources Bulletin Frederick, K.D., and J.C. Hansen. 1982. Water for 15(6):1666-1676. Western Agriculture. Washington, DC: Resources Hrezo, M.S., P.G. Bridgeman, and W.R. Walker. for the Future. 1986. Integrating drought planning into water Frederick, K.D. 1986. In: Frederick, K.D., ed. resources management. Natural Resources Journal 26:141-167. Scarce Water and Institutional Change. Washington, DC: Resources for the Future. Jacoby, H.D. 1989. Water Quality. In: Waggoner, Frederick, K.D., and A.V. Kneese. 1989. P.E., ed. Climate Change and U.S. Water Resources. New York: John Wiley and Sons. In Reallocation by markets and prices. In: Waggoner, P.E., ed. Climate Change and U.S. Water press. Resources. New York: John Wiley and Sons. In press. 184 Water Resources Kaplan, E., M. Rubino, and D. Neuhaus. 1981. Revelle, R.R., and P.E. Waggoner. 1983. Effects of Water Resource Development and Energy Growth the carbon dioxide induced climatic change on water in the Northeast. Brookhaven, NY: Brookhaven supplies in the western United States. In: National Laboratory, Report BNL 51522. Changing Climate: Report of the Carbon Dioxide Assessment Committee. Washington, DC: National Klemes, V., and J. Nemec. 1985. Assessing the Academy Press. impact of climate change on the development of surface water resources. In: Klemes V., ed. Rind, D., and S. Lebedeff. 1984. Potential Climatic Sensitivity of Water Resource Systems to Climate Impacts of Increasing Atmosphere CO2 With Variations. Geneva: World Meteorological Office. Emphasis on Water Availability and Hydrology in the United States. Washington, DC: U.S. Kneese, A.V., and G. Bonem. 1986. Hypothetical Environmental Protection Agency. shocks to water allocation institutions in the Colorado Basin. In: Weatherford, G., and F.L. Rosenberg, N.J. 1988. Global climate change holds Brown, eds. New Courses for the Colorado River: problems and uncertainties for agriculture. In: Major Issues for the Next Century. Albuquerque: Tutwiler, M.A., ed. U.S. Agriculture in a Global University of New Mexico Press, pp. 87-108. Setting. Washington, DC: Resources for the Future. Langbein, W.B. et al. 1949. Annual runoff in the United States. Geological Survey Circular 5. Saliba, B.C., D.B. Bush, W.E. Martinard, and T.C. Washington, DC: U.S. Department of the Interior. Brown. 1987. Do water market prices appropriately measure water values? Natural Linsley, R.K., and J.B. Franzini. 1979. Water Resources Journal 27:617-652. Resources Engineering. New York: McGraw-Hill, Inc. Schwarz, H.E. 1977. Climatic change and water supply: how sensitive is the Northeast? In: Matalas, N.C., and M.B. Fiering. 1977. National Research Council, ed. Climate, Climatic Water-resource systems planning. In: The National Change, and Water Supply. Washington, DC: Research Council, ed. Climate, Climatic Change, National Academy of Sciences, pp. 111-120. and Water Supply. Washington, DC: National Academy of Sciences, pp. 99-110. Sheer, D.P. 1985. Managing water supplies to increase water availability. In: National Water National Council on Public Works Improvement. Summary. Hydrologic Events and Surface-Water 1988. Fragile Foundations: A Report on America's Resources. Washington, DC: U.S. Geological Public Works. Final Report to the President and to Survey, U.S. Government Printing Office, pp. the Congress. Washington, DC: U.S. Government 101-112. Printing Office. Solley, W.B., C.F. Merck, and R.R. Pierce. 1988. National Research Council. 1977. Climate, Estimated use of water in the United States in 1985. Climatic Change, and Water Supply. Washington, U.S. Geological Survey. Circular 1004. Washington, DC: National Academy Press. DC: U.S. Government Printing Office. National Research Council. 1988. Estimating Stockton, C.W., and W.R. Boggess. 1979. Probabilities of Extreme Floods. National Research Geohydrological Implications of Climate Change on Council, Committee on Techniques for Estimating Water Resource Development. Fort Belvoir, VA: Probabilities of Extreme Floods, Water Science and U.S. Army Coastal Engineering Research Center. Technology Board; and Commission on Physical Sciences, Mathematics, and Resources. Washington, Strock, J.M. 1987. Adjusting water allocation law DC: National Academy Press. to meet water quality and availability concerns in a warming world. In: Preparing for Climate Change. Nemec, J., and J.C. Schaake. 1982. Sensitivity of Proceedings of the First North American water resource systems to climate variation. Conference on Preparing for Climate Change. Hydrological Sciences Journal 27:327-43. Rockville, MD: Government Institutes, Inc., pp. 382-387. 185 Chapter 9 Tarlock, A.D. 1987. Damning the dams and U.S. EPA. 1987b. U.S. Environmental Protection ditches: a review of D. Worster, rivers of empire: Agency. National Water Quality Inventory: 1986 water aridity and the American West. Natural Report to Congress. EPA-440/4-87-008. Resources Journal 27:477-490. Washington, DC: U.S. Environmental Protection Agency. Trelease, F.J. 1977. Climatic change and water law. In: National Research Council, ed. Climate, U.S. Geological Survey. 1984. National Water Climatic Change, and Water Supply. Washington, Summary 1983 - Hydrological Events and Issues. DC: National Academy of Sciences, pp. 70-84. Water Supply Paper 2250. Washington, DC: U.S. Government Printing Office. U.S. Army Corps of Engineers. 1977. Northeastern United States Water Supply Study, Summary U.S. Geological Survey. 1985. National Water Report. U.S. Army Corps of Engineers, North Summary 1984. Water Supply Paper 2275. Atlantic Division. Washington, DC: U.S. Government Printing Office. U.S. Army Corps of Engineers. 1988. Lessons U.S. Geological Survey. 1986. National Water Learned From the 1986 Drought. U.S. Army Corps Summary 1984. Water Supply Paper 2300. of Engineers, Institute for Water Resources. IWR Washington, DC: U.S. Government Printing Office. Policy Study 88 - PS - 1. Fort Belvoir, VA U.S. Water Resources Council. 1980. State of the U.S. Congress. 1987. U.S. Congress, Congressional States: Water Resources Planning and Management. Budget Office. Financing Municipal Water Supply Washington, DC: U.S. Water Resources Council. Systems; May. Washington, DC: Congressional Budget Office. U.S. Water Resources Council. 1978. The Nation's Water Resources 1975-2000: Second National Water U.S. Congress. 1984. U.S. Congress, Office of Assessment. Vol. 1: Summary. Washington, DC: Technology Assessment. Protecting the Nation's U.S. Government Printing Office. Groundwater From Contamination. Volume 1. Washington, DC: U.S. Government Printing Office. Viessman Jr., W., and C. DeMoncada. 1980. State and National Water Use Trends to the Year 2000. U.S. Department of the Interior. 1987. U.S. Washington, DC: Congressional Research Service, Department of the Interior, Bureau of Reclamation. The Library of Congress. Assessment '87...A New Direction for the Bureau of Reclamation. Washington, DC: U.S. Department of Wahl, R.W., and R.K. Davis. 1986. Satisfying the Interior. Southern California's thirst for water: efficient alternatives. In: Frederick, K.D., ed. Scarce Water U.S. EPA. 1987a. U.S. Environmental Protection and Institutional Change. Washington, DC: Agency. 1986 Needs Survey Report to Congress: Resources for the Future. Assessment of Needed Publicly Owned Wastewater Treatment Facilities in the United States. EPA- Wahl, R.W., and F.H. Osterhoudt. 1985. Voluntary 430/9-87-001. Washington, DC: U.S. Environmental transfers of water in the West. In: U.S. Geological Protection Agency. Survey. National Water Summary 1985. Hydrologic Events and Surface-Water Resources. Washington, DC: U.S. Government Printing Office, pp. 113-124. 186 CHAPTER 10 ELECTRICITY DEMAND FINDINGS increases in demand for electricity to run irrigation equipment. Global warming would increase electricity demand, These results are sensitive to assumptions about generating capacity requirements, annual generation, the rates of economic growth, technological and fuel costs nationally. The impacts could be improvements, and the relationship between significant within a few decades and would increase electricity use and climate. The potential savings substantially over time if global warming continues. in other energy sources (gas and oil) used for space heating and other end uses sensitive to The new generating capacity requirements climate and the potentially significant impacts on induced by climate change effects on electricity hydroelectric supplies and other utility demand estimated for 2010 show an increase of operations were not analyzed. 25 to 55 gigawatts (GW), or 9 to 19% above estimated new capacity requirements assuming Policy Implications no change in climate. Between 2010 and 2055, climate change impacts on electricity demand Utility executives and planners should begin to could accelerate, increasing new capacity consider climate change as a factor in planning requirements by 200 to 400 GW (14 to 23%) new capacity and future operations. The above what would be needed in the absence of estimated impacts of climate change in some climate change. These capacity increases would regions are similar to the range of other require investments of approximately $200 to uncertainties and issues utility planners need to $300 billion (in 1986 dollars). In the absence of consider over the 20- to 30-year period. climate change, population and economic Additional climate and utility analyses are growth may require investments of needed to develop refined risk assessments and approximately $2.4 to 3.3 trillion through 2055. risk management strategies. Estimated increases in annual electricity The increased demand for electricity induced by generation and fuel use induced by climate climate change also could exacerbate other change represent several thousand gigawatt- environmental problems, such as the hours by 2055. The estimated increases are 1 to implementation of "acid rain" strategies, 2% in 2010 and 4 to 6% in 2055. Annual fuel, adherence to the international nitrogen oxide operation, and maintenance cost to meet treaty, state implementation plans for ozone increased electricity demand would be several control, and thermal pollution control permit hundred million dollars in 2010 and several requirements. The Environmental Protection billion dollars in 2055. Without climate change, Agency should analyze the impacts of climate these annual costs would be $475 to 655 billion change on long-range policies and should in 2055. include climate change as an explicit criterion in making risk management decisions when Estimated regional impacts differ substantially. appropriate. The largest increases could occur in the Southeast and Southwest, where air-conditioning The increased demand for electricity could demands are large relative to heating. Northern make policies to stabilize the atmosphere border states may have a net reduction in through energy conservation more difficult to electricity generation relative to base case achieve. The estimated increases in electricity requirements assuming no change in climate. generation induced by climate change could These changes could be exacerbated by increase annual CO₂ emissions, depending upon reductions in hydropower production and future utility technology and fuel choice 187 Chapter 10 decisions. Assuming no change in efficiency of and refrigeration. These applications of electricity energy production and demand, reliance on may account for up to a third of total sales for some coal-based technologies to meet the increased utilities and may contribute an even larger portion demands could increase CO2 emissions by 40 to of seasonal and daily peak demands. 65 million tons in 2010 and by 250 to 500 million tons in 2055. Use of other, lower CO2- Changes in weather-sensitive demands for emitting technologies and fuels (e.g., efficient electricity can affect both the amount and the conversion technologies and nuclear and characteristics of generating capacity that a utility renewable resources) would reduce these must build and maintain to ensure reliable service. incremental additions. In addition, warmer These changes also can affect fuel requirements and winter temperatures could reduce the demand the characteristics of efficient utility system for oil and gas in end uses such as residential operations, particularly the scheduling and furnaces for heating, thereby lowering CO2 dispatching of the utility's generating capacity. For emissions from these sources. Future analyses example, electric energy used for air-conditioning of national and international strategies to limit exceeds that used for space heating nationwide, and greenhouse gases should include the changes in the temperature sensitivity associated with cooling energy demand created by global warming as a is higher than that associated with heating. This positive feedback. implies not only changes in seasonal electricity demands but also increases in annual electricity demands as a result of higher temperatures. CLIMATE CHANGE AND ELECTRICITY DEMAND Similarly, utilities in most regions experience their peak demands in the summer. A rise in air- conditioning and other temperature-sensitive Climate change could affect a wide range of summer loads would significantly increase peak energy sources and uses. In the near term, policies loads and, as a result, would step up utility aimed at reducing emissions of greenhouse gases investments in new generating capacity needed to from fossil fuel combustion could affect the level meet additional demands and to maintain system and mix of fuel consumption in various end-use reliability. technologies and in the generation of electric power. In the longer term, changes in temperature, Examples of other ways in which climate could precipitation, and other climatic conditions also affect electric utilities include the following. could affect energy resources. For example, warmer Changes in precipitation, evaporation, and runoff temperatures likely would reduce the demand for from mountain snowpack as well as changes in fuels used in the winter for space heating and water management practices in response to climate increase the demand for fuels used in the summer change could affect the annual and seasonal for air-conditioning; and reduced precipitation and availability of streamflow to generate hydropower. soil moisture in some regions could increase the use Reductions in hydropower would require utilities to of energy to pump water for irrigation. These rely upon other, possibly more costly and less effects could be particularly significant for planning environmentally benign generation sources to meet in the electric utility industry based upon the customer needs. Furthermore, reductions in water substantial amount of electric load accounted for by resources would adversely affect the availability weather-sensitive end uses, the variety of resources and/or cost of water for powerplant cooling. used to generate electric power, and the capital- intensivity of the industry. One major consideration Other direct impacts of climate change on is the potential impact of climate change on the electric utilities include the effects of temperatures demand for electricity and the implications of on powerplant operating efficiencies, the effects of changes in demand on utility capacity and sea level rise on the protection and siting of coastal generation requirements. facilities, and the effects of changes in various climate conditions on the supply of renewable Many electrical end uses vary with weather energy resources such as solar and wind power. conditions. The principal weather-sensitive end uses Also, legislation and regulations designed to limit are space heating, cooling, and irrigation pumping greenhouse gas emissions from utility sources could and to a lesser degree - water heating, cooking, significantly affect the supply and cost of electricity generation. 188 Electricity Demand Although some of these impacts could Linder et al. found that temperature increase significantly affect utility planning and operations could significantly heighten annual and peak (particularly on a regional basis), they have not been electricity demands by 2015, and that a temperature analyzed in detail and are not addressed in this rise would require construction of new generating report. Further research and analysis are needed to capacity and increases in annual generation. The develop a more complete assessment of utility southeastern utility had higher estimated increases impacts. in electricity demand, generation, and production costs than the New York utilities because of greater electricity demands for air-conditioning. In PREVIOUS CLIMATE CHANGE addition, streamflow used to generate hydropower STUDIES in New York could be reduced, requiring increased use of fossil fuel generation to meet customer demands for electricity. A number of utilities conduct analyses relating short-term variations in weather conditions with a need to "weather-normalize" historical demand data CLIMATE CHANGE STUDY IN and to test the sensitivity of system reliability and operations to these short-term variations. THIS REPORT Furthermore, some researchers have speculated regarding the potential effects of longer term climate changes on electricity demand (e.g., Stokoe Study Design et al., 1987). Linder and Inglis (Volume H) expanded the However, only one previous study has estimated case studies (Linder et al., 1987) of the sensitivity of the potential implications of longer term, global electricity demand to climate change and conducted warming-associated temperature changes on a national analysis of electricity demand. Relevant electricity demands and the effects of changes in regional results from the national studies of Linder demand on utility investment and operating plans. and Inglis are discussed in the regional chapters of Linder et al. (1987) used general circulation model this report. (GCM) results to estimate the potential impacts of temperature change on electricity demand (and on The analytic approach developed by Linder et the supply of hydropower) for selected case study al. (1987) formed the basis for estimating the utility systems in two geographical areas: a utility regional and national impacts described in this located in the southeastern United States and the report. The principal steps in the approach are major utilities in New York State, disaggregated summarized in Figure 10-1 (see Volume H for more into upstate and downstate systems. details). Estimated impacts were developed for the relatively near term (from the present to 2010, Climate Weather- Change Sensitivity of Scenarios Electricity Demand Utility Impacts on Planning Utility Investments, Model Operations, Costs Utility Planning Assumptions Figure 10-1. Analytic approach (Linder and Inglis, Volume H). 189 Chapter 10 within electric utility long-range resource planning and other utility characteristics. Linder and Inglis horizons of 20 to 30 years) and over the longer term assumed that future capacity and generation (to 2055), when the magnitude of temperature requirements will be met by investments either in changes is expected to approach equilibrium levels new coal-fired baseload capacity or in oil- and representative of a doubling of atmospheric natural gas-fired peaking capacity. Other sources, concentrations of CO₂. Linder and Inglis used such as nuclear energy and renewables or innovative Goddard Institute for Space Studies (GISS) A and fossil fuel-fired technologies (e.g., fluidized bed B transient estimates of temperature change in 2010 combustion), were not considered (for further and GISS A estimates for 2055 in their calculations. details, see Linder et al., 1987). The scenario changes in annual temperatures for the United States range from about 1.0 to 1.4°C in Demands for electricity in the absence of 2010 and are approximately 3.7°C by 2055. climate change can be related to the overall level of Regional temperature scenarios show greater economic activity as represented by the gross variation. national product (GNP). Because economic growth assumptions are critical to estimates of future Linder and Inglis used actual utility demand and electricity demands, alternative GNP growth rates temperature data from the case study utilities, and were assumed in developing the base cases; these from five other large, geographically dispersed utility ranged from 1.2 to 2.1% per year. 2 These systems, to develop a set of weather-sensitivity alternative assumptions are referred to as "lower parameters for utility areas. On a weighted-average growth" and "higher growth," respectively. basis (weighted by electricity sales), utility peak demands were estimated to increase by about 3.1% These assumptions served as inputs to a regional per change in degree Celsius (ranging from -1.35 to planning model called the Coal and Electric Utilities 5.40% across utility areas), and annual energy Model (CEUM). CEUM outputs include the demands were estimated to increase by about 1.0% amount and characteristics of new generating per change in degree Celsius (ranging from -0.54 to capacity additions, electricity generation by fuel 2.70%). type, and electricity production costs. A number of uncertainties associated with the Limitations data and assumptions used to develop these weather-sensitivity relationships suggested that the The study extrapolated temperature-sensitivity relationships may understate customer response to findings for some regions and did not include climate change, particularly at higher temperature specific analyses of temperature sensitivity for all change levels occurring in the future. For example, utility regions of the United States. It focused the approach did not explicitly account for probable narrowly on impact pathways, considering only the increases in the market saturation of air- potential effects of temperature change on changes conditioning equipment as temperatures rise over in electricity demand. Neither the potentially time. To address this possibility, an alternative case significant impacts of climate change on hydropower was designed in which the estimated weather- availability nor the impacts of reduced water sensitivity values were increased by 50%. This was supplies for powerplant cooling were included. designated as the "higher sensitivity" case. Furthermore, the study did not evaluate the Since this study is focused on estimating how sensitivity of the results to different, doubled-CO2 climate change may affect key utility planning factors, Linder and Inglis used a planning scenario assuming no change in climate (a "base case") to 1 Note that the development and use of a base case reflecting serve as a basis for comparison with planning changes in non-climate-related conditions over time was scenarios under alternative assumptions of climate undertaken only for the electricity demand study, not for other areas in this report. Changes in population and technology are change for 2010 and 2055. Thus, base case utility considered in Chapter 6: Agriculture. plans were developed for 2010 and 2055, using assumptions regarding future demands for electricity 2 These GNP growth rates are relatively conservative, but they in the absence of climate change (reflecting are comparable with GNP growth rates used by EPA in its report to Congress on Policy Options for Stabilizing Global population and economic growth), generating Climate. technology option performance and costs, fuel costs, 190 Electricity Demand GCM climate scenarios (GFDL and OSU), climate change scenarios (GISS A and GISS B) and although the use of the GISS transient experiment assumptions of the weather sensitivity of demand results for 2010 and 2055 indicates relative ("estimated sensitivity" and "higher sensitivity"). sensitivities to small and large temperature changes. Estimated increases in peak demand over the The study did not consider variations in base case on a national basis range from 2 to 6% by temperature changes and the occurrence of extreme 2010. Changes in estimated annual energy events, which affect powerplant dispatch and requirements by 2010 are more modest, ranging determinations of peak demands, respectively, and from 1 to 2%. In 2055, peak national demands are are important for utility planning. estimated to increase by 13 to 20% above base case values, and annual energy requirements are Many uncertainties exist regarding the concepts, estimated to increase by 4 to 6%. methods, and assumptions involved in developing and applying estimates of the temperature sensitivity By 2010, new climate change-induced generating of demand. For example, a key assumption is that capacity requirements increase by 6 to 19%, or the estimated sensitivities of demand to historical, about 24 to 55 GW, representing an average short-term variations in temperature are adequate increase of up to 1 GW per state (approximately the representations of future relationships between capacity of one to two large nuclear or coal-fired electricity demand and long-term changes in mean baseload powerplants). The majority of the capacity temperatures. increase is for peaking capacity rather than baseload capacity. The investment associated with these Uncertainties also exist regarding market, capacity increases is several billion dollars (in regulatory, technological, and other conditions that constant 1986 dollars). By 2055, the change in new will face the utility industry in the future. For capacity requirements increases in percentage terms example, technological changes that improve the and represents several hundred GW. Under high energy efficiency of weather-sensitive end-use GNP and higher weather-sensitivity assumptions, the equipment or electricity-generating equipment will estimated increase attributable to climate change is continue to evolve. These changes would likely lead almost 400 GW, or 23%. To put these results into to lower climate change impacts than estimated in perspective, it should be noted that current this report. On the other hand, regulatory changes generating capacity in the United States is about 700 aimed at reducing the emissions of greenhouse GW. The increase in new capacity requirements gases from electricity generation could limit a under the base case is 1,350 to 1,780 GW. utility's future fuel and technology investment options, leading to higher estimates of cost impacts Annual generation increases for the United than reported here. Because of these limitations, it States are not as large in percentage terms as those is important to recall that the results presented in estimated for new generating capacity requirements, the next section should not be considered as but nonetheless, they account for several hundred projections of actual powerplant investments and billion kWh by 2055. In the near term (i.e., to utility operations, but rather as comparisons 2010), increased levels and changing patterns of providing estimates of the magnitude of sensitivities climate change-induced electricity demand permit to alternative climate change assumptions. utilities in some areas having excess generating capacity to serve the growing needs of utilities in Results other areas through substitution of lower cost baseload generation for higher cost peaking The potential national impacts for 2010 and generation. On net, peaking generation would be 2055 are summarized in Table 10-1. The table lower as a result of climate change in 2010 (see presents base case values (i.e., assuming no change Linder et al., 1987, for further detail). In 2055, in climate) for each year and estimated impacts peaking generation is projected to increase along represented by changes from the base case values. with baseload generation, because all the excess The impacts for 2055 are presented for both the capacity that had existed in 2010 either would have lower growth GNP and the higher growth GNP been fully used by growing demands to 2055 or cases. Also, where ranges of impacts are presented, would have been retired. The estimated impacts of they summarize the estimates under alternative climate change on national new generating capacity 191 Chapter 10 Table 10-1. The Potential National Impacts of Climate Change on Electric Utilities 2010 2055 Lower GNP Higher GNP Base Increase Base Increase Base Increase Peak demand (GW) 774 20-44 1,355 181 1,780 238-357 New capacity requirements (GW)ᵃ Peaking 50 13-33 176 118 254 182-286 Baseload 226 11-22 1,011 67 1,423 74-98 Total 276 24-55 1,187 185 1,677 227-384 Annual sales (bkWh) 3,847 39-67 6,732 281 8,848 370-555 Annual generationᵇ (bkWh) Oil/gas 287 (12)-(29) 221 2 308 27-51 Coal 2,798 54-103 6,242 305 8,295 381-560 Other 1,092 1-(1) 846 (2) 1,003 (7)-0 Total 4,177 43-72 7,309 305 9,607 401-611 Cumulative capital costsᶜ,ᵈ 669 25-48 1,765 173 2,650 222-328 Annual costsd 162 3-6 474 33 655 48-73 ᵃIncludes reserve margin requirements; does not include "firm scheduled" capacity. ᵇIncludes transmission and distribution losses. ᶜ"Base" values include regional capital expenditures for utility-related equipment in addition to new generating capacity (e.g., new transmission facilities). ᵈIₙ billions of 1986 dollars. Abbreviations: GW = gigawatts; bkWh = billion kilowatthours. Source: Linder and Inglis (Volume H). requirements and annual generation are illustrated state-by-state basis. The state and regional in Figure 10-2. differences reflect differences in current climate conditions (e.g., seasonal temperature patterns), Table 10-1 also indicates that the increase in assumed future climate changes, and electricity end- annual costs for capital, fuel, and operation and use and utility system characteristics (e.g., market maintenance associated with climate change-induced saturation of weather-sensitive appliances and modifications in utility investments and operations equipment). are a few billion dollars in 2010 and are $33 to $73 billion by 2055, a 7 to 15% increase over base case Figure 10-3 shows that estimated reductions in values of $475 to $655 billion for 2055. new capacity requirements induced by climate change are limited to the winter-peaking regions of Figures 10-3 and 10-4 illustrate the diversity of the extreme Northeast and Northwest. The Great the estimated results for generating capacity on a Lakes, northern Great Plains, and Mountain States 192 Electricity Demand 2200 11000 New Capacity Requirements Annual Generation 2000 10000 1800 9000 1600 8000 Additional Climate Change Impacts: 1400 7000 Higher Sensitivity Glgawatts 1200 6000 Billion KWH Climate Change Impacts: Base Sensitivity 1000 5000 Base Case (No Climate Change) 800 4000 600 3000 400 2000 200 1000 0 0 2010 2055 2055 2010 2055 2055 Lower Higher Lower Higher GNP GNP GNP GNP Assumption Assumption Assumption Assumption Figure 10-2. Potential impacts of climate change on electric utilities, United States (Linder and Inglis, Volume H). 2055 % CHANGE IN NEW CAPACITY 20 to 30 10 to 20 0 to 10 10 to 0 Figure 10-3. Changes in electric utility capacity additions by state, induced by climate change in 2055 (derived from Linder and Inglis, Volume H). 193 Chapter 10 are estimated to experience increased new capacity may be for powerplants that are utilized heavily requirements by 2055 in the range of 0 to 10%. during only part of the year. Low annual utilization Increases greater than 20% are concentrated in the in the region would not justify construction of high- Southeast, southern Great Plains, and Southwest. capital and low-fuel cost baseload powerplants that can produce electricity more cheaply (per kWh) Figure 10-4 shows a somewhat similar than low-capital and high-fuel-cost peaking units. geographic pattern of impacts for electricity However, when considered across several regions, generation in 2055. Reductions in generation are the least-cost plan may be to construct baseload estimated in the North, and the greatest increases powerplants in certain regions, utilize them to an are concentrated in the Southwest. Despite extent greater than required by the region, and sell substantial use of air-conditioning in the Southeast, the "excess" electricity from these plants into other the estimated increases in generation are only in the regions. The location and amount of these 5 to 10% range. There is a relatively high market interregional sales would be subject to the transfer saturation of electric heat in the region, and the capabilities of transmission capacity in place. An increase in cooling is partly offset by a decrease in alternative to increased interregional bulk power heating as a result of warmer winters. sales would be the development and application of efficient and effective energy storage technologies. Because regions are affected differently, the results indicate potential changes in the patterns of interregional bulk power exchanges and capacity SOCIOECONOMIC AND sales over time and as climate changes. For example, under the assumption of increasing ENVIRONMENTAL temperatures, some regions may require significant IMPLICATIONS amounts of additional generating capacity to reliably meet increased demands during peak (cooling) Despite the limitations of the analysis and the seasons, but may experience lower demands in other need for more research to refine the data and (heating) seasons. As a result, the region's needs methods used, the results are judged to be 2055 % CHANGE GENERATION 10 to 15 5 10 to 5 -5 to 0 Figure 10-4. Changes in electricity generation by state, induced by climate change in 2055 (derived from Linder and Inglis, Volume H). 194 Electricity Demand reasonable estimates of the sensitivity of electricity new technological and market focus would demand to potential climate change. Key be directed toward this type of generating socioeconomic and environmental implications of plant. Related to this would be increased the results stem from the increases in electric research and development on electricity generating capacity and generation requirements storage technologies, which would allow associated with climate-induced changes in demand. lower cost, more efficient powerplants to The implications include the following: generate, at off-peak times, electricity for use during peak periods. Climate change could result in overall fuel mixes for electricity generation that differ Because increases in customer demands for from those expected in the absence of electricity may be particularly concentrated climate change. in certain seasons and at peak periods, conservation and especially load Climate change would not evenly affect management programs that improve the regional demands for electricity. Greater efficiency or change the patterns of impacts would occur in regions where customer uses of electricity could be more weather-sensitive end uses (particularly air- cost-effective when considered in the conditioning) are important sources of context of potential changes in climate. electricity demand. Substantially greater climate change impacts were estimated for Increased electricity generation implies the the Southeast and Southwest than for other potential for increased adverse regions, especially the northern tier of environmental impacts depending upon states. Other impacts not addressed in this generating technology and fuel-use study, such as the availability of water for assumptions. Potential adverse impacts hydropower generation and powerplant compared with the base case are associated cooling, also would be more important in with the following: some regions (e.g., the West) than in others. -- air quality (e.g., emissions of sulfur dioxide, NOx, and other pollutants); Regional differences in capacity and generation requirements suggest that -- land use for new powerplant sites, fuel important new opportunities for extraction, fuel storage, and solid waste interregional bulk power exchanges or disposal; capacity sales may arise as a result of climate change. -- water quality and use (e.g., for powerplant cooling and fuel The impacts of uncertain climate conditions processing); and over the long term could pose significant planning and economic risks. Because of resource depletion, especially of long lead times required to plan and build nonrenewable fuels such as natural gas. economic baseload generating capacity, the ability of utility planners to correctly Of particular concern would be additional anticipate climate change could result in water withdrawal and consumption lower electricity production costs. The requirements in areas where water supplies magnitude of these risks in some regions may be reduced by climate change. 3 (e.g., the Southeast and the southern Great Plains) could be similar to other uncertainties that utility planners and decisionmakers must face. ³For example, increased electricity generation induced by climate change in northern California could increase requirements for water withdrawal by 600 to 1,200 million cubic If the result is confirmed that the majority feet and for water consumption by 200 to 400 million cubic feet of new capacity requirements in response to in 2055. Comparable figures for the southern Great Plains in climate change are for peaking capacity, a 2055 would be water withdrawal of 5,800 to 11,500 million cubic feet and consumption of 1,800 to 3,500 million cubic feet. 195 Chapter 10 Increased electricity generation also implies policies should be considered with respect increased emissions of CO2 and other to their potential implications related to greenhouse gases compared with base case climate change issues. emissions. For example, if the estimated increases in climate change-induced Increases in electricity demands induced by generation reported in Table 10-1 were met climate change will make achievement of by conventional technologies, CO2 energy conservation goals more difficult. emissions could increase by 40 to 65 million For example, the conference statement tons per year by 2010 and by 250 to 500 from "The Changing Atmosphere: million tons per year by 2055. 4 Use of Implications for Global Strategy" lower CO2-emitting technologies and fuels (Environment Canada, 1988) calls for -- such as efficient conversion technologies reductions in CO₂ emissions to be achieved and nuclear or renewable resources -- in part through increased efforts in energy would lower these estimated impacts. efficiency and other conservation measures. An initial goal for wealthy, industrialized nations set by the conference is a reduction POLICY IMPLICATIONS in CO2 emissions through conservation of approximately 10% of 1988 emissions levels In general, the study results suggest that utility by 2005. The impacts of climate change to planners and policymakers should begin now to increase electricity demand should be assess more fully and to consider climate change as factored into the policies and plans a factor affecting their planning analyses and designed to achieve this conservation goal. decisions. If more complete and more detailed analyses support the socioeconomic and Similarly, climate change impacts may environmental implications of the climate change exacerbate the difficulties or costs effects described above, they should be explicitly associated with implementing acid rain addressed in planning analyses and decisions. mitigation strategies being considered by Specific policy implications related to the findings the Congress. However, these strategies include the following: center primarily on near-term solutions focusing on emissions reductions from In formulating future National Energy existing powerplants, and the impacts of Plans, the Department of Energy may wish climate change may not be large within that to consider the potential impacts of climate time frame. change on utility demands. Although not addressed directly in the The interactions of climate change and the analyses underlying this report, state and current efforts of the Federal Energy federal agencies should consider mitigation Regulatory Commission (FERC) to strategies that include energy conservation; restructure the electric utility industry are increased efficiency in the production, difficult to assess. For example, the conversion, and use of energy; and the industry's response to FERC policies could development and reliance on fuel sources either accelerate or reduce the rate of with low CO2 emissions. emissions of greenhouse gases, depending upon changes in the mix of generating fuels and effects on the efficiency of electricity RESEARCH NEEDS production. The possible alternative responses should be assessed, and FERC Important areas for further climate change research include improved methods for developing and disseminating climate change scenarios, with 4 Note, however, that these increases in emissions from particular emphasis on (1) improved estimates of climate variables (in addition to temperature) electricity production could be offset, at least in part, by reduced demand for space heating provided by natural gas and relevant to utility impact assessment (e.g., oil furnaces or by other direct uses of fossil fuels. 196 Electricity Demand hydrologic factors, winds); (2) estimates of the REFERENCES possible impacts of global warming on variations in weather conditions and the occurrence of extreme Environment Canada. 1988. The Changing events; (3) continued attention to estimates of the Atmosphere: Implications for Global Security rate of climate change over time; and (4) estimates Conference Statement. Environment Canada, of climate change at a more disaggregated regional Government of Canada. Toronto; June 27-30. or local level. Linder, K.P., M.J. Gibbs, and M.R. Inglis. ICF Follow-on research suggestions on the utility Incorporated. 1987. Potential Impacts of Climate side include (1) refinement of the analytical Change on Electric Utilities. Report 88-2. Albany, approach, in part through lessons learned from New York: New York State Energy Research and additional utility-specific analyses; (2) more detailed Development Authority. (Note: This report was and complete analyses of the weather sensitivity of also published by the Electric Power Research customer demand for electricity; (3) extension of the Institute, Palo Alto, California, in January 1989; approach to consider other pathways (including report no. EN-6249.) indirect and secondary effects) through which climate change could affect utility investments and Stokoe, P.K., and M. LeBlanc. P. Lane and operations; and (4) an assessment of the value of Associates, Ltd., and Discovery Consultants, Ltd. improved climate change information to utility 1987. Socio-economic Assessment of the Physical planners and managers. and Ecological Impacts of Climate Change on the Marine Environment of the Atlantic Region of Canada, Phase I. Halifax, Nova Scotia, Canada: School for Resource and Environmental Studies, Dalhousie University. 197 CHAPTER 11 AIR QUALITY FINDINGS ultimate effect on acid deposition is difficult to assess because of changes in clouds, winds, and Potential changes in regional temperatures, precipitation. precipitation patterns, clouds, windspeed and direction, and atmospheric water vapor that will -- Visibility may decrease because of the accompany global climate change will affect increase in hydrocarbon emissions and the future air pollution levels and episodes in the rate at which sulfur dioxide is oxidized to United States. sulfate. While uncertainties remain, it is likely that an -- The small increase in temperature will not increase in global temperatures would have the significantly affect carbon monoxide following effects on air quality, if other variables emissions. remain constant. These potential impacts should be interpreted as relative changes as Preliminary analyses of the effects of a scenario compared with air quality levels without climate of a 4°C temperature increase in the San change. This chapter does not predict what will Francisco Bay area, with no change in emissions happen to air quality without climate change or other climate variables, on ozone and does not consider changes in anthropogenic concentrations suggest that maximum ozone emissions or technology. concentrations could increase by approximately 20%, that the area in which the National -- Ozone levels in many urban areas would Ambient Air Quality Standard (NAAQS) would increase because higher global temperatures be exceeded would almost double, and that the would speed the reaction rates producing number of people-hours of exposure would ozone in the atmosphere. triple. The Midwest and Southeast also could incur high concentrations and an increase in the -- Natural emissions of hydrocarbons would area of high ozone by a factor of three. increase with a temperature rise. Natural emissions of sulfur would also change, but Increases in ambient ozone levels resulting from the direction is uncertain. The climate change could increase the number of hydrocarbons and nitrogen oxides nonattainment areas and make attainment more participate in reactions that produce ozone. expensive in many regions. Preliminary estimates suggest that an expenditure of several -- Manmade emissions of hydrocarbons, million dollars per year may be necessary for nitrogen oxides, and sulfur oxides may rise volatile organic compound (VOC) controls if more fossil fuel is used to meet higher above those needed to meet standards without electricity needs (see Chapter 10: Electricity climate change. The total costs for additional Demand) and if technology does not air pollution controls that may be needed improve. because of global warming cannot be estimated at this time. -- The formation of acidic materials (such as sulfates) would increase with warmer Because of the close relationship between air temperatures because sulfur and nitrogen pollution policies and global climate change, it oxides would oxidize more rapidly. The is appropriate for EPA to review the impact of 199 Chapter 11 global climate change on air policies and the emissions can enter their circulation patterns. In impact of air pollution regulations on global addition, they are frequently free of clouds, resulting climate change. in maximum sunlight and therefore more photochemical ozone production during the day. Also, during the evenings, the clear skies allow RELATIONSHIP BETWEEN surface-based (see below) inversion layers to form, CLIMATE AND AIR QUALITY concentrating pollutants in a small volume of air and often creating very high air pollution levels. The summer of 1988 provided direct evidence Climatologically, certain places in the country, of the importance of weather to pollution episodes such as the Great Plains and the Northeast (Figure in the United States. Despite significant progress in 11-1A), are frequently windy, and others, such as reducing emissions of many pollutants over the last the Southwest (Figure 11-1B), frequently have large decade, the extended stagnation periods and high mixing depths. These areas will have cleaner-than- temperatures caused ozone levels in 76 cities across average air if they do not contain too many the country to exceed the national standard by at pollutant sources. Areas, such as California, that least 25%. Whether this recent summer is an are frequently affected by high-pressure systems -- appropriate analog for the future cannot be causing lower windspeeds and smaller mixing depths determined with certainty, but scientists have will have more major air pollution episodes. recognized for some time that air pollution does vary with seasons and is directly affected by Circulation ventilation, circulation, and precipitation, all of which could be affected by future global climate Two semipermanent high-pressure systems are changes. important to the global circulation pattern and greatly influence U.S. air pollution climatology: the Ventilation large Pacific high, which is often situated between the Hawaiian Islands and the west coast of North Two major factors, referred to as "ventilation" America, and the Bermuda high, located over the when considered together, control the dilution of western Atlantic Ocean. pollutants by the atmosphere: windspeed and the depth of the atmospheric mixing layer (frequently The Pacific high often results in extended called the mixing depth). If windspeed is high, periods of air stagnation over the western United more air is available to dilute pollutants, thus States from Oregon and California to over the lowering pollutant concentrations. The mixing layer Rockies, and is responsible for many severe ozone (the distance between the ground and the first episodes in southern California. Air stagnation upper-layer inversion) tends to trap pollutants associated with the westward extension of the because the inversion above it acts as a barrier to Bermuda high occurs most often during the summer vertical pollutant movement. Thus, pollutant months and affects the eastern United States from concentrations decrease as mixing depth increases, southern Appalachia northward to New England. providing greater dilution. Within the Bermuda high, pollutants are slowly transported from the industrial areas of the Ohio The ventilation characteristics of an area River Valley into the populated areas of the change, depending on whether a high- or low- Northeast. The Bermuda high is also responsible pressure system is present. Low-pressure systems for the general southwest-to-northeast airflow in the usually produce good ventilation because they summer, carrying pollutants along the metropolitan normally have greater mixing depths and corridor from Richmond to Boston and exacerbating windspeeds, and precipitation is often associated the ozone problem in the Northeast. with them. High-pressure systems, on the other hand, generally produce poor ventilation conditions Precipitation because they frequently have smaller mixing depths on their western sides and lower windspeeds. They Atmospheric pollutants in both particulate and also tend to move more slowly than lows, so more gaseous forms are incorporated into clouds and 200 Air Quality A B 12 6 789 8 16 16 14 12 12 9 8 I2 10 8 4 6 16 IO 8 I2 8 5 5 7 I2 16 6 24 26 10 6 26 10 I2 8 8 24 20 16 9 9 I2 14 Figure 11-1. (A) Mean annual windspeed averaged through the afternoon mixing layer (speeds are in meters per second); (B) mean annual afternoon mixing height, in hundreds of meters (adapted from Holzworth, 1972). precipitation. These pollutants can then be a number of pollutants, including total suspended transported to the ground through rainfall (wet particulates (TSP), O₃, carbon monoxide (CO), deposition). Cloud-formation processes and the nitrogen dioxide (NO₂), lead, and sulfur dioxide consequent type of precipitation, together with the (SO₂). This section does not attempt to predict intensity and duration of precipitation, are future trends in emission levels. important in determining wet deposition of pollutants. Total Suspended Particulates Annual average TSP levels decreased by 23% PATTERNS AND TRENDS IN AIR between 1977 and 1986, and particulate emissions QUALITY decreased by 25% for the same period. The more recent TSP data (1982-86) show that concentrations are leveling off, with a 3% decrease in ambient TSP To protect the public health and welfare, the levels and a 4% decrease in estimated emissions U.S. EPA has promulgated National Ambient Air during that time. Quality Standards (NAAQS). In 1986, more people lived in counties with measured air quality levels In the future, air quality may decrease as the that violated the primary NAAQS for ozone (O₃) benefits of current pollution control measures are than for other pollutants (Figure 11-2). affected by increases in population and economic growth. Although millions of people continue to breathe air that is in violation of the primary NAAQS, Sulfur Dioxide considerable progress is being made in reducing air pollution levels. Nationally, long-term 10-year (1977- 86) improvements have been seen for Annual average SO₂ levels decreased 37% from 1977 to 1986. An even greater improvement 201 Chapter 11 Nationally, between 1979 and 1986, O₃ levels POLLUTANT decreased by 13%. Emissions of volatile organic compounds (VOCs), which are ozone precursors, TSP 41.7 decreased by 20% from 1979 to 1986. The estimated number of violations of the ozone SO₂ 0.9 standard decreased by 38% between 1979 and 1986. CO 41.4 The highest concentrations were in southern NO 7.5 2 California, but high levels also persisted in the Ozone 75 Texas gulf coast, the northeast corridor, and other Pb 4.5 heavily populated regions. O IO 20 30 40 50 60 70 80 90 Acid Deposition millions of persons Widespread concern exists concerning the effects of acid deposition on the environment. With Figure 11-2. Number of persons living in counties the present monitoring network density in eastern with air quality levels above the primary National North America, it is now possible to quantify Ambient Air Quality Standards in 1986 (based on regional patterns of concentration and deposition of 1980 population data) (U.S. EPA, 1988). sulfate, nitrate, and hydrogen ions, primary constituents of acid deposition. In Figures 11-3 through 11-5, isopleth maps show the geographic was observed in the estimated number of violations pattern of acid deposition, as reflected by the of the 24-hour standard for so, concentration, concentration and deposition of these three species which decreased by 98%. These decreases (Seilkop and Finkelstein, 1987). correspond to a 21% drop in sulfur dioxide emissions during this 10-year period. However, most For the relatively short period from 1980 and of the violations and the improvements occurred at 1984, evidence indicates the total deposition and source-oriented sites, particularly a few smelter average concentration of sulfate, nitrate, and sites. Additional reductions may be more difficult hydrogen ions in precipitation falling over eastern to obtain. The higher concentrations were found in North America decreased by 15 to 20%. The the heavily populated Midwest and Northeast. observed decreases correspond with reported reductions in the U.S. emissions of sulfur oxides Ozone (SOx) and nitrogen oxides (NOx), and sulfate and nitrate precursors. However, the emission figures A national standard for ambient levels of ozone are subject to estimation error and should be used was established with the original Clean Air Act in cautiously (Seilkop and Finkelstein, 1987). 1972, along with standards for five other pollutants. While headway has been made in meeting all these national air quality standards, progress in meeting STUDIES OF CLIMATE CHANGE the ozone standard has been particularly slow and frustrating for concerned lawmakers and AND AIR QUALITY environmental officials at all levels of government. At the end of 1987, the date anticipated in the act Some of the climate factors that could affect air for final attainment of the ozone standard, more quality are listed in Table 11-1. To explain these than 60 areas had not met the standard. In recent relationships, two projects were undertaken for this years, the number of nonattainment areas has report to identify the potential impacts of climate fluctuated with meteorology, often overwhelming the change on air quality: progress being made through reduced emissions. Thus "bad" weather (summertime conditions 1. Climate Change and Its Interactions with favorable to ozone formation) in 1983 led to an Air Chemistry: Perspectives and Research increased number of nonattainment areas, and Needs Penner, Connell, Wuebbles, and "good" conditions in 1986 led to a decreased number of areas. 202 Air Quality SULFATES CONCENTRATION DEPOSITION 30 13 4.0 1980 25 3.0 25 15 LS 20 20 1984 20 20 20 L5 20 15 Figure 11-3. Isopleth maps of average annual concentrations (mg/liter) and total annual deposition (g/m²) of sulfates in 1980-84 (Seilkop and Finkelstein, 1987). NITRATES CONCENTRATION DEPOSITION L5 LS 20 20 30 25 2.0 2.5 1980 L5 L5 LO L5 LS 20 20 2.0 1984 LO LO LO LO LO Figure 11-4. Isopleth maps of average annual concentration (mg/liter) and total annual deposition (g/m²) of nitrates in 1980-84 (Seilkop and Finkelstein, 1987). 203 Chapter 11 HYDROGEN IONS CONCENTRATION DEPOSITION p.os 0.00 0.00 0.00 1980 0.05 0.00 0.06 0.05 0.03 0.04 0.04 0.04 0.05 0.05 0.05 0.05 0.00 0.06 0.03 0.04 1984 0.07 0.04 0.03 Figure 11-5. Isopleth maps of average annual concentration (mg/liter) and total annual deposition (g/m²) of hydrogen ions in 1980-84 (Seilkop and Finkelstein, 1987). Table 11-1. Climate Change Factors Important for Regional Air Quality Changes in the following affect air quality: 1. the average maximum or minimum temperature 6. the vegetative and soil emissions of and/or changes in their spatial distribution hydrocarbons and NOₓ that are sensitive to leading to a change in reaction rates and the temperature and light levels, leading to changes solubility of gases in cloud water; in their concentrations; 2. stratospheric O₃ leading to a change in reaction 7. deposition rates to vegetative surfaces whose rates; absorption of pollutants is a function of moisture, temperature, light intensity, and other 3. the frequency and pattern of cloud cover factors, leading to changes in concentrations; leading to a change in reaction rates and rates of conversion of so₂ to acid deposition; 8. energy usage, leading to a change in energy- related emissions; 4. the frequency and intensity of stagnation episodes or a change in the mixing layer leading 9. aerosol formation, leading to changes in to more or less mixing of polluted air with reaction rates and the planetary albedo background air; (reflectivity); and 5. background boundary layer concentrations of 10. circulation and precipitation patterns leading to water vapor, hydrocarbons, NO, and O₃, a change in the abundance of pollutants leading to more or less dilution of polluted air deposited locally versus these exported off the in the boundary layer and altering the chemical continent. transformation rates; Source: Adapted from Penner et al. (Volume F). 204 Air Quality Covey Lawrence Livermore National maximum temperature. In 1988, the mean Laboratory (Volume F) maximum temperature was 77°F and there were 12 ozone excursions. In 1984, with a mean 2. Examination of the Sensitivity of a Regional temperature of 73.50°F, there was only one Oxidant Model to Climate Variations - excursion. Morris, Gery, Liu, Moore, Daly, and Greenfield - Systems Applications, Inc. Temperature-dependent modeling studies were (Volume F) conducted by Gery et al. (1987). For this modeling effort, Gery et al. used the OZIPM-3 trajectory The literature does not contain studies on the model, which is city specific. The scenarios for the effects of climate change on air quality. Thus, these different cities used actual observed mixing heights, studies should be considered as preliminary analyses solar radiation and zenith angle, and pollutant of the sensitivity of air quality to climate change. concentrations characteristic for the particular city considered for June 24, 1980. This base case was Climate Change and Its Interactions with chosen because it was a high-pollution day, and Air Chemistry ambient data were available. The increased temperature scenarios applied the increase Penner et al. conducted a literature review of throughout the day and were added to the base case studies on the relationship of climate and air quality. scenario. The light intensity increase was achieved They also organized a workshop on the issue. by increasing the photolyses rates for nitrogen dioxide, formaldehyde, acetaldehyde, hydrogen Effect of Climate Change on Ozone Formation peroxide, and ozone. Results for New York in June 1980 are shown in Table 11-2. In general, ozone Changes in ventilation, circulation, precipitation, concentration increased with increasing temperature. and other aspects of climate affect the concentra- The concentration of hydrogen peroxide (H₂O₂), a tions of the ozone precursors (VOCs and NOv). strong oxidant that converts SO₂ to sulfuric acid, Climate changes can also increase or decrease the was also observed to increase with higher rates at which these precursors react to form ozone. temperatures. This is compatible with the increase The effects of change in global temperature and in in ozone because the entire photochemical reaction stratospheric ozone concentration on tropospheric process is accelerated when temperature rises. As ozone precursor concentrations, reaction rates, and a result, cities currently violating the ozone NAAQS tropospheric ozone concentrations are discussed will be in violation to a greater degree in the future, below. and cities that are complying with the NAAQS now could be forced out of compliance just by a Temperature Change temperature increase. Figure 11-6 shows the predicted increase in low-level ozone for two Studies of the Effects of Temperature on temperature increases in Los Angeles, New York, Ozone. Smog chamber and modeling studies have Philadelphia, and Washington. shown that ozone levels increase as temperature increases. Kamens et al. (1982) have shown in an Modeling studies by Penner et al. have shown outdoor smog chamber study that the maximum that the effect temperature has on ozone formation ozone concentration increases as the daily maximum also depends on the ratio of volatile organic temperature increases (holding light intensity compounds to nitrogen oxides, both of which are constant). Their data show that there is no critical ozone precursors. Figure 11-7 shows that ozone "cut-off" temperature that eliminates photochemical levels will generally go up, except in areas where the ozone production. Instead, a general gradient is ratio of VOCs to NOₓ is low. observed as a function of temperature. Temperature change has a direct effect on Samson (1988) has recently studied ambient ozone concentrations because it increases the rates data for Muskegon, Michigan, and found that the of ozone-forming reactions. However, a number of ozone excursions above the standard temperature rise can also affect ozone formation by (0.12 ppm) is almost linearly related to mean altering four other aspects of climate or the atmosphere: cloud cover, frequency and intensity of 205 Chapter 11 Table 11-2. Maximum Hourly Concentrations and Percentage Changes for Ozone, H2O2, and PAN for the Future Sensitivity Tests Using an EKMA Model for the Simulation of June 24, 1980, New York Ozone Percent change Concentration (ppm) (from base) Change in Temp (°C) 0 +2 +5 0 +2 +5 Stratospheric Ozoneᵃ Base 0.125 0.130 0.138 -- 4 10 -16.6% 0.150 0.157 0.167 20 26 34 -33.3% 0.165 0.170 0.178 32 36 42 Hydrogen Peroxide (H₂O₂) Percent change Change in Temp (°C) Concentration (ppb) (from base) Stratospheric Ozoneᵃ Base 0.05 0.06 0.08 -- 20 60 -16.6% 0.43 0.58 0.84 760.0 1060 1580 -33.3% 3.08 3.31 3.60 6060.0 6520 7100 Peroxyacetyl Nitrate (PAN) Percent change Change in Temp (°C) Concentration (ppb) (from base) Stratospheric Ozoneᵃ Base 3.98 3.50 2.79 --- -12 -30 -16.6% 5.85 5.26 4.34 47 32 9 -33.3% 7.59 6.73 5.49 91 69 38 ᵃBase refers to the present stratospheric ozone column. The -16.6 and -33.3% refer to a depletion of the base value. Ultraviolet light will increase with the depletion (Gery et al., 1987). stagnation periods, mixing layer thickness, and cloud cover occurs. If clouds occur in the afternoon reactant concentrations. or evening, little effect is observed in the ozone production, but if clouds occur during the morning Effect of Changes in Cloud Cover. The hours, photochemical reactions are slowed, and less reduction in light intensity caused by increased ozone is produced. Jeffries et al. (1989) suggest cloud cover can reduce ozone production. Penner et that cloud cover can decrease ultraviolet radiation al. (Volume F) calculate that a reduction in light by 7 to 14% in their outdoor smog chamber located intensity of 50% throughout the day will reduce the in North Carolina. Although a global temperature ozone formation. However, the magnitude of ozone change would affect cloud cover, the type and reduction depends on the time of day when the direction of the change are unknown. 206 Air Quality 15.0 6 14.0 13.0 5 12.0 11.0 HC/NOx 28 10.0 PERCENT INCREASE IN OZONE 9.0 8.0 CONCENTRATION (ppm) 4 PEAK OZONE 3 HC/NOx=7 7.0 6.0 2 5.0 4.0 HC/NOx 7, INCREASED BL HEIGHT 3.0 HC/NOx=2 2.0 1.0 0 10 15 20 25 30 35 40 0.0 L.A. N.Y. Phil. Wash. L.A. N.Y. Phil. Wash. TEMPERATURE (°C) 2°C TEMPERATURE INCREASE 5°C TEMPERATURE INCREASE BL = Boundary Layer Figure 11-6. Percent increase in predicted O₃ over Figure 11-7. The effect of temperature on the peak future base case (0.12 ppm) for two temperature O₃ concentrations predicted in a box model increases in four cities (Gery, 1987). calculation of urban O₃ formation. Calculations are shown for three hydrocarbon to NOₓ ratios. The effect of increasing the boundary layer depth for the case with a hydrocarbon to NO ratio of 7 is also shown (Penner et al., Volume F). The Penner et al. study assumes that cloud Smog chamber studies have shown that at high cover causes an equal decrease in all wavelengths of pollutant levels, increases in water vapor can solar radiation. However, clouds are not expected significantly accelerate both the reaction rates of to cause an equal decrease at all wavelengths. Solar VOCs and the rate of oxidant formation (Altshuller radiation is needed to form ozone. Since Penner et and Bufalini, 1971). Walcek (1988) has shown with al. may have underestimated the intensity of some the use of a regional acid deposition model wavelengths of light, they may have overestimated (RADM) that the ozone, hydrogen peroxide, and the decrease in ozone production. sulfate production rates in the boundary layer of the troposphere all increase with increasing water vapor. Effect of Water Vapor. Water vapor is involved in the formation of free radicals (reactive Effect of Changes in Frequency and Intensity of compounds) and hydrogen peroxide, which are Stagnation Periods. As noted previously, high- necessary for the formation of ozone. Global pressure systems significantly enhance ozone increases in temperature are expected to raise formation potential. During a high-pressure tropospheric water vapor levels. episode, pollutants are exposed to high temperatures and prolonged irradiation (Research If sources of water vapor are not perturbed by Triangle Institute, 1975), resulting in high levels of vegetative changes, and if global circulation patterns ozone. If the intensity and frequency of high- do not significantly affect precipitation events (an pressure episodes increase with global warming, unlikely assumption), then global water vapor levels then ozone levels can be expected to be even higher. are expected to increase with increasing temperature. A temperature increase of 2°C could Effect of Changes in Mixing Layer Thickness. raise the water vapor concentration by 10 to 30% As shown in Figure 11-7, increases in the mixing (Penner et al., Volume F). This change should layer height decrease ozone formation, presumably affect both oxidant formation and sulfur dioxide because there are less ozone precursors per volume oxidation (acid deposition). of atmosphere. An increase of global temperature 207 Chapter 11 would probably lead to an increase in average mixing depths as a result of greater convection, which raises the mixing depth and increases mixing. 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 0.28 Effect of Changes in Reactant Concentrations. 0.24 0.24 The concentrations of ozone precursor pollutants 0.48 (VOCs, NOX) play a large part in determining the 0.20 0.20 amount of ozone produced. With increasing 0.36 NOx, PPM 016 34 0.16 temperature, natural hydrocarbon emissions are 32 0.12 030 expected to increase. Also, unless preventive 0.12 028 measures are taken, manmade emissions would 0.08 0.24 0.08 increase (vapor pressure of VOCs increases with 020 0.04 016 004 increasing temperature). If these ozone precursors 0.12 0.00 increased in concentration, ozone production would 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 increase. VOC, PPMC Lamb et al. (1985) have shown that natural hydrocarbon (VOC) emissions from deciduous forests would increase by about a factor of three Figure 11-8. Ozone isopleths as a function of NOₓ with a temperature change from 20 to 30°C. and volatile organic compounds (VOCs) (Dodge, However, as discussed in Chapter 5: Forests, the 1977). abundance of some deciduous forests could decline because of global warming. However, grasslands or shrubs that replace forests would still emit because stratospheric ozone regulates the amount of hydrocarbons. The net effect is probably uncertain. ultraviolet (UV) radiation available for producing Emissions of NOₓ from powerplants would grow ozone in the troposphere. Stratospheric ozone because of a greater demand for electricity during absorbs UV light from the sun and decreases the the summer months. Soil microbial activity is also UV energy striking the Earth's surface. When expected to increase with increasing temperature. stratospheric ozone is depleted by the This will increase natural emissions of NOx. chlorofluorocarbons (CFCs) generated by human Evaporative emissions of VOCs from vehicles and activity, more UV radiation reaches the Earth's refueling would also be expected to rise with surface, which increases the photolysis rates of warmer temperatures. However, exact predictions compounds that absorb solar radiation (NO₂, of the effects of all these factors on ozone formation formaldehyde, acetaldehyde, O₃, and H₂O₂). Faster are difficult to make because the relationship photolysis produces more free radicals (high-energy between precursor emissions and ozone is extremely species) that increase the amount of smog. Thus, complex and not fully understood, and because less stratospheric ozone will lead to enhanced ozone increases in emissions are difficult to quantify. formation in the troposphere. An example of this complex relationship Modeling results for New York from Gery et al. between ozone and its precursors is shown in Figure (1987) show that tropospheric ozone increased when 11-8 (Dodge, 1977). At high VOC levels and low stratospheric ozone decreased (see Table 11-2). NOx, adding or reducing VOCs has very little effect They also show that H2O2 and peroxyacetyl nitrate on ozone formation. Likewise, when NOX (PAN) yields increased. H2O2 is a strong oxidant concentrations are high and VOC concentrations that converts SO₂ to sulfuric acid, and PAN is an are low, increasing NOₓ reduces ozone formation air pollutant that damages plants and irritates eyes. while lowering NO increases ozone formation. The 16.6 and 33.3% decreases (Table 11-2) in Thus, VOCs and NOx must be examined together stratospheric ozone far exceed the expected when considering any ozone reduction strategy decrease resulting from the buildup of CFC based on controlling ozone-forming precursors. concentrations. This is especially true since the Stratospheric Ozone Change 1 Photolysis is the breakdown of chemicals as a result of the Changes in stratospheric ozone concentration absorption of solar radiation. can also affect tropospheric ozone formation 208 Air Quality Montreal Protocol agreement will limit CFC Organic acids, such as formic and acetic acids, production. These high values of stratospheric are also formed in the atmosphere. However, their ozone depletion are used only for illustrative relative importance to the acid deposition problem purposes. is unknown at present. Because they are weak acids (compared to H2SO4 and HNO3), their contribution Changes in Tropospheric Hydroxyl Radicals to the problem is expected to be much less than that of the inorganic acids (Galloway et al., 1982; Hydroxyl radicals (reactive compounds) are the Keene et al., 1983, 1984; Norton, 1985). most important free radicals found in the atmosphere. These reactive compounds are The acids produced in the atmosphere can be responsible for removing many atmospheric "dry deposited" to the Earth's surface as gases or pollutants (such as CH₄, VOCs, methyl chloroform, aerosols, or they can be "wet deposited" as acid rain. CO) from the atmosphere (Penner et al., Volume Changes in total acid levels depend on changes both F). Without these free radicals, pollutants would in atmospheric chemistry and changes in not be removed from the atmosphere and would precipitation. Wet deposition is affected most by build up to higher levels (global heating would be the amount, duration, and location of precipitation. greater). Hydroxyl radicals in the free troposphere Since the direction of regional precipitation are produced primarily by the decomposition of changes is unknown, it is not known whether acid ozone by sunlight and the subsequent reaction of rain will increase or decrease in the future. high-energy oxygen with water. In the urban However, many of the same factors that affect atmosphere, hydroxyl radicals are produced through ozone formation will also affect the total deposition a complex series of reactions involving VOCs, of acids. nitrogen oxides, and sunlight. The solar photolysis of hydrogen peroxide also gives rise to hydroxyl Temperature Change radicals. This occurs in both urban and rural areas. Higher temperatures accelerate the oxidation The effect of global climate changes on rates of SO₂ and NOx to sulfuric and nitric acids. hydroxyl radical abundance is unclear. In urban Gery et al. (1987) have shown that a temperature areas with increases in VOCs and NO,, a rise would also speed the formation of H2O2, temperature increase will increase hydroxyl radical increasing the conversion of SO₂ to sulfuric acid concentration. Also, if natural hydrocarbons and (see Table 11-2). Hales (1988) studied the NOX increase in rural areas, hydroxyl radicals are sensitivity to a 10°C temperature rise using the expected to increase. However, if methane, CO, storm-cloud model PLUVIUS-2. Considering only and natural hydrocarbons increase without an the chemistry occurring with a 10°C temperature additional increase in NO,, then hydroxyl radicals rise, sulfate production increased 2.5 times. No will be depleted. A definitive prediction on the modeling was performed at more modest effect of increasing temperature on global temperature increases (e.g., ~4°C); however, it is concentrations of hydroxyl radicals cannot be made likely that oxidation would also increase with a at this time. smaller increase in temperature. The limiting factor in the oxidation of SO₂ appears to be the availability Effect of Climate Change on Acid Deposition of H₂O₂. The model also suggested that a temperature increase would cause more sulfuric acid Rainwater and surface waters are more acidic to form near the sources where SO₂ is emitted. than natural background levels because of industrial and mobile emissions of so, and NO,, which form Effect of Global Circulation Pattern Changes. sulfuric and nitric acids in the atmosphere. In the Potential changes in global circulation patterns air, sulfuric acid (H₂SO₄) is produced primarily by would greatly affect local acid deposition, because the reaction of SO₂ with hydroxyl radicals (high- they would alter ventilation and precipitation energy species); in clouds, the oxidation of SO₂ to patterns. Galloway et al. (1984) have calculated that H2SO4 is more complex, involving reactions with over 30% of the sulfur emissions from the eastern hydrogen peroxide and other dissolved oxidants. United States are transported to the north and Nitric acid (HNO₂) is produced in air by the farther east. Changes in circulation patterns would reaction of hydroxyl radicals with NOₓ. 209 Chapter 11 affect this transport, although the direction or changes in climate. The scenario assumed that magnitude of the effect is unknown. temperatures would be 4°C warmer than in the base case, but all other climate variables were held Effects of Changes in Emissions. If electricity constant (relative humidity was held constant). The demand rises with rising temperatures (see Chapter scenario assumed no change in emission levels, no 10: Electricity Demand), if more fossil fuels are change in boundary layer, and no change in wind burned, and if technology is not improved, SO₂ and velocity. NOₓ emissions will increase. An approximate 10% growth in use of electricity in the summer could The RTM-III is a three-dimensional model that increase SO₂ emissions during the summer by represents point sources embedded in a grid approximately 30% if present-day technology is used framework. The model has three prognostic vertical in the future. This, in turn, would increase acid layers and a diagnostic surface layer. This means deposition. that the surface layer is represented by actual observations. The other three layers are predicted Effects of Reduced Stratospheric Ozone. A by using the surface layer data. The photochemical decrease in stratospheric ozone due to CFCs may reactions are based on the latest parameterized increase acid deposition because more UV radiation chemical mechanism. would be available to drive the chemical reactions. As discussed above, a modeling study by Gery et al. Limitations (1987) showed an increase in the yield of H2O2 when stratospheric ozone was reduced by 16 and Perhaps the most important limitation is that 33%. Because H2O2 is a strong oxidant, SO₂ would emission levels were held constant. It is likely that probably also be oxidized more quickly into sulfate future emission levels will be different, although this aerosols and acid rain, but this depends on the study did not estimate how. The results of this availability of water vapor (e.g., clouds, rain). study are useful for indicating the sensitivity of Implementation of the Montreal Protocol should ozone formation to temperature, but should not be help reduce CFC emissions. considered as a prediction of future ozone levels. The model ignored future increases in emissions Reduced Visibility. The growth in natural that would occur with increased temperatures. The organic emissions and increases in sulfates resulting estimates for ozone are only coarse approximations. from warmer temperatures should reduce visibility, Morris et al. used the National Acid Precipitation assuming that the frequency of rain events, wind Assessment Program (NAPAP) emissions data of velocity, and dry deposition rates remain the same. 1980. These data appear to underestimate actual If rain events increase, washout/rainout should ratios of VOCs to NOₓ as measured in urban areas. increase and visibility would be better than Ching et al. (1986) state that for most cities, the predicted (see Chapter 3: Climate Variability). NAPAP data underestimate VOC emission values by a factor of three or more. The model simplified some reactions of the hydrocarbons (VOCs) MODELING STUDY OF CLIMATE because the chemistry is not well known. AND AIR QUALITY This study did not estimate climate-induced alterations in most meteorological variables, except Study Design temperature and water vapor, which is an oversimplification. For example, this study assumed Morris et al. (Volume F) applied a regional that the mixing heights remain unchanged for the transport model RTM-III to an area covering temperature increase scenario; in reality, mixing central California and a region covering the heights could increase with rising temperature. midwestern and the southeastern United States. Holding the mixing heights constant probably The model was run for the present-day conditions overemphasized the importance of temperature in and for a future climate. For California, Morris et oxidant production, because an increased mixing al. used input data from August 5-10, 1981; for the layer depth might have had a dilution effect. Also, Midwest and the Southeast, they used input data as stated earlier, cloud cover will affect ozone from July 14-21, 1981. These were periods with production. If cloud cover increases, then ozone is high ozone levels and may be most sensitive to expected to decrease. Frequency and intensity of 210 Air Quality stagnation periods can also have profound effects on little change in ozone levels was observed with the ozone formation. This modeling exercise did not increased temperature. consider these factors. Midwest and Southeast Study Results The results from applying RTM-III to the Central California Study midwestern and southeastern areas are shown in Table 11-4. On one particular day (July 16), raising Table 11-3 summarizes the results from the the temperature caused maximum ozone to increase base case scenario and a climate sensitivity scenario from 12.5 pphm to 13.0 pphm (Figure 11-10). that used a 4°C temperature increase and an Although this is only a slight increase (0.5 pphm), attendant increase in water vapor concentration. the predicted area of exceedance of the ozone All of the days studied show a larger area exposed NAAQS increased by almost a factor of three, from to high levels of ozone. An increase in temperature 9,800 to 27,000 square kilometers. The differences may lengthen the duration of high ozone levels, occurred mainly in the upper Midwest. In general, although the maximum levels may be the same. the results range from a reduction of 2.4% to an Figure 11-9 illustrates the August 6 base case and increase of 8.0% in ozone levels. Although a climate sensitivity case. The temperature change temperature increase will generally increase ozone increased the August 6 maximum ozone formation, it is noted in Table 11-4 that on two concentration from 15 parts per hundred million days, July 14 and July 21, no ozone increases were (pphm) to 18 pphm, a 20% increase in ozone. The observed. This occurs when there are insufficient area in which the NAAQS was exceeded almost precursors to sustain ozone formation. Under these doubled from 3,700 to 6,600 square kilometers. conditions, ozone is produced more quickly with increasing temperature but the total amount The temperature increases in the two main produced need not be greater and could even be cities in the San Joaquin Valley (Fresno and less in some cases. Bakersfield) resulted in an approximate 0.5-pphm increase (approximately 8%) in maximum daily Both modeling exercises indicate that ozone concentration. In regions farther away from temperature change alone could increase ozone the emissions, such as the Sierra Nevada Mountains, levels over what they would be without climate change. Table 11-3. Maximum Daily Ozone Concentrations Predicted by the RTM-III for Each Day of the Central California Modeling Episodes for the Base Case and the Case of Climate Sensitivity to Increased Temperature of 4°C Maximum daily ozone concentrations (pphm) Date of Episode Base 4°C temperature Percent (1981) case increase increase August 5 11.8 12.1 3 August 6 15.0 18.0 20 August 7 11.7 13.1 12 August 8 13.5 13.7 2 August 9 10.5 11.2 7 August 10 9.1 9.18 8 Source: Morris et al. (1988). 211 Chapter 11 Exceeds Standard <6 >6 >8 >10 >12 >14 >16 Sacramento Sacramento San Fran. Oakland San Fran. Oakland Stockton Stockton San Jose San Jose Modesto Crow's Landing Modesto Crow's Landing Castle AFB Castle AFB Yosemite Yosemite Salinas Salinas Base Case Climate Sensitivity Scenario No. 1 August 6, 1981 4°C Temperature Increase Figure 11-9. Comparison of estimated maximum daily ozone concentrations (pphm) for the base case and climate sensitivity scenario No. 1 (temperature and water increase) for August 6, 1981 (Morris et al., 1988). Table 11-4. Maximum Daily Ozone Concentrations Predicted by the RTM-III for Each Day of the Midwestern/Southeastern Episode for the Base Case and the Case of Increased Temperature of 4°C Maximum daily ozone concentrations (pphm) Date of Episode Base 4°C temperature Percent (1981) case increase increase July 14 11.3 11.3 0.0 July 15 11.5 11.9 3.5 July 16 12.5 13.0 4.0 July 17 11.7 12.0 2.6 July 18 11.2 12.1 8.0 July 19 13.8 14.8 7.2 July 20 11.1 11.2 0.9 July 21 12.6 12.3 -2.4 Source: Morris et al. (1988). 212 Air Quality Exceeds Standard 7 7 L L L L L 'o 18 '6 be be Base Case Climate Sensitivity Scenario #1 Figure 11-10. Comparison of predicted estimated maximum daily ozone concentrations (pphm) for the base case and climate sensitivity scenario No. 1 (temperature and water increase) for July 16, 1980 (Morris et al., 1988). Population Exposure ECONOMIC, ECOLOGICAL, As discussed above, both the California and AND ENVIRONMENTAL Midwest/Southeast studies show a significant IMPLICATIONS increase in the area that is potentially exposed to higher levels of ozone when the temperature is increased as compared with base case conditions. Ozone Data taken from the 1980 census from central California and the midwestern and southeastern An increase in ozone levels due to climate areas were used to determine the number of people change is important for several reasons: exposed to ozone for the base case and a 4°C temperature rise scenario. Table 11-5 presents the Ozone itself is a radiatively important gas number of people-hours of exposure to ozone and contributes to climate change. Ozone concentrations exceeding 8, 12, and 16 pphm. These absorbs infrared energy much like carbon estimates of human exposure were generated by dioxide. It has been calculated that a 15% multiplying the number of people in the grid cells by increase in tropospheric ozone could lead to the total number of hours that the estimated hourly a 0.1°C rise in global temperature ozone concentration in those grid cells exceeded the (Ramanathan et al., 1987). 8-, 12-, or 16-pphm levels. Actual exposure levels may be less because indoor levels are generally Ozone levels in many areas are just below lower than ambient air levels. the current standard. If emissions are not reduced, any increase in ozone formation may push levels above the standard. 213 Chapter 11 Table 11-5. Number of People-Hours of Exposure to Ozone Concentrations in Excess of 8, 12, and 16 pphm for the Base Case and the Case of Climate Sensitivity to Increased Temperature Exposure to Exposure to Exposure to Scenario O₃ ≥ 8 pphm O₃ ≥ 12 pphm O₃ ≥ 16 pphm Central California Modeling Episode Base case 70,509,216 660,876 0 Increased temperature 102,012,064 2,052,143 92,220 Midwestern/Southeastern Modeling Episode Base case 1,722,590,208 29,805,348 0 Increased temperature 1,956,205,568 47,528,944 0 Source: Morris et al. (1988). Many inexpensive controls for ozone are peak ozone levels could affect a number of potential already in place in nonattainment areas. ozone violations. In the 1983-85 period for example, Increases in ozone levels would require 68 areas showed measured exceedances of the relatively expensive measures to sufficiently ozone air quality standards (for technical and legal reduce ozone precursors to attain the reasons, not all these areas were officially standard. designated nonattainment areas). A 10% increase in ozone levels in that period doubled the number The standard itself is defined in terms of of nonattainment areas to 136. This would include the highest levels of ozone experienced in 41 new metropolitan statistical areas (MSAs) added an area, not average levels. (As a yearly to the list and 27 non-MSAs. These new average, no area of the country would nonattainment areas would add most midsize and exceed the standard of 0.12 ppm.) Thus, a some small cities in the Midwest, South, and East to factor such as temperature that may have a the list of nonattainment areas. modest effect on average levels of ozone formation may have a much more significant The policy implications of this should be put effect on peak levels. into context because the full effect of climate change may not be felt until well into the next century. A rough estimate of each of these factors can Over the next several decades, various national illustrate the potential policy problems created by measures to reduce ozone precursors, such as a a rising temperature scenario. The data in Figure reduction in the volatility of gasoline, may go into 11-9 suggest that 4°C degree rise in temperature effect. These would provide a cushion to marginal may lead to an increase in peak ozone areas and could offset a temperature effect. concentrations of around 10%. A 10% increase in However, other factors suggest that rising temperatures could be a problem. 214 Air Quality Ozone levels and ozone precursors are closely First, emissions from fossil fuel powerplants related to economic expansion and population both influence acid rain and contribute to global growth. Consumer solvents (e.g., paints, sprays, and warming. In the future, global warming may even deodorants) are a major source of ozone increase energy demand and associated emissions. precursors. These are very difficult to control and Because the growth in demand for electricity in are likely to increase in the future in areas currently northern states (see Chapter 10: Electricity attaining the standards. Growth in other sources Demand) may be lower than in southern states, of ozone precursors would bring many areas regional shifts in emissions may occur in the future. relatively close to the limits of the ozone standard. Gradual increases in temperature would make Second, global climate change would influence remaining in compliance with the standard more atmospheric reaction rates and the deposition and difficult. Although any sudden change in the form of acidic material. It is conceivable that number of nonattainment areas as a result of a regions of high deposition may shift or that more secular trend toward increased temperature is acid rain may be transported off the North unlikely, a number of small to midsize cities American continent. Strategies that seek to control eventually may be forced to develop new control powerplants in regions near sensitive areas may or programs. may not be as effective, as global climate change occurs. The implications of warmer temperatures for existing nonattainment areas can also be estimated. Third, global climate change may alter the In these areas, existing and planned control impacts of acid rain on ecological and other systems measures may not be adequate to reach the in as yet unpredictable ways. For example: standard, if additional ozone forms. In the past, EPA has attempted to project the emission Changes in the amount of rainfall may dilute reductions and costs associated with the attempts the effect of acid rain on many sensitive of existing nonattainment areas to reach the ozone lakes. standard. Using the same modeling approach, the effects of a temperature increase were analyzed to Changes in clouds may alter the fertilization estimate the additional tons and costs associated of high-elevation forests. with a projected temperature rise. Extrapolations of existing inventories to the year 2000 suggest that Changes in humidity and frequency of rain higher temperatures could require an additional may alter degradation rates for materials. reduction of 700,000 tons of VOC from an inventory of about 6 million tons. Given that most current Increased midcontinental dryness would alter nonattainment areas already will have implemented the amount of calcium and magnesium in the most inexpensive measures, these additional dust, neutralizing impacts on soils. reductions may cost as much as $5,000 per ton per year. Their aggregate cost could be as much as $3.5 Increased numbers of days without frost billion each year. would decrease forest damage associated with frost and overfertilization by These conclusions should be viewed as atmospheric nitrogen. preliminary. Nonetheless, they demonstrate that the potential economic consequences could be Changes in snowpack and the seasonality of significant for an already expensive program to rainfall would change acid levels in streams combat ozone. and alter the timing and magnitude of spring shocks on aquatic species. Acid Rain Finally, solutions to both problems are The global climate change is likely to affect inextricably linked. Some solutions, such as SO₂ acidic deposition in the near future for several scrubbers and clean coal technologies, may abate reasons. acid rain levels, but they may do little to improve air quality or may increase global warming. Other 215 Chapter 11 solutions, including increased energy efficiency and determined, perhaps as each significant regulation switching fuels to natural gas or to renewable is proposed or reevaluated. energy sources, may provide positive solutions to both problems. The impact of EPA regulations, particularly the impact on energy use and greenhouse In summary, an examination of the time gases, should be a more important weight in horizons of importance to both acid rain and global future regulatory decisions. Since EPA climate change problems suggests that these two regulations often serve as models for other issues should not be viewed in isolation. Emissions, countries, the cost penalty for better energy atmospheric reaction rates, pollutant transport, and usage, while sometimes small in the United environmental impacts will likely be altered by States, may be important on a global basis. climate change. This suggests that a more holistic approach must be taken to air pollution problems Future reports to Congress and major and that proposed solutions should be evaluated on assessments of ecological effects, e.g., the the basis of their contributions to solving both 1990 Acid Deposition Assessment document, problems. should include sensitivity analyses of alternative climates. Risk management decisions of the Agency could then be made POLICY IMPLICATIONS with improved knowledge of climate impacts. The Environmental Protection Agency issues air pollution regulations to improve air quality and RESEARCH NEEDS to protect public health and welfare. In general, current regulations to reduce oxidant levels will also Some of the key questions that need to be provide positive benefits toward a goal of limiting resolved regarding climate change and air quality the rate of growth in global warming. Other include the following: How important will climate programs aimed at reducing carbon monoxide change be relative to other factors such as levels, particularly from mobile sources, or CFCs to population growth to future air pollution problems? protect the stratospheric ozone layer, also positively Is the impact of climate change likely to be affect greenhouse gases and the rate of global significant enough to require totally different air warming. However, the regulatory activities of the pollution strategies? What mix of control strategies Agency have not been retrospectively reviewed to could be most cost effective in reducing acid rain, determine their impacts on global warming. In global warming, tropospheric ozone, and other some cases, there may be important benefits; for pollution problems? The research elements needed example, current emission standards for automobiles to address these issues include basic research, do not encourage more efficient use of gasoline. A sensitivity analyses, full-scale atmospheric modeling, different form of standard, while potentially and cost-effectiveness studies. Examples are disruptive to air pollution efforts, might produce presented below: positive greenhouse gas benefits via reduced energy consumption. These issues will have to be analyzed Basic Research - There is an important need to in the future. understand how manmade and natural emissions of hydrocarbons and other pollutants might change in Because of the climate change issue, the the future when temperature, CO₂, and UV-B following are some of the more important policy radiation increase and other climate parameters issues: vary. Air pollution control agencies such as EPA Sensitivity Analyses - Analyses of ozone should undertake a broad review to concentrations are dependent on boundary layer determine the impact of global climate height, clouds, water vapor, windspeed, UV-B change on air pollution policies. In radiation, and other parameters. Sensitivity tests particular, the cost of added controls using single models could improve our resulting from climate change should be understanding of the relative importance of these 216 Air Quality variables and could provide important information Park, NC: U.S. Environmental Protection Agency. for general circulation modelers. October. Full-Scale Modeling Complete understanding of Hales, J. 1988. In: Sensitivity of Urban/Regional the interactions of climate change and air quality Chemistry to Climate Change: Report of the will ultimately require that general circulation Workshop, Chapel Hill, NC. Wuebbles, D.J., and models and mesoscale chemistry models be linked J.E. Penner, eds. Livermore, CA: Lawrence in some direct or indirect manner. This will require Livermore National Laboratory. Feb. 17-18. the development of innovative approaches between the general circulation and air pollution modeling Holzworth, G.C. 1972. Mixing Heights, communities. Windspeeds and Potential for Urban Air Pollution Throughout the Contiguous United States. Research Cost-Effectiveness Studies There are currently a Triangle Park, NC: U.S. Environmental Protection number of congressional proposals to improve the Agency, Office of Air Programs. Publication No. Clean Air Act and to reduce global climate change. AP-101. To assume that both air quality and global climate change goals are achieved, analyses of the cost- Jeffries, H.E., K.G. Sexton, J.R. Arnold, and T.L. effectiveness of alternating strategies will be Kole. 1989. Validation Testing of New necessary. Mechanisms with Outdoor Chamber Data, Volume 4: Appendixes to Photochemical Reaction Photolysis Rates in the UNC Outdoor Chamber. EPA/600/3- REFERENCES 89/010d. Research Triangle Park, NC: U.S. Environmental Protection Agency. Altshuller, A.P., and J.J. Bufalini. 1971. Photochemical aspects of air pollution. Kamens, R.M., H.E. Jeffries, K.G. Sexton, and A.A. Environmental Science and Technology 5:39-64. Gerhardt. 1982. Smog Chamber Experiments to Test Oxidant-Related Control Strategy Issues. EPA Ching, J.K.S., J.H. Novak, K.L. Schere, and F.A. 600/3-82-014. Research Triangle Park, NC: U.S. Schiermeier. 1986. Reconciliation of Urban Environmental Protection Agency. August. Emissions and Corresponding Ambient Air Concentrations Using Mass Flow Rate Technique. Keene, W.C., J.N. Galloway, and J.D. Holden Jr. Draft report prepared for EPA/ORD; April. 1983. Measurement of weak organic acidity in precipitation from remote areas of the world. Dodge, M.C. 1977. Effect of Selected Parameters Journal of Geophysical Research 88:5122-5130. on Predictions of a Photo-Chemical Model. EPA 600/3-77/048. Research Triangle Park, NC: U.S. Keene, W.C., and J.N. Galloway. 1984. Organic Environmental Protection Agency. June. acidity in precipitation of North America. Atmospheric Environment 18: 2491-2497. Galloway, J.N., G.E. Likens, W.C. Keene, and M.M. Miller. 1982. The composition of precipitation in Lamb B.K., H.H. Westberg, T. Quarles, and D.L. remote areas of the world. Journal of Geophysical Flyckt. 1985. Natural Hydrocarbon Emission Rate Research 87:8771-8788. Measurements From Selected Forest Sites. EPA-600/3-84-001. Research Triangle Park, NC: Galloway, J.N., D.M. Whelpdale, and G.T. Wolff. U.S. Environmental Protection Agency. October. 1984. The flux of sulfur and nitrogen eastward from North America. Atmospheric Environment Norton, R.B. 1985. Measurements of Formate and 18:2595-2607. Acetate in Precipitation at Niwot Ridge and Boulder, Colorado. Geophysical Research Letters Gery, M.W., R.D. Edmond, and G.Z. Whitten. 1987. 12:769-772. Tropospheric Ultraviolet Radiation: Assessment of Existing Data and Effect on Ozone Formation. Ramanathan, V., L. Callis, R. Cess, J. Hansen, I. EPA Report 600/3-87/047. Research Triangle Isaksen, W. Kuhn, A. Lacis, F. Luther, J. Mahlman, 217 Chapter 11 R. Reck, and M. Schlesinger. 1987. Climate- Seilkop, S.K., and P.L. Finkelstein. 1987. Acid chemical interactions and effects of changing precipitation patterns and trends in eastern North atmospheric trace gases. Review Geophysics America, 1980-84. Journal of Climate and Applied 25:1441-1482. Meteorology 26(8):980-994. Research Triangle Institute. 1975. Investigation of U.S. EPA. 1988. U.S. Environmental Protection Rural Oxidant Levels As Related to Urban Agency, Office of Air Quality Planning and Hydrocarbon Control Strategies. EPA-450/3-75- Standards. National Air Quality and Emissions 036. Research Triangle Park, NC: U.S. Trends Report, 1986. Research Triangle Park, NC: Environmental Protection Agency. March. U.S. Environmental Protection Agency. EPA report No. 45014-88-001. Research Triangle Park, NC: Samson, P.J. 1988. Linkages Between Global U.S. Environmental Protection Agency. Climate Warming and Ambient Air Quality. Paper presented at the Global Climate Linkages Walcek C. 1988. In: Sensitivity of Urban/Regional Conference, Washington, DC; Nov. 15-16. Chemistry to Climate Change: Report of Workshop, Chapel Hill, NC. Wuebbles, D.J. and J.E. Penner, eds. Livermore, CA: Lawrence Livermore National Laboratory. Feb. 17-18. 218 CHAPTER 12 HUMAN HEALTH FINDINGS in the winters) have been observed in several areas in the United States. The longer and Global warming may lead to increases in human hotter summers that may accompany climate illness (morbidity) and mortality during summer. change could increase infant mortality rates, Populations at particular risk are the elderly and although changes in variability may be more very young (age 1 year and below), particularly important than average changes in those who are poor and/or homeless. These effects temperature. may be more pronounced in some regions than in others, with northern regions more vulnerable to Vector-borne diseases, such as those carried the effects of higher temperature episodes than by ticks, fleas, and mosquitoes, could increase southern regions. Milder winters may offset in certain regions and decrease in others. In increases in morbidity and mortality, although net addition, climate change may alter habitats. mortality may increase. Mortality in southern cities For example, some forests may become currently shows a lesser effect from heat waves, grasslands, thereby modifying the incidence of presumably because populations have acclimatized. vector-borne diseases. If northern populations show this same acclimatization, the impact of global warming on While uncertainties remain about the summer mortality rates may be substantially lower magnitude of other effects, climate change than estimated. The full scope of the impacts of could have the following impacts: climate change on human health remains uncertain and is a subject for future research. -- If some farmland is abandoned or some forests become grasslands, a result could be Although there may be an increase in weather- an increased amount of weeds growing on related summer deaths due to respiratory, cultivated land, and a potential increase in cardiovascular, and cerebrovascular diseases, the incidence of hay fever and asthma. there may be a decrease in weather-related winter deaths from the same diseases. In the -- If humidity increases, the incidence and United States, however, our studies suggest that severity of skin infections and infestations an increase in weather-related deaths in such as ringworm, candidiasis, and scabies summer would be greater than the decrease in may also rise. weather-related deaths in winter. To draw firm conclusions, however, this area needs additional -- Increases in the persistence and level of air study. pollution episodes associated with climate change may have adverse health effects. Sudden changes in temperature are correlated with increases in deaths. So if climate variability increases, morbidity and mortality CLIMATE-SENSITIVE ASPECTS may also increase. Conversely, a decrease in OF HUMAN HEALTH the frequency or intensity of climate extremes may be associated with a decrease in mortality and morbidity. Human illness and mortality are linked in many ways to the environment (Figure 12-1). Seasonal variation in perinatal mortality and Mortality rates, particularly for the aged and very ill, preterm birth (higher in the summers, lower 219 Chapter 12 H HEAT STRESS C U L M REPRODUCTIVE EFFECTS M A I 0 N M NUTRITION R A B T Food Production Organic Disease I D E FISHERIES I COMMUNICABLE DISEASES T AGRICULTURE Y Pollen Production ALLERGENS ALLERGIC DISEASES AND C FORESTS H M VECTORS VECTOR-BORNE DISEASES 0 A WETLANDS R N CHRONIC DISEASES T G Habitat A L E I S T AIR POLLUTION Y Figure 12-1. Schematic showing how climate change can affect human health. are influenced by the frequency and severity of summer when the mosquitoes that transmit it are extreme temperatures. The life cycles of disease- active. In addition, adverse effects on reproduction, carrying insects, such as mosquitoes and ticks, are such as increased incidence of premature births, affected by changes in temperature and rainfall, as show a summertime peak in some cities. Table 12- well as by modifications in habitat that result from 1 lists the number of deaths and the number of climate change. Air pollution, frequently associated physician visits (used to estimate the incidence of with climate change, is known to increase the illness associated with a given effect) associated with incidence or severity of respiratory diseases such as major causes of mortality and illness in the United emphysema and asthma. A variety of human States. illnesses show sensitivity to the changes in temperature (and/or humidity) that accompany General Mortality and Illness changes in season. Stroke and heart attacks increase with very cold or very warm weather. The relationship between mortality and Allergic diseases such as asthma and hay fever weather has been studied for over a century increase in spring and summer when pollens are (Kutschenreuter, 1959; Kalkstein, Volume G), with released. Diseases spread by insects such as St. the relationship between mortality and temperature Louis encephalitis increase in the warmth of receiving the most attention. Kutschenreuter (1959) observed "mortality is higher during cold winters 1 and hot summers and lower during warm winters St.Louis encephalitis is an example of a vector-borne disease. and cool summers." The people most sensitive to Such diseases are spread to humans or animals by arthropods (e.g., mosquitoes or ticks). The disease-causing organism, such temperature extremes are the elderly (White and as a virus, is carried and transmitted by the vector, also known Hertz-Picciotto, 1985). One explanation is the as the agent. Some vectors, such as ticks, live on other animals, increased susceptibility of the elderly is that for such as deer and birds, which are called intermediate hosts. individuals already stressed by the circulatory For example, Lyme disease is caused by a bacteria (the agent), which is carried by a certain type of tick (the vector), which problems associated with vascular and heart disease, lives on deer and mice (the intermediate hosts). heat waves (temperatures above 100°F for 5 220 Human Health Table 12-1. Major Causes of Illness and Mortality in the United States (1984)ᵃ Estimated number of Estimated mortality physician Cause of illness or mortality contacts Number Rate/100,000 Accidents and adverse effects 70,000,000 93,520 39.6 Cerebrovascular diseases b 9,100,000 154,680 65.5 Chronic liver disease and cirrhosis 1,400,000 26,690 11.3 Chronic obstructive pulmonary diseases and allied conditions 20,500,000 70,140 29.7 Congenital anomalies 4,300,000 12,900 5.5 Diabetes mellitus 35,600,000 35,900 5.2 Heart diseases 72,400,000 763,260 323.2 Malignant neoplasms 20,300,000 453,660 192.1 Pneumonia and influenza 14,500,000 58,800 24.9 Suicides, homicides -- 47,470 20.1 Total for potentially weather- sensitive diseases 152,100,000 1,082,780 448.5 Total all causes 248,100,000 1,717,020 717.1 a Causes are presented in alphabetical order and therefore are not ranked by severity. b Conditions that can be influenced by changes in weather and climate are indicated in bold type. Source: CDC (1986). consecutive days) "overload" the thermoregulatory total mortality from all causes, a growing body of system, which is struggling to maintain the literature evaluates the relationship of weather to appropriate body temperature. This results in heat specific causes of death. For example, changes in stress, heatstroke, and often mortality as well weather have been associated with impacts on the (White and Hertz-Picciotto, 1985). cardiovascular, cerebrovascular, and respiratory systems. As previously shown in Table 12-1, In addition to the elderly, people working in diseases of these three systems cause the majority of hot environments, such as steel mills and deaths observed on a yearly basis in the United construction sites, are at special risk from heat States, as well as significant illness. Incidences of waves (Dukes-Dobos, 1981). These workers face these diseases rise as climate extremes increase. even greater risk if they have underlying medical problems such as impaired circulation; higher than The relationships of weather variables to normal body temperature due to disease; chronic diseases of these systems are diverse and diseases such as alcoholism, diabetes, and obesity; complicated. Weather is not the main causative or other problems. factor in these diseases but, rather, changes in weather have an impact because they add stress to Cardiovascular, Cerebrovascular, and systems that have already been compromised for Respiratory Diseases some other reason(s). For example, although it has been observed that deaths in individuals with Although much of the earlier information diseases of the cardiovascular system go up with characterized the relationship between weather and heat waves, the precise reason for this relationship is not known. 221 Chapter 12 To understand the relationship between Mountain spotted fever and Lyme disease, induce weather and these diseases, one must examine the similar initial symptoms: high fever, chills, headache, specific diseases that come under broad categories backache, and profound fatigue. Rocky Mountain such as "cardiovascular disease." For instance, heart spotted fever can eventually result in hemorrhagic attack, coronary heart disease, and possibly coronary areas that ulcerate, and Lyme disease may cause arteriosclerosis and rheumatic heart disease are permanent neurologic, cardiac, and rheumatologic apparently sensitive to changes in temperature abnormalities (APHA, 1985). The ticks that spread (particularly cold and heat waves), whereas ischemic these diseases, and therefore the geographic heart disease is not (Vuori, 1987). distribution of the diseases themselves, are affected both directly and indirectly by climate variables. That these different relationships exist is not Such environmental factors as temperature, unexpected given that different parts of the system humidity, and vegetation directly affect tick are compromised (e.g., the arteries in populations and the hosts of the tick populations, arteriosclerosis and the heart muscle in rheumatic e.g., deer, mice, and birds. heart disease), and that different causes are also likely (e.g., an infection-related process in rheumatic Mosquito-borne diseases, such as malaria and heart disease and diet and heredity in certain types of encephalitis (inflammation of the arteriosclerosis). What this information does brain), are not a major health problem in the indicate, however, is that these relationships are United States today because occurrences are very complex and that unraveling them to predict relatively rare. However, mosquitoes are also the effects of global warming will require weather-sensitive insects favoring a warm, humid considerable analysis (Lopez and Salvaggio, 1983). climate. The spread of mosquito populations and the diseases they carry depends in part upon such The relationship between temperature changes climate factors as temperature and humidity, and and illness (morbidity) from diseases such as heart upon vegetation, which is also influenced by the attack and stroke is not as well defined as the climate. relationship reported for mortality. Mortality has national reporting procedures, whereas morbidity Human Reproduction must be estimated from such data as hospital admission figures. A few studies have evaluated the Preterm delivery and perinatal mortality (i.e., relationship of weather to hospital admissions from death just before, during, or just after birth) are two cardiovascular or cerebrovascular disease. These adverse reproductive outcomes that are associated have shown a relationship to weather changes, e.g., with particular seasons and, thus, might be affected an increase in admissions for cardiovascular effects by climate change. Statistically significant increases with heat waves, similar to that observed for in preterm births and in perinatal mortality in the mortality (Sotaniemi et al., 1970; Gill et al., 1988). summer months have been documented (Keller and Nugent, 1983; Copperstock and Wolfe, 1986) (see Morbidity from respiratory diseases is Figure 12-2). The data on total perinatal deaths somewhat easier to estimate, principally because correspond closely with those on perinatal deaths two such diseases, asthma and hay fever, affect as associated with infection in the mother or infant, much as 3 and 6% of the U.S. population, suggesting that the observed seasonality in perinatal respectively, causing significant losses of work time. death is linked to a seasonality of reproductive The most common seasonal pattern for the allergic infections (Keller and Nugent, 1983). type of asthma and for hay fever is an increased springtime occurrence in response to grass pollens. A nonseasonal form of allergic asthma may also POTENTIAL HUMAN HEALTH occur in response to allergens such as molds, which are affected by changes in precipitation and EFFECTS OF CLIMATE CHANGE temperature. To assess the effects of climate change on Vector-Borne Diseases human health, EPA sponsored three studies for this report (Table 12-2). Longstreth and Wiseman Two tick-borne diseases currently posing a (Volume G) reviewed the literature on the role of public health problem in the United States, Rocky climate, season, and weather variables in the 222 Human Health infectious diseases in the United States. Following A. the workshop, Haile (Volume G) conducted 5.1 modeling studies of the potential impact of climate 5.0 change on (1) the distribution of the American dog 4.9 tick, the vector of Rocky Mountain spotted fever; Probability of a Perinatal Death By Month Per 1,000 at Risk 4.8 and (2) the potential for malaria transmission in the 4.7 United States. The third study, by Kalkstein 4.6 (Volume G), as an extension of an earlier modeling 4.5 Deaths study that assessed the potential effects of global 4.4 climate change on the elderly and on total mortality 4.3 in New York (Kalkstein et al., 1986). Kalkstein J F M A M J J A S O N D (Volume G) expanded the New York analysis to Month include 14 other cities. A detailed review of these three studies, supplemented with other information from the literature, is presented in this section. B. 60 General Mortality 59 Preterm Probability of a Preterm Birth by Births Month Per 1,000 at Risk Preliminary analyses suggest that unless the 58 U.S. population becomes fully acclimatized² to higher temperatures, climate change will be 57 associated with a sharply rising number of summer deaths. With full acclimatization to the warmer 56 summers, heat-related mortality might increase less dramatically or not at all. In winter, the number of J F M A M J J A S O N D weather-related deaths will probably decline Month regardless of acclimatization. It is not clear what the net effect of these two offsetting trends may be. Figure 12-2. Probabilities of (A) perinatal death or (B) preterm delivery (Keller and Nugent, 1983). Only a few studies have evaluated the effects of global climate change on human mortality. Kalkstein et al. (1986) developed a regression Table 12-2. Studies Conducted for This Report equation involving nine weather elements, such as temperature, windspeed, and humidity, to give the best algorithm for describing the current impact of The Impact of CO₂ and Trace Gas-Induced weather on mortality. The algorithm used mortality Climate Change Upon Human Mortality data from New York City for 1964-66, 1972-78, and Kalkstein, University of Delaware (Volume G) 1980. Computer Simulation of the Effects of Changes The analysis revealed the existence of a in Weather Pattern on Vector-Borne Disease summertime "threshold temperature" -- the Transmission - Haile, U.S. Department of maximum temperature above which mortality Agriculture (Volume G) increases -- for New York City of 92°F for total deaths. This information was then used to assess The Potential Impact of Climate Change on the potential impact of climate change under the Patterns of Infectious Disease in the United assumption that the population would not States - Longstreth and Wiseman, ICF/Clement Associates, Inc. (Volume G) 2 Estimations of the impact of warming on future mortality must address the question of whether humans will acclimatize incidence of, and mortality due to, vector-borne (socially, psychologically, or physiologically adapt) to changing weather. How quickly humans may become acclimatized is a diseases. In November 1987, they also conducted a topic of considerable controversy, so it is difficult to predict workshop of scientists to evaluate the potential whether the climate changes due to global warming will occur impacts of global climate change on vector-borne slowly enough to permit acclimatization. 223 Chapter 12 acclimatize, as well as under the assumption that it For this report, Kalkstein (Volume G) would acclimatize. Unacclimatized impacts were extended the New York analysis to cover 14 estimated by combining the climate scenarios and additional metropolitan areas and to evaluate the the historical weather algorithm described above, impact of two climate scenarios: the GISS doubled and acclimatized impacts were estimated by CO₂ scenario, and the GISS transient A scenario, analyzing analog cities that have values of weather evaluated at 1994 to 2010 and at 2024 to 2040. variables today that look like those New York is Threshold temperatures were calculated for each estimated to have under climate change. city for summer and winter. Historical relationships between mortality and temperature were derived Assuming full acclimatization and a scenario independently for each of these 15 cities for both predicting that New York will be 3 to 4°C (5 to 7°F) summer and winter. Table 12-3 summarizes the warmer than it is today, no additional deaths were results for total mortality, by city and by season predicted. However, assuming no acclimatization, (summer or winter), for the doubled CO, scenario the number of summertime deaths attributable to with and without acclimatization. The cities temperatures above the threshold (hereafter called with the highest estimated number of suprathreshold summer deaths) increased seven- to suprathreshold summer deaths historically were tenfold. Changes in winter weather, i.e., more New York City, Chicago, and Philadelphia; each subthreshold temperatures, were not estimated to averaged over 100. All of the cities with the highest affect mortality. average number of summer deaths are in the Table 12-3. Estimated Future Mortality Under Doubled CO2 Climate Conditions without and with Acclimatization Number of deaths per season Summer Winter City Current Without With Current Without With Atlanta 18 159 0 2 2 0 Chicago 173 412 835 46 2 96 Cincinnati 42 226 116 14 6 0 Dallas 19 309 179 16 1 0 Detroit 118 592 0 16 2 37 Kansas City 31 60 138 21 5 0 Los Angeles 84 1,654 0 0 0 0 Memphis 20 177 0 0 0 0 Minneapolis 46 142 235 5 1 0 New Orleans 0 0 0 0 0 0 New York 320 1,743 23 56 18 25 Oklahoma City 0 0 47 0 0 0 Philadelphia 145 938 466 10 1 1 St. Louis 113 744 0 47 7 0 San Francisco 27 246 159 10 7 0 Total 1,156 7,402 2,198 243 52 159 Source: Kalkstein (Volume G). 224 Human Health Midwest or Northeast, and those with the lowest The direction of predicted change, i.e., an increase, number are in the South. is probably much more solid than the magnitude of change. In addition, this research has concentrated As would be expected, generally more deaths on mortality occurring above a particular threshold were predicted for populations that do not temperature for summer or below a particular acclimatize. However, for certain cities, e.g., threshold temperature for winter. Consideration of Chicago, Kansas City, and Minneapolis, more deaths a broader range of temperatures could conceivably were predicted with acclimatization than without. result in different conclusions being drawn. Exactly why this occurred is uncertain. The results appear to be very sensitive to the choice of the Cardiovascular, Cerebrovascular, and analog city. For example, Chicago appears to have Respiratory Diseases more deaths if its population becomes acclimatized than if it does not. It may be that the analog city Overall global warming and climate change chosen to represent a particular acclimatized city, may exacerbate the effects of cardiovascular, Chicago for instance, is more sensitive to weather cerebrovascular, and respiratory diseases. Data effects on mortality than Chicago currently is. More from these studies show an inverse relationship research is planned to investigate this apparent between mortality and temperature (i.e., deaths go anomaly to refine the estimates of what global down as temperature goes up) for the range warming will mean in terms of mortality. Thus, between -5°C and about +25°C, with sharp Kalkstein's results should not be used as predictions increases at temperatures above and below this of individual city behavior, but as illustrations of range, particularly for the elderly and for hot sensitivities. weather (White and Hertz-Picciotto, 1985); the exact range appears to depend on the city. In the absence of any acclimatization, Illustrations of this relationship for coronary heart suprathreshold summer mortality in the United disease and stroke are shown in Figures 12-3 and Stated under conditions of doubled CO2 is 12-4, respectively (Rogot and Padgett, 1976). This estimated to rise from an estimated current total of complex relationship precludes simple prediction of 1,156 deaths to 7,402 deaths, with deaths in the the net effect of climate change. For example, it is elderly (aged 65 or over) subset contributing about possible that hot weather-associated mortality from 60% of each figure (727 and 4,605, respectively). these diseases may increase in some localities, but Currently, the percentage of elderly in the U.S. this trend may be offset, at least in part, by a population is increasing. Thus, the mortality decrease in cold weather-associated mortality. estimated to result from climate change may be larger than that found by Kalkstein because his Just as higher summer temperatures are analysis is predicated on today's age distribution. associated with increases in mortality from Even with full acclimatization, the number of cardiovascular, cerebrovascular, and respiratory weather-associated summer deaths almost doubles diseases, they are also likely to be associated with to 2,198, possibly because hot weather increases increases in morbidity from these diseases through physiological stress. Kalkstein's analysis also increases in the number or duration of hospital estimates a drop in the number of subthreshold admissions. Particular stress may be put on the winter deaths. Historically, however, the number of respiratory system because climate change can these deaths during the winter in the United States potentially increase pollen, urban smog (discussed is much smaller (243) than that observed for the below), and heat stress, all of which have an adverse summer, and subthreshold winter deaths were effect on the respiratory system. estimated to fall to 52 without acclimatization and to 159 with acclimatization. The net result for the For example, if, as has been suggested in the United States is an increase in yearly mortality chapter on forests, climate change encourages a associated with doubled CO2. transition from forest to grassland in some areas, grass pollens could increase. This, in turn, may This study is exploratory research in the field increase cases of pollen-induced hay fever and of the potential impacts of climate change on allergic asthma. (However, the switch from forest human health. Some aspects of the analyses that to grassland would reduce the amount of tree led to these estimates need further investigation; pollens that also cause allergic responses in some thus, the estimates should be accepted with caution. 225 Chapter 12 200 50 8 NEW YORK 9 NEW YORK 40 100 NEW ORLEANS 90 30 a LOS ANGELES 80 70 CHICAGO 60 LOS ANGELES 50 20 40 DETROIT PHILADELPHIA AVERAGE DAILY NUMBERS OF DEATHS FROM CHD (LOG SCALE) 30 SAN FRANCISCO 20 AVERAGE DAILY NUMBERS OF DEATHS FROM STROKE (LOG SCALE) CHICAGO 9 ST. LOUIS PHILADELPHIA ST. LOUIS 10 DETROIT MINNEAPOLIS 9 10 9 8 SAN FRANCISCO 8 DALLAS 7 a 7 6 ATLANTA 6 MINNEAPOLIS 5 4 MEMPHIS 5 3 MEMPHIS 4 ATLANTA 2 3 DALLAS NEW ORLEANS 1 2 -20 0 20 40 60 80 100 F -20 0 20 40 60 80 100 F -29 -18 -7 4 16 27 38 C -29 -18 -7 4 16 27 38 C AVERAGE TEMPERATURE ON DAY OF DEATH AVERAGE TEMPERATURE ON DAY OF DEATH Figure 12-3. Relationship of temperature to heart Figure 12-4. Relationship of temperature to disease mortality (adapted from Rogot and Padgett, mortality from stroke (adapted from Rogot and 1976). Padgett, 1976). individuals.) Rises in humidity also may affect the ozone, adverse consequences could result for adult incidence of mold-induced asthma and hay fever. asthmatics and people who suffer from acute or chronic bronchitis. As indicated in Chapter 11: Air Quality, global warming may modify global and regional air Vector-Borne Diseases pollution because it may increase concentrations of ozone and may also have impacts on acid deposition Potential changes in humidity and temperature and general oxidant formation. The increasing could alter the geographic ranges and life cycles of occurrence of numerous respiratory diseases, such plants, animals, insects, bacteria, and viruses. (For as lung cancer, emphysema, bronchitis, and asthma, further discussion of forestry and agriculture, see has been attributed to the pollutants in urban smog Chapters 5 and 6, respectively.) For example, the (Lopez and Salvaggio, 1983). Many of the trace range of many plant pests may move northward by gases implicated in global warming contribute to several hundred miles. Such changes could occur these problems; other pollutants are created from for insects that spread diseases to both humans and the interaction of ultraviolet light with these and animals. Vector-borne diseases that affect humans other chemicals present in the atmosphere. are relatively rare in the United States. The incidence of most of those found, however, is The component that causes the greatest increasing. The incidence of some, such as Lyme concern in urban smog is ozone (Grant, 1988). If disease, is increasing dramatically (CDC, 1986). global warming causes an increase in tropospheric 226 Human Health Tick-Borne Diseases CO2 scenarios (GISS, GFDL, and OSU) to estimate population dynamics, growth rate, and generation Both Rocky Mountain spotted fever and Lyme time. Haile assumed that habitats and host density disease are considered to be public health problems did not change in response to global warming. in the United States. Although these two diseases Sample results for six cities representing the most are spread by different species of ticks, some southern, the most northern, and the two middle overlap exists in their geographic distribution latitudes are presented in Figure 12-6. The results (Figure 12-5). Because tick populations appear to indicate that under all scenarios, tick populations be limited by the size of their intermediate host would shift from south to north and would be populations (such as white-tailed deer), the spread virtually eliminated from the most southern of tick-borne diseases may be particularly sensitive locations (Jacksonville and San Antonio). However, to any change that may affect the geographic range in the middle latitude cities, the results are mixed of these hosts and, consequently, the range of the and depend on the scenario evaluated. The model vector, or carrier. does not estimate changes in incidence of the disease. In addition to the presence of the host, tick populations also depend upon the seasonality of In this analysis, the only model inputs that environmental factors such as temperature, were changed to simulate climate change were the humidity, and vegetation. Optimally, climate must weather inputs. Other important parameters in the be warm enough to promote progression through model are the distribution of habitat between forests the life cycles, humid enough to prevent the drying and meadows and the presence of suitable hosts. out of eggs, and cold enough in winter to initiate the Both parameters are likely to be changed relative to resting stage. current conditions under climate change. As indicated in Chapter 5: Forests, a change from As for many tick-borne diseases, the forests to meadows may occur in certain areas of opportunity for a tick to acquire the infective agent the country; this would depress the tick population. from an infected animal is limited to the short However, the distribution of small mammals also period when the level of the agent in the blood of may change. If small mammal populations the host is high enough for the tick to receive an increased, tick populations would grow. In addition, infective dose. Higher temperatures may increase this study did not consider changes in climate the amount of the agent (the organism that is variability, which may have a major effect on the transmitted by the carrier, such as a virus) and the outbreak of diseases. time it remains lodged on the host animal. Both these mechanisms would increase the rate of In a sensitivity analysis of their model, Mount infection of the carrier. However, although higher and Haile (1988) found that the model predictions temperatures may favor the presence of the agent, could vary sixteenfold, depending on the inputs used there is some indication that they could disrupt the for host density, whereas the variability conferred by life cycle of some tick species. In these cases, changes in the weather inputs is about fourfold. warmer temperatures would reduce both tick Based on the sensitivity analysis, host densities are survival and the spread of diseases they carry. extremely important to these predictions. Keeping them constant, as was done in this analysis, could Tick populations also vary with the natural have underestimated or overestimated the impact of vegetation of an area. The incidence of Rocky climate change on the density of the American dog Mountain spotted fever, in particular, has been tick. linked to natural vegetation and changes in climate. Mosquito-Borne Diseases In examining the potential impact of climate change in the United States on Rocky Mountain A second category of vector-borne diseases spotted fever, Haile (Volume G) used a weather- that can be affected by climate change consists of based model, ATSIM, to evaluate the impact of the diseases carried by mosquitoes. Climate changes scenario climate changes on the distribution of the resulting in more days between 16 and 35°C (61 to American dog tick, the primary carrier of this 95°F), with humidity between 25 and 60%, are likely disease (Haile, Volume G; Mount and Haile, 1988). to favor the growth of mosquitoes (White and The model uses data inputs from the three doubled Hertz-Picciotto, 1985). Mosquito populations are 227 Chapter 12 LYME DISEASE ROCKY MOUNTAIN SPOTTED FEVER No. of Cases 0 - 10 10 25 25 100 States With Highest 100 226 Incidence (Cases Per 100K) NC 3.6, SC 3.2, OK 2.3 Figure 12-5. Geographic distribution of Lyme disease and Rocky Mountain spotted fever (Longstreth and Wiseman, Volume G). 228 Human Health OSU GISS GFDL Richmond, VA BASE Columbus, OH Jacksonville, FL San Antonio, TX Halfax, N.S. Missoula, MT 0 5 10 15 20 25 30 35 DENSITY (ADULT TICKS ON HOSTS/HECTARE) Figure 12-6. Simulated tick densities for selected cities under various scenarios of climate change (Haile, Volume G). also sensitive to the presence of standing water. It At a recent workshop, five of the numerous is not clear whether standing water will generally mosquito-borne diseases were considered to pose a increase or decrease (see Chapter 9: Water potential risk to U.S. populations if the status quo Resources). is disturbed by climate change (Longstreth and Wiseman, Volume G). Malaria, dengue fever, and Worldwide, mosquito-borne diseases are arbovirus-induced encephalitides were considered to associated with significant illness and mortality. In be significant risks, and yellow fever and Rift Valley the United States, however, vector control programs fever were considered to be possible risks. and improved hygiene have virtually eliminated endogenously transmitted cases of these diseases, Malaria with the exception of sporadic outbreaks of arbovirus-encephalitis. (Imported cases are seen Malaria is an infectious disease transmitted by occasionally.) Numerous mosquito species are mosquitoes and induced by parasites (Plasmodia). present in the United States, however. Recent The symptoms are highly variable, depending on the restrictions on pesticide use, coupled with the influx species of the agent. They include chills, sweats, of visitors and immigrants who can serve as sources and headache, and in severe cases, may progress to of infectious agents, as well as the lack of available liver damage and even liver and renal failure. vaccines for many of the potential diseases, suggest the potential for reintroduction and establishment of As a result of effective vector control and these diseases in the United States -- particularly if treatment programs, malaria is no longer indigenous global warming provides a more suitable climate for to the United States. However, imported cases their growth and development (Longstreth and occur regularly, and occasionally indigenous Wiseman, Volume G). transmission has been documented (Longstreth 229 Chapter 12 and Wiseman, Volume G). Current U.S. South (e.g., Miami, Key West, and Orlando), under demographic trends, including a large number of current climate conditions, are very favorable for legal and illegal immigrants from locations where malaria transmission. 3 Using the climate change malaria is endemic, could present a pool of infected scenarios in MALSIM did little to affect the individuals that, in conjunction with climate changes, estimated transmission potential of malaria in the may create sufficient conditions for increased United States (Figure 12-7). In a few cities, e.g., disease incidence. Richmond, Nashville, and Atlanta, the model estimated large increases in one scenario relative to Haile used the weather-dependent model those that would occur normally. However, the MALSIM to evaluate the potential impact of results varied with different climate scenarios, did climate change on malaria in an infected population not occur at all locations, and should be considered living in an area where a competent carrier is to be inconclusive. present. The model was originally developed to help predict malaria outbreaks in tropical countries such as Kenya. This is the first application of the model to the United States. This analysis did not ³The MALSIM estimates of malaria incidence by city under consider changes in climate variability, which may current conditions were based on two assumptions: that there were 100,000 female mosquitoes in the vicinity of each city and be important for the spread of malaria. The that 100 infected people were added to the cities' populations. MALSIM model showed that several cities in the Under those assumptions, infection of virtually the entire population of Miami was predicted to be possible unless protective measures were taken. Miami, FL Key West, FL Orlando, FL Jacksonville, FL San Antonio, TX Atlanta, GA Nashville, TN Tulsa, OK Dallas, TX Richmond, VA XXXXX OSU Baltimore, MD GISS GFDL BASE Indianapolis, IN Boston, MA 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 INCIDENCE (CASES/10,000 POPULATION) Figure 12-7. Simulated incidence of malaria for selected cities under various scenarios of climate change (Haile, Volume G). 230 Human Health Dengue Fever Outbreaks of encephalitis attributable to these viruses are normally limited to specific geographic Dengue fever is an arbovirus-induced⁴ illness locations and seasons for several reasons. First, characterized by fever, rash, and severe pain in the warm temperatures are normally required for the joints. The dengue virus has four different types viruses to multiply and to be transmitted to a new (DEN 1 through DEN 4). Sequential infection by host. Higher temperatures may quicken the different types is possible and has been suggested transmission process and promote epidemic disease. to lead to an increased risk of developing a more However, the extent of this effect depends largely severe, hemorrhagic form of the disease that can be on the particular virus. Some viruses require cooler fatal in the very young and the elderly. Like weather and higher moisture conditions. Thus, malaria, it is not currently endemic in the United higher temperatures may reduce their prevalence. States, although potential carriers are present and Second, environmental conditions that favor the the disease is imported here regularly by people presence of carriers and hosts must prevail. For who have traveled abroad. example, relative humidity may affect plant life necessary for the feeding of hosts. The ability of the vector to transmit the agent appears to depend on temperature, and current Other Diseases conditions do not appear to be favorable for this process. Climate changes that raise temperatures, The incidence of a variety of other U.S. however, may reduce the required incubation period diseases appears to be sensitive to changes in and increase the infectivity of the carrier, increasing weather. If humidity is higher, an increased the potential transmission of the disease. incidence and severity of fungal skin diseases (such as ringworm and athlete's foot) and yeast infections Arbovirus-Related Encephalitides (such as candidiasis) may be observed. Studies on soldiers stationed in Vietnam during the war Arbovirus-related encephalitides are a group indicated that outpatient visits for skin diseases (the of acute inflammatory diseases that involve parts of largest single cause of outpatient visits) were the brain, spinal cord, and meninges. In mild cases, directly correlated to increases in humidity but these infections result in feverish headaches or showed a 4-month lag with relationship to aseptic meningitis; in more severe cases, those temperature increases (Figure 12-8). In addition, symptoms can be accompanied by stupor, coma, excessively high temperatures can lead to such skin convulsions (in infants), and occasionally spastic diseases as prickly heat and heat rash, which impair paralysis (APHA, 1985). the ability of the skin to breathe and thus place additional stress on people already suffering from At least seven types of viruses causing overexposure to heat from other causes. encephalitis are present in the United States. These include the three forms that also infect horses (the Several diseases appear to be associated with western, eastern, and Venezuelan equine the acquisition of winter infections. If a reduction encephalitis viruses) as well as four that are named in winter severity is also accompanied by a decrease after the location of their discovery (the La Cross, in wintertime infections, these diseases could be St. Louis, Powassan, and California encephalitis reduced under global warming. viruses). Cases range in severity depending on the type of virus, with yearly fatality rates between 0.3 For example, birth in cold winter months has and 60%. These infections are rare. In 1984, 129 been associated with a higher risk of schizophrenia cases were reported to the Centers for Disease in individuals whose schizophrenia is without an Control, which maintains an active surveillance apparent genetic component (Kovelman and program for them (CDC, 1986). Scheibel, 1983). In addition, juvenile-onset diabetes, which has been reported to be increasing over the past several decades, has been shown to be associated with a seasonal variation in that the ⁴An arbovirus is a virus transmitted by an arthropod. month of first admission peaks in the winter Arthropods are a group of animals that includes insects and arachnids. Examples of arthropods that transmit disease (Glatthaar et al., 1988; Patterson et al., 1988). It is include mosquitoes and ticks. a common clinical experience that a minor viral illness precedes the onset of symptoms. 231 Chapter 12 1200 1100 OUTPATIENT VISITS 1000 900 800 OUTPATIENT VISITS/1000/YEAR 700 RELATIVE HUMIDITY 29 88 MEAN 600 TEMPERATURE 28 86 A 500 27 84 400 26 82 300 25 80 200 24 78 100 23 76 0 22 74 JAN JUN JAN JUN JAN JUN JAN JUN 1967 1968 1969 1970 MONTH AND YEAR TEMPERATURE (°C) RELATIVE HUMIDITY Figure 12-8. Relationship of skin infections to humidity and temperature (Longstreth and Wiseman, Volume G). SOCIAL AND ECONOMIC Climate change may affect regional and IMPLICATIONS national health care. For instance, the treatment requirements for asthma may increase or decrease as locations experience changes in the distribution Demographic and technological trends (the and intensity of pollen concentrations. Increased aging of the population, an influx of immigrants, resources may be needed to treat premature infants advances in treatment techniques) make it difficult if the number of preterm births increases. If heart to analyze the potential impacts of climate change attacks, stroke, and respiratory problems increase, on human health. Although this chapter attempts hospitalization costs and costs due to days lost from to identify those human health effects at risk from work may also increase. Higher health care costs climate change, the analyses were not designed to might be particularly obvious in Medicaid and consider adaptive responses and should not be Medicare because those below the poverty line treated as predictions of what will happen with would be less able to take adaptive measures (e.g., climate change but as illustrations of sensitivities. air-conditioning), and the elderly are more Rather, the analyses presented here represent susceptible to the ill effects of extreme weather possible scenarios, in the absence of consideration conditions. of demographic trends or adaptive responses, that may either exacerbate or ameliorate the impact of The United States is already experiencing an climate change on human health. Societies possess infant mortality higher than that of any other considerable ability to adapt to change. The industrialized nation (World Bank, 1987). Some potential for climate to affect human health may be studies have found that perinatal mortality is higher considerably modified by adaptive responses, such in the summer; consequently, the increased as immunizations, modification of the environmental temperatures expected with global warming may temperature (e.g., use of air conditioners), and well exacerbate infant mortality (or at least neonatal control of disease carriers. mortality). 232 Human Health The need for irrigation may increase in many In the future, a cadre of trained professionals regions of the United States (see Chapter 6: may be needed to respond to outbreaks of diseases. Agriculture). Irrigation may result in greater A shift in the distribution of carriers of human amounts of standing water and can therefore disease may necessitate regional shifts in increase mosquito populations. Arbovirus surveillance and eradication programs. States that encephalitis may become a greater problem than at do not have these programs may need to develop present, and other mosquito-borne diseases, such them. as dengue or yellow fever, could be more easily spread if introduced. RESEARCH NEEDS One health impact of climate change not assessed in this report is the morbidity and mortality Although information evaluating the associated with certain kinds of extreme events, e.g., relationship of weather and season to various health tornadoes and hurricanes. These currently cause effects is plentiful, research into the significance of some mortality in the United States; however, it is these relationships in the context of global warming difficult to say whether there will be a change in the is scarce. A number of areas requiring further mortality induced by these events with global research are described below. warming. As indicated in Chapter 3: Variability, changes in the frequency of such extreme events A number of studies have identified cannot be predicted on the basis of an analysis of relationships between temperature changes and the general circulation model (GCM) output, mortality from diseases of the heart, respiratory although an increase in severity of some kinds of system, and cerebrovascular system. These studies storms, e.g., hurricanes, is not inconsistent with show slightly different relationships depending on current theories and more detailed models of storm the city that provided the data, although some behavior. common elements exist. A statistical analysis of this information might be warranted to determine if one The impact of global change on human health general relationship (across the United States, or will most likely be greater in the lesser-developed perhaps related to latitude) could be developed for countries (LDCs) that do not have the resources to each of these categories. Such a relationship could take the adaptive or preventative measures available then be used to estimate the impact of global to the United States. Impacts on agriculture and warming by specific disease category. water resources in the LDCs could result in poor nutrition and water shortages that may make A companion study to that proposed above populations more susceptible to disease. Changes should identify the top 10 causes of deaths in insect (arthropod) habitats may allow diseases to associated with changes in weather in the Kalkstein flourish where they never have before. Changes in study. The results could then be compared with the extreme events such as monsoons or floods could information derived above to determine other significantly affect mortality in the developing world. causes of mortality that show great sensitivity to the Such external impacts on health might have an weather. impact on the United States not only via the potential for introduction of diseases already The Kalkstein analysis did not look at deaths discussed but also via our participation in occurring in the very young (aged 1 year and international aid and relief programs. below). Given the seasonality of perinatal mortality and preterm death observed in several studies, an investigation of the relationship between POLICY IMPLICATIONS temperature and mortality in the very young probably would be worthwhile. More baseline The full impacts of climate change on human information is needed for the latter study. Related health will require more research. Agencies such as studies on perinatal mortality could examine the the Department of Health and Human Services following: should consider conducting studies on potential impact. 233 Chapter 12 Whether the South has a higher per capita changes associated with global warming incidence of perinatal mortality. may affect famine development. Similarly, the Department of Defense is Whether infections, which have been using a number of models comparable to suggested as a potential cause of the those used by Haile to attempt to predict perinatal mortality observed, show a where infectious diseases are likely to seasonality in parallel to perinatal pose problems for U.S. troops. It might mortality, and whether more such be interesting to evaluate how the climate infections occur in the South. variables from the GCM-generated scenarios would affect these predictions, The principal causes (e.g., bacteria, particularly in the LDCs where these viruses) for perinatal infections, and diseases present a very real problem to whether they are the same as those for the health care systems. skin infections that increase with increases in humidity. Introduction of infectious diseases into the United States via immigrants. Anecdotal Whether the incidence of preterm birth or information indicates that many perinatal mortality is related to weather immigrants are not served by the health parameters such as temperature, humidity, care system; consequently, they could or high-pressure systems. become a population where diseases might develop into full-blown epidemics The following additional research areas are before initiation of treatment. suggested: Determining whether or not global warming will affect this process, either Synergism between stratospheric ozone directly via the provision of a more depletion (due to increases in UV-B hospitable environment for the disease or radiation) and global warming. Increased indirectly via an increased number of UV-B radiation and global warming might immigrants and refugees, will require a be expected to exacerbate infectious better characterization of the current diseases. UV-B radiation may have an situation. impact on the ability of an individual to respond to a disease, and global warming Intermediate hosts and their habitats. In may change the incidence of certain the models used by Haile, two important infectious diseases. For example, input parameters that were held constant leishmaniasis is an important disease in were the size of the intermediate host many African countries. In animal population and the distribution of habitat models, UV-B irradiation adversely affects between forest and meadow. It is likely the development of immunity to that both of these parameters will Leishmania. If climate change creates themselves be affected by climate change. more favorable habitats for the sand-fly A better grasp of how climate change will vector of this disease, then a double insult affect these parameters needs to be to the system could occur: a higher developed and integrated into the incidence, and a worse prognosis. infectious disease models. The impacts on LDCs. The Agency for Irrigation and incidence of vector-borne International Development is supporting disease. An increase in irrigation is the development of a Famine Early possible, which could have a significant Warning System (FEWS) that will use a impact on mosquito development and variety of inputs (many of them weather therefore on mosquito-borne diseases. related) to help predict when conditions The importance of such water is time- leading to famine may be occurring. dependent, however (i.e., it must occur at Appropriate GCM outputs could be input the right moment in the insect's life- into this system to evaluate how the cycle). Thus an analysis of how the growing season overlaps transmission of 234 Human Health diseases such as La Cross encephalitis Grant, L.D. 1988. Health effects issues associated might provide an indication of whether with regional and global air pollution problems. changes in irrigation practices should be a Prepared for World Conference on the Changing concern in terms of public health. Atmosphere, Toronto. Draft document. Mortality from extreme events. Another Harris, R.E., F.E. Harrell, K.D. Patil, and R. Al- issue that might warrant investigation is Rashid. 1987. The seasonal risk of pediatric/juvenile how climate change may affect the acute lymphocytic leukemia in the Unites States. mortality associated with extreme events, Journal of Chronic Diseases 40:915-923. such as hurricanes and floods. Kalkstein, L.S., R.E. Davis, J.A. Skindlov, and K.M. Air pollution and respiratory disease. Air Valimont. 1986. The impact of human-induced pollution is already a major contributing climate warming upon human mortality: a New factor in the incidence and severity of York case study. Proceedings of the International respiratory disease in the United States. Conference on Health and Environmental Effects of An analysis of the extent that global Ozone Modification and Climate Change, warming will exacerbate air pollution is Washington, DC; June. critical to an assessment of the potential health effects of climate change. Kalkstein, L.S., and K.M. Valimont. 1987. Effect on human health. In: Tirpak, D., ed. Potential Effects of Future Climate Changes on Forest and REFERENCES Vegetation, Agriculture, Water Resources, and Human Health, Vol. V, pp. 122-152. Washington, APHA. 1985. American Public Health Association. DC: U.S. Environmental Protection Agency. EPA In: Berenson, A.S., ed. Control of Communicable 400/1-87/101E. Diseases in Man. Springfield, VA: John D. Lucas Printing Company. Keller, C.A., and R.P. Nugent. 1983. Seasonal patterns in perinatal mortality and preterm delivery. CDC. 1986. Centers for Disease Control. Annual American Journal of Epidemiology 118:689-98. Summary 1984. MMWR 33:54. Kovelman, J., and A. Scheibel. 1986. Biological Cooperstock, M., and R.A. Wolfe. 1986. Seasonality substrates of schizophrenia. Acta Neurologica of preterm birth in the collaborative perinatal Scandinavica 73:1-32. project: demographic factors. American Journal of Epidemiology 124:234-41. Kutschenreuter, P.H. 1959. A study of the effect of weather on mortality. New York Academy of Dukes-Dobos, F., 1981. Hazards of heat exposure. Sciences 22:126-138. Scandinavian Journal of Work and Environmental Health 73-83. Lopez, M., and J.E. Salvaggio. 1983. Climate- weather-air pollution. In: Middleton, E., and C.E. Gill, J.S., P. Davies, S.K. Gill, and D.G. Beevers. Reed, eds. Allergy, Chapter 54. St. Louis, MO: 1988. Wind-chill and the seasonal variation of C.V. Mosby Company. cerebrovascular disease. Journal of Clinical Epidem- iology 41:225-230. Mount, G.A., and D.G. Haile. 1988. Computer simulation of population dynamics of the American Glatthaar, C., P. Whittall, T. Welborn, M. Giboon, dog tick, Dermacentor variabilis (Acari: ixodidae). B. Brooks, M.M. Ryan, and G. Byrne. 1988. Journal of Medical Entomology. In press. Diabetes in Western Australian children: descriptive epidemiology. Medical Journal of Patterson, C., P. Smith, J. Webb, M. Heasman, and Australia 148:117-123. J. Mann. 1988. Geographical variation in the incidence of diabetes mellitus in Scottish children Glenzen, W.P. 1982. Serious illness and mortality during the period 1977-1983. Diabetic Medicine 5:160-165. associated with influenza epidemics. Epidemiological Reviews 4:25-44. 235 Chapter 12 Rogot, E., and S.J. Padgett. 1976. Associations of White, M.R., and I. Hertz-Picciotto. 1985. Human coronary and stroke mortality with temperature and health: analysis of climate related to health. In: snowfall in selected areas of the United States 1962- White, M.R., ed. Characterization of Information 1966. American Journal of Epidemiology 103:565- Requirements for Studies of CO2 Effects: Water 575. Resources, Agriculture, Fisheries, Forests, and Human Health. Washington, DC: Department of Sontaniemi, E., U. Vuopala, E. Huhta, and J. Energy. DOE/ER/0236. Takkunem. 1970. Effect of temperature on hospital admissions for myocardial infarction in a subarctic World Bank. 1987. World Development Report area. British Medical Journal 4:150-1. 1987. New York: Oxford University Press. Vuori, I. 1987. The heart and the cold. Annals of Clinical Research 19:156-162. 236 CHAPTER 13 URBAN INFRASTRUCTURE FINDINGS -- As sea level rises, some coastal cities would require levees to hold back the sea or fill to raise the land surface area. In the Global climate change could require U.S. cities to make major changes in capital investments and case of Miami, the cost of these activities operating budgets. Areas most likely to be affected might exceed $500 million over the next include water supplies, roads, and bridges; storm 50 to 75 years, necessitating an average sewers and flood control levees; and energy demand increase of 1 to 2% in annual capital in municipal buildings and schools. spending in Greater Miami. Most urban infrastructure in the United States Water Supply and Demand will turn over in the next 35 to 50 years. If potential changes in climate are considered, Climate change will influence the supply and this turnover will allow cities to prepare for demand for water in many cities. A lengthened climate change at lower costs. In some cases, summer season and higher temperatures would the risk of climate change should be increase the use of water for air conditioners, incorporated into decisions beginning today, lawns, and gardens. Changes in rainfall such as coastal drainage systems that are likely patterns, runoff, and flood control measures to last for 50 to 100 years. may alter water supplies. In the Hudson River Basin, summer water demand could increase by Northern and Southern Cities 5% over the demand for water without climate change, while supplies might fall. Such a Northern cities, such as Cleveland, may incur change would require new institutional and a change in the mix of their expenditures. In management approaches for both the Delaware such locations, increased electricity costs for and Hudson Rivers. air-conditioning could be offset by reductions in expenditures for heating fuel, snow and ice Policy Implications control, and road maintenance. Southern cities could see increases in operating budgets Climate change has implications for many due to the demand for additional air- national programs and policies, including the conditioning. following: Coastal Cities -- The National Flood Insurance Program may react to climate change by redrawing Coastal cities, including 12 of the 20 largest floodplain maps and adjusting insurance metropolitan areas, may face somewhat larger rates to account for sea level rise and impacts, such as the following: changes in riverflows. This program might consider discouraging development that -- Sea level rise or more frequent droughts would be vulnerable to sea level rise. would increase the salinity of shallow coastal aquifers and tidal surface waters. -- Because of the key role federal programs Cities that rely on water from these sources play in the development of cities, the should examine water supply options. Such Department of Housing and Urban areas as Dade County, Florida, or New Development should examine the York City would probably be vulnerable. implications of climate change on long-term policies. A minimum response might be to 237 Chapter 13 provide guidance on the certainties and of $45 billion to the capital stock (National Council uncertainties of climate change to groups on Public Works Improvement, 1988). such as the National League of Cities, the U.S. Conference of Mayors, and the Of the 20 most populated U.S. urban areas, 18 American Planning Association. have access to oceans, major lakes, or rivers and have invested in infrastructure for port facilities and -- Because water supply infrastructure may last flood control. 1 The expenditure required to for several centuries, improved planning is construct coastal defense structures which prevent important. The U.S. Geological Survey inundation by the sea, slow oceanfront erosion, should study the probable impacts of global control storm surges, slow saltwater advance up climate change and sea level rise on the rivers, and reduce saltwater intrusion into aquifers water supplies of major cities. The U.S. -- is now minimal. Army Corps of Engineers should factor climate change into the design of major projects. Table 13-1. Value of the Nation's Stock of Selected Infrastructure (billions of -- Given the assumption that modest changes 1984 dollars) in the design and location of many transportation systems may facilitate an accommodation to climate change, the Component Valueᵃ Department of Transportation should factor climate change into the design of roads, bridges, and mass transit facilities. Water supply $108 -- Voluntary standards organizations, such as Wastewater 136 the American Society of Civil Engineers, the Building Officials and Code Administrators Urban drainage 60 International, and the American Society of Heating and Refrigerating and Air Streets 470 Conditioning Engineers should examine the need for changes in existing energy and Public airports 31 safety factors to account for the possibility of climate change. Mass transit 34 Electric power 266 RELATIONSHIP BETWEEN (private only)ᵇ URBAN INFRASTRUCTURE AND Public buildings unknown CLIMATE Total $1,105+ Three-quarters of the U.S. population is a Based on a useful life of 35 to 50 years for most concentrated in urban areas (Statistical Abstract, assets, and 10 to 20 years for transit vehicles. 1988). The majority of the nation's investment in b About 77% of electric power is privately produced. water supply, wastewater transport and treatment Source: Statistical Abstract (1988); National Council facilities, drainage, roadways, airports, mass transit, on Public Works Improvement (1988). electric power, solid waste disposal sites, and public buildings serves these urban areas. The current value of selected infrastructure nationwide, displayed in Table 13-1, provides insight into the 1 Of the 20 most populated urban areas in the United States, 12 aggregate investment at stake if climate changes. are tidal waterfront cities (Baltimore, Boston, Houston, Los Angeles, Miami, New York, Philadelphia/Wilmington, San Most of these items could be considered part of Francisco/Oakland, San Diego, Seattle, Tampa/St. Petersburg, urban infrastructure; their locations and designs and Washington, DC), 3 are located on the Great Lakes have been based on historic meteorologic (Chicago, Cleveland, and Detroit), 3 are on navigable rivers information. Annually, governments add an average (Minneapolis, Pittsburgh, and St. Louis), and 2 are not on a navigable waterway (Atlanta and Dallas). 238 Urban Infrastructure Although actual practice varies, the nominal protection in Dade and Broward Counties, Florida, replacement cycle for most infrastructure is 35 to 50 and concluded that the effects might be substantial. years (National Council on Public Works Linder et al. (1987) estimated that CO₂ doubling Improvement, 1988). Some water supply might require raising electric capacity by 21% in a investments have 100-year cycles between planned southeastern utility and by 10 to 19% in New York replacement; however, sea level rise, temperature State. Hull and Titus (1986) analyzed the potential change, and changes in precipitation patterns could impact of sea level rise on water supply in the alter the balance between water supply and demand. Philadelphia-Wilmington-Trenton area and found The nature and pattern of precipitation could affect that a rise of 0.3 meters could require adding 140 drainage requirements as well as highway design million cubic meters of reservoir capacity, about a and maintenance. 12% increase, to prevent saltwater from advancing past water intakes on the Delaware River. The heat wave of 1988 illustrated some of the Additional investment would be required to prevent potential impacts. Hundred-degree weather or respond to saltwater infiltration into underground distorted railroad tracks, forcing Amtrak to cut aquifers. Cohen (1987) estimated that large speeds from 200 to 128 kilometers per hour between municipalities along the Great Lakes might increase Washington and Philadelphia (Bruske, 1988) and water withdrawals by 5.2 to 5.6% during May to possibly contributing to a train wreck that injured September because of increased lawn watering. 160 people on a Chicago-Seattle run (The Washington Post, 1988). A U.S. Army Corps of Two recent studies illustrate the importance of Engineers contractor worked around the clock for considering sea level rise in urban coastal 2 weeks to build a 170-meter-wide, 9-meter-high silt infrastructure planning and the uncertain nature of wall across the bottom 40% of the Mississippi River the decisions involved. Wilcoxen (1986) examined channel, 48 kilometers below New Orleans the impact of sea level rise on a portion of San (Sossaman, 1988a,b). This $2 million wall, designed Francisco's sewage transport system buried near the to wash away when spring snowmelt demands the shoreline. The study estimated that if sea level rose full capacity of the channel, slowed an advancing 0.6 meters by the year 2100, an expenditure of wedge of saltwater that threatened the water supply roughly $70 million on beach nourishment might be in New Orleans and nearby parishes. In Manhattan, required to prevent damage to a structure that cost heat exacerbated the effects of longstanding leaks in $100 million to build in the late 1970s. The author 256 kilometers of steam pipes, causing the asphalt suggested that consideration (at no additional cost) to soften. As vehicles kneaded the soft asphalt, of sea level rise in siting the structure could have thousands of bumps formed on city streets, prevented these expenses. Titus et al. (1987) requiring extensive repairs (Hirsch, 1988). In the examined the impact of sea level rise on a proposed suburbs of Washington, DC, steel expansion joints coastal drainage system in Charleston, South bubbled along a 21-kilometer stretch of Interstate Carolina, and estimated that a 0.3-meter sea level 66 (Lewis, 1988). rise by 2025 would require almost $2.5 million in additional investments to maintain the target level The following sections of this chapter will of flood protection. The present value of these examine such issues as the portions of the investments is $730,000. In contrast, only about infrastructure that will be significantly affected, and $260,000, one-third of the cost of responding in anticipated costs and who will bear them. 2025, would be required to add this level of protection at initial construction. Thus, the investment would be worthwhile if the probability of PREVIOUS CLIMATE CHANGE sea level rising this rapidly exceeds 35%. STUDIES ON URBAN INFRASTRUCTURE URBAN INFRASTRUCTURE STUDY IN THIS REPORT Available literature on the potential effects of global climate change on urban infrastructure is Several studies undertaken for this report sparse. Rhoads et al. (1987) examined the potential impacts of sea level rise on water supply and flood examined some of the implications of climate change in relationship to urban infrastructure. One 239 Chapter 13 study comprehensively examined the impacts on Study Design infrastructure in several cities: The study was based on a critical review of Impact of Global Climate Change on Urban existing infrastructure studies in the three cities, Infrastructure Walker, Miller, Kingsley, discussions of likely impacts with local infrastructure and Hyman, The Urban Institute (Volume experts, analyses undertaken by these experts, and H) preliminary calculations of probable impacts. Experts were presented with GCM scenarios for The following studies, referenced in this chapter, CO2 doubling, and scenarios were used to calculate covered issues relating to urban infrastructure: effects on energy demand, roadways, and other systems. The study also derived conclusions based The Potential Impacts of Climate Change on experiences in other cities where current on Electric Utilities: Regional and National temperatures are analogous to temperatures Estimates - Linder and Inglis, ICF Inc. projected for the cities under study, using the (Volume H) analogs identified by Kalkstein (Volume G). Impacts of Extremes in Lake Michigan Limitations Levels Along the Illinois Shoreline: Low Levels - Changnon, Leffler, and Shealy, The principal limitation of the overall study is University of Illinois (Volume H) the limited use of hydrologic and other modeling. In addition, experts were asked to derive Methods for Evaluating the Potential conclusions regarding conditions beyond their Impacts of Global Climate Change: Case experience. Since only three cities are included, the Studies of the Water Supply Systems of the full range of effects on urban infrastructure was not State of California and Atlanta, Georgia covered. The authors did not perform engineering Sheer and Randall, Water Resources analyses of cost-effective responses, and they did not Management Inc. (Volume A) assess the potential for reducing impacts through technological change. Thus, these results should be National Assessment of Beach Nourishment considered as approximations of the costs of impacts Requirements Associated with Sea Level and as illustrative of the sensitivity of urban Rise Leatherman, University of Maryland infrastructure to climate change. (Volume B) Results and Implications The Costs of Defending Developed Shorelines Along Sheltered Waters of the Miami's Infrastructure United States from a Two-Meter Rise in Mean Sea Level - Weggel, Brown, Greater Miami is bounded by water on all Escajadillo, Breen, and Doheny, Drexel sides during the rainy season. An extensive network University (Volume B) of canals and levees has been built to control ocean and freshwater flooding and to recharge the aquifer Effect of Climate Change on Shipping beneath the area. Miami has one of the world's Within Lake Superior and Lake Erie - most porous aquifers, which lies less than 1.5 meters Keith, DeAvila, and Willis, Engineering below the surface in one-third of the developed Computer Optecnomics (Volume H) area. Federal law requires that roughly 15% of Miami's freshwater be released into the Everglades National Park. RESULTS OF THE INFRASTRUCTURE STUDY The Miami case study examined the probable impacts of climate change and sea level rise on Dade County's water supply, water control and Impacts on Miami, Cleveland, and New drainage systems, building foundations, roads, York City bridges, airports, solid waste disposal sites, and sewage transport and treatment systems, assuming Walker et al. examined three cities distinctly that a gradual sea level rise would be managed affected by climate change to determine a range of through strategies such as raising the land in impacts on urban infrastructure. 240 Urban Infrastructure low-lying areas, upgrading levees and dikes with Table 13-2. Probable Infrastructure Needs and pumped outflows, retreating selectively from some Investment in Miami in Response to a areas, and increasing the freshwater head roughly in Doubling of CO₂ (millions of 1987 proportion to sea level rise to prevent saltwater dollars) infiltration into the aquifer. As Table 13-2 shows, global climate change Infrastructure need Cost could require more than $500 million in capital investment in Greater Miami over the next century. Because needed investments in many systems could Raising canals/levees 60ª not be estimated and because a complete Canal control structures 50 engineering analysis was not performed, these Pumping not estimated results should be considered only as rough Raising streets 250 added to estimates. They imply an increase of 1% to 2% in reconstruction Greater Miami's capital spending for the next 100 cost years, no more than $20 per household per year at Raising yards and houses not estimated 1985 population levels (Metropolitan Dade County Pumped sewer connections not estimated Planning Department, 1988). Raising lots at reconstruction not estimated Drainage 200-300 Because the south Florida aquifer extends Airport 30 under the ocean, the typical urban response to a Raising bridges not estimated rising sea diking the water at the surface and Sewer pipe corrosion not estimated pumping out the seepage from ditches behind the Water supply uncertain dikes appears to be unworkable. Unless the dike Electric generation 20-30% capacity extended downward more than 45 meters, rising increase seawater pressure would cause the sea to rush into the aquifer below the surface and push freshwater Costs are partially based on Weggel et al., upward, almost to the surface. Volume B. Source: Walker et al. (Volume H); Linder and The one-time capital costs for upgrading Inglis (Volume H). existing canals and levees in response to a 1-meter sea level rise could be about $60 million. However, almost $50 million in new control structures, would have to be raised or risk collapse. If sea level including extensive pumping capacity, might be rose gradually, thereby permitting raising of streets required for the canals used to maintain the and related sewer mains during scheduled recon- freshwater head. Large-scale pumping along canals struction, the added public cost might be also could involve substantial operating costs, but approximately $250 million. Building owners would these have not been estimated. Storm sewers and incur substantial costs to improve drainage, raise drainage would need upgrading, requiring invest- yards, raise lots at reconstruction, and pump sewage ment of several hundred million dollars above to mains. Miami's airport also would need better normal replacement costs. drainage, requiring an approximately $30 million investment. Building foundations generally should remain stable if the freshwater head rises 1 meter because A 1-meter rise in sea level would require houses are built on concrete slabs, most buildings in raising most bridges to ensure adequate clearances newer areas already are built on raised lots to meet and reduce vulnerability to storm surges during Dade County's flood control ordinance, and the hurricanes. foundations of many larger buildings are designed to extend into the water table. It is unclear what effect climate change will have on hurricanes. Without increased hurricane Conversely, the water table could infiltrate the activity, climate change probably would exacerbate base of about a third of Dade County streets, which water shortages that are expected to result from population growth in Greater Miami. Thus, climate 241 Chapter 13 change could accelerate Miami's long-range plan Table 13-3. Estimated Impacts of a CO2 Doubling for large-scale production of desalinated water at on Cleveland's Annual Infrastructure three times current water prices. If hurricanes Costs (millions of 1987 dollars) increase, Miami's added expense for water supply might be roughly $100 million to move some wells farther inland. Conversely, increased hurricane Annual frequency and intensity could cause billions of Infrastructure category operating costs dollars in property damage. Analysis of Miami's coastal defense and water Heating -2.3 supply options provides insight into the impacts of Air-conditioning +6.6-9.3 sea level rise on cities built on coral reefs, but not Snow and ice control -4.5 into the response of most mainland cities on the Frost damage to roads -0.7 U.S. coastline. Dade County is unusual because Road maintenance -0.5 readily extracted fill is extensively available on Road reconstruction -0.2 public lands having easy access to a canal system Mass transit summer increase that can be navigated by flat-bottomed barges. offsets winter Nevertheless, this case study suggests that global savings climate change could cause large coastal cities to River dredging less than 0.5 invest billions of dollars over the next 50 to 75 years Water supply negligible to add and upgrade infrastructure. Storm water system negligible Total -1.6 to + $1.1 Cleveland's Infrastructure Source: Walker et al. (Volume H); Keith et al. The Cleveland case study examined impacts of (Volume H). climate change on snow and ice control costs, road construction and maintenance, heating and cooling costs and equipment needs, water supply, and storm and wastewater transport. The study also included conditioning costs seemed likely to rise by $6.6 to a preliminary analysis of the effects of a drop in the $9.3 million. The impacts on the transit operating level of Lake Erie as estimated by Croley (see budget seemed likely to mirror the impacts on the Chapter 15: Great Lakes). The impact on the snow general budget, with reduced mishaps and traffic and ice control budget was estimated by analogy to delays in ice and snow offsetting increased fuel costs the budget in Nashville, Tennessee. for vehicle cooling. Results are displayed in Table 13-3, which The study suggested Cleveland might spend shows that the net impact of climate change on about $65 to $80 million to add air-conditioning to Cleveland's annual infrastructure costs could be older schools and to large nonoffice spaces such as negligible, although expenditures probably would gyms and repair garages. Much of this expenditure shift between categories. In addition to the costs would occur as buildings were replaced or shown in Table 13-3, a one-time capital expenditure refurbished and might have occurred even without of $68 to $80 million could be required to add air climate change. conditioners in public buildings. Also, many private residences probably would install air conditioners. The rise in winter temperatures associated with a doubling of CO2 might allow Cleveland to use Walker et al. estimated that global climate thinner pavement, resulting in possible savings of change could cause annual snowfall in Cleveland to about 3% in road resurfacing costs and 1% in drop from 1.25 to roughly 0.2 meters (4.1 to 0.7 reconstruction costs. The net savings could average feet), reducing annual snow and ice control costs by about $200,000 per year or 1.3% of the city's current about $4.5 million. Decreased frost damage to capital budget. Engineering standards (AASHTO, roads and bridges could yield further savings 1987) suggested that the rate of pavement estimated at $700,000 per year. A drop of $2.3 deterioration probably also should decline as winter million per year in heating costs for public buildings temperatures rise, saving roughly $500,000 per year. also was estimated. Conversely, annual public air- 242 Urban Infrastructure A climate-induced drop in the level of Lake Table 13-4. Probable Impacts of a CO2 Doubling Erie probably would not adversely affect Cleveland, on Selected Infrastructure in the New although some dredging might be required in the York Metropolitan Area (millions of Cuyahoga River and port area (Keith et al., Volume 1987 dollars) H). Upgrading of the city's combined storm and wastewater collection system appeared to be unnecessary, although this would depend upon Infrastructure rainfall variability. category Costs If temperature rises several degrees, most northern cities probably could anticipate savings in Upgrading levees 120 snow and ice control, heating, and roadway construction and maintenance costs similar to those Drainage increased flooding in low- described for Cleveland. These savings might lying areas, minimal approximately offset the increase in air-conditioning sewer system changes costs. More southern cities could experience modest budget increases. Sewer outflows more frequent inspection Cleveland could become a more attractive Water supply 3,000 location for water-intensive industry if water supplies in other areas become less reliable. Snow and ice control reduced substantially Resulting in-migration could bring further growth- related infrastructure costs. Lower Great Lakes Road maintenance and winter savings, offset by levels could require dredging, modification to ports, reconstruction melting asphalt in and relocation of some water intakes. (For a Manhattan further discussion of these issues, see Chapter 15: Great Lakes.) Mass transit summer increase offsets winter savings New York City's Water Supply Electricity production 65-150 New York City's infrastructure may be affected in many ways by global climate change. Heating reduced Temperature change could affect the same capital expense categories in both New York City and NOTE: Impacts on underground infrastructure, Cleveland. In addition, the city may have to airports, and ports have not been probed, gradually raise its dikes and better protect but a discussion of these impacts among underground infrastructure from seawater Port Authority representatives and other infiltration. Interpolating from Weggel et al. experts at the Second North American (Volume B), approximately $120 million might be Conference on Preparing for Climate invested to protect shorelines from a sea level rise Change, Washington, DC, December 7, of 1 meter. The most pressing, and perhaps largest, 1988, suggested they might be small. problem facing the city may be the effects of global Source: Walker et al. (Volume H); Weggel et al. climate change on the adequacy of the city's water (Volume B); Linder et al. (1987); Schwarz and supply. The New York City study focused on that Dillard (1989). issue. Table 13-4 provides estimates drawn from a number of studies about possible infrastructure impacts on New York City. The water supply network is in deficit. The The New York metropolitan area draws water Mayor's Task Force (1987) has recommended from the adjoining Hudson and Delaware River remedying New York City's portion of the deficit Basins and from underground aquifers serving through better management of water demand and coastal New Jersey and Long Island. Figure 13-1 detailed study of the possibility of reactivation of a shows the region and its water supply sources. 243 Chapter 13 DELAWARE SYSTEM GILBOA CATSKILL CATSKILL SCHOHARIE SYSTEM BRANCH RESERVOIR CREEK CANNONSVILLE RESERVOIR WEST WALTON PEPACTON 125MILES 125 RESERVOIR DEPOSIT WEST SHANDAKEN TUNNEL <<<<<<<<<<<<<<<<<<<<<<<<< SCREEK CREEKE CATSKILL DOWNSVILLE NEW YORK CONNECTICUT KAST ASHOKAN RESERVOIR NEW YORK EAST PENNSYLVANIA HANCOCK BRANCH BEAVER KILL DELAWARE TUNNEL DELAWARE LAWARE RONDOUT RESERVOIR ONE ESOPUS CREEK HUDSON RIVER CREEK DELAWARE RIVER LIBERTY LACKAWACK CROTON 100MILES RONDOUT NEVERSINK NEVERSINK SYSTEM RESERVOIR TUNNEL POUGHKEEPSIE ELLENVILLE CHELSEA NEVERSINK RIVER PUMP STATION 75MILES WALLKILL RIVER COTTSKITY WEST BRANCH RESERVOIR PORT JERVIS AQUEDUCT MILFORD MONTAGUE CROTON LAKE PENNSYLW VANIA JERSEY SOMES NEW NEW JERSEY YORK NEW CROTON AQUEDUCT OLD CROTON KENSICO AQUEDUCT RESERVOIR HILL VIEW NEW CROTON AQUEDUCT HILL VIEW RESERVOIR LONGOUND ISLAND RESERVOIR JEROME PARK K MES HALL RESERVOIR JEROME PARK RESERVOIR RESERVOIR OLD CROTON FROM CENTRAL PARK AQUEDUCT LONG ISLAND CITY TUNNEL NO. CITY TUNNEL NO. 1 CITY TUNNEL NO. 2 CENTRAL PARK RICHMOND RESERVOIR TUNNEL SIL VER LAKE CITY TUNNEL PARK NO 2 (UNDERGROUND STATE STORAGE TANKS ISLAND ATLANTIC OCEAN Figure 13-1. The sources of New York City's water supply (New York City Municipal Water Finance Authority, 1986). 244 Urban Infrastructure water intake at Chelsea, a $223 to $391 million Implications Arising from Other EPA investment that would yield 375 to 750 million liters of water daily. Studies in This Report Walker et al. estimated changes in water Linder and Inglis (Volume H; Chapter 10: demand using design standards for commercial Electricity Demand) suggest that increased air- cooling-tower demand, changes in electricity conditioning use could raise peak electricity demand demand estimated by Linder et al. (1987), and by 10 to 30% in the southern half of the United historic residential summer water use. The impact States. Nationally, utilities supplying the of sea level rise on water supply was estimated by northernmost cities could experience decreased analogy using Hull and Titus (1986), which analyzes demand, while those supplying cities in the possible saltwater advance up the Delaware River. remainder of the country could experience The impact on reservoir supply also was estimated electricity needs higher than they have anticipated. by analogy, using a Great Lakes water balance Sheer's study of California (see Chapter 14) water model (Linder et al., 1987). Walker et al. assumed supply suggests that new surface water that baseline demand would not increase above impoundments may be needed to meet urban water projected demand in 2030, potentially needs and other demands. The coastal defense underestimating the increased demand for water. strategies suggested in Chapter 7: Sea Level Rise would apply to most urban coastal areas, especially Walker et al. estimated that a rise in those along the Atlantic and Gulf coasts. temperatures consistent with the GISS and GFDL scenarios would mean about a 20% increase in Changnon et al. (Volume H) conclude that a cooling degree days. In response, average daily falling lake level might prompt investment of $200 demand for water used in cooling large buildings to $400 million to adapt recreational and could increase by 190 million liters during the commercial harbors and beach facilities, and an summer, and increased lawn watering could raise investment of $20 million to adjust water supply demand by 110 million liters per day, thereby intakes and sewer outfalls along the Illinois generating a 5% rise in annual demand. shoreline of Lake Michigan, with similar costs likely on the other Great Lakes. The Keith study (see Higher temperatures could increase Chapter 15: Great Lakes) suggests that each evaporation and evapotranspiration, decreasing the commercial harbor on Great Lakes Erie and ability to store water efficiently in surface Superior could spend $5 to $30 million on dredging to maintain harbor access. impoundments. The water balance model indicated the supply loss could range from 10 to 24%. Saltwater infiltration due to rising sea level RESULTS OF RELATED STUDIES would further reduce supply. The study suggested that a 1-meter sea level rise could place the proposed $300 million Chelsea intake below the salt Metropolitan Water Supply line during the peak summer demand period in mild drought years, reducing supply another 13%. Schwarz and Dillard (1989) conducted Larger sea level rise or greater droughts might telephone interviews with local infrastructure prevent use of the existing Poughkeepsie intake managers to identify the probable impacts of global during severe droughts, further reducing supply. In climate change on water supply and drainage in addition, subsurface infiltration could reduce the several metropolitan areas. Results from some cities are discussed here. supply available from the Long Island aquifer. In summary, a doubled CO₂ atmosphere could Washington, DC produce a shortfall equal to 28 to 42% of planned supply in the Hudson River Basin. Longer hot spells could warm the Potomac River and cause trihalomethane formed during chlorination to rise above allowable limits. Remedying this could require a capital investment of roughly $50 to $70 million and could increase 245 Chapter 13 treatment costs. Also, lawn watering probably roughly offsetting gains and losses. Others would increase during long spells of hot, dry especially those along the coastlines and in water- weather. Although a substantial decrease in runoff short areas, could bear increased infrastructure could reduce supply in parts of the system, the costs. The costs would be especially high if changes availability of additional storage capacity would came through abrupt "sawtooth" shifts or increases make a shortage unlikely. in extreme events, making it difficult to adapt infrastructure primarily during normal repair and New Orleans replacement. The likely impacts of an effective doubling of atmospheric CO₂ could affect a wide Sea level rise could necessitate moving the range of infrastructure. Additional climate change water intakes considerably farther up the Mississippi effects beyond doubled CO2 or sea level rise above and replacing cast iron water mains that would 1 meter could result in even greater costs. corrode if exposed to saltwater. Reduced riverflow also could increase settling and treatment Water requirements. Rising sea level could increase saltwater infiltration into the water system and could Hotter temperatures could cause faster require increased pumping capacity. evaporation of groundwater and raise the demand for water to support commercial air-conditioning New York City systems and lawn watering. Earlier snowmelt in the West could force a lowering of dam levels to ensure This study raised many of the same concerns availability of enough capacity to control flood regarding water supply and demand as the study by waters. At the same time, sea level rise could cause Walker et al. (Volume H) and indicated that even saltwater to advance up rivers and to infiltrate into a 0.25-meter sea level rise would mean the proposed coastal aquifers. In droughts, many existing water Chelsea intake was too far downstream. The intakes might deliver brackish water. sanitary and storm sewage system capacity and design probably would not need revision. The solution to these problems could involve Nevertheless, in a few low-lying areas, higher sea strong conservation measures, such as miles of level could increase sewer backups, ponding, and aqueducts from new water intakes at higher river basement flooding when high tides coincided with elevations, new reservoirs, sewage effluent recycling high runoffs. systems to support commercial cooling or lawn watering, and perhaps desalinization efforts along Tucson the coasts. The solution for the New York- Philadelphia corridor alone is likely to cost $3 to $7 Tucson is depleting its aquifer despite substan- billion. Communities in the Delaware River Basin, tial conservation efforts and lawn watering with northern New Jersey, the lower Hudson, and Long treated wastewater. Higher temperatures would Island might well form a multistate water supply increase demand and tighten supply, possibly and management district of unprecedented size and jeopardizing the city's ability to draw on water from complexity to handle financing and capital the Central Arizona Project on the already strained construction. Colorado River. While modest savings might be achieved through stricter conservation measures and Drainage and Wastewater Systems more wastewater use, purchase of water in the regional market most likely would be the only Increased storm size and intensity could tax practical response to climate-related shortfalls. many storm sewer systems. Sea level rise also could reduce coastal flood protection levels in low-lying areas. The resulting increases in flooding and IMPLICATIONS FOR URBAN releases of untreated waste into watercourses from INFRASTRUCTURE combined storm and wastewater systems probably would motivate new sewer investments. In Dade County alone, costs to maintain flood protection at The implications of climate change for urban America vary spatially. Some localities, especially existing levels could be $200 to $300 million if sea level rose 1 meter. those along the Great Lakes, might experience 246 Urban Infrastructure Temperature rise could increase hydrogen Electricity and Air-Conditioning sulfide formation in sewer pipes, leading to internal corrosion and eventual failure. In coastal areas with Hotter temperatures could increase air-con- increased ocean flooding, storm sewers would carry ditioning use. Consequently, peak load capacity to corrosive saltwater with increased frequency. Sea generate electric power might have to increase in level rise also could cause more pipes in coastal response to global climate change. Fortunately, air- areas to face the external risk of corrosive seawater. conditioning equipment is replaced frequently, so More frequent inspection and earlier replacement of increased loads on existing equipment could be much existing pipe, as well as a gradual shift to accommodated incrementally. Some houses and more corrosion-resistant pipe with plastic lining, public buildings in northern climates might need to might be required. add air-conditioning, but such retrofitting has been performed since the first window air conditioners Coastal Defenses were introduced. Protection from a rising sea could require periodic investment in many major coastal POLICY IMPLICATIONS communities. In urban areas, a common approach might be the New Orleans solution, where The possibility of global climate change extensively developed coastal areas are protected by increases the risks of infrastructure investment. dikes, and covered drainage ditches behind the dikes Application of design standards and extrapolation are pumped to keep out the saltwater. from historical data still may not provide reasonable assurance that water and power supply, dam Roads strength and capacity, bridge clearances, or storm sewerage capacity will be adequate for the 35-, Rising temperatures could reduce the costs of 50-, and 100-year design cycles of these facilities. road construction and maintenance. Snow and ice For example, the National Flood Insurance control costs might drop dramatically. A decrease Program's maps identifying the historical 100-year in deep freezes and freeze-thaw cycles also would floodplain and 500-year floodway may no longer mean fewer potholes. Warmer temperatures and provide a reliable basis for local building and zoning the improved drainage resulting from higher ordinances designed to minimize flood losses to life evaporation rates could permit use of thinner and property. pavements in many areas, but could require enhanced expansion capabilities. Investment Analysis Methods Bridges Especially in coastal areas, the possibility of global climate change may soon require careful Sea level rise and increased storm intensity decisions regarding how and when to adapt the could require upgrading of many bridges either infrastructure. A strong emphasis on lifecycle through costly retrofit or as part of normal recon- costing and upgrading during reconstruction in struction. The range of temperature accommodated anticipation of future changes could yield large, by expansion joints also might need to be increased. long-term cost savings. To accomplish this goal, The costs might be modest if bridge planners such institutions as the Department of Housing and upgraded in anticipation of climate change. Urban Development might work with the American Public Works Association, the National League of Mass Transit Cities, the U.S. Conference of Mayors, the American Planning Association, and similar groups In the North, buses and railcars could to educate their constituencies regarding the experience fewer snow-related delays. Conversely, uncertainties and ways to incorporate them into the slight increases in fuel costs could result from decisionmaking process. increased use of air conditioners. 247 Chapter 13 Water Supply geographically based standards -- for example, on roadbed depth and home insulation levels -- and Water supply is of particular concern because provide significant savings. Thus, the standard- decades are required to plan and complete projects, making organizations might beneficially establish which then might last 100 years. Dams, reservoirs, policies concerning how and when their committees and water intakes currently being planned and built should account for global climate change or educate could become obsolete or inadequate as a result of their committees about the prospects. global climate change. Elsewhere, communities might be allowing development of land needed for reservoirs to meet the water shortages that would RESEARCH NEEDS result from climate change. The following are recommended for further Such federal agencies as the U.S. Geological research: Survey, U.S. Army Corps of Engineers, and EPA may wish to work with states and municipalities to 1. More case studies of urban impacts, with study the possible impacts of climate change on the priority on a west coast city and an inland city. water supply of major metropolitan areas. Issues of particular interest include the effects on subsidence problems in cities similar to Water supply investments frequently affect Phoenix, the implications for sewage treatment multistate areas, creating a need for federal capacity in areas where more frequent and coordination. The Supreme Court has been forced intense periods of low riverflow could reduce to settle previous water rights disputes concerning acceptable effluent discharge rates, the impact many major rivers, and global climate change might on bridge replacement costs, and the potential well generate new disputes. Cost-effective response for and probable consequences of saltwater to climate change also might require new multistate infiltration into pipes in older coastal water projects. For example, a major project on the communities. Hudson River that allowed New York City to reduce its use of Delaware River water might be the 2. The probable impacts of global climate change least costly way to increase water supply in on domestic and international migration flows Philadelphia. The upcoming state debates over and the infrastructure demands these flows water supply financing should be informed by the produce. Heat and high water prices might lesson of past infrastructure crises: water piping drive jobs and people away from some regions, and pumping costs resulting from global climate while others might flourish. Infrastructure change should be fully recovered from the water investment in new water supply, for example, users to avoid stimulating artificial demand for might be unnecessary in areas that would lose bargain water. population, but extra capacity might be needed in areas where population would grow. Infrastructure Standards Similarly, as climate change shifts the best growing areas for specific crops, new farm-to- Voluntary standards organizations, such as the market transportation networks might need to American Society of Civil Engineers, the Building be developed. Rights-of-way for these systems Officials and Code Administrators International, might best be set aside now, before land prices and the American Association of State Highway and rise in response to climate change. Transportation Officials, may wish to educate their committees on global climate change. Growing uncertainty concerning future temperature, REFERENCES precipitation, and sea levels might dictate a reassessment of existing standards and safety factors AASHTO. 1987. American Association of State for ventilation, drainage, flood protection, facility Highway and Transportation Officials. Manual for siting, thermal tolerances, resistance to corrosion, the Design of Permanent Structures, Appendix A. and so forth. Conversely, prompt detection of Treatment of Roadbed Swelling and/or Frost Heave lasting changes could allow adjustment of 248 Urban Infrastructure in Design. Washington, DC: American Association New York City Municipal Water Finance Authority. of State Highway and Transportation Officials. 1986. Water and Sewer System Revenue Bonds, Fiscal 1986, Series A. Prospectus. New York. Bruske, E. 1988. 104 (phew!) degrees hottest in 52 years. The Washington Post 111(225):A1, A6. July Rhoads, P.B., G.C. Shih, and R.L. Hamrick. 1987. 17. Water resource planning concerns and changing climate: a Florida perspective. In: Proceedings of Cohen, S.J. 1987. Projected increases in municipal the Symposium on Climate Change in the Southern water use in the Great Lakes due to CO2-induced United States: Future Impacts and Present Policy climatic change. Water Resources Bulletin Issues. Norman, OK: University of Oklahoma, pp. 23(1):91-101. 348-363. Hirsch, J. 1988. As streets melt, cars are Schwarz, H.E., and L. Dillard. 1989. Urban water. flummoxed by hummocks. The New York Times Chapter III-D. In: Waggoner, P.E., ed. Climatic 137(47599):B1, B5. August 16. Variability, Climate Change, and U.S. Water Resources. New York: John Wiley and Sons. In Hull, C.H.J., and J.G. Titus, ed. 1986. Greenhouse press. Effect, Sea Level Rise, and Salinity in the Delaware Estuary. Washington, DC: U.S. Environmental Sossaman, B.A. 1988a. News release. U.S. Army Protection Agency. Publication No. 230-05-86-010. Corps of Engineers, New Orleans District. June 28. Lewis, N. 1988. Two more heat records fall as Sossaman, B.A. 1988b. News release. U.S. Army summer of 1988 boils on. The Washington Post 111 Corps of Engineers, New Orleans District. July 15. (257):A1, A10, A11. August 18. Statistical Abstract of the United States. 1988. Linder, K.P., M.J. Gibbs, and M.R. Inglis. ICF Washington, DC: U.S. Government Printing Office. Incorporated. 1987. Potential Impacts of Global Climate Change on Electric Utilities. Albany, NY: Titus, J.G., C.Y. Kuo, M.J. Gibbs, T.B. LaRoche, New York State Energy Research and Development M.K. Webb, and J.O. Waddell. 1987. Greenhouse Authority. Publication No. 824-CON-AEP-86. effect, sea level rise, and coastal drainage systems. Journal of Water Resources Planning and Mayor's Intergovernmental Task Force on New Management 113(2):216-227. York City Water Supply Needs. 1987. Managing for the Present, Planning for the Future; December. The Washington Post. 1988. Warped rails checked in Amtrak wreck. 111(246):A5. August 7. Metropolitan Dade County Planning Department. 1988. Comprehensive Development Master Plan for Wilcoxen, P.J. 1986. Coastal erosion and sea level Metropolitan Dade County, Florida. July 1979, July rise: implications for Ocean Beach and San 1985, June 1987, and April 1988. Francisco's Westside Transport Project. Coastal Zone Management Journal 14(3):173-191. National Council on Public Works Improvement. 1988. Fragile Foundations - Final Report to the President and Congress, Washington, DC; February. 249 CHAPTER 14 CALIFORNIA FINDINGS maintained, the estuary could still increase in area and volume by 30 and 15%, respectively, as a result of a 1-meter sea level rise alone. Global warming could cause higher winter runoff and lower spring runoff in California and increase Sea level rise of 1 meter could cause saline the difficulty of meeting water supply needs. It (brackish) water to migrate inland between 4 could also increase salinity in the San Francisco Bay and 10 kilometers (2.5 and 6 miles, and the Sacramento-San Joaquin Delta and increase respectively) if the levees fail and if tidal the relative abundance of marine species in the bay; channels do not erode. Freshwater releases degrade water quality in subalpine lakes; raise into the delta might have to be doubled to ambient ozone levels; increase electricity demand; repel saline water near the major freshwater and raise the demand for water for irrigation. pumping facilities. Water Resources Wetlands and Fisheries Higher temperatures would lead to higher The wetlands in the San Francisco Bay estuary winter runoff from the mountains surrounding would be gradually inundated as sea level rises the Central Valley, because less precipitation faster than the wetlands accrete sediments. would fall as snow, and the snowpack would The amount of wetlands lost would be a melt earlier. Runoff in the late spring and function of the rate of sea level rise and of summer consequently would be reduced. whether shorelines are protected. If sea level rises 1 meter by 2100, the rate of rise will be As a result, the amount and reliability of the greater than wetland vertical accretion by the water supply from reservoirs in the Central middle of the next century. If sea level rises 2 Valley Basin would decrease. Annual water to 3 meters by 2100, wetland inundation will deliveries from the State Water Project (SWP) begin early in the 21st century. could be reduced by 200,000 to 400,000 acre- feet or 7 to 16%. In comparison, the statewide If salinity increases within the San Francisco increase for water from the SWP, due to non- Bay estuary, wetland vegetation will shift from climate factors such as population growth, may brackish and freshwater species to more salt- total 1.4 million acre-feet by 2010. Even if tolerant plants. This shift could severely operating rules were changed, current reduce waterfowl populations that depend on reservoirs would not have the capacity to store freshwater habitats. The timing, magnitude, the heavier winter runoff and at the same time and location of phytoplankton production could retain flood control capabilities. shift. Marine fish species could increase in abundance, while saltwater species that breed Rising sea level could increase the possibility of in freshwater areas would most likely decline. levee failure. If the delta and bay levees failed and sea level rose 1 meter (40 inches) by 2100, Higher temperatures in subalpine lakes could agriculture in the delta region would be almost increase annual primary production (such as eliminated, the pumping of freshwater out of algae) by between 16 and 87%, which could the delta to users to the south could be degrade lake water quality and change the jeopardized by increasing salinity, and the area composition of fish species. and volume of the estuary could triple and double, respectively. Even if the levees were 251 Chapter 14 Agriculture By 2010, 2 to 3 gigawatts (GW) would be needed to meet the increased demand. By The impacts of climate change on agriculture 2055, 10 to 20 GW would be needed a 14 to in California are uncertain. The effects of 20% increase over baseline additions that may changes in temperature and precipitation alone occur without climate change. The additional would most likely reduce yields by 3 to 40%, capital cost by 2055 would be $10 to $27 billion depending on the crop. However, with the (in 1986 dollars). combined effects of climate and higher CO₂ levels, yields for all modeled crops, except corn Policy Implications and sugarbeets, might increase. Water management institutions, such as the The potential growth in irrigation in some U.S. Bureau of Reclamation and the California parts of the state may require increased Department of Water Resources, should extraction of groundwater because of current analyze the potential impacts of climate change full use of surface water supplies. This would on water management in California. They decrease water quality and affect water should consider whether and how the Central management options. Valley Project and State Water Project should be modified to meet increasing demands in the Yields in California may be less adversely face of diminishing supplies due to climate affected than those in most parts of the change. They may also consider whether to country. Crop acreage could increase because change water allocation procedures to of the shifts in yields and the presence of encourage more efficient use of water. irrigation infrastructure. The California Water Resources Control Natural Vegetation Board should consider the impact of climate change on surface and groundwater quality. Drier climate conditions could reduce forest density, particularly pine and fir trees, and State and local entities should consider the timber productivity. (The full impacts on impacts of climate change on levee and California forests were not assessed in this wetland management in San Francisco Bay and report.) the delta. Air Quality The California Air Quality Board should review the long-term implications of climate If today's emissions exist in a future warmer change on air quality management strategies. climate, ozone levels in central California could increase and could change location because of The California Energy Commission should higher temperatures. As a result, the area in consider the impacts of climate change on the central California with ozone levels exceeding energy supply needs for the state. EPA standards (0.12 parts per hundred million (pphm)) on a given day could almost double unless additional steps are taken to control CLIMATE-SENSITIVE emissions. These additional controls would RESOURCES OF CALIFORNIA increase the cost of pollution control. California's Central Valley is the most Electricity Demand productive and diverse agricultural region of its size in the world. The Central Valley Basin, which The annual demand for electricity in California includes the drainages of the Sacramento and San could rise by 3 to 6 billion kilowatthours (kWh) Joaquin Rivers, encompasses several large (1 to 2%) over baseline demand in 2010 and by metropolitan areas, dispersed manufacturing, major 21 to 41 billion kWh (3 to 5%) over baseline port facilities, important timber reserves, heavily demand in 2055. used recreational areas, and diverse ecosystems. 252 California Much of the region's economic and social importance is derived from its water resources. Over 40% of California's total surface water runoff drains from the Central Valley Basin into the San 120° Francisco Bay area (Miller and Hyslop, 1983). The basin supplies water for irrigated agricultural, municipal, and industrial uses, and for a host of 124° other resources and activities. 42° 42° The Central Valley Basin encompasses approximately 40% of California's land area (Figure RED BLUFF 40° 40° 14-1). Elevations range from just below sea level 124° on leveed islands in the Sacramento-San Joaquin River Delta to peaks of over 4,200 meters (14,000 SACRAMENTO feet) in the Sierra Nevada (Figures 14-2 and 14-3). CENTRAL VALLEY DRAINAGE BASIN 38° 38° Mountains ring most of the basin: the Sierra Nevada along the eastern side and the Coast Ranges on the west. The only outlet to the Pacific Ocean is FRESNO 116° via the San Francisco Bay estuary (Figure 14-2). N 36° 36° 122° Current Climate CENTRAL VALLEY 0 40 80 Miles California's climate is characterized by little, if 34° 0 50 100 Kilometers * LOS ANGELES 34° BLYTHE any, summer precipitation and by generally wet GISS 120° winters (Major, 1977). Both temperature and GFDL OSU precipitation vary with elevation and latitude in the 116° Central Valley Basin. Extremes in mean annual precipitation range from about 15 centimeters (6 inches) in the southern San Joaquin River Basin to about 190 centimeters (75 inches) in the mountains of the Sacramento River Basin. While almost all Figure 14-1. The Central Valley (shaded) and valley floor precipitation falls as rain, winter Central Valley Drainage Basin of California. precipitation in the high mountains often falls as Symbols refer to locations of general circulation snow. Storage of water in the snowpack controls model (GCM) gridpoints. (See California Regional the seasonal timing of runoff in the Central Valley Climate Scenarios section of this chapter for details rivers and has shaped the evolution of strategies for on GCMs). water management and flood protection. Under current climatic conditions, peak runoff occurs between February and May for individual rivers 70% of its population and 80% of its total demand within the Central Valley Basin (California for water lie to the south (California Department of Department of Water Resources, 1983; Gleick, Water Resources, 1985). In addition, about 85% of 1987b). the Central Valley Basin's total annual precipitation occurs between November and April, whereas peak Water Resources water use occurs during the summer. In working to solve these water distribution Water Distribution problems, the U.S. Government and California have built two of the largest and most elaborate water California's water resources are poorly development projects in the world: the Federal distributed relative to human settlement patterns in Central Valley Project (CVP) and the California the state. Over two-thirds of the state's surface State Water Project (SWP). Both are essentially water supply originates north of Sacramento, and designed to move water from water-rich northern 253 Chapter 14 N Scaramento 50 28 36 SUISUN BAY SAN PABLO BAY Carquinez Strait 14 CENTRAL BAY Delta 0 Golden Gate Pumping Oakland Plant Tracy Pumping Plant San Francisco SOUTH BAY PACIFIC OCEAN Figure 14-2. The San Francisco Bay estuary and locations of the freshwater pumping plants in the delta. The numbered bars indicate distance (in miles) from the Golden Gate. The dotted line indicates the maximum area affected by a 100-year high tide with a 1-meter (40-inch) sea level rise. 254 California o SACRAMENTO N RIVER SEA LEVEL SA N o STOCKTON D LEGEND Pumping Delta Waterways Plant Above Sea Level Sea Level to 10 Feet Pumping Plant 10 to 15 Feet Pumping SKY - 15 Feet and Deeper Plant 2 0 2 4 6 Scale In Miles Figure 14-3. The Sacramento-San Joaquin River Delta. Shaded areas indicate land below sea level. See Figure 14-2 for location of the delta in the San Francisco Bay estuary. 255 Chapter 14 California to the water-poor south, and to supply Commerce water for agricultural, municipal, and industrial purposes. Currently, the CVP has a water surplus The San Francisco Bay estuary includes the and the SWP has a shortage, especially in largest bay on the California coast (see Figure 14- relationship to users' projected requirements. Thus, 2). The bay's northern reach between the Golden the SWP is particularly susceptible to dry years. Gate and the Sacramento-San Joaquin River Delta is a brackish estuary dominated by seasonally Flood Control and Hydroelectric Power varying river inflow (Conomos et al., 1985). The southern reach between the Golden Gate and the Another objective of the CVP and SWP is southern terminus of the bay is a tidally oscillating flood control. By 1984, CVP facilities had lagoon-type estuary. The port facilities of the San prevented almost $500 million in flood damages Francisco Bay area are vital to California's internal (U.S. Bureau of Reclamation, 1985). Flood control, trade, to Pacific coast commerce, and to foreign however, comes at the expense of water storage trade, particularly with Asian countries. The ports (and hence water deliveries), because reservoir of Oakland and San Francisco, combined, ranked levels must be kept low to absorb high riverflows fourth in the United States in tonnage of during the rainy season. containerized cargo handled in 1983 (U.S. Maritime Administration, 1985). These facilities and Hydroelectric power generation is also an operations are sensitive, in varying degrees, to both objective of the CVP and SWP, and surplus power sea level change and fluctuation in freshwater is sold to utility companies. CVP powerplants runoff. produce an average of 5.5 to 6 billion kWh per year. In 1976 and 1977, precipitation was 35 and 55% Agriculture below normal, respectively, and hydroelectric power generation fell to 50 and 40%, respectively, of target production. California annually produces about 10% of the cash farm receipts in the United States and Sacramento-San Joaquin River Delta produced $14.5 billion in farm income in 1986 (U.S. Department of Agriculture, 1987). Central Valley The delta at the confluence of the Sacramento farms make up significant proportions of total U.S. and San Joaquin Rivers is the focal point of major production of many crops, including cotton, apricots, water-related issues in California (Figure 14-3). For grapes, almonds, tomatoes, and lettuce. example, most islands in the delta lie below sea level and are protected by levees, some of which are Agriculture, the primary land use and the made of peat and are relatively fragile. These largest consumer of water in the Central Valley islands would be vulnerable to inundation from Basin, accounts for 87% of total net water use in rising sea level associated with climate warming. the region. Furthermore, the region accounts for 72% of total net water use for the entire state and The deep peat soils on these islands support highly productive agriculture that would be lost if almost 80% of net agricultural use (California inundated. Department of Water Resources, 1987a). In addition to agricultural importance, the Forestry delta is also the source of all CVP and SWP water exports to points farther south, and in this regard Silviculture is extensively practiced in basically functions as a transfer point of water from California's mountains. The nine national forests the north to the south. The freshwater pumping substantially within the Central Valley Basin plants (see Figure 14-3) in the delta are the largest recorded over $88.6 million in timber sales in fiscal freshwater diversions in California (Sudman, 1987). year 1986 (U.S. Department of the Interior, 1986). Delta outflow must be maintained at a required Forest productivity is sensitive to climate variation. level to prevent saltwater intrusion into the pumping For example, the drought of 1976-77 contributed to plants. The volume of water released from significant tree mortality because of large upstream reservoirs to achieve this level is known as infestations of bark beetles (California Division of carriage water. Forestry and Fire Protection, 1988). 256 California Natural Vegetation freshwater areas for breeding), and the delta is an important nursery for these species. Chinook Approximately one-fourth of all the threatened salmon also constitute an important commercial fish and endangered plants in the United States are species, and Central Valley rivers support about found in California. About 460 species, or about 75% of California's chinook salmon catch, valued at 9% of the California species listed by Munz and $13.4 million at 1981 prices. The populations of Keck (1959), are either extinct or in danger of these species are affected by water quality in the becoming extinct. estuary. California contains about 5,060 native vascular To protect aquatic organisms in the delta, the plant species; of these, about 30% occur only in State Water Resources Control Board (SWRCB) California (Munz and Keck, 1959; Raven, 1977). adopted water right Decision 1485 in 1978 that sets These species are more numerous than those water quality standards to protect the delta and present in the entire central and northeastern Suisun Marsh. The standards vary from year to United States and adjacent Canada, a region about year, with less stringent requirements in dry years. eight times larger than California (Fernald, 1950). The standards are achieved by meeting minimum delta outflow requirements. If delta outflow falls Within the Central Valley Basin, terrestrial below the required level, then releases from vegetation may be grouped into the following broad upstream state and federal reservoirs must be classes, listed according to decreasing elevation: increased so that the outflow requirement is met. alpine, subalpine forest, montane forest, mixed The water quality standards take precedence over evergreen forest, chaparral and oak woodland, and water export from the delta. valley grassland (Barbour and Major, 1977). Recreation and Nature Preservation Wetlands Recreation and nature preservation are The San Francisco Bay estuary includes important in California. Major recreational areas in approximately 90% of the salt marsh area in the Central Valley Basin include four national parks California (Macdonald, 1977). Nichols and Wright (Lassen Volcanic, Sequoia, Kings Canyon, and (1971) documented a 60% reduction in San Yosemite) and nine national forests that lie either Francisco Bay marsh between 1850 and 1968. This completely or largely within its boundaries. Two reduction was largely the result of reclamation for national recreation areas and 13 designated wildlife salt ponds, agriculture, expanding urbanization, refuges and management areas also are situated in shipping facilities, and marinas. Further loss of the region. Downhill skiing and other winter sports wetlands could result in substantial ecological and are economically important in the state. Water economic losses for the region. For example, the projects throughout the Central Valley Basin managed wetlands north of Suisun Bay support a provide significant recreational opportunities. hunting and fishing industry producing over $150 million annually (Meyer, 1987). Tourism, rare and endangered species, and heritage values also could PREVIOUS CLIMATE CHANGE be harmed. STUDIES Wildlife and Fisheries Two of the few studies previously undertaken to assess the potential effects of climate change on The San Francisco Bay estuary provides vital the region are discussed in this section. habitat for many bird and fish species (California Department of Water Resources, 1983). The Forests estuary is an important wintering area for waterfowl of the Pacific flyway. Important sport fish include Leverenz and Lev (1987) estimated the striped bass, chinook salmon, sturgeon, American potential range changes, caused by CO2-induced shad, and steelhead rainbow trout. These species climate change, for six major commercial tree are anadromous (i.e., saltwater species that enter species in the western United States. Two of the 257 Chapter 14 species, ponderosa pine and Douglas-fir, have Water is a key limiting resource in both significant populations in California. Leverenz and managed and unmanaged ecosystems in the Central Lev based their estimates of range changes on the Valley Basin, and freshwater is important in species' response to increased temperature, estuarine ecosystems in the delta region. decreased water balance, and higher CO₂ Consequently, the California studies were organized concentrations. The scenario of climate change so that the impacts of climate warming on the entire used was based on a simulation using the hydrologic system could be examined, starting at Geophysical Fluid Dynamics Laboratory (GFDL) subalpine lakes in the mountains surrounding the model (a different run from that used for this valley and finishing at the freshwater outflow into study), with CO2 concentrations double their the delta region and estuary (Figure 14-4). The present levels. Their results suggest that in individual projects examined the potential impacts California, ponderosa pine could increase in range of climate change and sea level rise on particular and abundance because of its ability to withstand ecosystems and water-delivery systems in the long summer drought. Douglas-fir could be Central Valley (see Chapter 4: Methodology). One eliminated from coastal lowlands in California but of the major goals of this regional study was to might occur in coastal areas at higher elevations. determine how much runoff would flow into the Central Valley from the surrounding mountains Water Resources under different scenarios of climate change, how much of that runoff would be available for delivery Gleick (1987a,b) applied 18 general circulation to the water users in the state, and how much would model (GCM)-based and hypothetical scenarios of reach the delta. climate change to a hydrologic model of the Sacramento River Basin. He used a two-part water Analyses Performed for This Study balance model to estimate monthly runoff and soil moisture changes in the basin. His results suggest The following analyses were performed for this that winter runoff could increase substantially, and study. summer runoff might decrease under most of the scenarios. Summer soil-moisture levels might also Interpretation of Hydrologic Effects of decrease substantially. These changes are driven by Climate Change in the Sacramento-San higher temperatures, which decrease the amount of Joaquin River Basin - Lettenmaier and winter precipitation falling as snow and cause an Gan, University of Washington, and Dawdy, earlier and faster melting of the snowpack that does consultant (Volume A) form. The Lettenmaier et al. project is the first of a series of four projects designed to determine the CALIFORNIA STUDIES IN THIS impact of climate change on runoff and water REPORT deliveries within the Central Valley Basin (Figures 14-4 and 14-5). Their project was designed to estimate changes in runoff from the mountains to Seven studies were completed as part of this the water resource system in the floor of the valley. regional study of the possible impacts of climate Lettenmaier et al. used data from climate scenarios warming on California (Figure 14-4). These studies supplied by EPA as input to their modeling studies. were quantitatively integrated as much as possible (See Chapter 4: Methodology, and the following within the overall timeframe of this report to section, California Regional Climate Change Congress to obtain as complete a picture of those Scenarios). impacts as possible. Also, several of the national studies have results pertaining to California. At the Methods for Evaluating the Potential outset, it should be emphasized that most of these Impacts of Global Climate Change: Case studies used existing models, and most evaluated Studies of the Water Supply Systems of the potential climate change in terms of present State of California and Atlanta, Georgia demands, values, and conditions (including the Sheer and Randall, Water Resources current population and water delivery system). Management, Inc. (Volume A) 258 California Hydrologic Water Quality CLIMATE Impacts of Subalpine Lakes SCENARIOS (Lettenmaier et al.) (Byron et al.) CALIFORNIA TERRESTRIAL Effects on AGRICULTURAL VEGETATION Water Deliveries EFFECTS EFFECTS (Sheer and Randall) (Dudek) (Davis) National Studies: Salinity Effects in San Francisco Agriculture Bay Estuary Air Quality (Williams) Electricity SEA LEVEL RISE SCENARIOS Sensitivity of San Francisco Bay Wetlands (Josselyn and Callaway) Figure 14-4. Organization of the study, showing paths of data input from scenarios and between projects (solid lines). Dashed lines indicate some important linakges between projects that were not quantitatively made in this study. Sheer and Randall used the projected runoff level rise effects in the delta (see Figure 14-3). from the mountains determined by Lettenmaier et al. to simulate the response of the Central Valley Ecological Effects of Global Climate and State Water Projects to climate change. Output Change: Wetland Resources of San from this study includes estimated total water Francisco Bay Josselyn and Callaway, San deliveries to State Water Project users. Francisco State University (Volume E) The Impacts of Climate Change on the Josselyn and Callaway used results from Salinity of San Francisco Bay - Williams, Williams and Park (see Chapter 7: Sea Level Rise) Philip Williams and Associates (Volume A) to assess the impact of changing salinity and sea level rise on the wetlands within San Francisco Bay. The main goal of Williams' project was to determine the impact of sea level rise and changing Climate Change Impacts upon Agriculture freshwater outflow into the delta on salinity within and Resources: A Case Study of California the bay. Williams also determined how much - Dudek, Environmental Defense Fund carriage water might be required to hold back (Volume C) salinity intrusions from the delta pumping plants after sea level rise. The new carriage water Dudek simulated the impact of changing requirements were then factored into Sheer and climate on California agriculture. Besides using the Randall's simulation of the water resource system, climate data from the different climate scenarios to and they represent an important feedback between estimate crop productivity impacts, Dudek used the hydrologic effects of climate change and sea estimates of mean annual water deliveries for 259 Chapter 14 Ancient Analogs for Greenhouse Warming of Central California - Davis, University of Arizona (Volume D) McCLOUD RIVER BASIN Davis reconstructed the vegetation present in CASTLE LAKE the Sierra Nevada during warm analog periods of THOMES the Holocene to estimate the potential impact of CREEK BASIN NORTH FORK warming on the present-day vegetation in these AMERICAN RIVER Secramento River BASIN mountains (see Figure 14-5). National Studies That Included Results for FOSSIL POLLEN California SITES MERCED RIVER BASIN The Economic Effects of Climate Change on U.S. Agriculture: A Preliminary N Assessment - Adams and Glyer, Oregon State University, and McCarl, Texas A&M CENTRAL VALLEY University (Volume C) DRAINAGE BASIN 0 40 80 Miles 0 50 100 Kilometers Adams et al. conducted a national study of agriculture to estimate shifts in land and water use. Results pertaining to California are discussed in this chapter. The Potential Impacts of Climate Change Figure 14-5. The Central Valley Drainage Basin of on Electric Utilities: Regional and National California. Shaded areas refer to the four study Estimates - Linder and Inglis, ICF, Inc. catchments used by Lettenmaier et al. Dots (Volume H) indicate the positions of the Castle Lake study site (Byron et al., Volume E) and the five fossil pollen As part of a national study, Linder and Inglis sites (Davis, Volume D). estimated future California electrical demands in response to climate change. deliveries for irrigation under the different climate Examination of the Sensitivity of a Regional scenarios as input to a regional economic model to Oxidant Model to Climate Variations - estimate shifts in land and water use. This Morris, Gery, Liu, Moore, Daly and information was qualitatively used to compare Greenfield, Systems Applications, Inc. available future water supplies and future water (Volume F) demand (see Figure 14-4). The ability of water policy changes to compensate for climate impacts Morris et al. describe possible interactions of was also evaluated. climate change and air pollution. Results pertaining to California are discussed in this chapter. The Effects of Global Climate Change on Water Quality of Mountain Lakes and Streams - Byron, Jassby, and Goldman, CALIFORNIA REGIONAL University of California at Davis (Volume E) CLIMATE CHANGE SCENARIOS Byron et al. studied the impact of climate Results from two GCM gridpoints were used change on the water quality of a subalpine lake in to drive the effects models used in most of the northern California (see Figure 14-5). 260 California California studies. (For a discussion of how the remains virtually unchanged in the GFDL and OSU scenarios were developed and applied, see Chapter scenarios. Seasonal changes are more varied. For 4: Methodology.) Both gridpoints lie at 120°W, with instance, spring rainfall in GFDL is 0.35 millimeters the northern gridpoint near the Oregon-California per day (0.41 inches per month) lower, while spring border and the southern gridpoint south of rainfall in the OSU and GISS scenarios is higher. Sacramento (see Figure 14-1). Average The scenarios also show a large difference in fall temperature and precipitation changes for both precipitation (Figure 14-6). gridpoints are displayed in Figure 14-6. Generally large seasonal increases in mean temperature are Overall, the OSU scenario represents a smaller projected by the models. Winter temperatures are change from the present climate, and GFDL and between 1.7°C (OSU) and 4.9°C (GISS) warmer, GISS show larger temperature changes. The GISS and summer temperatures are between 2.6°C scenario has higher precipitation than the other two (OSU) and 4.8°C (GFDL) warmer. The OSU scenarios. Generally, temperature increases are model generally projects less warming than the larger in the northern gridpoints than in the other two GCM models. southern gridpoints. Changes in annual precipitation are greater in the north in GISS and Annual precipitation increases in GISS by 0.28 show little regional difference for the other models. millimeters per day (4.02 inches per year) and A. Temperature B. Precipitation 6 0.6 GISS 0.5 5 0.4 GFDL 0.3 OSU 4 0.2 CHANGE (°C) CHANGE (mm/Day) 0.1 3 0 -0.1 2 -0.2 -0.3 1 -0.4 -0.5 0 Winter Spring Summer Fall Annual -0.6 Winter Spring Summer Fall Annual Figure 14-6. General circulation model (GCM) scenario results showing seasonal and annual (A) temperature and (B) precipitation changes between GCM model runs at doubled CO₂ and current CO₂ concentrations. The values are averages of the two gridpoints used by the water resource modelers. (See Figure 14-1 for the location of the gridpoints.) 261 Chapter 14 RESULTS OF THE CALIFORNIA Limitations STUDIES Results would be different if geographic and temporal variability were not held constant within Hydrology of Catchments in the Central each grid. Several assumptions made in this study Valley Basin are important considerations in terms of limitations of the results. The intensity of rainfall is the same. Fewer rainfall events of higher intensity could Changes in mountain snowpack and runoff increase runoff relatively more than a greater could have a major impact on water supply and number of rainfall events of lower intensity. One quality in the Central Valley Basin. Lettenmaier et implicit assumption is that no long-term changes in al. used a hydrologic modeling approach to simulate runoff under different climate scenarios; these vegetation cover and composition would occur, when in fact such changes are virtually certain (but estimates then served as input to the simulation of their hydrologic manifestations are difficult to the Central Valley Basin water resource system predict). If vegetation cover decreases, runoff could response to climate change (Sheer and Randall, increase, since less precipitation would be used by Volume A). plants. Study Design Lettenmaier et al. assumed that the flows into the water resource system were adequately The approach taken was to model the hydrologic response of four representative medium- estimated from the study catchment flows using their statistical model. One limitation of this model sized catchments in the Central Valley Basin. Then was that the study catchments are at high elevations streamflows for 13 larger subbasins in the Central and their runoff is strongly affected by changes in Valley Basin were estimated using the results from the four catchments. The four catchments chosen snowfall, whereas some of the areas contributing runoff to the water resource system are at lower (see Figure 14-5) for modeling range in size from 526 to 927 square kilometers (203 to 358 square elevations with runoff driven primarily by rainfall under present climatic conditions. Since the miles). Outflows for each basin were determined using two hydrologic models that estimate snow principal change under the scenarios was a change in snowfall accumulation patterns, the statistical accumulation, ablation, and daily runoff. The model was biased toward these effects and may have models were calibrated using a subset of the historic somewhat overestimated the total effect of snowfall record and were verified using an independent subset of the data. change on the water resource system. However, because basins at lower elevations have a relatively small impact on the total hydrology, this bias Lettenmaier et al. developed an additional minimally affected the results. climate scenario besides those specified by EPA to test the sensitivity of their results to changes in the Despite these limitations, the results from this scenarios. The scenario they developed included study are qualitatively robust. Any improvement in only the GISS doubled CO₂ temperature estimates; the hydrologic modeling probably would not alter precipitation was kept unchanged from the current the general nature of the results, although their values. The purpose of this scenario was to precision probably would increase. determine the sensitivity of runoff to temperature changes alone. Results To provide input for the water resource Total annual runoff from the four subbasins simulation model of Sheer and Randall (Volume would remain about the same or increase slightly A), Lettenmaier et al. developed a statistical model that relates historic flows in the four study under the doubled CO₂ scenarios, but major changes occur in the seasonality of the runoff. catchments to historic flows in 13 larger subbasins Runoff could be higher in the winter months than it in the Central Valley Basin. This statistical model was then used to estimate flows in the 13 subbasins is today, because less of the precipitation would fall as snow and the snowpack could melt earlier under the different climate scenarios. (Figure 14-7A). As a consequence of higher early winter snowmelt, spring and summer runoff would 262 California A. Merced B. Merced 150 150 Base Base GFDL 2xCO2 GISS Temp. Only 120 GISS 2xCO2 120 GISS 2xCO2 OSU 2xCO2 1930 Analog Flow (x1000 Acre-Feet) 90 60 Flow (x1000 Acre-Feet) 90 60 30 30 0 0 Oct Dec Feb Apr Jun Aug Oct Oct Dec Feb Apr Jun Aug Oct Month Month Figure 14-7. Mean monthly streamflows under difference climate scenarios for the Merced River Basin, one of the the four study catchments modeled (see Figure 14-5 for locations of the study catchments): (A) results from the three doubled CO₂ scenarios; and (B) results from the scenario incorporating only the temperature change projected in the GISS model run, and from the 1930s analog scenario (Lettenmaier et al., Volume A). substantially decrease under these scenarios. The base case (Figure 14-7B). The reason for this variability of the runoff could substantially increase difference is that the 1930s drought was mainly in the winter months. Winter soil moisture could caused by a reduction in precipitation, rather than increase; evapotranspiration could increase in the by an increase in temperature. spring; and late spring, summer, and fall soil moisture could decrease. A major shift in the These results are consistent with those of seasonality of runoff could occur in 50 to 75 years, Gleick (1987b), in that higher temperatures cause a according to the transient scenario GISS A. major change in the seasonality of runoff. Since two different modeling approaches using many climate When only temperature changes were change scenarios produced similar results, these incorporated into the climate scenario and results can be viewed as relatively robust. precipitation was held equal to the base case, total annual runoff was estimated to be lower in all four Implications catchments than in the scenario in which both temperature and precipitation were changed (Figure The potential change in seasonality of runoff 14-7). However, the seasonal shift in runoff, which could have significant implications for stream is the dominant effect of a general warming, would ecosystems and the water resource system in the be similar. Central Valley Basin. Reduction in streamflows in the late spring and summer could negatively affect The scenario producing results that differed aquatic organisms simply because of decreased the most from the other scenarios was the 1930s water volume. Wildlife using streams for food and analog. In this case, runoff was estimated to be water also could be harmed. Water quality lower in most months in the four subbasins, but the probably could be degraded because pollutants seasonal distribution of runoff was similar to the would become more concentrated in the streams as 263 Chapter 14 flows decrease. The possible impacts on the water determine the effects of doubling the carriage water resource system are discussed in the next section. requirement on water deliveries. Both simulations were run with a monthly time step, with water The decrease in spring, summer, and fall soil deliveries summarized on a yearly basis. moisture could have a strong impact on the Interannual variation was used as an indicator of vegetation in the basin, with plants adapted to drier delivery reliability. conditions becoming more abundant at the expense of plants adapted to higher moisture conditions. Sheer held a meeting with representatives of These potential vegetation changes also could affect the California Department of Water Resources and wildlife, and perhaps water quality, through changes the U.S. Bureau of Reclamation to discuss the in the nutrient composition of upland runoff and results of his analyses and to obtain their responses changes in erosion rates. on how the water resource system would handle the changes in runoff. Water Resources in the Central Valley Basin Limitations Changes in runoff under the different climate The limitations to Lettenmaier's study carry scenarios could have a major impact on water over to this one. Thus, interpretation of the results resources in the Central Valley. The study by Sheer of the simulation of the water resource system's and Randall (Volume A) used estimates from response to climate change should focus on how the Lettenmaier et al. of streamflows into the Central system deals with the change in seasonality of Valley to simulate how the water resource system runoff, rather than on the absolute values of the would perform under the various climate scenarios. model output. Also, the model was run using 1990 Particular emphasis was given to how water conditions, and changes in future management deliveries to users would be affected by climate practices, operating rules, physical facilities, water change. marketing, agriculture, and demand were not considered in the simulation. Study Design Results To estimate the climate scenarios' impact on water deliveries, Sheer and Randall used an existing The simulation results suggest that both the model of the California water resource system amount and reliability of water deliveries could currently used by the southern California decrease after global warming. The decreases in Metropolitan Water District (MWD) (Sheer and mean annual SWP deliveries were estimated to Baeck, 1987). The model emulates the State of range from 7% (OSU) to 14% (GISS) to 16% California's Department of Water Resources (GFDL) (200,000 to 400,000 acre-feet) (Figure 14- Planning Simulation Model (California Department 8). In some years, the decreases would be over of Water Resources, 1986). The model used 20% for all three doubled CO2 scenarios. The hydrologic inputs to project water-use demands, projected decrease in water deliveries occurs despite instream and delta outflow requirements, and a slight increase in precipitation over current levels reservoir operating policies. Water requirements in the climate scenarios and greater total outflow were set at levels projected for 1990. from the delta. Deliveries to the CVP are not reduced under the scenarios. Average monthly Two different sets of runs were made with the outflow from the delta increases in the late fall and model. The first involved running the model for winter under the climate scenarios and is lower in the different climate scenarios using current carriage the spring (Figure 14-9). In comparison, the state water requirements. Williams (see the following estimates that population growth and other factors section of this chapter, Salinity in San Francisco will increase demand for SWP deliveries by 1.4 Bay) determined that in response to rising sea level million acre-feet by 2010 (California DWR, 1983). and levee failure, carriage water might have to be doubled to maintain the water quality at the delta The driving factor behind this decrease is the pumping plants (see Figure 14-2). Consequently, change in seasonality of runoff. Higher winter Sheer and Randall ran the model a second time to temperatures could lead to more of the winter 264 California 100 120 0 Base GISS GFDL 100 OSU -100 80 4 KAF -200 Average Flow (CFS) (Thousands) 60 -300 40 20 -400 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep -500 Month GISS GFDL OSU Figure 14-8. Mean annual change in SWP deliveries Figure 14-9. Projected monthly delta outflows (base case minus scenario). KAF = thousands of under different general circulation model climate acre-feet (Sheer and Randall, Volume A). scenarios (adapted from Sheer and Randall, Volume A). precipitation in the mountains falling as rain rather The consensus of the meeting of the than snow, and also to an earlier melt of the representatives from the state DWR and the Bureau snowpack. Consequently, more water would flow of Reclamation concerning the potential changes in into the system during the winter, and less during seasonality of runoff was that the magnitude of this the spring and summer. Given current operating change would be such that operational changes rules and storage capacity, much of the higher alone would not markedly improve the system's winter runoff would be spilled from the reservoirs to performance. One factor limiting the potential for maintain enough storage capacity to capture heavy adjusting the system to the projected changes is the runoff later in the rainy season and thus prevent likely need to provide for additional flood control downstream flooding. When the threat of floods storage during the winter months because of higher decreases at the end of the rainy season in the peak flows. spring and the reservoirs could be filled, runoff into the system would be reduced because of the smaller Implications snowpack. Thus, total storage would be lower at the end of spring and water deliveries would be Under the three doubled CO2 climate lower during the dry summer months. With system scenarios, water deliveries would be less than the changes, the extra runoff could be stored. The shift base case and could fall short of 1990 requirements. in the seasonality of runoff and the response of the Moreover, if carriage water requirements are water resource system to that shift determine the doubled, shortages during a prolonged drought changes in monthly delta outflow (Figure 14-9). could become more significant. In comparison to these projected changes, the severe drought of 1977 Doubling the carriage water requirement in reduced water deliveries by over 50% from the the model run for the GFDL scenario would only previous year. This decrease is over three times minimally affect SWP deliveries. This is because greater than those projected by Sheer and Randall. the base period (1951-80) does not include a lengthy However, their study produced estimates of average drought period, during which the doubled carriage changes, while the 1977 value reflects an extreme water requirement could have a substantial impact event over a short time period, which would have on deliveries. to be dealt with less frequently and in a potentially 265 Chapter 14 different manner than a more persistent shortfall in would affect the shape of the bay by establishing the average supply. Also, Sheer and Randall did not elevation/area and elevation/volume relationships consider future increases in water requirements for all areas below +3 meters (+10.0 feet) caused by population increases and changes in the according to National Geodetic Vertical Datum state's economy, which would exacerbate the (NGVD). In the second part of the study, the bay's projected water shortages. For instance, users and tidal exchange characteristics were determined for managers project a 55% (1.3 million acre-feet) its future shape by using a tidal hydrodynamic increase in water required by SWP users in 2010 model (Fischer, 1970). over the amount the system can reliably supply to them today (California Department of Water Finally in the third part of Williams' study, the Resources, 1983). bay's salinity under the combined impacts of sea level rise and changing delta outflows was calculated The potential decrease in water deliveries using a mixing model developed by Denton and could affect urban, agricultural, and industrial water Hunt (1986). This model was first run with nine users in the state. How the potential decrease different constant delta outflows (all months the should be managed has many policy implications, same) to establish new carriage water requirements which are discussed at the end of this chapter. after sea level rise. (These requirements will also meet the state water quality standards for Suisun On a positive note, the increase in delta Marsh, as detailed in Water Rights Decision 1485.) outflow shows that more water could flow through Once these were established, and Sheer and Randall the Central Valley Basin under these scenarios, and (Volume A) had run their simulation model with water deliveries could be increased if major new the new requirements, the mixing model was run storage facilities were constructed. However, this again to determine the salinity regime in the estuary would be an environmentally and politically after climate change. Included in the model output controversial option (see Policy Implications section were average monthly and average annual salinities of this chapter). in different parts of the estuary under the different scenarios. Salinity in San Francisco Bay Limitations Climate change could affect the San Francisco Bay estuary in two ways: first, changes in Because of the short time available for precipitation and temperature could affect the analysis, Williams used some old and inaccurate amount of freshwater runoff that will flow into the surveys in the morphometric analysis instead of bay; and second, global warming could cause sea making new surveys. These could produce errors of level to rise because of thermal expansion of the plus or minus 20% in the estimates of the estuary's water and glacial melting, which could in turn affect volume. In addition, some levees probably would be a wide range of physical characteristics in the bay. maintained under any delta management plan, and The major objective of the study by Williams thus the flooding of the delta islands would not be (Volume A) was to estimate the implications of as extensive as assumed in the levee failure scenario. global warming and rising sea level on the size and Williams did not consider changes in siltation and shape (morphometry) of the San Francisco Bay erosion of sediments that would likely occur under estuary and on salinity in the estuary. the different climate change scenarios. However, erosion would probably have a significant impact on Study Design water flow in the delta. For instance, deepening of the tidal channels in the delta could lead to Williams' project was conducted in three parts, intrusion of salinity farther upstream than projected using two sea level rise scenarios and delta outflows in this study. In addition, more sophisticated estimated by Sheer and Randall (Volume A). The models of salinity and tidal ranges and exchanges sea level rise scenarios are a 1-meter (40-inch) rise might improve the accuracy of the results. Finally, with the levees in the Sacramento-San Joaquin the new carriage water requirements were based on Delta and San Francisco Bay maintained, and a 1- a steady-state analysis (e.g., constant delta outflows). meter sea level rise with levee failure. The first part Changes in the hydraulics of the Sacramento-San of this study involved estimating how sea level rise Joaquin Delta and Suisun Bay with sea level rise 266 California could increase these requirements. Williams' results should be viewed as a preliminary estimate of estuarine changes, with emphasis placed on the direction of change, rather than on the absolute amount of change. Results The morphometric analyses suggested that given a 1-meter (40-inch) sea level rise and failure of the levees, the total area of the estuary might Suisun Bay triple, and its volume could double. If the levees are maintained, the increases in area and volume could be about 30 and 15%, respectively. The amount of sea level rise would be less important to OSU the physical size of the bay than whether or not the GFDL Carquinez GISS levees are maintained. Strait BASE Under the sea level rise scenarios with levees maintained, tidal ranges would not change significantly from current conditions. If the levees failed, downstream constrictions at Carquinez Strait and to the east of Suisun Bay (see Figure 14-2) would limit tidal transport and reduce tidal range in Figure 14-10. Movement of mean annual salinity the delta, assuming that erosion does not alter the of 10 parts per thousand under different hydrology tidal characteristics of the delta. scenarios. Other salinity levels move similar distances (see Figure 14-2 for location of Suisun The results from the initial application of the Bay; Williams, Volume A). salinity model to constant delta outflows indicate that monthly carriage water requirements might have to be doubled to repel saline water from the for all months as compared with the base case, upper part of the delta. Also, whether or not the except for winter and early spring months in the levees are maintained would have little effect on the GISS scenario. The greatly increased runoff of the salinity regimes in the bay according to the model's GISS scenario (see Figure 14-9) during these results. However, because possible scouring of tidal months kept the salinity at the same level as the channels was not incorporated into the model, the base case. Williams additionally modeled the predicted salinity after levee failure is probably frequency of a given salinity value in any month. In underestimated. June, for example, salinities that were exceeded in 50% of the years in the base case might be Using Sheer and Randall's estimated delta exceeded in 80% of the years in both the GISS and outflow with double carriage water, Williams also OSU scenarios because of the lower outflows estimated annual salinity in the bay. The results predicted under these scenarios. suggest that after a climate warming, a 1-meter sea level rise, and failure of the levees, water of a given Implications average annual salinity could migrate inland between 4 kilometers (2.5 miles) (GISS scenario) Rising sea level could place the delta islands and 9.6 kilometers (6 miles) (OSU scenario) (Figure under increased risk of inundation, not only because 14-10). of higher water levels but also because the larger area and volume of the San Francisco Bay estuary Williams also calculated the average monthly could result in greater wave energy and higher salinity for Suisun Bay for the three climate erosion rates of the levees. Improving the levees scenarios, levee failure, and double carriage water just to protect them against flooding at the current requirements. Monthly salinities would be higher sea level could cost at least $4 billion (California 267 Chapter 14 Department of Water Resources, 1982). With between sedimentation rates required for marsh higher sea levels, the cost of maintaining the levees maintenance and sea level rise rates was examined. would increase. The effects of salinity changes on the distributions and abundances of organisms were related to The large body of water created if all the various freshwater outflow scenarios developed by levees failed would have a longer water residence Sheer and Randall (see Figure 14-9). In the time. This means that any contamination (salt or absence of appropriate quantitative models, biotic other pollutant) would be more difficult to flush out changes in the estuary in response to changing of the delta region. Also, if saline water fills the salinity were qualitatively determined based on islands when levees fail, significant amounts of literature review and expert judgment. freshwater would be needed to flush out the salt. Limitations Increasing salinity could necessitate increases in carriage water to maintain freshwater at the Circulation and sedimentation in the estuary export point in the delta or could require developing could change dramatically as sea level rises and if a different method to convey freshwater from levees fail. The specific characteristics of these reservoirs to users. Assuming the current water biologically important changes are unknown at management system is not expanded, the increase in present and were not considered in this study. The carriage water coupled with the decrease in sea level rise scenarios did not consider the reservoir storage would most likely mean reduction possibilities of sudden changes in sea level. in water deliveries to at least some of the system's Increased water temperature, which may directly users during extended droughts. With higher future affect the reproduction, growth, and survival of water requirements, shortages caused by the higher estuarine organisms, or may have an indirect effect carriage water requirements may not be limited to through changes in oxygen availability, also was not extended droughts. An increase in sea level could considered. Although specific impacts on plant and make navigation easier, temporarily reducing the animal species in the estuary are difficult to assess, need for dredging of navigation channels. On the the general impacts would most likely be similar to other hand, a rising sea level could threaten fixed those reported here. port terminals and piers. Results Wetlands in the San Francisco Bay Estuary Rates of sea level rise from 1990 to 2040 for the three scenarios are presented in Figure 14-11. Climate warming could alter two important Once the rate of sea level rise exceeds the rate of physical factors that affect wetland distribution: sea sediment accretion, tidal marsh habitats would level and freshwater outflow. Major impacts of sea become inundated and erosion of the marsh edge level rise could include erosion and marsh could increase. For the 1-meter rise scenario, the inundation. Changes in freshwater outflow can rate of rise was not estimated to exceed maximum change the distribution and productivity of estuarine accretion rates (7 to 8 millimeters per year) until plants and animals. Josselyn and Callaway (Volume about the year 2040. For the 2- and 3-meter (80- E) estimated the possible effects of climatic and 120-inch) rise scenarios, the rate of sea level warming on deep-water and wetland habitats of the rise could exceed accretion rates after 2010 and San Francisco Bay estuary (see Figure 14-2). 2000, respectively (Figure 14-11). Study Design Peak primary productivity, at present, occurs in early spring in San Pablo Bay and in the summer in Josselyn and Callaway examined the impacts of Suisun Bay. These maximum productivity levels a 1-, 2-, and 3-meter (40-, 80-, and 120-inch) sea could be substantially reduced, particularly for level rise by the year 2100. Of the three scenarios, brackish and freshwater plant species, under the a 1-meter rise by the year 2100 is regarded as the higher salinities of the OSU scenario (see Figure most probable (NRC, 1987). Models were used to 14-10). Peak spring production might also shift estimate rates of sea level rise from 1990 through upstream into the delta if levees fail. However, 2100 under these three scenarios. The relationship under the higher freshwater outflows of the GFDL 268 California 25 25 20 20 3-m rise SEA-LEVEL RISE RATE (mm/yr) 15 15 rise 2-m 10 10 max. sedimentation rate SEDIMENT ACCRETION RATE (mm/yr) 5 1-m rise 5 0 0 1990 2000 2010 2020 2030 2040 YEAR Figure 14-11. Estimated sea level rise at San Francisco for three scenarios by the year 2100 (Josselyn and Callaway, Volume E). and GISS scenarios, the locations of maximum fish (saltwater fish that enter freshwater areas for production levels might remain in their present spawning). Lower outflows could result in declines positions if the levees are maintained. If the levees among these populations (Kjeldson et al., 1981). fail, primary production could increase in the extensive shallow water and mudflat areas created. If levees failed, a large inland lake with fresh to brackish water quality could be created in the Since many areas currently protected by levees delta. Striped bass and shad spawn in essentially are 1 to 2 meters (40 to 80 inches) or more below freshwater conditions and their spawning could be sea level, levee failure would cause them to become reduced under increased salinity, especially if they deepwater areas rather than marshes (see Figure did not move upstream to relatively fresh water. 14-3). Eventually, enough sediment might be Marine fish species could increase in abundance in deposited in these formerly leveed areas to support the Suisun and San Pablo Bays in response to the marsh development. Inundation of marshes and projected higher salinities, and freshwater and salinity impacts on freshwater and brackish-water anadromous species could decrease. plant species could reduce sources of food and cover for waterfowl. Loss of emergent vegetation Implications could significantly reduce the numbers of migratory waterfowl using the managed wetlands along Suisun The loss of wetlands could result in substantial Bay's north shore. ecological and economic losses for the region. For example, the managed wetlands north of Suisun Bay If levees are maintained under conditions of support a hunting and fishing industry valued at sea level rise, salt may build up behind them from over $150 million annually (Meyer, 1987). Tourism, the evaporation of standing water. This salt would hunting, fishing, rare and endangered species, and cause marsh vegetation to die back and reduce the heritage values also could suffer. value of these wetlands to wildlife. California Agriculture Freshwater outflows estimated during springtime under the climate change scenarios (see Figure 14-9) may be too low to support anadromous California's agricultural production is highly dependent on irrigation, which accounts for approximately 80% of the state's net annual water 269 Chapter 14 use. Dudek (Volume C) used existing consideration of the direct effects of CO₂). agroecological models to explore potential responses Generally, the greatest impacts are estimated under of California agriculture to climate change. the hotter GISS scenario. Table 14-1 presents regional yield changes for sugarbeets, corn, cotton, Study Design and tomatoes. These projections were generated by the agricultural productivity model and did not Climate changes from the GISS and GFDL consider economic adjustments or water supply doubled CO2 scenarios were linked to an limitations. Tomatoes might suffer the least agricultural productivity model adapted from damage, with yields reduced by 5 to 16%. Doorenbos and Kassam (1979). Growth responses Sugarbeets could be hardest hit, with declines of 21 to both climate change and climate change plus to 40%. Yield reductions in sugarbeets were direct effects of carbon dioxide were modeled. estimated to be greatest in the relatively hot interior These productivity responses were then introduced southern regions. Differences in growth response into the California Agriculture and Resources between the two climate scenarios are greatest for Model (CARM) (Howitt and Mean, 1985), which corn and least for tomatoes. estimates the economic and market implications of such changes. Mean surface water supplies under Without economic adjustments, corn yields are the base, GISS, and GFDL scenarios, calculated estimated to decline by 14 to 31%, based on the from the simulations of Sheer and Randall (Volume agricultural productivity model under the GISS A), were also used as inputs into CARM. scenario (Table 14-1). With economic adjustments, declines of roughly 15% were estimated, a result at Limitations the lower end of the direct productivity impacts. The CO2 direct effects results should be When the direct effects of CO2 on crop yields viewed as preliminary, since they are based on data were considered, yields of cotton and tomatoes from growth chamber experiments that may poorly generally increased over the 1985 base (Table 14- represent field conditions. This study did not 1). Corn and sugarbeets were generally estimated consider changes in crop varieties, planting dates, to be unable to increase growth in response to energy costs, water-use efficiency, changes in the increases in CO2 concentration, although yield status of groundwater resources under a changed reductions were not as great as with climate change climate, or possible changes in delta agricultural alone (Table 14-1). Cotton could benefit the most acreage caused by flooding after levee failure. Also, from inadvertent CO₂ fertilization, with yields new crop/location combinations were not increasing in most cases by 3 to 41% (although considered, nor were changes in soil quality such as under the GISS scenarios in the Sacramento Valley, increases in salinity. The interaction between they were estimated to decrease by 2%). climate change and direct CO₂ effects on productivity were not examined but may significantly Potential increases in yields in response to limit potential growth increases. The effects of CO₂ fertilization might be achieved only at a cost of climate changes on other agricultural production increased groundwater extraction in many areas. regions in the nation and the rest of the world were For example, when surface water use was projected not considered. These could be major factors in at 100% of capacity, as in the Central Coast regions, determining how California farmers respond to higher water requirements would necessitate climate change. Given these limitations, realistic increased groundwater usage (Figure 14-12). estimates of agricultural responses to climate change However, increased crop yields may offset increased may be difficult to obtain. The results may be more economic costs of water. valuable as indications of sensitivity than as specific impacts. Regionally, across all scenarios (not considering potential changes outside California) the Results largest reductions in crop acreage were projected in the Imperial Valley, while the delta region showed Relative to the 1985 base, yields could be the largest gains in acreage (Figure 14-12). This significantly reduced for California crops in expansion of agriculture in the delta region would response to climate changes alone (i.e., without 270 California Table 14-1. Regional and Statewide Percentage Yield Changes (relative to 1985) Under Different General Circulation Model Climate Scenariosᵃ Crop sugarbeets corn cotton tomatoes Regionᵇ Scenario CC Net CC Net CC Net CC Net South Coast Los Angeles GISS -27 -3 -22 -18 -22 11 -8 17 GFDL -21 5 -3 3 -4 41 -5 20 North Interior Red Bluff GISS -34 -11 -17 -12 -30 3 -16 10 GFDL -26 0 -14 -9 -26 9 -14 12 Sacramento Valley Sacramento GISS -29 -3 -14 -9 -34 -2 -14 13 GFDL -24 3 -8 0 -32 2 -12 15 Southern San Joaquin Fresno GISS -34 -14 -19 -14 -29 6 -15 10 GFDL -32 -13 -13 -7 -26 11 -15 10 Southern Deserts Blythe GISS -40 -2 -31 -27 -28 6 -13 13 GFDL -39 0 -14 -8 -19 21 -12 15 CARM Statewide GISS -31 -8 -15 -10 -29 6 -14 12 GFDL -25 -1 -10 -4 -26 11 -13 13 a Regional changes were projected by the Doorenbos and Kassam agricultural productivity model, while statewide production changes were projected by the California Agriculture and Resources Model (CARM). The latter estimates included economic adjustment. "Net" includes the direct effects of increases in CO₂ and climate change (CC). Refer to Figure 14-12 for locations. Source: Dudek (Volume C). depend on maintenance of levees protecting the of existing nonpoint source pollution and farmland. Without a consideration of CO₂ accelerated rates of groundwater overdraft with fertilization, statewide crop acreage was estimated ensuing environmental impacts. to be reduced by about 4 to 6% from the 1985 base. When CO2 direct effects were considered, statewide Changing water supply requirements may result crop acreage was estimated to be approximately in increased conflicts between water users. In equal with 1985 base levels. addition, shifts in the location of agricultural production could affect the future viability of natural Implications systems. Such shifts could also have a significant impact on the economic health of small agricultural Regional changes in cropping locations and communities. patterns of water use imply potential exacerbation 271 Chapter 14 140 140 Delta Sacramento 120 120 100 100 Resource Use Index 80 Resource Use Index 80 60 60 40 De 20 20, 0 crop acreage groundwater surface water 0 crop acreage groundwater surface water 140 140 Central Coast Northern San Joaquin 20 120 100 100 Resource Use Index 80 Resource Use Index 80 80 60 40 20 20 0 crop acreage groundwater surface water crop acreage groundwater surface water 140 South Coast 120 $ Southern San Joaquin 100 120 Resource Use Index 80 rod 60 40 Resource Jae Index so 80 40 20 20 0 crop acreage groundwater surface water 0 crop acreage groundwater surface water 140 LEGEND Imperial Valley 120 100 GISS Climate Change Resource Use Index 80 GFDL Climate Change 60 GISS Net Effect 40 20 GFDL Net Effect 0 crop acreage groundwater surface water Figure 14-12. Regional crop acreage, groundwater use, and surface water use under different GCM climate scenarios. Net effect includes the direct effects of increases in CO₂ and climate change. The resource use indices represent the ratio (as percentages) of scenario results to the 1985 base period (Dudek, Volume C). 272 California Regional Implications of National Study Design Agriculture Changes Goldman et al. (1989) correlated an index of Adams et al. conducted a national agricultural water quality, primary production (i.e., the amount study that included results relevant to California of biomass produced by algae in the lake) with (Adams et al., Volume C). The results of the study climate variability at Castle Lake. Subsequently, are not directly comparable with the results from Byron et al. (Volume E) were able to develop Dudek's study (discussed above), since Adams et al. empirical models relating primary production with considered national agricultural impacts and various climate parameters. aggregated California into a Pacific region with Limitations Oregon and Washington. Further, the two studies did not examine the same set of crops and modeled productivity differently. (For a description of the Their model was limited to estimating annual study's design and methodology, see Chapter 6: values of primary production; seasonal variability Agriculture.) was not calculated. The model also did not project changes in species composition and nutrient Results dynamics, which could have important consequences for water quality. Changes in upland vegetation and Adams et al. (Volume C) estimated that nutrient cycling, which could also affect the lake's national crop acreage could decline by 2 to 4% in water quality, were not part of the model. response to climate change, but Pacific Coast State acreage could increase by 18 to 20%. This increase The estimates of annual primary production in the Pacific region is attributable to the region's produced by this model are precise, although the extensive use of irrigated agriculture. In contrast, results are general in the sense that no species- most other regions of the United States specific projections are made. predominantly use dryland farming, and crop Results acreage might decline in response to moisture stress. The Adams et al. approach was based on maximizing farmers' profits and indicates that Byron et al. estimate that mean annual primary higher yields associated with direct CO2 effects production could increase under all three doubled might result in further declines in crop acreage (or CO₂ scenarios, with increases ranging from 16% in the case of the Pacific Coast States, a smaller (OSU scenario) to 87% (GISS scenario) (Figure 14- increase), since fewer acres might be required to 13). The OSU results are within one standard error produce the necessary crops. of present production. Thus, under this scenario, there would be no significant decrease in water Water Quality of Subalpine Lakes quality. The increase in annual primary production in the transient scenario was only statistically significant in the last decade of the transient Subalpine lakes are common in the California scenario (2050-59). Primary production in the last mountains, and many of these are the source of decade was estimated to be 25% greater than the streams and rivers flowing down into the lowlands. base case. Changes in the water quality of these lakes could significantly alter their species composition and The increase in annual primary production is nutrient dynamics and also could have an impact on attributed principally to the temperature increase downstream water quality and ecosystems. The projected by the scenarios. The higher sensitivity of California's subalpine lakes to weather temperatures would result in less snow variability and climate change has not been accumulation, which is correlated with an earlier extensively studied. Consequently, Byron et al. melting of the lake ice and a longer growing season. studied how climate controls the water quality of Castle Lake, a subalpine lake in northern California Implications (see Figure 14-5). Higher primary production could result in climatic effects being indirectly felt at higher points 273 Chapter 14 Summary of Effects on Water Resources In terms of economic and social importance, changes in water resources are among the most 800 important possible effects of climate change in Mean 700 California. A wide variety of factors related to 95% Confidence Interval 600 climate change could affect water resources, ranging Primary Production (mg C/m2/yr) from those factors changing water supply to those 500 affecting water requirements. All the individual 400 projects discussed above addressed some aspect of 300 climate impacts on water resources in the state. However, these studies did not consider all the 200 major factors that could affect California water 100 resources in the next century, mainly because of the 0 complexity and inherent difficulties in forecasting Measured Model GFDL GISS OSU future requirements for water. This section discusses other factors that would affect future water demands not directly considered by the individual studies, including future changes in agriculture, population, water-use efficiency, and Figure 14-13. Annual primary production estimates sources of water, including groundwater. for Castle Lake showing actual and model values for present conditions and model values for three Dudek's study used estimates of water GCM climate scenarios (see Figure 14-5 for the deliveries from Sheer and Randall's study, but location of Castle Lake). Solid bars show the 95% changes in agriculture that he determined, and confidence interval for each estimate (Byron et al., hence changes in agricultural demand for water, are Volume E). not factored back into the water simulation model. For instance, Dudek's results indicate that because of climate conditions, crop acreage in the Imperial in the Castle Lake food web and could affect the Valley decreases, freeing water used there for lake's nutrient dynamics. irrigation to be used elsewhere in the state if water institutions permit such transfers. Also, as cropping Extrapolating these results to other subalpine patterns change, so does the pattern of needed lakes suggests their water quality could decrease and water transfers via the water resource system, thus their species composition might change after climate affecting water deliveries. Finally, Dudek found that warming. Increased primary production could groundwater usage can increase when the direct provide additional food for other aquatic organisms, effects of CO2 are included in his model. Estimated such as fish, but could also degrade water quality by groundwater usage is projected to increase when ultimately causing a decrease in dissolved oxygen full use of surficial water sources does not meet and by blocking light filtration to lower levels. agricultural demands estimated in the model. Thus, Fisheries in unproductive lakes may be enhanced, Dudek's results suggest that agricultural demand for although trout populations may suffer in lakes water could exceed surficial supplies after climate where temperatures rise past a threshold value and warming, further exacerbating water shortages. oxygen levels drop too low. Not considered in the overall California study, Changes in production and concomitant but critical to determining the magnitude of changes in nutrient dynamics could affect potential water shortages in the next century, are downstream river and reservoir water quality. population growth and accompanying changes in However, since the streams draining subalpine lakes water demands. Projections of population growth are well oxygenated, the increased biomass entering place the state's population at about 35 million in them would most likely be rapidly decomposed and 2010 as compared with 24 million in 1980, an probably would not affect the water quality of lower increase of 45% (California Department of Water reaches of streams and rivers. Resources, 1983). As mentioned earlier, 274 California requirements for SWP deliveries by urban, California could present a possible analog to a agricultural, and industrial users could increase by warmer future climate. 50% over what the system can reliably supply today. This shortfall by itself is significantly greater than Study Design the decrease in deliveries caused by the climate scenarios as determined by Sheer and Randall. The composition of the vegetation that existed in the central Sierra Nevada over the last 12,000 If water shortages become more common, years was determined using fossil pollen analysis. agricultural, industrial, and residential users will Fossil pollen samples were collected from five lakes probably change their water-use efficiency. Changes situated along an east-west transect (see Figure 14- in efficiency could moderate possible future 5) passing through the major vegetation zones of the shortages. Any change in water pricing or water Sierra Nevada. Dissimilarity values were calculated law also could affect water demand and supply, but between modern and fossil pollen samples to these changes are very difficult to project far into determine the past vegetation at a particular site. the future. Limitations Groundwater usage is discussed by Dudek, but the overall impacts of climate change on The climate estimated in the three doubled groundwater are not addressed in this project. As CO₂ scenarios is different from the climate that demand for water increases beyond the capability of probably existed between 6,000 and 9,000 years ago the water resource system to deliver the needed in the Sierra Nevada, according to Davis's water, mining of groundwater (as Dudek shows for interpretation of the region's vegetation history. agriculture) is one option users could adopt to meet Davis suggests that 9,000 years ago, the climate was their demand. Using groundwater could lessen the drier than it is today. Whether it was warmer or severity of water shortages in the short term but cooler is uncertain. The climate 6,000 years ago presents environmental problems, such as land was not much different from the modern climate. subsidence, over the long term.. Thus, the analog climates are in marked contrast to the warmer climate estimated by all three GCMs In general, given the current water resource for the gridpoint closest to the western slope of the system, qualitative considerations of future changes Sierra Nevada. Also, the models suggest that total in water requirements suggest that future water annual precipitation will not significantly change. shortages could be significantly greater than Consequently, the results of this study do not estimated here for climate change alone. provide an indication of how the present-day vegetation could respond under the climate Vegetation of the Sierra Nevada scenarios constructed from the GCMs. Nevertheless, they do present a possible analog for To better understand the sensitivity of natural how Sierra Nevada vegetation could respond to an vegetation in California to climate change, Davis overall warmer Northern Hemisphere climate that (Volume D) studied changes that have occurred produces a drier but not significantly warmer Sierra over the past 12,000 years in terrestrial vegetation Nevada climate. growing in the California Sierra Nevada. Changes in vegetation that occurred during this period Furthermore, the warming 6,000 to 9,000 years suggest how the vegetation that currently exists in ago occurred over thousands of years, as opposed to the mountains could respond to future climate the potential warming within a century. Thus, the changes. The middle latitudes of the Northern analog does not indicate whether vegetation would Hemisphere are believed to have been warmest (1 be able to migrate and keep up with a relatively to 3°C warmer than today) about 6,000 years ago rapid warming. (Budyko, 1982), and parts of western North America were apparently warmest 9,000 years ago Another constraint associated with using the (Ritchie et al., 1983; Davis et al., 1986). Thus, the past as an analog to trace gas-induced warming is period between 6,000 and 9,000 years ago in that carbon dioxide levels were lower during the past 12,000 years than those projected for the next 275 Chapter 14 century. Higher carbon dioxide concentrations If future forests west of the Sierra crest could partially compensate for adverse effects of become similar to current forests east of the crest, higher temperatures and lower moisture levels on timber production could significantly decline. Based tree growth. The extent of this compensating effect on inventory data from national forests, timberlands is uncertain at this time. Nevertheless, the east of the crest currently support only about 60% possibility exists that the magnitude of the of the wood volume of timberlands west of the crest vegetation change in the past to a warmer (U.S. Forest Service, Portland, Oregon, personal hemispheric climate could have been less if carbon communication, 1988). Different future climates dioxide concentrations had been higher. could also necessitate changes in timber practices (e.g., reforestation techniques). A relatively small set of modern pollen samples was available for comparison to the fossil samples; Vegetation change in response to climate therefore, the precision of the vegetation change could produce additional stress for reconstruction is uncertain. Also, the precision of endangered animal species as their preferred the estimated elevational shifts in the vegetation habitats change. Populations of nonendangered zones is low because of the limited number of fossil wildlife also could be affected as vegetation changes. sites available for the analysis. Nevertheless, this study provides a good general summary of the Since the GCMs estimate a different future vegetation changes in the Sierra Nevada during the climate than the climate reconstructed for the past 12,000 years. analog period, it is important to consider how the vegetation in the Sierra Nevada could respond Results under the GCM-based climate scenarios as compared with the way it responded during the The forests existing in the western Sierra analog period. Recall that the climate in the GCMs Nevada 9,000 years ago resembled those found east is estimated to be significantly warmer than today's of the crest today (Figure 14-14), with lower forest climate, with similar amounts of precipitation, while cover and tree density. Pine and fir densities, in the analog climate was significantly drier with particular, were lower. Between 9,000 and 6,000 similar temperatures. One major difference in the years ago, the vegetation gradually became similar impact of the two types of climate scenarios could to the modern vegetation in the same area, and by be in the response of species at higher elevations in 6,000 years ago the modern vegetation zones were the Sierra Nevada. Since growing season length and established on both sides of the Sierra crest. The warmth are generally considered to control the vegetation 6,000 years ago was subtly different from position of timberline (Wardle, 1974; Daubenmire, that in the area today, with less fir and more sage. 1978), warmer temperatures under the GCM The forests may have been slightly more open than scenarios could be expected to raise the timberline. today. The timberline was not significantly higher during the analog period. Higher temperatures could also Implications increase the elevation of other vegetation zones in the Sierra Nevada. If climate conditions of the Sierra Nevada in the next century become similar to those that Another effect of higher temperatures in the existed 9,000 years ago, major changes could occur GCM scenarios that would probably affect in forest composition and density. The vegetation vegetation at all elevations is a reduction in effective changes could generate significant environmental moisture during the growing season. Lettenmaier et impacts, ranging from changes in evapotranspiration al. (Volume A), in fact, estimate such a decrease as and related hydrogeological feedbacks to changes soil moisture decreases in late spring, summer, and in nutrient cycling and soils, which could degrade fall compared with the base case. Furthermore, for the water quality of mountain streams. Fire lower elevations at least, the growing season could frequency could increase as a function of changes in be effectively shortened because of the earlier onset fuel loads and vegetation. If dead wood rapidly of moisture stress after winter rains. One result of builds up because of the decline in one or more tree this could be the extension of grasslands and species, large catastrophic fires could occur. chaparral higher up the slopes of the Sierra Nevada. 276 California 4000 MODERN 3000 Tioga Pass Pond 0 SA ES 0 Barrett UM Starkweather 0 Exchequer Baisam 0 PF 2000 Sierra Montane 1000 Great Basin Grassland 0 (West) (East) 4000 6K Tioga Pass Pond O SA 3000 ES 0 Barrett UM Starkweather O Exchequer 2000 Baisam 0 PF Sierra Montane 1000 Great Basin Grassland 0 (West) (East) 4000 9K 3000 Tioga Pass Pond 0 ES a Barrett Starkweather 0 Exchequer 2000 Baisam o Pine Forest 1000 Great Basin (?) 0 (West) (East) Elev. (m) Figure 14-14. Vegetation zonation in the central Sierra Nevada at present; 6,000 years (6K) before present; and 9,000 years (9K) before present. (See Figure 14-5 for approximate locations of fossil pollen sites.) The dashed lines indicate uncertainty in the placement of vegetation zone boundaries (Davis, Volume D). SA = subalpine; UM = upper montane; ES = eastern subalpine; and PF = pine forest. Also, reduced moisture availability could alter the Electricity Demand outcome of competition between plant species with different growth forms and longevity, thus changing Electric power demand is sensitive to potential the composition of the vegetation zones. Plant climate change. As part of a national study, Linder species with drought-resistant characteristics would and Inglis estimated California's energy demand for probably increase in relative abundance. One the years 2010 and 2055. (For a description of the possible consequence of this shift in species study's design and methodology, see Chapter 10: abundance is the formation of plant communities Electricity Demand.) that resemble in some aspects plant communities that occurred 9,000 years ago. However, the Results complicating factor of more direct effects of higher temperatures makes such a projection uncertain, as In California, climate change scenarios result does the lack of consideration of the direct effects in only small changes in estimated electrical utility of increasing concentrations of carbon dioxide. generation and costs by the year 2010. Annual 277 Chapter 14 power generation is estimated to increase by 1 to climate warming in central California. The National 2% (over the 345 billion kWh estimated to serve the Ambient Air Quality Standard (NAAQS) for ozone California population and economy in 2010), and is 12 ppm. Morris et al. estimated that the number new generation capacity requirements would be less of August days that exceed this standard could than 1% greater than increases without climate increase by 30%. Furthermore, the area exceeding change. By the year 2055, annual power generation the NAAQS could increase by 1,900 square is estimated to increase by 3% under lower growth kilometers (730 square miles), and the number of of electricity demand (604 billion kWh base) to 5% people exposed to these elevated ozone levels could under higher growth (794 billion kWh base). New increase by over 275,000. generation capacity requirements would be 14 to 20% greater than non-climate-induced needs. Then Implications cumulative investments in new capacity could cost $10 to $27 billion (in 1986 dollars). Trace gas-induced climate change may significantly affect the air's chemistry on local and Implications regional scales. These changes may exacerbate existing air quality problems around California More powerplants may be required. These metropolitan areas and agricultural areas of the would need more cooling water, further depleting Central Valley, causing health problems and crop the water supply. Climate-induced changes in losses. Increases in air pollution may directly affect hydrology may reduce hydropower generation and the composition and productivity of natural and increase dependence on fossil fuels and nuclear managed ecosystems. power. Increased use of fossil fuels may provide positive feedback for the greenhouse effect and may deteriorate local air quality. The increased utility POLICY IMPLICATIONS rates that may be required to pay for new power generation capacity may limit groundwater pumping for agriculture. An overall question applies to resource management in general: What is the most efficient Air Pollution way to manage natural resources? Currently, management is based on governmental jurisdiction with, for example, forests managed at the local, Morris et al. (Volume F) studied possible state, or federal level. Management of hydrologic interactions of climate change and air pollution in systems is also based on governmental jurisdiction. California. They estimated the impacts of climate An alternative would be to manage these systems change on ozone concentrations using a regional using natural boundaries as the criteria for transport model. The values they calculated should determining management jurisdiction. The pros and be viewed as coarse approximations because of the cons of such a management strategy deserve at least limitations in the application of the model. For some preliminary research. instance, the study looked only at changes in temperature and water vapor and kept as Water Supply and Flood Control unchanged many other important meteorological variables. An important unchanged variable was Water supply is the basis for most economic mixing height. Instead of remaining unchanged, development in California. Yet, almost all the mixing height could increase with rising water available in the SWP is allocated for use. A temperatures. This would have a dilution effect on air pollution. (The study's design limitations and major problem is to accommodate rising demand for water, interannual climate fluctuations, and the methodology are discussed in Chapter 11: Air need to export water from northern to southern Quality.) California. Results In addition, the results from these studies Morris et al. estimated that ozone suggest that climate change over the next 100 years concentrations could increase up to 20% during could cause earlier runoff, thus reducing water deliveries below their projected 1990 level. This some days in August in response to a 4°C (7°F) 278 California situation (together with increasing requirements for avoid a significant loss of flood safety would most water caused by increasing population) would create likely bring about little improvement in the system's a set of major policy problems for the water performance under the given climatic scenarios. managers and land-use planners in California. Detailed study of this point is needed, however. Two major policy questions can be raised The second approach to maintain or increase concerning the possible reduction in water water deliveries might be to construct new water deliveries: How can the water resource system be management and storage facilities. However, trends changed to prevent a decrease in water deliveries over the past decade have shifted away from caused by climate change? If water deliveries fall planning large physical facilities (e.g., the Auburn short of demand, how should potential water Dam and Delta Peripheral Canal). Building new shortages be allocated? facilities is expensive and raises serious environmental concerns about such issues as wild Approaches for Modifying the Water Resource and scenic rivers. Another option is to use smaller System facilities, such as the proposed new offstream storage facility south of the delta, and to improve Several possible approaches can be attempted the delta's pumping and conveyance facilities. With to increase water deliveries. First, system the help of these facilities, the SWP plans to achieve management can be modified. For instance, the a 90% firm yield (the amount that can be delivered most recent SWP development plan suggests the in 9 out of 10 years) of about 3.3 maf by 2010 possibility of state management of both SWP and (California Department of Water Resources, 1987a). CVP facilities (California Department of Water Another relatively inexpensive option for off-line Resources, 1987a). Complete joint management storage is artificial recharge of groundwater during could produce more than 1 million acre-feet (maf) wet years. The SWP is currently pursuing a proposal additional reliable yield in the system. Steps toward to deliver surplus water to groundwater recharge greater cooperation have been taken. The areas in the southern Central Valley to provide Coordinated Operating Agreement (H.R. 3113) stored water for dry years. between the SWP and the CVP, ratified in 1986, allows the SWP to purchase water from the CVP. The third approach to increase water deliveries Using conservation techniques and improving the is to turn to other sources of water. For instance, efficiency of transfer might also increase water use of groundwater could be increased. However, deliveries. in many metropolitan areas, groundwater bodies are currently being pumped at their sustainable yields. Operating rules for the reservoirs also could be Any increase in pumping could result in overdraft. modified to increase allowable reservoir storage in Furthermore, decisions to use groundwater are April, which would increase water storage at the made by local agencies and/or individual property end of the rainy season and deliverable water during owners, and groundwater is not managed as part of the peak demand season in midsummer. However, an integrated regional water system. Whether or an increase in storage in the late winter and early not to include it in the system is an important policy spring would likely reduce the amount of flood issue. protection (increase the risk of flooding) in the region; this in itself could negatively affect owners Another option is for southern California to of floodplain property. Floods also place the delta choose to fully use its allotment of Colorado River islands at risk because of higher water levels. The water (which could lead to conflicts between tradeoff between water supply and flood control in California and other users of that water, especially northern California represents a potentially serious Arizona). Other possibilities include desalinization policy conflict affecting all levels of government in plants, cloud seeding over the Sierras, and reuse of the region. In fact, the meeting between wastewater. However, desalinization plants are representatives of the State DWR and Bureau of energy intensive and may exacerbate air quality Reclamation, which was held to discuss Sheer and problems. Also, cloud seeding is controversial, since Randall's results (Volume A), concluded that any downwind users may not be willing to lose some of likely changes in reservoir operation that would their precipitation. 279 Chapter 14 Options for Allocating Water Shortages The individual delta islands have a significant range of values. For example, some islands contain The second major policy question is how best communities and highways, and others are strictly to allocate potential water shortages. One way agricultural. The property value of the islands is would be to allow greater flexibility in water about $2 billion (California Department of Water marketing. The adverse effects of this policy change Resources, 1987b). The islands also help repel (e.g., perhaps water becoming too expensive for saline water from the delta pumping plants (see agriculture and possible speculative price increases) Figure 14-2). could be ameliorated through a variety of governmental policies. Yet, even with regulation, The levees have been failing at an increasing any changes in the current system along these lines rate in recent years, and further sea level rise could would most likely be very controversial. increase failure probability. Improving the levees to protect the islands from flooding at the existing sea A second way to allocate the shortages is to level and flood probability would cost approximately rely on mechanisms used in the past to deal with $4 billion (California Department of Water droughts and water shortages, specifically Resources, 1982). governmental restrictions on water use. In the past, these mechanisms have included increased use The issue of levee failure raises three efficiency, transfers of agricultural water to important policy questions. First, will some or all of municipal and industrial uses, and restrictions on the levees be maintained? The range of options "nonessential" uses of water (e.g., watering of concerning the levees includes inaction, maintenance lawns). Increased efficiency of water usage through of the status quo, strategic inundation of particular various conservation techniques could effectively islands, and construction of polder levees. increase the number of water users without actually increasing the amount of water delivered. If climate Inaction, meaning the levees would not be gradually changed and water shortages became improved with time, could eventually lead to the more common, these restrictions could become formation of a large brackish-water bay as all of the virtually permanent. levees failed. Williams (Volume A) suggests that the area of the San Francisco Bay estuary could Sacramento-San Joaquin River Delta triple if all the levees failed. The delta area of the Sacramento and San Currently, the general policy is to maintain the Joaquin Rivers in the San Francisco Bay estuary delta's configuration. One important policy favoring receives great attention from governmental bodies the maintenance of the levees is the Delta Levee at all levels because of its valuable agricultural land, Maintenance Subventions Program, in which state its crucial role in the state's water resource system, financial assistance is available for maintaining and and its sensitive environment. The results of the improving levees. The value of the islands for studies in this overall project suggest that this region agriculture and maintenance of water quality (see could be significantly affected by climate change. below) has created additional institutional support Major changes could occur in delta island land use for maintaining the levees, even though the and in the water quality of the San Francisco Bay cumulative cost may exceed the value of the land estuary. The policy implications of these possible protected. Future funding decisions for this and changes are discussed below. related programs should consider the possibility of climate change. If the levees are maintained, an Delta Island Land Use important policy question must be considered: Who will pay for the maintenance? A critical land use issue is whether to maintain the levees surrounding islands threatened by Not all the islands are equal with regard to inundation. Much of the land present on these their value in protecting the freshwater delivery islands is below sea level and is usable for system. A possible future policy response to rising agriculture, recreation, and settlement only through sea level would be to maintain only certain levees levee protection. and not reclaim other islands as they became flooded. In essence, this would be a strategic 280 California inundation policy. Some precedence exists for this delivery. Current policy does not explicitly take into policy, as Mildred Island was flooded in 1983 and account the potential for future climate change. not reclaimed; the high cost of reclaiming the island Thus, D-1485 could be interpreted as requiring relative to its value was cited as a rationale. maintenance of delta water quality standards even if sea level rises and causes further penetration of Construction of large levees similar to the saline water into the delta. Delta water quality polders in Holland is an option for protecting the standards are currently being reviewed at the Bay- islands and maintaining shipping channels. Delta Hearing in Sacramento, which began in mid- However, this approach would be expensive and, 1987 and is expected to continue for 3 years. The although it has been discussed, has not attracted choice of future options will be greatly affected by much serious attention. decisions made at the hearing. The second policy question concerns failure of Possible methods of combating the impacts of the levees. If all or some levees are allowed to fail, saltwater intrusion include maintaining levees, will landowners be compensated? If so, where will increasing freshwater outflows, reducing the money come from? The delta islands contain withdrawals, enlarging channels, constructing a some of the most valuable agricultural land in the barrier in the Carquinez Strait or lower delta, state. Loss of this land would be a severe economic and/or constructing a canal around the delta's hardship for the local farmers and for the associated periphery. Alternatively, the freshwater pumping business community. Whether these farmers should plants could be moved to less vulnerable sites. be compensated for their loss is an important public Decisions regarding response options will not be policy issue. easily made. Levee maintenance and construction are costly. The water delivery agencies might be A final policy question remains: How will reluctant to increase delta outflows or to reduce management of the delta islands be coordinated? withdrawals. Enlargement of delta channels, Four government bodies have jurisdiction over the construction of saltwater barriers, and construction islands at the local, state, and federal levels. These of a peripheral canal are extremely controversial bodies will need to coordinate activities to reach environmental issues. Another possible response to decisions regarding the future of individual delta these climatic impacts would be a gradual, planned islands. retreat from the delta, devoting resources to options compatible with the absence of a freshwater delta. Water Quality of the San Francisco Bay Estuary This response would also be very controversial, both politically and environmentally. The intrusion of saline waters into the upper reaches of the San Francisco Bay estuary could be Water Quality of Freshwater Systems a major problem in a warmer climate. Climate change is projected to cause increased salinity in the The water quality of lakes, streams, and rivers estuary, largely as a result of sea level rise, levee could change as climate changes. Results from the failure, and the inadequacy of freshwater outflow to Castle Lake study indicate that primary production offset the increase in salinity. Furthermore, land of subalpine lakes could increase, with the potential subsidence due to groundwater extraction could for changes in the water quality of mountain augment sea level rise. In some areas of the streams (Byron et al., Volume E). Reduction in estuary, subsidence up to 1.5 meters (59 inches) has summer flows of streams and rivers in the Central occurred within the past 40 years (Atwater et al., Valley Basin could concentrate pollutants in these 1977). aquatic systems. A major policy question relates to these potential changes: How will potential Maintenance of current salinity levels is reductions in water quality below levels mandated in addressed in the water right Decision 1485 (D-1485) the current Water Quality Act of 1987 (Public Law of 1978. This decision requires that water quality 100-4) be prevented? standards in the delta be maintained. If they are not, additional water must be released from Maintaining water quality despite decreased reservoirs to improve delta water quality, which summer flows could be difficult and expensive. could reduce the amount of water available for Controlling nonpoint source pollution is a goal of 281 Chapter 14 the Water Quality Act of 1987, and meeting this One major step in response to possible future goal in the future could be more difficult and climate change is to incorporate climate expensive because of the lower summer flows. considerations into current planning processes. Changes in land use near streams and rivers may be Federal planning for the effects of climate change required to prevent runoff from agricultural land on forests is discussed in Chapter 5: Forestry. from reaching them. Reducing herbicide and Similar changes in the planning process could be pesticide use could also be another response, but considered at other levels of government. this could harm agricultural production. Another Coordinating the actions of government agencies option for preventing increased concentrations of involved with land management to climate change in pollutants in river reaches below reservoirs is to California is another possible response. increase releases from reservoirs during summer months; this strategy would dilute the pollutants. The flora and fauna in California are highly However, this strategy would also have obvious diverse and include many rare and endangered negative impacts on water deliveries. species. Climate could change faster than some species could adapt, leading to local extinction of Municipalities that release treated sewage into these species. Species conservation (as mandated by rivers also could face increased difficulties in the Rare and Endangered Species Act of 1973) meeting water quality standards. Options include might require habitat reconstruction and/or expanding sewage treatment facilities, which is transplanting in some situations. Monitoring expensive; releasing water from reservoirs to dilute programs may need to be instituted to track trends the pollutants, as discussed above; or controlling the in populations and communities. Extensive production of wastewater. Any municipalities programs have been developed for currently planning for new sewage treatment plants should endangered species in the state (e.g., the California include climate change as one factor in the design condor), and similar efforts probably could be criteria. mounted in the future for other highly valued species. Reductions in summer flows could harm populations of aquatic organisms and terrestrial Agriculture organisms that use riparian habitats. To the extent that these species become threatened with Changes in water availability and temperature extinction, laws requiring preservation of stresses are projected to affect agricultural endangered species (e.g., Endangered Species Act of production. How will changes in agricultural 1973) may be invoked as a legal basis for increasing production and crop types be managed, and how reservoir releases to preserve these species. This will California agriculture respond in national and could place into conflict the governmental agencies international settings? (For further discussion, see and public constituencies concerned with preserving Chapter 6: Agriculture.) biodiversity and those concerned with the economic impacts on agriculture and industry. Historically, agriculture has quickly adapted to climate fluctuations. New technology and Terrestrial Vegetation and Wildlife reallocation of resources might offset the impact of changed climatic conditions and water availability. Changing species composition and productivity Improved farm irrigation efficiency, such as might alter the character of forestry operations and extensive use of drip irrigation, could mitigate the the esthetic appeal of currently popular recreational impact of water-delivery shortages. Water areas. Climate-induced reductions in growth and marketing may provide a cost-effective means of regeneration rates, and increases in losses from meeting water demands and providing market wildfire and insect damage, could decrease the size opportunities for conserving water (Howitt et al., and value of industrial forests in the state. How 1980). For example, water marketing may provide these changes would be managed is a complex rights holders with the financial ability to invest in question involving all levels of government as well as water conservation programs to cope with climate private landowners. warming impacts on water availability. 282 California Changes in cropping locations and patterns of level rise. (For further discussion of these issues, water use could exacerbate nonpoint source see Chapter 7: Sea Level Rise.) pollution and accelerate rates of groundwater overdraft. Furthermore, changing water supply The accumulation of sediment behind water demands may heighten the conflicts between water project dams and the effects of diversion structures, allocation strategies and ecosystem and wildlife dredging operations, and harbor developments have values. limited the sources of sediment for beach maintenance (particularly along the southern It is uncertain how agricultural effects would California coast). Individual landowners and be manifest in California's evolving economic and institutions constructing such infrastructures should policy environment. For example, increased consider their effects on sedimentation processes. commodity prices could mitigate the financial Only through artificial deposition of sand (primarily impacts of potential reductions in crop acreage and from offshore sources) have southern California production. beaches been maintained. Beaches provide recreational areas and storm buffers, and their Wetland Vegetation and Fisheries maintenance will require a major and continued commitment. Wetland species are valuable ecologically, esthetically, and economically (photography, Energy Demand hunting, fishing, etc.). With rising sea level, areas supporting shallow-water vegetation might be A warmer climate could affect both energy inundated and converted to deep-water habitats demand and supply. For instance, higher supporting different species. New shallow-water temperatures could cause increased cooling sites could be created by artificially adding demands, and changes in runoff could affect sediment. This option features its own hydroelectric power generation. Institutions in environmental impacts and would most likely be California that are involved with energy planning, expensive. However, maintaining shallow-water such as the State Energy Resources Conservation vegetation is important not only to the conservation and Development Commission, should begin to of plant species but also to migratory birds, which consider climate change in their planning efforts so feed on such vegetation. that future energy demands can be met in a timely and efficient fashion. Salinity impacts on phytoplankton and fisheries might be controlled via levee maintenance coupled Air Quality with increases in delta outflow. Increasing temperatures could exacerbate air Shoreline Impacts of Sea Level Rise pollution problems in California, increasing the number of days during which pollutant levels are The California coast includes a diverse array of higher than the National Ambient Air Quality shorelines ranging from cliffs to sandy beaches. Standards. Devising technological and regulatory Erosion along these coastlines may increase as a approaches to meet ambient air standards is consequence of sea level rise. Such erosion could currently a major challenge in certain regions of the substantially damage shoreline structures and state, and these efforts must be continued. Under recreational values. Preventing the erosion would a warmer climate, achieving air quality standards be very costly. For example, protecting the sewer may become even more difficult. To ensure that air culvert of the San Francisco Westside Transport quality standards are met under warmer conditions, Project from potential damage caused by sea level policymakers, such as EPA and the California Air rise may cost over $70 million (Wilcoxen, 1986). Quality Board, may wish to consider possible Sound planning for shoreline structures should climate changes as they formulate long-term consider future erosion that may be caused by sea management options for improving air quality. 283 Chapter 14 REFERENCES California Division of Forestry and Fire Protection. 1988. California's Forest and Rangelands: Growing Atwater, B.F., C.W. Hedel, and E.J. Helley. 1977. Conflict Over Changing Uses. Sacramento, CA: Late Quaternary Depositional History, Holocene Forest and Rangeland Resources Assessment and Sea Level Changes, and Vertical Crustal Movement, Policy Act Committee. Southern San Francisco Bay. U.S. Geological Survey Professional Paper 1014. Menlo Park, CA: U.S. Conomos, T.J., R.S. Smith, and J.W. Gartner. 1985. Geological Survey. Environmental setting of San Francisco Bay. Hydrobiologia 129:1-12. Atwater, B.F., S.G. Conard, J.N. Dowden, C.W. Hedel, R.L. MacDonald, and W. Savage. 1979. Daubenmire, R. 1978. Plant Geography With History, landforms, and vegetation of the estuary's Special Reference to North America. New York: tidal marshes. In: Conomos, T.J., ed. San Francisco Academic Press, Inc. Bay: The Urbanized Estuary. San Francisco, CA: Pacific Division, American Association for the Davis, O.K., J.C. Sheppard, and S. Robertson. 1986. Advancement of Science, pp. 347-386. Contrasting climatic histories for the Snake River Plain result from multiple thermal maxima. Barbour, M.G., and J. Major. 1977. Terrestrial Quaternary Research 26:321-339. Vegetation of California. New York: John Wiley and Sons. Denton, R.A., and J.R. Hunt. 1986. Currents in San Francisco Bay. Final report. Berkeley, CA: Budyko, M.I. 1982. The Earth's Climate: Past and University of California. Future. New York: Academic Press. Doorenbos, J., and A.H. Kassam. 1979. Yield California Department of Water Resources. 1982. Response to Water. FAO Irrigation and Drainage Delta Levees Investigation. Bulletin 199. Paper No. 33. Rome: Food and Agriculture Sacramento, CA: California Department of Water Organization. Resources. Fernald, M.L. 1950. Gray's Manual of Botany, 8th California Department of Water Resources. 1983. ed. New York: American Book Company. The California Water Plan: Projected Use and Available Water Supplies to 2010. Bulletin 160-83. Fischer, H.B. 1970. A Method for Predicting Sacramento, CA: California Department of Water Pollutant Transport in Tidal Waters. Contribution Resources. No. 132. Berkeley, CA: University of California Water Resources Center. California Department of Water Resources. 1985. California State Water Project. Typewritten brief. Gleick, P.H. 1987a. Regional hydrologic Sacramento, CA: California Department of Water consequences of increases in atmospheric CO, and Resources. other trace gases. Climatic Change 10:137-161. California Department of Water Resources. 1986. Gleick, P.H. 1987b. The development and testing of Operations Criteria Applied in DWR Planning a water balance model for climate impact Simulation Model. Memorandum report. assessment: modeling the Sacramento Basin. Water Resources Research 23:1049-1061. Sacramento, CA: California Department of Water Resources. Goldman, C.R., A. Jassby, and T. Powell. 1989. California Department of Water Resources. 1987a. Interannual fluctuations in primary production: California Water: Looking to the Future. Bulletin impact of climate and weather at two subalpine 160-87. Sacramento, CA: California Department of lakes. Limnology and Oceanography. In press. Water Resources. Howitt, R.E., D.E. Mann, and H.J. Vaux, Jr. 1980. California Department of Water Resources. 1987b. The economics of water allocation. In: Englebert, Sacramento-San Joaquin Delta Atlas. Sacramento, E.A., ed. Competition for California Water. CA: California Department of Water Resources. Berkeley, CA: University of California Press. 284 California Howitt, R.E., and P. Mean. 1985. Positive Quadratic Raven, P.H. 1977. The California flora. In: Programming Models. Working Paper No. 85-10. Barbour, M.G., and J. Major, eds. Terrestrial University of California, Department of Agricultural Vegetation of California. New York: John Wiley Economics. Davis, CA: University of California. and Sons, pp. 109-137. Kjeldson, M.A., P.F. Raquel, and F.W. Fisher. 1981. Ritchie, J.C., L.C. Cwynar, and R.W. Spear. 1983. Influences of freshwater flow on chinook salmon in Evidence from north-west Canada for an early the Sacramento-San Joaquin Estuary. In: Cross, Holocene Milankovitch thermal maximum. Nature R.D., and D.L. Williams, eds. Proceedings of the 305:126-128. National Symposium on Freshwater Inflow to Estuaries, Vol. 2. Washington, DC: U.S. Sheer, D.P., and M.L. Baeck. 1987. Documentation Department of the Interior, pp. 88-108. of the CVP/SWP Simulation Models Developed by WRMI. Columbia, MD: Water Resources Leverenz, J.W., and D.J. Lev. 1987. Effects of Management, Inc. carbon dioxide-induced climate changes on the natural ranges of six major commercial tree species Sudman, R.S. 1987. Layperson's Guide to the Delta. in the western United States. In: Shands, W.E., and Sacramento, CA: Western Water Education J.S. Hoffman, eds. The Greenhouse Effect, Climate Foundation. Change, and U.S. Forests. Washington, DC: The Conservation Foundation, pp. 123-155. U.S. Bureau of Reclamation. 1985. Summary Statistics, 1984; Volume 1: Water, Land and Related Macdonald, K.B. 1977. Coastal salt marsh. In: Data. Denver, CO: U.S. Bureau of Reclamation, Barbour, M.G., and J. Major, eds. Terrestrial Division of Water and Land Technical Services. Vegetation of California. New York: John Wiley and Sons, pp. 263-294. U.S. Department of Agriculture. 1987. Agricultural Statistics 1987. Washington, DC: U.S. Government Major, J. 1977. California climate in relation to Printing Office. vegetation. In: Barbour, M.J., and J. Major, eds. Terrestrial Vegetation of California. New York: U.S. Department of the Interior. 1986. National John Wiley and Sons, pp. 11-74. Forest Statement of Receipts, Fiscal Year 1986. Washington, DC: U.S. Government Printing Office. Meyer, P.A. 1987. The value of wildlife in San Francisco Bay. Exhibit 38. Entered by the Bay U.S. Environmental Protection Agency. In Institute of San Francisco to the State Water preparation. Ecological Effects of Global Climatic Resources Control Board in Sacramento, CA. Change. Chapter 4. In: U.S. EPA Global Climatic Change Program. Miller, C.S., and R.S. Hyslop. 1983. California: The Geography of Diversity. Palo Alto, CA: Mayfield. U.S. Maritime Administration. 1985. Containerized cargo statistics, 1983. Washington, DC: U.S. Munz, P.A., and D.D. Keck. 1959. A California Government Printing Office. Flora. Berkeley, CA: University of California Press. Wardle, P. 1974. Alpine timberlines. In: Ives, J.D., National Research Council. 1987. Responding to and R.G. Barry, eds. Arctic and Alpine Changes in Sea Level. Committee on Engineering Environments. London: Metheun and Company, Implications of Changes in Relative Mean Sea pp. 371-402. Level. Washington, DC: National Academy Press. Wilcoxen, P.J. 1986. Coastal erosion and sea level Nichols, D.R., and N.A. Wright. 1971. Preliminary rise: implications for ocean beach and San map of historic margins of marshland, San Francisco's Westside Transport Project. Coastal Francisco, California. U.S. Geological Survey Open Zone Management Journal 14:173-191. File Map, San Francisco Bay Region Environment and Resource Planning Study. Basic Data Contribution 9. 285 CHAPTER 15 GREAT LAKES FINDINGS dredging, shipping costs could rise 2 to 33% as a result of reduced cargo capacity. However, reduced ice cover would lengthen the shipping Global climate change could affect the Great Lakes season by 1 to 3 months. Under scenarios of by lowering lake levels, reducing ice cover, and relatively smaller lake level drop (0.7 to 1 degrading water quality in rivers and shallow areas meter), the shipping season would be of the lakes. It could also expand agriculture in the lengthened sufficiently to allow for the northern states, change forest composition, decrease transport of at least the same amount of cargo. regional forest productivity in some areas, increase Under a scenario of larger lake level drops open water fish productivity, and alter energy (1.65 meters) and no dredging, total annual demand and supply. cargo shipments could be reduced. Lakes Water Quality and Fisheries Average lake levels could fall by 0.5 to 2.5 Higher temperatures could change the thermal meters (1.7 to 8.3 feet) because of higher structure of the Great Lakes. The result would temperatures under the doubled CO₂ be a longer and greater stratification of the scenarios in this report. A drop of 1 meter lakes and increased growth of algae. This would leave average levels below historic lows. result is very sensitive to changes in windspeed Even if rainfall increases, the levels would fall and storm frequency two areas of relative because higher temperatures would reduce the uncertainty. These two factors would combine snowpack and accelerate evaporation. The to reduce dissolved oxygen levels in shallow estimates of lake level drop are sensitive to areas of lakes such as Lake Erie. A study of assumptions about evaporation; under certain southern Lake Michigan indicated that annual limited conditions, lake levels could rise. turnover of the lakes could be disrupted. As a result of higher temperatures, the Climate change could increase concentrations duration of ice cover on the lakes would be of pollutants in the Great Lakes Basin. reduced by 1 to 3 months. Ice could still form Dredging of ports could suspend toxic in near-shore and shallow areas. Changes in sediments in near-shore areas. Potential windspeed and storm intensity would affect the reductions in riverflow in the basin would duration of ice cover. create higher concentrations of pollutants in streams. The disposal of toxic dredge spoils Shoreline communities would have to make was not studied in this report. adjustments to lower lake levels over the next century. Hundreds of millions of dollars may The effects on fisheries would be generally have to be spent along the Illinois shoreline beneficial. Higher temperatures may expand alone, dredging ports, harbors, and channels. fish habitats during fall, winter, and spring, and Water intake and outflow pipes may have to accelerate the growth and productivity of fish be relocated. On the other hand, lower levels such as black basses, lake trout, and yellow would expose more beaches, which would perch. On the other hand, fish populations enhance shoreline protection and recreation. could be hurt by decreased habitats and lower dissolved oxygen levels during the summer. Climate change could have both good and bad The effects of potential changes in wetlands effects on shipping. Lower lake levels may due to lower lake levels, reductions in ice cover, necessitate increased dredging of ports and introduction of new exotic species, and increase channels or reduced cargo loads. Without in species interaction were not analyzed, 287 Chapter 15 although they could offset the positive results of Electricity Demand these studies. There could be little net change in annual Forests electricity demand. In northern areas, such as Michigan, reduced heating needs could exceed The composition and abundance of forests in increased cooling requirements, while in the Great Lakes region could change. Higher southern areas, such as Illinois, cooling needs temperatures and lower soil moisture could may be greater than heating reductions. The reduce forest biomass in dry sites in central annual demand for electricity in the entire Michigan by 77 to 99%. These mixed region could rise by 1 to 2 billion kilowatthours hardwood and oak forests could become oak (kWh) by 2010 and by 8 to 17 billion kWh (less savannas or grasslands. In northern areas such than 1%) by 2055. This study did not analyze as Minnesota, boreal and cedar bog forests the reduced use of other fuels such as oil and could change to treeless bogs, and mixed gas in the winter, changes in demand due to northern hardwood and boreal forests in higher prices, and the impacts on hydroelectric upland areas could become all northern supplies. Previous studies have suggested that hardwoods. Productivity could decrease on reduced lake levels and river flows could lead dry sites and bogland sites, but it could to reductions in hydroelectric power production. increase on some well-drained wet sites. Softwood species that are currently By 2010, approximately 2 to 5 gigawatts (GW) commercially important could be eliminated could be needed to meet the increased demand, and replaced by hardwoods, such as oak and and by 2055, 23 to 48 GW could be needed maple, which are useful for different purposes. an 8 to 11% increase over baseline additions that may be needed without climate change. Depending on the scenario, changes in forests These additions could cost $23 to $35 billion by could be evident in 30 to 60 years. These 2055. results do not reflect additional stresses, such as pests and increased fire frequency, nor do Policy Implications they reflect the possible beneficial impacts of increased CO2 levels. U.S. and Canadian policymakers, through such institutions as the International Joint Agriculture Commission, should consider the implications of many issues for the region. This study raises Considering climate change alone, corn and additional issues concerning the following: soybean yields in northern areas, such as Minnesota, could increase by 50 to 100% and -- The water regulation plans for Lake could decline in the rest of the region by up to Ontario and possibly for Lake Superior lake 60%. The combined effects of climate and levels. higher CO₂ levels could further increase yields in the north and result in net increases in the -- The potential increased demands for rest of the region, unless climate change is diverting Great Lakes water for uses severe. outside the basin. Before such a potential demand could be accommodated, additional Agricultural production in the northern part of analysis would be required. This is not the region may expand as a result of declines currently allowed by federal statutes. elsewhere. However, the presence of glaciated soils in northern states could limit this -- Long-range industrial, municipal, and expansion. Acreage in the Corn Belt states agricultural water pollution control may change little. Wider cultivation in the strategies. Agencies such as EPA may wish north could increase erosion and runoff, and to examine the implications for long-term degrade surface and groundwater quality. point and nonpoint water pollution control Increased agriculture would require changes in strategies. the infrastructure base, such as in transportation networks. 288 Great Lakes -- The research, planting, and land purchase soils, moderate temperatures, and abundant rainfall decisions in northern forests by federal, have made the southern part of the region a major state, and private institutions. agricultural producer. Forests are abundant in the north and support commercial and recreational uses. The basin has become the home of over 29 CLIMATE-SENSITIVE NATURAL million Americans and produces 37% of U.S. RESOURCES IN THE GREAT manufacturing output (U.S. EPA and Environment Canada, 1987; Ray et al., Volume J). LAKES REGION Current Climate The Great Lakes region¹ is highly developed, largely because of its natural resources. The steel The Great Lakes region has a midlatitude industry developed along the southern rim of the continental climate. Winter is sufficiently cold to lakes, in part because iron ore from the north could produce a stable snow cover on land and ice on the be inexpensively transported over the lakes. Rich lakes. The average January temperature over Lake Superior is -15°C (5°F), and the average July temperature in the southern part of the region is 1 This chapter will cover only the U.S. side of the Great Lakes 22°C (72°F). The average rainfall varies from 700 and the eight states bordering them (see Figure 15-1). CREAT LAFES DRAINAGE BASIN LIMIT Virginia LAKE SUPERIOR Twin Harbors Duluth White Fish Bay Sault St. Marie MINNESOTA WISCONSIN Georgian Cornwall Bay LAKE HURON Mount Pleasant Green Bay LAKE MICHIGAN ONTARIO LAKE ONTARIO MICHIGAN Buffalo NEW YORK Flint ERIE LAKE GREAT Toledo LAKES Cleveland DRAINAGE BASIN LIMIT PENNSYLVANIA Fort Wayne Pittsburg ILLINOIS INDIANA OHIO Forest Sites Compensating Works Agriculture Sites Shipping Sites Figure 15-1. Map of the Great Lakes study sites. 289 Chapter 15 to 1,000 millimeters (27 to 39 inches), depending and along the St. Lawrence River and Seaway. Two on location (Cohen, in Glantz, Volume J). regulatory plans (Plan 1977 for Superior and Plan 1958D for Ontario) set ranges of levels between The Lakes which Lakes Superior and Ontario must be maintained. Diversion out of the lakes is also The Great Lakes consist of a system of five limited by law. Flow through the Chicago diversion major lakes that contain approximately 18% of the was limited by the Supreme Court to 90 cubic world supply of surface freshwater and 95% of the meters per second (3,200 cubic feet per second) surface freshwater in the United States (U.S. EPA (Tarlock, 1988), and the 1986 Water Resources and Environment Canada, 1987) (see Figure 15-1, Development Act forbids diversion out of the lakes' Map of the Great Lakes). The natural flow of the basin without the consent of all Great Lakes lake system begins in Lake Superior, the largest of governors (Ray et al., Volume J). the lakes, which drains via the St. Mary's River into Lakes Michigan and Huron (considered a single Climate-Sensitive Uses of the Lakes hydrologic unit because they are connected by the Straits of Mackinac). Water from Lakes Michigan Shipping and Huron flows out through the St. Clair River into Lake St. Clair. From there, the water flows The U.S. Great Lakes fleet, which consists of through the Detroit River and into Lake Erie, the approximately 70 ships, transported over 171 million shallowest lake. The Niagara River connects Lakes tons of cargo in 1987 (The New York Times, 1988). Erie and Ontario, and the system ultimately empties The tonnage of U.S. shipping consists of iron ore, into the Atlantic Ocean via the St. Lawrence River coal, and limestone, all primary inputs for steel and Seaway. (77%); lake grain (13%); and petroleum products, potash, and cement (10%) (Nekvasil, 1988). Cargo The greatest influence on lake levels is nature. volumes are displayed in Table 15-1. Most of the Seasonal fluctuations are on the order of 0.3 to 0.5 goods are shipped within the Great Lakes, with only meter (1 to 1.7 feet), with the lakes peaking in late 7% of the tonnage (mainly grains) shipped to summer because of condensation over the northern overseas markets (Ray et al., Volume J). Although lakes and reaching minimum levels in late winter. shipping activity had declined as a result of Interannual lake level changes have been much reductions in U.S. steel production, recent increases larger, approximately 2 meters (6.6 feet). in steel output have led to additional demand for shipping (The New York Times, 1988). Lake Regulation Great Lakes ships last over half a century and The flow between the lakes is controlled by are designed to pass within a foot of the bottom of dams at two points: (1) the St. Mary's River to control levels of Lake Superior; and (2) Iroquois, Ontario, to control Lake Ontario. The major Table 15-1. 1987 U.S. Great Lakes Shipping Cargo diversion out of the lakes is the Chicago diversion, (thousands of tons) which transfers water from Lake Michigan through the Illinois River into the Mississippi River. Human Cargo Weight Percentage influence on lake levels is relatively small. Doubling the flow down the Chicago diversion would lower lake levels only by 2.5 inches in 15 years (F. Quinn, Iron ore 61,670 36 Great Lakes Environmental Research Lab., 1987, Coal 37,731 22 personal communication). Stone 33,164 19 Grain 22,338 13 Joint control of lake supply was codified in the Petroleum 11,491 7 Boundary Waters Treaty of 1909 between Canada products and the United States, which created the Cement 3,806 2 International Joint Commission (IJC) consisting of Potash 1,702 1 representatives from both countries. The IJC Total 171,902 100 regulates flow through the control structures and diversions by balancing the needs of shipping, Source: Nekvasil (Lake Carriers Association, 1988, hydropower, and consumptive uses among the lakes personal communication). 290 Great Lakes channels and locks. Cargo capacity is quite sensitive near the water's edge. Shoreline property owners to lake and channel depth because of this low have riparian rights to use adjoining waters. The clearance. The presence of ice usually shuts down shoreline property owners cannot substantially Great Lakes shipping up to 4 months each year. diminish the quantity or quality of surface waters (Ray et al., Volume J). Hydropower Climate and Water Quality The eight Great Lakes States use the connecting channels and the St. Lawrence River to Water quality is directly affected by climate. obtain 35,435 gigawatt hours of hydropower each Lower stream runoff increases concentrations of year, which is about 5% of their electricity pollutants. Every summer, the lakes stratify into a generation. About four-fifths of the hydropower is warmer upper layer and a cooler lower layer. This produced in New York State, which derives over stratification can limit biological activity by 26% of its electricity from hydropower (Edison restricting the flow of nutrients between layers. In Electric Institute, 1987). addition, warm temperatures and an excess supply of nutrients (phosphorous and other chemicals from Municipal Consumption agricultural runoff and sewage effluent) can lead to algal blooms that decay and cause a loss of oxygen Most water used for the domestic and (eutrophication) and reduction in aquatic life in the industrial consumption in the basin is taken from lower layers of lakes such as Lake Erie. Cool the lakes. Surface waters supply 95% of the basin's weather and the formation of ice help to deepen the water needs. By the year 2000, consumption is mixed layer, break up the stratification, and estimated to increase by 50 to 96% (Ray et al., thoroughly mix the lakes in the winter. Volume J; Cohen, 1987b; IJC, 1985). Development, industrialization, and intensive Fisheries agriculture in the Great Lakes Basin have created serious pollution in the lakes, especially Lake Erie. In 1984, the value of the harvest to the U.S. In the early 1970s, nutrient loadings were so high commercial fishing industry was approximately $15 that Lake Erie experienced significant million (U.S. EPA and Environment Canada, 1987; eutrophication problems for several years (DiToro U.S. Department of Commerce, 1987). Although et al., 1987). most fishing in the Great Lakes is for recreation, fisheries are managed by the states; the Great Lakes Two measures have helped improve water Fishery Commission coordinates activities among quality. The U.S.-Canada Great Lakes Water the states. Quality Agreement of 1972 called for controlling nutrient inputs and eliminating the discharge of Tourism toxic chemicals, and the Clean Water Act mandated construction of sewage treatment plants and Three national and 67 state parks are located controls on industrial pollutants. The United States along the shores of the lakes, as are numerous local and Canada spent a total of $6.8 billion on sewage parks. Over 63 million people visited these parks in treatment in the Great Lakes. By 1980, nutrient 1983 (Ray et al., Volume J; Great Lakes Basin loadings into Lake Erie had been cut in half (Ray et Commission, 1975). In 1984, lake-generated al., Volume J; DiToro et al., 1987), and water recreation yielded revenues of $8 to 15 million. quality had markedly improved. Fishing, boating, and swimming are very popular. Fluctuating Lake Levels Shoreline Development Recent high and low lake levels have Over 80% of the U.S. side of the Great Lakes significantly affected users of the lakes. In 1964, shoreline is privately owned. One of the most Lake Michigan was 0.92 meters (3 feet) below developed shorelines is the 101-kilometer Illinois average, making some docks and harbors unusable. shoreline, where many parks and residential Shipping loads were reduced by 5 to 10% and more structures, including apartment houses, are built shipments were required, subsequently raising the 291 Chapter 15 cost of raw materials and supplies by 10 to 15%. In pasture. The Great Lakes States encompass most addition, many water intakes had to be extended or of the Corn Belt. In 1983, roughly 59% of all U.S. lowered (Changnon, Volume H). Flow through the cash receipts for corn and 40% of the receipts for Niagara hydropower project fell by more than 20%, soybeans came from this region. Overall, the Great with electricity generation off by more than 35%. Lakes States produced 26% of the total U.S. Flow through New York's St. Lawrence hydro agricultural output, or $36 billion (Federal Reserve project was more than 30% below its mean, with Bank of Chicago, 1985). Most crops are grown on electricity generation decreased by 20% (Linder, dryland, as only about 1% of the region's croplands 1987). However, low lake levels also provided were irrigated in 1975 (U.S. Department of benefits, for example, beaches became larger. Commerce, 1987). In the mid-1980s, a series of cool and wet Livestock are also important to the agricultural years caused the lakes to rise to record heights. economy of the region. Approximately 18% of U.S. Apartment houses that were built too close to the cattle are raised in these eight states; of these, 52% shoreline during the low levels of the 1960s were are dairy COWS (USDA, 1987). (The sensitivity of flooded, as were roadways built close to the shore. livestock to climate change is discussed in Chapter The low water levels in the 1960s exposed the 6: Agriculture.) supporting structures along Chicago's shoreline to air, causing dry rot. When lake levels rose, the Forests wood pilings and sections of the revetment collapsed. The estimated construction cost for The forests in the region have commercial, rebuilding the damaged shoreline protection system recreational, and conservation uses. The forests in is $843 million (Changnon, Volume H). The last 2 the south are mainly oak and northern hardwoods, years have been relatively hot and dry, causing lake such as maple. The north has almost 21 million levels to recede to average levels. The lower levels hectares (52 million acres) of forests consisting have forced shippers to reduce tonnage just as the mostly of northern hardwoods, such as maple, birch, steel industry in the region is undergoing a and beech, and boreal forests, such as spruce and fir resurgence. trees. The federal and state governments own, respectively, 11 and 13% of the forests in Michigan, Land Around the Lakes Minnesota, and Wisconsin, while over half are privately owned (USDA, 1982). The pulp, The land in the Great Lakes region is construction, and furniture industries are major extensively used for industry, agriculture, and consumers of such species as aspen, pines, balsam forestry. Many of the uses are sensitive to climate. fir, spruce, maples, paper birch, and oak. The forest industry is a major employer in the northern Land Uses part of the region. In Wisconsin, for example, 283,000 jobs are in timber harvesting and Urban Development manufacturing related to forestry (Botkin et al., Volume D; U.S. EPA and Environment Canada, Approximately 29 million people live in the 1987). Forestry is considered to be a growth Great Lakes Basin, mostly in the urban areas industry in the region, since Michigan has identified around the cities on the southern edge of the Great forest products as one of the three key industries Lakes: Chicago, Detroit, Cleveland, Toledo, and targeted for expansion in the state (Ray et al., Buffalo. Many of the residents work in Volume J). manufacturing industries, which despite recent declines, still provide 23% of payroll employment (Ray et al., Volume J). PREVIOUS CLIMATE CHANGE STUDIES Agriculture Agriculture is the single largest user of land: The impacts of climate change on many of the 42% of all land in the eight Great Lakes States is systems in the Great Lakes have been analyzed in devoted to crops, and an additional 10% is used for previous studies, mainly by Canadian researchers. 292 Great Lakes These studies are summarized in Cohen and decline between 6 and 8.5% as a result of Allsopp (1988). Several Canadian studies have reductions in streamflow. examined the potential impacts of climate change on Great Lakes levels and concluded that levels would Impacts on managed and unmanaged vegetation fall. Southam and Dumont (1985) used the have also been studied. The Land Evaluation Goddard Institute for Space Studies (GISS) scenario Group examined the potential impacts of climate to estimate that lake levels would fall by 0.2 to 0.6 change on agriculture in Ontario and found that meters (0.7 to 2 feet). Cohen (1986) used yields could decrease in southern Ontario and hydrologic calculations to estimate that the lakes farming could become feasible in northern Ontario. might fall between 0.2 and 0.8 meters. More The study also indicated that the direction of change recently, Marchand et al. (1988) also used a for yields depends on whether rainfall increases or hydrologic model of the lakes to estimate that the decreases (Land Evaluation Group, 1986). Solomon lakes would drop by an average of 0.2 to 0.6 meters. and West (1986) used a stand simulation model (see Cohen (1987a) found that changes in lake levels are this chapter, Forests) to estimate the impacts of very sensitive to humidity and windspeed. It is not doubling and quadrupling of CO₂ levels on a known how climate change would affect these northwest Michigan coniferous-deciduous parameters on a regional scale. Wall (1985) transitional forest. They found that doubled CO₂ concluded that lower lake levels could reduce would lead to an eventual disappearance of boreal ecological diversity and dry up enclosed marshes. In forests and an increase in deciduous trees. Total another study, Cohen (1987b) estimated that biomass would decline at first and rebound in about withdrawals of water from the lakes for municipal two centuries. consumption would increase by about 2.5% on an annual basis and would only marginally affect lake Two studies by Canadian researchers examined levels. the possible impacts of climate change on tourism and recreation in Ontario. Both studies used Assel et al. (1985) studied the extent of ice climate change scenarios based on the GISS and cover during the winter of 1982-83, which had Geophysical Fluid Dynamics Laboratory (GFDL) temperatures 3.3 to 4.4°C warmer than the 30-year models (although these may have been earlier mean. They found that ice cover on Lake Superior model runs). Crowe (1985) estimated that snowfall was reduced from a normal 75% coverage to 21%. would decrease by 25 to 75%, and the ski season On Lake Erie, ice coverage was down to 25% from would be cut by 75 to 92% (7 to 12 weeks) in the normal 90%. Meisner et al. (1987) conducted southern Ontario and by 13 to 31% (2 to 4 weeks) a literature review on the possible effects of global in northern Ontario. Wall found similar results. warming on Great Lakes fish. Results are discussed He concluded that reduced snowfall could eliminate in the fisheries section of this chapter. skiing in southern Ontario and would shorten the northern Ontario ski season by 30 to 44%. A Marchand et al. (1988) (see also Sanderson, longer summer season could increase such summer 1987) estimated the combined effects of lower lake tourism activities as camping. Wall (1985) also levels and reduced ice cover due to climate change, thought that lower lake levels could decrease and higher water consumption and shipping tonnage ecological diversity and dry up enclosed marshes. due to population and economic growth of Canadian shipping and hydropower production. They found that without economic changes, lower GREAT LAKES STUDIES IN THIS lake levels would increase shipping costs by 5%. REPORT After consideration of economic growth, lower lake levels and reduced ice cover could increase shipping costs by 12%. Unlike previous studies, the studies for this report used common scenarios to address some of Linder (1987) used the transient scenarios to the potential impacts of climate change on a estimate impacts on electricity demand and number of natural and societal systems in the Great hydropower generation in 2015 in upstate New Lakes region. The studies address the direct effects York. He found total energy demand declining by of climate change on the resources and some of the 0.21 to 0.27%, but peak demand increasing by 1 to indirect effects on infrastructure and society. They 2%. Meanwhile, hydropower production could focused on the lakes themselves, examining such 293 Chapter 15 issues as lake levels, ice cover, thermal structure, Water Quality and fisheries. They also looked at the effects of these changes on shipping and shoreline properties, The following studies focus on water quality and examined the sensitivities of agriculture and and the effects on aquatic life in the lakes. The first forest to climate change. Finally, the studies two studies examined the direct effects of climate on examined the implications of climate change for the thermal structure of some of the lakes. Great Lakes policies and institutions. Some of the studies were linked quantitatively, but most were Potential Climatic Changes to the Lake conducted independently of each other. Michigan Thermal Structure - McCormick, Great Lakes Environmental Research The studies involved either new topics or Laboratory (Volume A) approaches that were not used in previous studies. For example, the analysis of lake levels used a more The Effects of Climate Warming on Lake complex hydrologic model than was used previously. Erie Water Quality - Blumberg and The agriculture analysis complements the Land DiToro, Hydroqual, Inc. (Volume A) Evaluation Group's study of Ontario by using a different model to examine impacts on the U.S. side The results from these studies were used in the of the lakes. The potential impacts of climate following: change on thermal structure were examined for the first time. Also for the first time, models were used Potential Responses of Great Lakes Fishes to analyze impacts on fisheries. This study and Their Habitat to Global Climate complements previous studies on forests by using a Warming - Magnuson, Regier, Hill, combination of modeling techniques to test the Holmes, Meisner, and Shuter, Universities similarity of results. of Wisconsin and Toronto (Volume E) The following analyses were performed for this Forests report: A series of studies on forests was commissioned Direct Effects on Lakes to examine shifts in ranges, transient impacts, and the potential for migration of some Great Lakes Effects of Climate Changes on the forests. Basically, these are different analytic Laurentian Great Lakes Levels - Croley and techniques for understanding how climate change Hartmann, Great Lakes Environmental may affect the composition and abundance of Research Laboratory (Volume A) forests in the region. Impact of Global Warming on Great Lakes Transient Effects on Great Lakes Forests - Ice Cycles - Assel, Great Lakes Botkin, Nisbet, and Reynales, University of Environmental Research Laboratory California at Santa Barbara (Volume D) (Volume A) Hard Times Ahead for Great Lakes Impacts of Lake Changes on Infrastructure Forests: A Climate Threshold Model Predicts Responses to CO2-Induced The results from the first two studies were Climate Change - Zabinski and Davis, used in the following studies: University of Minnesota (Volume D) Effect of Climatic Change on Shipping Assessing the Response of Vegetation to Within Lake Superior and Lake Erie - Future Climate Change: Ecological Keith, DeAvila, and Willis, Engineering Response Surfaces and Paleoecological Computer Optecnomics, Inc. (Volume H) Model Validation - Overpeck and Bartlein, Lamont-Doherty (regional results were Impacts of Extremes in Lake Michigan taken from this study) (Volume D) Levels Along Illinois Shoreline Part 1: Low Levels - Changnon, Leffler, and Shealy, Illinois State Water Survey (Volume H) 294 Great Lakes Agriculture GREAT LAKES REGIONAL CLIMATE CHANGE SCENARIOS The potential changes in agriculture in the Great Lakes were analyzed by studying changes in crop yields in the region and integrating the results All three general circulation models (GCMs) in a national analysis of production changes. That that provide the basis for the climate change national analysis was used to determine if scenarios show rather large increases in temperature production in the region could increase or decrease. for the Great Lakes region under the doubled CO₂ The results of these studies were used to examine climate. The seasonal and annual temperatures and potential farm level adjustments. precipitation are displayed in Figure 15-2. The Oregon State University (OSU) scenario has an Effect of Global Climate Change on annual temperature rise of 3.5°C, with no change in Agriculture: Great Lakes Region Ritchie, seasonal pattern. The Goddard Institute for Space Baer, and Chou, Michigan State University Studies (GISS) scenario is about a degree warmer (Volume C) on average and has the largest warming in the winter and fall. The Geophysical Fluid Dynamics Farm Level Adjustments by Illinois Corn Laboratory (GFDL) scenario has the largest Producers to Climatic Change - Easterling, warming of the three models, about 6.5°C annually, Illinois State Water Survey (Volume C) with the largest warming in the summer. All three scenarios have annual increases in precipitation. This chapter will use regional results from the OSU has an increase of approximately 0.1 following: millimeters per day (0.1 inches per year), with precipitation rising in all seasons. GISS has an The Economic Effects of Climate Change increase of approximately 0.2 millimeters per day on U.S. Agriculture: A Preliminary (0.03 inches per year), with precipitation declining Assessment - Adams, Glyer and McCarl, slightly in the fall. GFDL has an annual Oregon State University (Volume C) precipitation increase of only 0.05 millimeters per day (0.07 inches per year), but rainfall drops by 0.5 millimeters per day (0.02 inches per day) in the Energy summer. The large temperature increase and small rainfall increase combine to make GFDL the most This project analyzed potential changes in the severe scenario. This is especially true in summer national demand for electricity and estimated months, when GFDL has the largest temperature changes in regional demands. Results for the Great rise of any scenario and is the only scenario that Lakes region are presented in this chapter. reduces rainfall. OSU is the mildest scenario owing to the smaller temperature increase. (Other runs Electric Utilities Linder and Inglis, ICF, of the GFDL model have lower temperature Inc. (Volume H) increases, although they still estimate a decline in summer rainfall.) GISS is in the middle in terms Policy of severity, and OSU is the mildest of the three scenarios. The potential policy implications of the changes indicated by these and previous studies for One limitation related to using the GCMs as a local, state, federal, and international basis for climate change scenarios for the Great decisionmaking are examined. This project Lakes region is that the lakes are not well provided information for the background and policy represented in the GCMs. The relatively large size implications sections. of the GCM grid boxes results in little feedback from the lakes to the regional climate estimates Effects of Global Warming on the Great from the GCMs. Lakes: The Implications for Policies and Institutions - Ray, Lindland, and Brah, The Center for the Great Lakes (Volume J) 295 Chapter 15 A. Temperature B. Precipitation 0.7 8 GISS 0.8 7 0.6 GFDL 0.4 6 0.3 OSU NC 0.2 6 CHANGE (°C) CHANGE (mm/Day) 0.1 4 O -0.1 3 -0.2 -0.3 2 -0.4 -0.5 1 -0.6 0 -0.7 Winter Spring Summer Fall Annual Winter Spring Summer Fall Annual NC No Change Figure 15-2. Average change in temperature (A) and precipitation (B) over Great Lakes gridpoints in GISS, GFDL, and OSU models (2xCO₂ minus 1xCO₂). RESULTS OF THE GREAT Croley and Hartmann simulated runoff in each of LAKES STUDIES the subbasins, overlake precipitation, and evaporation. Lake levels are very sensitive to Lakes evaporation; therefore, Croley and Hartmann ran each GCM scenario with different assumptions Lake Levels about evaporation. 2 Finally, they used the current plans (Plan 1977 for Superior and Plan 1958-D for Geologic records indicate that Great Lakes Ontario) and hydraulic routing models of outlet and levels have fluctuated as paleohistoric climates have connecting channel flow and estimated water levels been wetter and drier (Larson, 1985). Recent short- on each of the Great Lakes. term variations have been the result of short-term changes in precipitation patterns. Croley and The regulation plan for Lake Superior failed Hartmann examined the potential impacts of global under the GFDL scenario. To obtain an estimate warming on average lake levels. of changes in levels for Superior-Huron, St. Clair, and Erie, Croley and Hartmann assumed that over Study Design a 30-year period, total inflows into Lake Superior (runoff + overlake precipitation + diversions - Croley and Hartmann used a water supply and evaporation) would equal total outflows, and Lake lake level model of the Great Lakes Basin developed by the Great Lakes Environmental Research Laboratory to estimate the potential ²Iₙ Volume A, Croley focuses on results from his latest run. impacts of climate change on levels of the Great This run includes assumptions that lead to relatively high Lakes (Croley, 1983a,b; Croley, 1988; Quinn, 1978). amounts of evaporation and larger drops in lake levels. Earlier This model is the most detailed hydrologic model of runs had less evaporation and larger drops in lake levels. the Great Lakes Basin and includes a separate Results in this chapter include the latest run and an earlier run. model for each of the 121 watersheds in the basin. 296 Great Lakes Superior levels would not change. No figures are were estimated to continue fluctuating on an annual presented for changes in the level of Lake Superior basis. Specific estimates of fluctuation are not in the GFDL scenario. The levels of Lake Superior discussed here, since variability was assumed not to would probably fall. Only 30-year average lake change. levels were calculated for the other lakes. Croley and Hartmann also found that the flow Limitations in the St. Mary's could increase by less than 1% in the GISS high rainfall scenario and drop by 13% in The relationships in this model were the drier OSU scenario for Lake Superior. The developed for a cool and wet climate. The analysis flow in the Niagara River was estimated to be 2 to did not account for changes in the consumptive uses 30% lower. Croley and Hartmann did not estimate of the lakes (due to population and economic the flow of these rivers for the GFDL scenario. growth or climate change), and it did not consider changes in the regulation plans, or increases in or The lowering of lake levels appears to be additions to diversions into or out of the lakes. The correlated with increased temperatures in the analysis also used the difference in vector winds scenarios. Under all the doubled CO2 scenarios, from the GCMs as a proxy for the difference in there could be declines in runoff to the lakes and scalar winds because GCM estimates of changes of increases in evaporation from the lakes. The scalar winds were not available. Thus, the wind reduction in runoff would be largely the result of estimates probably underestimate changes in changes in snowpack accumulation and ablation. windspeed (David Rind, Goddard Institute for Snowpack in the Lake Superior Basin could be Space Studies, 1988, personal communication). The reduced by one-third to two-thirds, and in the other uncertainty on winds is complicated by the basins, farther to the south, the snowpack could be uncertainties concerning evaporation. Different almost entirely absent. The reduction in runoff assumptions of evaporation in this analysis affect the would reduce average streamflow in the basin. magnitude of lake level drop, but they do not affect These results appear to be driven mainly by the the direction of change lake levels fall under all temperature increase, since precipitation rises in all evaporation assumptions. Cohen (1987a) found that scenarios. potential changes in Great Lakes levels are very sensitive to estimates of changes in windspeed and humidity. He concluded that with the right Table 15-2. Doubled CO2 Scenarios: Reduction in combination of conditions, even with higher Average Great Lakes Levels from 1951 temperatures, it is possible for lake levels to rise. to 1980 (meters) Results Scenario Superior Michigan Erie Ontario Lake levels were estimated to fall significantly under all three scenarios (see Table 15-2). The lake level changes are displayed in ranges from low to GISS -0.43 to -1.25 to -0.95 to NA high evaporation. -0.47 -1.31 -1.16 Average levels for Lake Superior would be GFDL NA -2.48 to -1.65 to NA about 0.4 to 0.5 meters (1.3 to 1.7 feet) below -2.52 -1.91 average levels for the 1951-80 period under the OSU and GISS scenarios. These average levels OSU -0.39 to -0.86 to -0.63 to NA would be generally lower than recorded lows of -0.47 -0.99 -0.80 recent history. The lakes would likely still fluctuate around these average levels, so levels during some Transient Scenario years would be lower. Even though precipitation (average rate of change per decade 1980-2060) rose in all three scenarios, lake levels were estimated to fall, primarily as a result of the higher GISS-A -0.006 -0.055 -0.04 NA temperatures. Apparently, only a large increase in rainfall or humidity or a large decrease in NA = Not applicable. windspeeds could offset these changes. Lake levels Source: Croley and Hartmann (Volume A.) 297 Chapter 15 Evaporation would increase under all three other lake levels fall 0.04 to 0.055 meter (1.6 to 2.2 scenarios. The increase in evaporation varied under inches) per decade. An extrapolation of the different assumptions about the relationship of transient results to the decade of the 2060s (when evaporation to change in climate variables and the GISS A transient run reaches doubled CO₂ ranged from 20 to 48%. For a given assumption climate conditions) results in lake level reductions about evaporation, higher temperature scenarios less than for the doubled CO₂ GISS scenario. This would generally cause more evaporation. Lake level is because lake levels may not respond immediately reductions could also be higher or lower, depending to climate change, but must catch up. The results on these assumptions. may also be affected by the variability assumptions in the transient scenarios (see Chapter 4: All of these changes could cause a reduction Methodology). By the end of the transient scenario, in net basin supply (the sum of overlake the 2050s, lake levels fall at a faster rate by more precipitation and runoff minus evaporation) by 14 to than 0.05 meters (2.0 inches) per decade. Thus, 68%. The exception to this is the GISS scenario for these studies do not clearly indicate the length of Lake Superior. In that scenario, annual rainfall time required for the lakes to drop by the amounts increased by 18%, which could lead to a 1% shown in Table 15-2. increase in net basin supply. Croley and Hartmann found that enough heat The Ontario regulation plan would fail under could reside in Lakes Superior, Michigan, Huron, all scenarios, including the transient run. Under and Ontario to maintain water surface temperatures these conditions, the system would not contain at a sufficiently high level throughout the year, so enough water to keep the level of Lake Ontario and that buoyancy-driven turnovers of the water column the flow in the St. Lawrence River within ranges may not occur at all. This could significantly affect currently specified by the plan. The Lake Superior lakewater quality and aquatic life (see this chapter, regulation plan was estimated to fail under the Thermal Structure of Southern Lake Michigan). GFDL scenario. Although net basin supply in Lake Croley estimated that average surface water Superior increased under GISS, the regulation plan temperatures in the winter would be above 0°C and would require increased flow through the St. Mary's would significantly reduce ice concentrations. River to the water-short lower lakes, resulting in a net drop in Lake Superior levels. Implications These results are consistent with other studies Hydropower production could be reduced, as done on lake levels and climate change. Both flows through the St. Mary's, the Niagara, and the Cohen and Sanderson agree with Croley and St. Lawrence Rivers fall. Losses to hydropower Hartmann that lake levels would drop under various were not estimated for the EPA study, although climate change scenarios. The other two studies, Linder's earlier work on hydropower losses by 2015 however, estimated lake levels would drop less than in New York State showed potential loss of 1500 to 1 meter. Croley and Hartmann may have estimated 2066 gigawatt-hours (6 to 9%) (Linder, 1987). greater changes because they used a more Sanderson (1987) estimated that under a doubled sophisticated runoff, evaporation, and routing model CO₂ scenario, Canadian hydroelectric power and because of different assumptions made about production on the St. Mary's River could rise by evaporation. Croley and Hartmann also used a 2.5% (because the level of Lakes Michigan-Huron more integrated approach and more variables from falls more than that of Lake Superior) and power the GCMs. The estimates for GFDL may also be production on the Niagara River could fall by 13 to higher because the GFDL scenario used in this 18% as a result of a drop in flow. The impacts of study had a higher temperature rise than the GFDL lower lake levels on wetlands were not estimated, scenarios used by Cohen and Sanderson. and the impacts on shipping and on shoreline infrastructure are discussed later in this chapter. The results of the transient run (GISS A) are expressed as the average change in lake level per Lower lake levels and reduced riverflow would decade and are not indicative of what would happen likely adversely affect water quality in the basin. in any particular decade. Lake Superior levels drop Less water would reduce dilution of pollutants. only 0.006 meters (0.2 inches) per decade, while the Forty-two "hot spots" occupy many bays and harbors 298 Great Lakes along the Great Lakes. These are contaminated only the costs of rebuilding infrastructure and did with a wide variety of halogenated organics and not examine ecological impacts. heavy metals, as well as remobilizable nutrients. Lower lakes may cause emergence and near Results emergence of these toxic sediments through erosion, leaching, oxidization, or volatilization. The largest costs appear to accrue to recreational and commercial harbors (see Table 15- Higher temperatures may lead to increased 3). The major expenses are associated with withdrawals of water from lakes for municipal dredging harbors and lowering bulkheads, which consumption. Climate change may also result in could cost approximately $200 to $400 million. If more calls for diversion of water out of the Great lake levels fall enough, keeping some harbors open Lakes Basin for use elsewhere. However, lake (e.g., Waukegan, Illinois) may not be a cost-effective levels may be lowered even more as a result of choice. higher demand for withdrawals for use in the basin as a result of population and economic growth. Changnon et al. concluded that slips and docks would be only slightly affected. Many of these Effects of Lower Lake Levels probably would have been replaced anyway and could be set at lower levels as the lakes fall. (The Coastal infrastructure around the Great Lakes impacts on commercial shipping in Lakes Superior has generally been built assuming average lake and Erie are discussed below.) levels would not change. A drop in levels could make much of the current infrastructure unusable Intake valves for municipal and industrial and necessitate reconstruction. Changnon et al. consumption could be exposed and may have to be examined the potential impacts and adjustments to lowered or moved farther offshore. Outfalls for infrastructure along the 101-kilometer (63-mile) stormwater would have to be extended. Changnon Illinois shoreline. This study and the shipping et al. estimated that extending urban water intakes analysis used the lower range of the lake level drops and stormwater outfalls could cost $16 to 17 million. from Table 15-2 because subsequent analyses that gave different lake levels were performed too late to Although the exposure of more land could be incorporated. present some erosion problems, it could also enlarge many beaches. An additional 1 to 2.2 Study Design square kilometers (0.3 to 0.8 square miles) of beaches would be added to the Illinois shoreline. In Changnon et al. interviewed experts about the all, Changnon et al. estimated that the costs of possible impacts and costs of adjustment along the adjusting to lower levels of 1.25 to 2.5 meters along Illinois shoreline to the lower lake level estimates the Illinois shoreline, excluding normal replacement described above. Results are expressed in current of docks and piers, would be $220 to $430 million. dollars. If normal replacement costs do not account for lower lake levels, costs could be $30 to $110 million Limitations higher. To put these figures into context, the City of Chicago may spend over $800 million to repair This analysis did not use economic models, shorelines damaged by high water levels in recent used current prices, and did not consider changes in years. population, GNP, or technology. Results are based on expert judgment. Changnon et al. also assumed Walker et al. (Volume H; for a discussion of that lakes would reach the levels described above by methodology and results, see Chapter 13: Urban 2030. The change in lake levels may not be reached Infrastructure) examined the potential capacity of until decades later (by the year 2060 or later) so climate change on Cleveland's infrastructure. They costs may be borne over a longer period than found that savings in such areas as snow removal Changnon estimated, allowing for more routine and bridge repair could offset increased cooling and replacement of infrastructure. This study examined dredging costs. Cities on the Illinois shoreline would also have savings due to reduced winter expenditure. 299 Chapter 15 Table 15-3. Estimated Economic Impacts of Lowerings of the Levels of Lake Michigan Over a 50-Year Period (1990-2040) Costᵃ Type of expense 1.25 meters lower 2.5 meters lower Recreational harbors Dredging 30-50 75-100 Sheeting 15 35 Slips/docks 20ᵇ 40ᵇ Commercial harbors Dredging 108 212 Sheeting/bulkheads 38 38 Slips/docks 40ᵇ 90ᵇ Water supply sources Extending urban intakes 15 15 Wilmette Harbor Intake 1 2 Beaches Facility relocations 1-2 1-2 Outfalls for stormwater Extensions and modifications 2 4 Totals $270-292 $512-540 a Costs in millions of 1988 dollars to address future lake levels at indicated depths below average (1951-80) levels of Lake Michigan. b Some costs could be partly covered by normal replacement expenditures over the period of changing levels. Source: Changnon et al. (Volume H). Ice Cover Eastern Basins, and for Whitefish Bay in Lake Superior. Whitefish Bay was included because it Warmer winters would reduce ice cover on has the longest period of ice cover and acts as a the Great Lakes. Some analysts have speculated choke point on shipping in and out of Lake that ice would be completely eliminated. Assel used Superior. Lakes Superior and Erie represent a model to estimate the potential extent and extremes in terms of air temperature regimes, lake duration of ice cover. depth, and heat storage capacity, and bound the range of potential ice cover changes. Study Design Limitations Assel developed a statistical relationship between temperature and ice cover for this study. Assel's study did not consider the effects of The models were developed for the three basins of wind and other variables on ice formation. Lake Erie, for the Lake Superior Western and Implicitly, the analysis assumed that winds stay the 300 Great Lakes same. Stronger winds would make the ice season shorter than estimated, and weaker winds (and calmer waters) would make it longer. The three 100 GCMs estimate that windspeeds over the two lakes BASE drop by 0.0 to 0.3 meters per second (see Croley, GISS 1981-2009 GISS 2010-2039 Volume A). Inclusion of windspeed changes would GISS 2xCO2 80 have lowered ice cover reduction results. The GFDL 2xCO2 OSU 2xCO2 model was built based on the relatively cool years of the 1960s and 1970s; therefore, the doubled CO₂ scenario temperatures are outside the range of winter temperatures in those years. However, the model simulated ice duration within 3 weeks of actual ice duration for the warm winter of 1982-83. Percentage of Basin Ice Covered 60 40 Results 20 Assel found that although average ice cover might be significantly reduced, ice would still form on the lakes (Table 15-4). Results for the central 0 Dec Jan Feb Mar Apr basin of Lake Erie are displayed in Figure 15-3. It Month now averages 83 days of ice cover. In the 1981-2009 transient scenario, ice cover was estimated to be 71 days; in the 2010-2039 scenario, it was estimated to Figure 15-3. Changes in duration and extent of ice decline to 41 days. Under the doubled CO₂ cover in central basin of Lake Erie under transient climate, ice cover could be reduced to a total of 6 to and doubled CO₂ scenarios (Assel, Volume A). 19 days, and ice formations would be generally limited to near-shore and shallow areas. Whitefish Bay in Lake Superior currently averages about 115 The temperature rise in the scenarios may not days of ice cover. Under the doubled CO₂ be warm enough to eliminate ice cover on the Great scenarios, ice duration would be reduced to 69 to 80 Lakes, but many winters could have no ice at all. days. Also, the maximum percentage of Whitefish The Lake Erie Central Basin is estimated to be ice- Bay covered by ice would be reduced from close to free from 11 to 22 years out of 30 years, rather than 100% to 70-20%. 1 out of 30 years, as estimated for base climate Table 15-4. Reduction in Ice Cover in Lakes Erie and Superior (average annual days of cover) Base GISS Transient A Doubled CO₂ Analog Lake 1951-80 1981-2009 2010-2039 GISS GFDL OSU 1930s Erie West 93 84 54 26 23 35 85 Erie Cent 83 71 41 8 6 19 61 Erie East 97 82 43 6 5 13 70 Supr West 112 108 88 46 24 75 106 Supr East 108 103 84 43 19 69 103 Supr WFB 115 109 92 55 26 80 112 Abbreviations: Supr = Superior; WFB = Whitefish Bay; Cent = Central. Source: Assel (Volume A). 301 Chapter 15 conditions. This result appears to be sensitive to Whitefish Bays in Lake Superior; and Toledo, depth, as estimates indicate that the deeper Lake Cleveland, and Buffalo in Lake Erie. They used the Erie East Basin would be ice-free 60 to 84% of the "ECO Great Lakes Shipping Model," which includes time, and the shallow West Basin would be ice-free current data on major ports and commercial ships in 7 to 17% of the winters. Since it is colder, Lake in the Great Lakes, types of cargo, costs of Superior would have ice cover in virtually all winters transport, and operating costs. Keith et al. used under the scenarios. lake level reductions from Croley and Hartmann to study the change in cargo capacity and costs per ton, Assel found that ice cover reductions during and they used the change in cargo capacity to the first 30 years of the transient scenario (model estimate how many days of shipping would be years 1981-2010) may not be significantly different needed to transport the same amount of cargo as than under current conditions. The length and transported at present. The latter figure was extent of ice cover noticeably decline, beginning in compared to ice duration reductions estimated by the second 30 years of the transient scenario (2011- Assel to determine whether the shipping season was 40). By the last decade of the transient scenario, sufficiently extended to allow for transport of the the 2050s, the extent of ice cover was almost same amount of annual cargo as currently identical to the GISS doubled CO₂ coverage. transported. Croley also found that ice cover would be Limitations reduced. His analysis found that average surface temperatures on all the lakes in the winter could be The analysis did not consider changes in the above 0°C. Even if average temperatures are that composition of the fleet or in the mix and amount high, water temperatures in near-shore and shallow of cargo. It also assumed that demand for shipping areas, the areas to which Assel said ice would be of goods did not change, even in response to limited, would be sufficiently cold to cause ice changes in availability of shipping. The analysis did formation. not examine whether goods would shift to or from alternate ports or means of transportation and how Implications changes in the costs of shipping and in the shipping season would affect users. Keith et al. also assumed Ice cover reductions could have positive and that channels were not dredged to be deeper. Thus, negative effects. On the positive side, the shipping analysis is useful for estimating the direction and season would be extended (see below). Water approximate magnitude of change, but quantitative would flow more freely through rivers and results should be interpreted with caution. connecting channels, allowing for more hydropower production in the winter. On the other hand, ice Results protects some aquatic life, such as whitefish, and protects shorelines against the erosive impact of The costs of shipping were estimated to high-energy waves (Meisner et al., 1987). increase as a result of lower lake levels. The effect on the cargo load for ships using the Port of Buffalo Shipping are displayed in Figure 15-4. Under drops of 0.7 to 1.0 meter in Lake Erie, which are the lake level With lower lake levels, ships would have to reductions estimated by Croley for the OSU and reduce their cargo, or ports and channels would GISS scenarios, cargo capacity would decrease by have to be dredged. However, the shorter duration about 5 to 13%, and costs per ton would rise by the of ice cover would allow for a longer shipping same amount. Croley's estimate from the GFDL season. The additional days of transport may make scenario was that Lake Erie would fall 1.65 meters up for the loss of capacity on each voyage. (5.4 feet), but the shipping model does not include lake level drops of more than 5 feet. A drop of 5 Study Design feet would decrease cargo capacity per voyage by 27% and increase costs by 33%. Thus, the drop in Keith et al. studied the potential impacts of lake levels estimated under the GFDL scenario changes in lake levels and ice cover on shipping in could increase costs by more than 33%. Since lake six ports: Two Harbors; Duluth/Superior and levels in Lake Superior were not estimated to fall as 302 Great Lakes GFDL 35 100 30 GISS 80 25 OSU 20 60 15 40 10 PERCENT CHANGE 20 5 0 0 DAYS -5 -20 -10 Increase in Costs -40 -15 -20 Reduction in Tonnage/Voyage -60 -25 Additional Days Required to Transport Same Amount of Cargo -80 -30 -35 -100 0 1 2 3 4 5 WATER LEVEL REDUCTION (Feet) Figure 15-4. Impacts of lower lake levels and reduced ice cover on shipping, cargo capacity, costs, and days of transport for the Port of Buffalo (Keith et al., Volume H). much, the corresponding reduction in cargo capacity to 3-foot drop of the wetter and relatively cooler for ships on those ports would be in the range of 2 OSU and GISS scenarios, another 15 to 40 days of to 8%. shipping would be needed. Assel estimated that under those scenarios, ice duration in eastern Lake Sanderson estimated that lake level reduction Erie would be reduced by 84 to 91 days. Thus, of 0.2 to 0.6 meters would increase total Canadian under these scenarios, even with reduced capacity shipping costs by 5%, assuming the current fleet and per voyage, there would be enough additional days mix stayed the same. Although results are not of travel to transport even more goods. If lake directly comparable, since Keith et al. examined levels fell 5 feet, which is less than estimated by U.S. flagships and ports while Sanderson studied GFDL, an additional 100 days of transport would be Canadian ships and ports, the estimates are of the needed to handle the same amount of cargo. Ice same magnitude. duration in eastern Lake Erie could be reduced by 92 days under this scenario, which would not allow Whether the same amount of annual cargo can enough time to transport the same amount of cargo, be transported, assuming no dredging to deepen assuming the current fleet and demand for channels, depends mostly on how much lake levels transport. The results appear to be more sensitive drop. If the drop is sufficiently large, annual to changes in lake levels than to reductions in ice tonnage could be reduced. The following discussion cover. assumes that lake level declines occur at the same time as ice cover reductions. It is not clear from Keith and Willis used current dredging costs to these studies whether lake levels will respond more estimate the cost of dredging the ports to restore slowly to climate change than ice cover. Figure 15- current channel depths. The total costs of dredging 4 also displays the additional days needed to the three ports in Lake Erie range from $7 to $31 transport the same amount of cargo as is currently million per port (1987 dollars). Current annual shipped through Buffalo. Under the approximate 2- dredging costs for those ports range from $800,000 303 Chapter 15 per year in Buffalo to $2.5 million per year in limited baseline climate variability, although these Toledo (J. Hasseler, U.S. Army Corps of Engineers years include cold and warm periods. The results Buffalo District, 1988, personal communication). are most sensitive to changes in windspeed. Since the scenario may underestimate reductions in Implications windspeed from the GCMs (see the discussion of the limitations of the lake level study), this analysis Reduction in the tonnage per voyage or may overestimate wind-driven mixing in the upper increased costs for dredging would raise shipping layer and underestimate changes in the length of costs. However, with a longer shipping season, time and degree of stratification. On the other users of shipping such as powerplants would not hand, if the intensity of summer storm increases, have to carry large inventories to last through the then stratification may be weakened and shortened. winter and own enough land to store those The analysis assumed there was no change in the inventories. Besides reducing costs, this could allow frequency of storms. More summer storms may current lakefront storage areas to be used for other weaken stratification, while fewer storms could purposes. Whether these savings would offset strengthen stratification. higher shipping costs was not examined. Results Dredging the ports and channels could degrade the water quality of the lakes. The McCormick estimated that the length of the sediments in many of these ports are toxic, and stratified season could increase under all three disposal of the sediments could be complicated by scenarios. Figure 15-5 displays the mixed-layer their toxicity and by the reduced disposal areas depth over an average year. The higher heat resulting from lower lake levels. content may cause the lake to begin to thermally stratify, on average, about 2 months earlier than in Water Quality the base case (in April as opposed to June). The stratified layers were estimated to begin to deepen Two studies estimated the temperatures and around late fall, as under current climate conditions. thermal structures of southern Lake Michigan and the Lake Erie Central Basin. The Lake Erie study estimated biological activity, such as algal production and changes in dissolved oxygen levels. The Michigan and Erie analyses were used by Magnuson et al. to study changes in the thermal 0 habitats of fish. Thermal Structure of Southern Lake Michigan 45 Study Design DEPTH (m) 90 McCormick used a one-dimensional thermal structure model (Garwood, 1977) to estimate the BASE heat content and structure of a site in south-central 135 GISS GFDL Lake Michigan. The model has been successfully OSU applied to oceans and inland seas and was used by 180 McCormick to analyze a site 150 meters (500 feet) J F M A M J J A S o N D deep. GCM data for windspeed, temperature, MONTH humidity, solar radiation, and cloud cover were applied to hourly data from 1981 to 1984. Limitations Figure 15-5. Average annual mixed-layer depth in southern Lake Michigan (McCormick, Volume A). McCormick used the years 1981-84 as his base case because hourly water temperature data are not available for 1951-80. Three years provide very 304 Great Lakes Surface lake temperatures were estimated to model is similar to the one used by McCormick for be up to several degrees higher than in the base southern Lake Michigan. case. The increase in surface temperatures was greater than the increase in subsurface Blumberg and DiToro then examined the direct temperatures. There appears to be a larger effects of changes in the thermal structure on warming of the entire water column in the winter, aquatic life in the basin. The outputs from the about 2 to 3°C, than in the summer, which has a thermal model were fed into a eutrophication model warming of about 2°C. The warmer lake that had been previously developed by DiToro temperatures are consistent with the studies of (DiToro and Connolly, 1980). The latter model Croley and Assel, which suggest that midlake water estimates what would happen to dissolved oxygen would generally be ice-free. The earlier onset of levels in the lakes by simulating the interactions stratification, reduced winds in the scenarios, and between nutrient availability and biological (e.g., greater temperature differences between lake layers plankton) activity. could yield stronger density differences between upper and lower layers. The models were run using only two base years, 1970 and 1975. In 1970, the thermocline (density McCormick detected a significant decrease in gradient between the upper and lower layers) was the frequency of complete mixing of the lakes. The deep, and over 60% of the hypolimnion (lower surface layer could be warmer and more buoyant, level) in the Lake Erie Central Basin was anoxic making it more difficult for entrainment and mixing (depleted of oxygen). In 1975, the thermocline was to occur. Temperatures were too warm in the shallow, and less than 10% of the lower layer was winters of some years to allow the lake to become anoxic (DiToro et al., 1987). isothermal (the mixed layer would stay above the bottom of the lake all year), leading to a year-long Limitations stratification. This result is consistent with Croley's analysis. Although the two base years encompass a wide range of baseline anoxic conditions, they do not Implications represent a full range of climate variability. In addition, as in the Lake Michigan study, the Reduced turnover of the lakes could have scenario assumed no change in the frequency of serious implications for aquatic species in the lakes. storms. More summer storms would weaken Mixing of oxygen and nutrients could be disrupted, stratification and increase dissolved oxygen levels, possibly affecting the abundance of life in the lower while fewer storms would have the opposite effect. and upper layers of the lakes. The analysis did not incorporate the actual reduction in nutrient loadings from the base years, Eutrophication of the Lake Erie Central Basin or the estimated drop in lake levels from Croley's work. Lower lake levels would reduce the volume Nutrient loadings have made many areas of of the lower layer in Lake Erie, possibly increasing the shallow Lake Erie eutrophic at times. The eutrophication. The models were not run for the shallow western and central basins of the lake are winter, but Blumberg and DiToro tested the particularly vulnerable to eutrophication. sensitivity of results to higher water column Installation of pollution controls in recent years has temperatures (due to warmer winter air improved water quality. Blumberg and DiToro temperatures) in the spring and found no significant analyzed whether climate change would have an difference in results. Blumberg and DiToro used effect on eutrophication in the Lake Erie Central the vector wind estimates from the GCMs, which Basin. may overestimate mixing in the upper layer. Study Design Pollution loadings in 1970 and 1975 were much higher than they are today. Use of current pollution Blumberg and DiToro modeled the thermal loadings would have resulted in higher estimates of structure of the Lake Erie Central Basin. They dissolved oxygen levels and lower estimates of the developed a thermal model for the basin, using a area of the basin that could become anoxic. The modeling framework previously designed by Blumberg (Blumberg and Mellor, 1983). This 305 Chapter 15 direction of change estimated by Blumberg and DiToro would not have been affected. Results AUGUST 1970* AUGUST 1975* Blumberg and DiToro estimated that the Lake BASE CASE Erie Central Basin could remain stratified about 2 to 4 months longer than under current conditions, with the stratified season starting 2 to 6 weeks 40.6% 0.0% sooner and ending 2 to 7 weeks later. The GISS temperature differences between the upper and lower layers of the basin were estimated to be W greater under all scenarios, leading to less exchange 80.5% 0.0% of nutrients across the thermocline. The depth of GFDL the thermocline appears to be most sensitive to estimated changes in windspeeds. In two scenarios, W GISS and GFDL, windspeeds were generally lower, 94.4% 5.9% and the thermocline was estimated to be about 2 OSU meters higher than current depths. Under the OSU scenario, windspeeds were estimated to increase and the thermocline was estimated to be approximately 100% 28.8% 1 meter deeper than current levels. A lowering of the thermocline depth by 2 meters in the 25-meter- deep Lake Erie Central Basin can reduce the * Base Case Years ///// Area That Is Anoxic (Has No Oxygen) volume of the lower layer by 20%, limiting total oxygen availability. All three scenarios generally led to decreases in dissolved oxygen levels compared with base case Figure 15-6. Area of central basin of Lake Erie that conditions despite differences in thermocline depth. becomes anoxic (Blumber and DiToro, Volume A). The increase in area of the Lake Erie Central Basin that was estimated to become anoxic is shown in Figure 15-6. Dissolved oxygen levels were estimated longer in Lake Erie and begins earlier and breaks to increase only in the July 1970 case, and this up at the same time as the present stratification in occurred because the levels were near zero to begin Lake Michigan. It is not clear whether this with. Blumberg and DiToro concluded that the difference is attributable to different lake depths, to difference in oxygen content was caused by warmer surface meteorology used to force the models, or to lake temperatures, which raise biological activity surface boundary conditions in the calculations. enough to increase oxygen demand. The enhanced biological activity was combined with a more intense Implications and longer stratified season to further lower dissolved oxygen levels. Lower thermocline depths, Decreased dissolved oxygen levels could make such as in the OSU scenario, result in even greater the Lake Erie Central Basin less habitable for decreases in dissolved oxygen levels. finfish and shellfish during the summer. This could reduce recreational uses of the lake such as The estimated changes in the thermal swimming, fishing, and boating. It also could put structure of Lake Erie are comparable to more pressure on reducing sources of pollutants, McCormick's results for southern Lake Michigan. especially such nutrients as phosphorous, from point Both estimated that average temperatures in the and nonpoint sources. water column would rise, that there would be greater differences in temperature between the Fisheries epilimnion and hypolimnion, and that stratification would last longer. One major difference in the The Blumberg and McCormick studies show results is that stratification begins earlier and lasts that climate change would probably raise lake 306 Great Lakes temperatures and reduce oxygen levels in certain Limitations areas. To get an initial sense of what these changes might mean for Great Lakes fish, Magnuson et al. The study did not examine the combined effects examined the potential ecosystem, organism, and of reduced habitat and greater need for forage in population responses to warmer temperatures. the summer, which would combine to intensify species interactions. The analysis did not Study Design incorporate impacts resulting from lower lake levels, such as possible loss of wetlands, and it did not Magnuson et al. estimated changes in fish analyze the aquatic effects of the potential reduction habitat, growth, prey consumption, and population in the frequency of lake turnover or the impacts of for sites in Lakes Erie, Michigan, and Superior. The a reduction in ice cover. The introduction of new work used several approaches and models to species, which could have negative impacts on examine the following: existing fish, was not examined. Changes in ecosystem activity, such as Any uncertainties associated with the changes in phytoplankton populations, were McCormick and Blumberg studies would be carried estimated by using a community "Q₁₀" rule over into the analysis on habitat. These changes in (Ruttner, 1931), which approximates the the lakes and littoral systems may have negative higher biological activity associated with impacts on Great Lakes fish. These uncertainties higher temperatures. could reverse the direction of results and lead to more declines in fish populations than indicated Magnuson et al. used the Blumberg and here. McCormick thermal structure studies to estimate the potential effects on thermal Results habitats - the niche in which temperatures are optimum for fish. To estimate changes Phytoplankton production, zooplankton in habitats, the study used laboratory biomass, and maximum fishery yields were estimates of the temperature regimes estimated to increase 1.3- to 2.7-fold, with the preferred by fish (Magnuson et al., 1979; largest increase in phytoplankton production (1.6- Crowder and Magnuson, 1983) and to 2.7-fold) (Figure 15-7). The larger increases in assumed that the lower layer of the Lake biological activity were generally associated with Erie Central Basin is uninhabitable. In larger temperature increases. The increase in addition, using a thermal model for streams phytoplankton provides more forage for (Delay and Seaders, 1966), the study zooplankton, which, in turn, provides more forage calculated the change in habitat for brook for fish. The increase in phytoplankton can also trout in a southern Ontario river. enhance eutrophication, as was estimated by Blumberg and DiToro. Magnuson et al. used a food consumption and conversion model (Kitchell et al., 1977) Magnuson et al. found that the average annual to estimate the changes in annual growth thermal habitat for all fishes would increase. This and prey consumption at three near-shore was especially apparent for lake trout, which is a sites in Lakes Superior, Michigan, and Erie. coldwater fish with a preference for very cold water, This analysis assumed that consumption and which could have more than a 100% increase in rates increase with climate warming. habitat (see Figure 15-8). The major reason for the Growth simulation for Lake Michigan using increase in habitat is that more habitable waters water temperature scenarios from would be found in the fall, winter, and spring. On McCormick assumed that prey availability the other hand, hotter temperatures could decrease did not increase. This study assumed that summer habitats for certain species by 2 to 47%, fish migrate to habitable sites when inshore depending on the temperature rise and species. The temperatures are too warm. length of stream suitable for brook trout in the summer could be reduced by 25 to 33% because of higher temperatures. 307 Chapter 15 0 COLD REGION COOL REGION WARM REGION 50 3 BASE CLIMATE PHYTOPLANKTON PRODUCTION 100 2 0 50 PRODUCTIVITY INCREASE (2XCO₂ BASE) 1 OSU 3 ZOOPLANKTON BIOMASS DEPTH (M) 100 0 2 50 GISS 1 100 3 FISHERY MAXIMUM SUSTAINED YIELDS 0 2 50 GFDL 100 1 JAN MAR JUN SEP DEC OSU GISS GFDL MONTH HABITAT: + 2°C OF OPTIMUM TEMPERATURE + 5°C OF OPTIMUM TEMPERATURE Figure 15-7. Increases in Great Lakes aquatic Figure 15-8. Increase in lake trout habitat productivity (Magnuson et al., Volume E). (Magnuson et al., Volume E). Fishes were generally estimated to have recruitment, and Meisner et al. concluded that loss increased body size under the scenarios. Cool and of wetlands due to lower lake levels could reduce cold coldwater fishes could have 20 to 70% more spawning, nursery, and feeding grounds for fish in growth, and warmwater fishes in warm areas could shallow areas, reducing fish populations (Meisner et have 220 to 470% more growth. This assumes that al., 1987). prey availability increases. If prey availability does not increase, fish growth would also decrease owing Implications to an inability to compensate for the increased metabolic costs of living in higher temperatures. Fish populations could increase, with beneficial Magnuson et al. calculated that if prey availability implications for commercial and recreational fishing, does not increase, fish growth in Lake Michigan although certain species, such as brook trout in could decrease by 10 to 30%. Warmwater fish streams, may be reduced. A net increase in would have larger decreases if prey did not increase. fisheries would lead to more employment in Furthermore, the increased demand for forage may commercial fishing and tourism industries, but intensify species' interactions and alter the food web would increase the need for maintaining water structure. quality in the lakes. Increased demand on the forage base by predators and the introduction of The effects of reduced ice cover and possible new species and reduced ice cover could have reduction in wetlands on Great Lakes fishes was not negative effects, but these cannot be predicted and investigated, although Freeberg (1985) suggests that must be considered as surprises of unknown a reduction in ice cover would reduce whitefish probability. 308 Great Lakes Forests al., 1972, 1973). This model, which is known as a stand simulation model, can be used to estimate the Climate change could affect the distribution transitional changes in composition and abundance and abundance of forests in the Great Lakes region. of forest species in response to environmental Overpeck and Bartlein examined the equilibrium changes such as higher temperature and range shift of forests, Botkin et al. studied precipitation. transitional impacts on composition and abundance, and Zabinski and Davis analyzed the ability of trees Botkin et al. studied two diverse sites in the to migrate along with a rapidly changing climate. Great Lakes region. The first is in Mt. Pleasant, Michigan, a heavily settled area dominated by Potential Range Shifts northern hardwoods and oaks, where commercial forests are an important resource. The other site is Study Design in Virginia, Minnesota, an undeveloped area dominated by boreal forests that have commercial Overpeck and Bartlein studied the potential and recreational uses. shifts in ranges of forest types over eastern North America. This analysis suggests where trees are Limitations likely to grow in equilibrium doubled CO₂ climate conditions after allowing for migration of tree The model includes all dominant tree species species to fully catch up with climate change (see in the northern United States and assumes that Forest Migration). It indicates only the seeds from all these trees are universally available approximate abundance of different species within throughout the region. Species with predominantly a range, not what the transitional effects of climate southern distributions are not included; therefore, on forests might be, or how fast trees will be able to the model does not estimate whether they could migrate to the new ranges. (For a discussion of the grow in the region under the warmer climate. study's methodology and limitations, see Chapter 5: (Overpeck found that southern pines may migrate Forests.) into the southern part of the region.) Thus, the stand simulation model does not accurately estimate Results migration of trees, either within the region or from other areas. Furthermore, the results do not assess Under all three doubled CO2 scenarios, the whether transplantation by humans of more range of spruce, a major component of the boreal southern species would be successful. In addition, forests, could shift almost entirely out of the region. the model does not account for fertilization effects Northern hardwoods, such as birch and northern of CO2, although CO₂ may not have positive effects pine species, would shift to the north but may still in the competitive environment of unmanaged be in the region. Oak trees, which are mostly found ecosystems (see Botkin et al., Volume D). Botkin in the southern part of the region, would be found et al.'s analysis did not account for introduction of all over the region in the warmer conditions. The new pests into the region, for the possibility of abundance of prairie forbs (shrubs) would increase increased frequency of fires, or for the combined in the region, and southern pines could eventually impact of changes in tropospheric air pollution migrate to the southern part of the region. levels and UV-B radiation. Transitional Effects Results In contrast to Overpeck and Bartlein, Botkin Botkin et al. estimated the doubled CO2 et al. examined the transitional effect of climate climate would cause major changes in forest change on forests as well as doubled CO₂ effects. composition throughout the region. Results from the Mt. Pleasant site indicate that tree biomass at Study Design dry sites, which now have oak and sugar maple, could be reduced by 73 to 99% and could convert to Botkin et al. used a model of forest species oak savannas or even prairies. Relatively wet soil growth and competition to estimate the effects of sites might be converted from sugar maple to mostly climate change on Great Lakes forests (Botkin et 309 Chapter 15 oak woodlands with some red maple. Biomass at that more southern species could be transplanted to these sites could be reduced by 37 to 77%. these sites, although this was not studied. In the Minnesota site, the boreal forests could In both sites, the biggest decline is seen in the be replaced by northern hardwood forests, now hotter and drier GFDL scenario. Decreased soil characteristic of areas to the south (see Figure 15- moisture, which is a result of higher temperatures 9). Relatively dry areas, such as the Boundary and reduced rainfall, appears to be the most Waters Canoe Area where balsam fir dominates, significant factor reducing biomass. and upland areas where white birch and quaking aspen dominate, could be replaced by forests Botkin et al. found that the abundance of consisting mainly of sugar maples. Where currently species could significantly change in three to six saturated soils in these upland areas become drier decades. Figure 15-10 displays results from the and better sites for tree growth, wood production transient scenarios for balsam fir and sugar maple may increase. However, bogs that now contain at the Minnesota site. The basal area of balsam fir white cedar could become treeless. This is because could start to decline in three to six decades. no species that could tolerate warmer bog Potential declines in several decades are also seen conditions are currently in the region. It is possible in simulations of white cedar and white birch in the NORTHERN MINNESOTA IN 1980 BALSAM FIR White Birch Balsam Fir 8000 Quaking Aspen NO CLIMATE CHANGE GISS A 6000 White Cedar 120 Very Dry 100 Wetland 80 60 Soil Centimeters BASAL AREA (cm sq/100 m sq.) GISS B 4000 40 2000 20 Bog Water Table 0 Base of Soil 1980 2000 2020 2040 2060 2080 Bedrock YEAR NORTHERN MINNESOTA 2xCO2 Sugar Maple SUGAR MAPLE 8000 Sugar Maple NO CLIMATE CHANGE GISS A 6000 GISS B 100 80 40 Centimeters BASAL AREA (cm sq./100 m sq.) 4000 120 Soil 60 2000 Treeless Bog 20 0 Water Table 1980 2000 2020 2040 2060 2080 Base of Soil YEAR Bedrock Figure 15-9. Changes in composition of northern Figure 15-10. Change in forest composition during Minnesota forests (Virginia, Minnesota; soil depth the next century for a deep, wet, sandy soil in = 1.0 meter; water table depth = 0.8 meter) northern Minnesota (Botkin et al., Volume D). (Botkin et al., Volume D). 310 Great Lakes Minnesota site. Sugar maple, which has negligible Results basal area in the current climate, was estimated to start to exhibit significant growth within three Under the wetter GISS scenario, the potential decades in both transient scenarios. ranges of sugar maple, yellow birch, hemlock, and beech move markedly northward to central Canada. Forest Migration The results for hemlock and sugar maple are displayed in Figure 15-11. The stippled area shows Both Overpeck and Botkin assumed that trees the potential range, and the black area shows how would be able to migrate to new locations (although far the trees could migrate by 2090. Zabinski and Botkin did not assume southern species would be Davis found that hemlock, yellow birch, and sugar able to migrate into the Great Lakes region). maple could become much less abundant in the Zabinski and Davis examined the potential range parts of Wisconsin and Michigan where they shifts of sugar maple, yellow birch, hemlock, and currently grow. Beech may be completely beech currently found in the Great Lakes region eliminated from the lower peninsula of Michigan and compared that shift with potential rates of where it is presently abundant. In addition, the rate migration. of migration would be slower than the climate change. The trees would not migrate as far as the Study Design northern boundary of the climate range (the stippled area). The southern boundary would be Zabinski and Davis assumed that tree species driven northward by climate change. Since the shift grow only in climates with temperatures and in climate zones is faster than the assumed rate of precipitation identical to their current range. They migration, the southern boundary would move north determined the location of potential species ranges faster than the northern migration rates. The total under the GISS and GFDL scenarios. The climate range of all four species would be reduced. values were determined by extrapolating between gridpoints. Zabinski and Davis examined the Under the GFDL scenario, which is the hottest potential migration of the species by assuming that and driest, all four species are eliminated from the the doubled CO₂ climate would not occur until Great Lakes region. Northern hardwood tree 2090, and that these species could migrate into new species might be replaced by trees characteristic of regions at the rate of 100 kilometers (62 miles) per more southern latitudes or by prairie or scrubland. century. Since the southern range of the trees moves farther north than in GISS, the inhabited range would be Limitations much smaller than under GISS. Zabinski and Davis found that all four tree species would be confined to The study did not consider human an area in eastern Canada having a diameter of only transplantation of seedlings to speed migration. several hundred kilometers. The analysis did not consider competition among species or whether migratory routes would be The ability of the four species to survive in blocked. It also did not analyze whether species more northern latitudes may depend on whether could survive in the soil conditions, nutrient they could adapt to different day lengths and soils. availability, sunlight, and other relevant factors in northern areas. Doubled CO₂ climate conditions Implications of Forest Studies could occur sooner than 2090, resulting in greater range reductions. The rate of forest migration used All three studies, through different analytic is double the maximum rate ever recorded for approaches, agree that the scenarios of climate temperate trees. A faster warming and slower change would produce major shifts in forest migration would make it more difficult for forests to composition and abundance. Boreal forests would keep up with shifts in range attributable to climate most likely no longer exist in the region. Northern change. Zabinski and Davis did not consider hardwood forests might still be present, especially in whether higher atmospheric CO₂ concentrations the north. Uncertainty exists concerning whether would mitigate the decline of forests along southern forests in the southern part of the region will die boundaries of their ranges. back leaving grasslands or whether new species will be able to migrate or will be transplanted and flourish. 311 Chapter 15 Hemlock Present Range Range After 2050: GISS Range After 2050: GFDL Sugar Maple Present Range Range After 2050: GISS Range After 2050: GFDL Potential Range Scale 0 400Km Inhabited Range Figure 15-11. Shifts in range of hemlock and sugar maple (Zabinski and Davis, Volume D). The rapid rate of climate change, coupled with the for production of pulp, paper, and construction presence of urban areas and extensive farmland in materials. These species would decline and would the southern Great Lakes States, may impede be replaced by oaks and maples, which are useful migration of southern species into the region. Such for furniture but take longer to become fully grown. a shift could result in increased soil erosion and Red maple, which may be more abundant in the decreased water quality. In addition, higher tree southern area, is not currently used commercially. mortality and drier soils could increase fire Changes in forest abundance may also affect frequency. There also may be an increase in tourism and recreation. pathogen-related mortality in trees. Shifts in forest composition and abundance may have implications Agriculture for wildlife in the region. The agriculture studies combined analyses of This shift in species also could have significant impacts on the region and across the country. impacts on the commercial forest industry in the Ritchie et al. studied the potential impacts of region. The industry currently harvests softwoods climate change on crop yields in the region. Adams 312 Great Lakes et al. then used the results from this study and other very large because current yields are very low regional crop yield analyses to estimate economic relative to other sites. adjustments by farmers. Easterling studied how a typical Illinois corn farmer would try to adapt to Results climate change. Ritchie et al. found that temperature and Crop Yields precipitation changes alone could reduce crop yields everywhere in the region, except in the Study Design northernmost latitudes, such as Duluth, where yields could increase depending on rainfall availability. Ritchie et al. used crop growth models to Results from selected sites are displayed in Table estimate the impacts of climate change on yields for 15-5. Corn yields could decrease from 3 to 60%, corn and soybeans in the Great Lakes States (Jones depending on climate and water regime (dryland or and Kiniry, 1986; Wilkerson, 1983). The two irrigated). However, Duluth, the most northern physiological models examine the direct effects of site, could see increases of 49 to 86%. Current temperature and precipitation on crop yields. dryland and irrigated corn yields are lower in Ritchie et al. also used simple estimates of Duluth than in the more southern sites. Dryland increased photosynthesis and decreased yields in Duluth under climate change could be transpiration to conduct a sensitivity analysis of the equal to other sites, and irrigated yields could combined impacts of change in weather and CO₂ exceed the other locations. fertilization on crop yields. In addition, they studied whether crop varieties currently in southern areas Dryland soybean yields are estimated to drop may mitigate climate effects. by 3 to 65% in the region, except in the north. There, dryland yields may decrease by 6% under Limitations GFDL but increase by 109% under the wetter GISS. Under irrigated scenarios, soybean yields in the The direct effects of CO₂ in the crop modeling north increase by 96 to 153%. Even with the study results may be overestimated for two reasons. increase in output, the soybean yields in Duluth may First, experimental results from controlled still be lower than in areas to the south. environments may show more positive effects of CO2 than would actually occur in variable, windy, The reduction in yields in the south would be and pest-infested (weeds, insects, and diseases) field due mainly to the shorter growing period resulting conditions. Second, because other radiatively active from extreme summer heat. Production in the trace gases, such as methane (CH₄) also are north is currently limited by the long winter, so a increasing, the equivalent warming of a doubled longer frost-free season results in increased yields. CO2 climate may occur somewhat before an actual doubling of atmospheric CO₂. A level of 660 ppm Ritchie found that the demand for irrigation CO₂ was assumed for the crop modeling would rise between 20 and 173% under the GFDL experiments, while the CO₂ concentration in 2060 is scenario and up to 82% under GISS, although some estimated to be 555 ppm (Hansen et al., 1988). sites under GISS were estimated to have reductions in demand of up to 21%. All the scenarios assumed that by having low salinity and no compaction, soils would be relatively The combined effects of higher concentrations favorable for crops, and there were would be no of CO2 and climate change could increase yields if limits on the supply of all nutrients. In addition, the sufficient rainfall is available. If it is not, yields analysis assumed farmers would make no could rise or fall. Dryland corn and soybean yields technological adjustments to improve crop yields or may rise up to 135% under the GISS scenario and introduce new crops. Possible negative impacts due up to 390% in Duluth. In the dry GFDL scenario, to changes in storm frequency, droughts, and pests however, yields could fall up to 30% or rise up to and pathogens were not factored into this study. 17%, again except for Duluth, which has an increase The results could be significantly affected by such of 66 to 163%. Irrigated yields for corn rise and fall changes. The percentage changes for Duluth are under both scenarios, but irrigated soybean yields 313 Chapter 15 Table 15-5. Effects of Climate Change Alone on Corn and Soybean Yields for Selected Sites in Great Lakes States (ranges are GISS-GFDL and are % change from base) Corn Soybeans Site Dryland Irrigated Dryland Irrigated Duluth, MN +49 to -30 +86 to + 36 + 109 to -6 + 153 to +96 Green Bay, WI -7 to -60 -3 to -44 -3 to -65 +3 to -26 Flint, MI -17 to -48 -14 to -38 -6 to -51 +6 to -11 Buffalo, NY -26 to -47 -18 to -38 -21 to -53 +6 to -6 Fort Wayne, IN -11 to -51 -15 to -48 -2 to -58 0 to -19 Cleveland, OH -26 to -50 -19 to -43 -16 to -59 -1 to -14 Pittsburgh, PA -22 to -55 -19 to -45 -13 to -59 0 to -13 Source: Ritchie et al. (Volume C). could rise 43 to 72% in the south and up to 465% Regional Shifts in Duluth. The combined effects lead to an estimated reduction in demand for irrigation for Ritchie et al.'s analysis only estimates changes corn of 26 to 100% under both scenarios, whereas in potential yields for the Great Lakes region. How irrigation needs for soybeans under GFDL rise by much farmers actually grow will depend in part on 65 to 207% and range in GISS from a reduction of what happens elsewhere. If the relative productivity 10% to an increase of 32%. of agriculture rises, farmers will probably increase output. If relative productivity falls, they would Ritchie found that use of a longer season corn most likely cut back. Adams et al. examined how variety could reduce the negative effects of climate different regions of the United States may react to alone, under the GFDL scenario, but would still potential productivity changes. Results are result in net losses. presented here for the Great Lakes region only. It is not clear whether crop yields would rise Adams et al. modeled potential nationwide or fall in the region. Among other factors, this will shifts in crops using the Great Lakes analysis and depend upon how CO₂ and climate change combine analyses of shifts in other regional crop yields. He to affect crop growth and on how hot and dry the did the analysis for yields attributable to climate climate becomes. Yields and the potential demand change alone, and for the combined effects of for irrigation appear to be quite sensitive to rainfall, climate and enhanced CO₂ concentrations. Adams being higher under relatively drier scenarios. If et al.'s analysis did not account for the effects of climate change is severe enough, as under the climate on agriculture in other countries. How U.S. GFDL scenario, yields could fall. In general, and regional agriculture respond to climate change irrigation demand would rise, but some significant may be strongly influenced by changes in relative exceptions exist. global productivity and demand. The study did not consider introduction of new crops such as citrus. Implications (For a discussion of the study's design and limitations, see Chapter 6: Agriculture.) The potential shifts of agriculture northward are discussed below. Since the demand for Results irrigation is generally higher, it could become a more attractive option for farmers in the region. Adams et al.'s estimates of acreage changes for Whether more irrigation is actually used will depend the Great Lakes States are shown in Table 15-6. It on its costs and the price of crops. appears that land devoted to agriculture in the 314 Great Lakes Great Lakes region would not change significantly change scenarios and with estimates of corn yields in response to climate change. The results indicate and prices for climate effects alone from the Ritchie a slight tendency to increase acreage in the northern et al. and Adams et al. studies. Based on the Great Lakes States, although only by small amounts. interviews, a set of decision rules was established to Results for the Corn Belt States are inconclusive. estimate how a typical Illinois corn farmer would alter farming practices in response to the climate and agriculture scenarios. Table 15-6. Percentage Change in Acreage for Great Lakes States After Doubled Limitations CO2 Climate Change (Corn Belt States include Iowa and Missouri) The climate change scenarios involve climate conditions not experienced by the experts. Their estimates of how farmers would respond are not Climate change Climate and based on experience with similar conditions but on alone CO₂ speculation. The results of the combined climate Area GISS GFDL GISS GFDL and CO2 sensitivity analyses were not presented to the experts. The analysis is specifically for Illinois corn farmers and cannot be extrapolated to other Lake States +3 0 +1 +10 areas or crops. Corn Belt +2 -6 -1 -6 Results Easterling found that the degree of adjustment depends on how much climate changes. Under the Implications wetter GISS scenario, farmers could make adjustments to help mitigate the impacts of higher The results of Adams et al. and Ritchie et al. temperatures. Such adjustments could include suggest that northern regions could become more planting earlier in the spring to avoid low soil attractive for agriculture, although more extensive moisture levels in the summer, using full-season analysis is needed to confirm this result. The corn varieties for earlier planting, and changing presence of thin, glaciated soils may limit this tillage practices and lowering planting densities to expansion. If it occurs, such an expansion could better conserve soil moisture. Under the hotter and have significant implications for development of the drier GFDL scenario, corn production might not be north. Additional acreage could be converted from feasible. Farmers would likely install irrigation current uses, such as forests, to agriculture. systems; switch to short-season corn, soybeans, and Increased erosion and runoff from this additional grain sorghum; and perhaps remove marginal lands acreage would pollute groundwater and streams and from production. This last conclusion is consistent lakes in relatively pristine areas. Enhanced with the Adams et al. study. agriculture may increase the need for more shipping as lower lake levels raise shipping costs. Implications Adjustments by Illinois Corn Producers Although farmers have a variety of adjustment options to help cope with climate change, they may Farmers may make many adjustments to have great difficulty coping with extreme changes climate change such as planting different crop such as the dry climate implied by the GFDL varieties, planting earlier in the season, irrigating, scenario. Use of more irrigation would have and using different fertilizers. Easterling examined negative implications for water quality, although this how a typical corn farmer in Illinois would react to would be partly counterbalanced by any retirement climate change. of marginal lands. Study Design Easterling presented several professional crop consultants with the GISS and GFDL climate 315 Chapter 15 Electricity Demand Table 15-7. Estimated Changes in Electricity Demand Induced by Transient Study Design Climate Change Scenarios for Great Lakes Utilities (%) Linder and Inglis used the GISS transient scenarios to estimate the national changes in demand for electricity for the years 2010 and 2055. Utility Annual (2010) Annual (2055) The temperature change for 2055 is almost as high as the GISS doubled CO₂ estimate of 4.2°C. They first estimated the change in electricity demand due Minnesota -0.2 to -0.3 -1.2 to gross national product (GNP) and population Wisconsin 0.4 to -0.5 -2.3 growth, and then factored in demand changes based Michigan -0.2 to -0.3 -1.2 on change in climate. The results for the Great Upstate Lakes States are displayed here in terms of the New York -0.2 to -0.5 -1.3 percentage change from the non-climate-related Ohio, north -0.2 to -0.3 -1.3 growth. The Great Lakes analysis did not consider Ohio, south 0.4 to -0.5 2.1 any reductions in hydropower production resulting Pennsylvania 0.4 to -0.5 2.2 from drops in lake levels. (For a description of the Illinois 0.5 2.0 study's design and limitations, see Chapter 10: Indiana 0.4 1.9 Electricity Demand.) Total Negligible <1 Results Source: Linder and Inglis (Volume H). Estimates of changes in annual demand induced by climate change arę displayed in Table 15-7. The results for 2010 are a range based on million. By 2055, costs would rise to $23 to $35 GISS transient scenarios A and B, and the results billion under GISS A. However, Linder and Inglis for 2055 are just for GISS A. A latitudinal estimated that the cost to build additional capacity difference exists within the Great Lakes region. In to meet GNP and population growth without the northern states of Minnesota, Wisconsin, climate change would be $488 to $715 billion. Michigan, northern Ohio, and upstate New York, annual demand falls. The reduced demand for Implications winter heating apparently offsets the increased demand for summer cooling. This is true in 2010 Increased capacity requirements could place and 2055, when scenario temperatures are, additional stress on the region. Fossil fuel plants respectively, 1 and 4°C higher than the base case. could add more pollutants to the air. The lake level Annual demand in the southern part of the region analysis indicates that hydropower production from (in Illinois, Indiana, southern Ohio, and the lakes would be reduced, further increasing the Pennsylvania) was estimated to rise because demand for energy from other sources. increased cooling needs are apparently greater than reductions in heating. POLICY IMPLICATIONS Although annual demand could fall in some areas, new generation capacity requirements for all Climate change could raise many issues to be utilities in the region would be higher than they are addressed by policymakers in the region. now because of increased summer cooling needs. Fundamentally, decisionmakers may have to cope New generation capacity requirements needs are with water use, water quality, and land management estimated to rise by 3 to 8% in 2010 and by 8 to issues. They could have to respond to a decline in 11% in 2055. Whether costs would rise in the next water availability, increased demand for water, two decades is not clear. Linder and Inglis poorer water quality, and shifts in land use, estimated that under the gradual warming of GISS including the possibility of expanded agriculture in B, cumulative capital costs in the region would be the north. reduced by $1.3 billion, while under the more rapid warming of GISS A, costs would increase by $300 316 Great Lakes Most likely, many of the decisions in response makers will have to balance these demands with the to climate change, especially issues concerning water needs of people in the basin. management, would be made on an international basis. Both Canada and the United States oversee Shipping the regulation of the lakes, water quality, and diversions of water out of the basin. Any response to the potential impacts on the shipping industry may be costly. Possibilities Water Supply Issues include dredging of both ports and connecting channels. Dredging could cost tens, if not hundreds, Lake Regulation of millions of dollars. In addition to the high capital costs of dredging, substantial environmental costs One important issue to be faced by both could be incurred in disposing of dredge soils countries may be regulation of the lakes. Lower contaminated with toxic chemicals. If dredging were lake levels may require altering regulation plans for not undertaken, cargo loads would be lower and Lakes Superior and Ontario. This would involve would possibly impair Great Lakes commerce. tradeoffs among the needs of shippers, hydropower, shoreline property owners, and infrastructure, and Pollution Control downstream needs, in deciding how high to keep the lakes and rivers. For example, maintaining Climate change could lead to stricter pollution highwater levels in the lakes to support shipping, control to maintain water quality. Reduced hydropower, consumption, and improved water riverflow, lower lake levels, changed thermal quality would be at the expense of shipping, structure, and potentially reduced groundwater hydropower, municipal and industrial consumption, supplies may necessitate stricter standards and and water quality in the St. Lawrence River. additional controls on sources of pollution. A need Additional structures to control the flow on the may exist for better management of nutrient runoff lakes may be an option. The International Joint from farms into shallow areas, such as the Lake Commission should begin to consider in its long- Erie Western and Central Basins. Many pollution term planning the potential impacts of climate control institutions, such as EPA and state and local change on lake regulations. water quality agencies, would have the authority to impose appropriate controls on polluters. Withdrawals The water quality problems directly caused by Even without climate change, population climate change could be exacerbated by other growth would increase demand for water for responses to climate change. Intensified agriculture municipal and industrial consumption and power in the region could increase runoff, necessitating generation. Climate change would most likely more control of nonpoint sources of pollution. If intensify the demand for withdrawals from the lakes agriculture in northern areas expands, surface and for even more uses within and outside the basin. groundwater quality in relatively pristine areas may Municipal consumption would rise (Cohen, 1987b), be degraded. Pollution control authorities such as and farmers in the region may need more water for the U.S. EPA may need to impose more irrigation. comprehensive controls for those areas and should consider this in their long-term planning. Others outside the Great Lakes may demand diversion of water from the basin. The 1986 Water Fisheries Resources Development Act prohibits such diversion without the agreement of all Great Lakes Although the analysis on fisheries indicates that governors and prohibits the federal government fish populations in the Great Lakes would generally from studying this issue. Increased diversion increase, maintaining fisheries may require intensive through the Chicago Ship Canal was requested in management. In productive areas, the possibility of the summer of 1988 to raise water levels on the introduction of new species could mean major drought-starved Mississippi River. The U.S. Army changes in aquatic ecosystems. Fisheries Corps of Engineers rejected the request. Policy- management may be needed to maintain commercially and recreationally valuable species. 317 Chapter 15 The Great Lakes Fishery Commission may wish to of such plantings. The forestry industry may consider the possible implications of climate change consider growing different types of species and on valuable fisheries and management strategies to producing wood for different uses, such as for handle these possible changes. Additional pollution furniture rather than for pulp and paper. controls may be needed to help maintain fisheries in such areas as western and central Lake Erie. Agriculture Land Use Although forests may decline, demand for more land for agriculture in northern areas may grow; Shorelines however, Adams et al. indicated this demand may be small and will depend on market forces and The potential changes in land availability and policies. Federal and state land managers as well as uses present opportunities and challenges. Lower local zoning laws may need to consider that the lake levels would open up new beaches and demand for land use may change. Rules on these potential areas for recreation and development, lands could have a major influence on how, if at all, although high capital costs may be associated with the north is developed. developing them. These lands could be kept undeveloped to serve as recreational areas and as Demographic Shifts protection against fluctuating lake levels and erosion. Conversely, they could be developed to This report did not study the demographics provide more housing and commercial uses. associated with climate change and cannot say Building structures closer to the shorelines would whether people will migrate north along with make them more vulnerable to short-term rises in warmer climates. A workshop on climate change lake levels. and the Great Lakes region, conducted by Ray et al. and attended by government representatives, How these lands will be used will be decided academics, and citizens group representatives who by local and state governments as well as private have studied climate-related Great Lakes resources, shoreline property owners. Under the Coastal Zone concluded that populations from other regions of Management Act, states may identify coastal zone the United States could migrate to the Great Lakes. boundaries and define permissible land (and water) The region could have a more favorable climate uses (Baldwin, 1984). Thus, the act could be used than more southern areas. Although lake levels to help manage the use of exposed shorelines. may fall, the lakes will still contain a large amount of freshwater while other areas have more severe Lower lake levels and less ice cover may also water availability problems. Consequently, the increase shoreline erosion, decreasing the value of Great Lakes region may be relatively more shorelines and degrading water quality. The Great attractive than other regions. Lakes Basin is not included in the U.S. coastal barrier system, a program that denies federal funds Like lower lake levels, an in-migration could for development of designated erosion or flood- present opportunities and challenges. Such a prone coastal barriers (Ray et al., Volume J). migration could revitalize the region, reversing population and economic losses of recent decades. Forestry However, it also could exacerbate some of the problems associated with climate change. More The potential decline in forests and northward people and industries would require more water and shift in Great Lakes agriculture raise many land-use add more pollution, further stressing water supplies issues. One important issue may be how to manage and quality. Population growth could increase potentially large and rapid shifts in forest pressure to develop exposed shorelines along the composition. To speed northward colonization, lakes. plantings of the species might be recommended along the advancing front of suitable climate. However, unsuitable soils and day lengths shorter than the species can tolerate might limit the success 318 Great Lakes REFERENCES Croley, T.E. II. 1983a. Great Lakes basins (U.S.A. to Canada) runoff modelling. Journal of Hydrology 64:135-158. Assel, R.A., C.R. Snider, and R. Lawrence. 1985. Comparison of 1983 Great Lakes winter weather and ice conditions with previous years. Monthly Croley, T.E. II. 1983b. Lake Ontario Basin (U.S.A. to Canada) runoff modeling. Journal of Hydrology Weather Review Vol. 113. 66:101-121. Baldwin, M.F. 1984. The Federal Government's role in the management of private rural land. Croley, T.E. II. 1988. Lumped Modeling of Laurentian Great Lakes Evaporation, Heat Storage Unpublished paper. p. 22. and Energy Fluxes for Forecasting and Simulation. NOAA Technical Memo, ERL GLERL-XX, Blumberg, A.F., and G.L. Mellor. 1983. Diagnostic Environmental Research Laboratories, Boulder, CO. and prognostic numerical circulation studies of the South Atlantic Bight. Journal of Geophysical In press. Research 88:4579-4592. Crowder, L.B., and J.J. Magnuson. 1983. Cost- Botkin, D.B., J.F. Janak, and J.R. Wallis. 1972. benefit analysis of temperature and food resource Rationale, limitations and assumptions of a use: a synthesis with examples from fishes. In: northeast forest growth simulator. IBM Journal of Aspey, W.P., and S.I. Lustick, eds. Behavioral Research and Development 16:101-116. Energetics. Columbus, OH: Ohio State University Press, pp. 189-221. Botkin, D.B., J.F. Janak, and J.R. Wallis. 1973. Crowe, R.B. 1985. Effect of Carbon Dioxide Estimating the effects of carbon fertilization on forest composition by ecosystem simulation. In: Warming Scenarios on Total Winter Snowfall and Woodwell, G.M., and E.V. Pecan, eds. Carbon and Length of Winter Snow Season in Southern Ontario. Atmospheric Environment Service - Canada. the Biosphere. Brookhaven National Laboratory Symposium No. 24. Oak Ridge, TN: USAEC, pp. Unpublished manuscript. 328-344. Delay, W.H., and J. Seaders. 1966. Predicting Cobb, C.E. Jr. 1987. The Great Lakes' troubled temperatures in rivers and reservoirs. Journal of waters. National Geographic 172(1):2-31. Sanitary Engineering 92:115-134. DiToro, D.M., and J.P. Connolly. 1980. Cohen, S.J. 1986. The effects of climate change on Mathematical Models of Water Quality in Large the Great Lakes. In: Titus, J.G., ed. Effects of Lakes. Part 2: Lake Erie. Duluth, MN: U.S. Changes in Stratospheric Ozone and Global Environmental Protection Agency. USEPA-600/3- Climate, Vol. 3. Washington, DC: United Nations 80-065. Environment Programme, U.S. EPA. pp. 163-183. DiToro, D.M., N.A. Thomas, C.S. Herdendorf, R.P. Cohen, S.J. 1987a. Sensitivity of water resources in Winfield, and J.P. Connolly. 1987. A post audit of a the Great Lakes region to changes in temperature, Lake Erie eutrophication model. Journal of Great precipitation, humidity, and wind speed. Lakes Research 134(4):801-825. Proceedings of the Vancouver Symposium, August, 1987. IAHS Publ. No. 168. Edison Electric Institute. 1987. Statistical Yearbook of the Electric Utility Industry/1986. No. Cohen, S.J. 1987b. Projected increase in municipal 54. water use in the Great Lakes due to CO₂ induced climate change. Water Resources Bulletin Eliot, J.L. 1985. A North Woods Park Primeval: 23(1):91-101. Isle Royal. National Geographic 167(4):534-550. Cohen, S.J., and Allsopp, T.R. 1988. The potential Environment Canada. 1987. Implications of impacts of a scenario of CO,-induced climatic change on Ontario, Canada. Journal of Climate Climate Change for Navigation and Power Generation in the Lakes. Climate Change Digest 1(7):669-681. CCD 87-03. 319 Chapter 15 Federal Reserve Bank of Chicago and The Great Land Evaluation Group. 1986. Implications of Lakes Commission. 1985. The Great Lakes Climatic Change and Variability for Ontario's Agri- Economy: A Resource and Industry Profile of the food Sector. Prepared for Atmospheric Great Lakes States. Boyne City, MI: Harbor House Environment Service, Environment Canada. LEG- Publishers, Inc. 26. Freeberg, M.H. 1985. Early life history factors Larsen, C.E. 1985. Lake Level, Uplift, and Outlet influencing lake whitefish (Coregonus clupcatormis) Incision, the Nipissing and Algoma Great Lakes. year-class strength on Grand Traverse Bay, Lake Quaternary Evolution of the Great Lakes: Michigan. Master of Science Degree Thesis. Geological Association of Canada, Special Paper 30. Michigan State University, Department of Fisheries and Wildlife, East Lansing, MI. Linder, K.P., M.J. Gibbs, and M.R. Inglis. 1987. Potential Impacts of Climate Change on Electric Garwood, R.W., Jr. 1977. An oceanic mixed layer Utilities. New York State Energy Research and model capable of simulating cyclic states. Journal Development Authority, Albany, NY. Report 88-2. of Physical Oceanography 7:455-468. Magnuson, J.J., L.B. Crowder, and P.A. Medvicle. Great Lakes Basin Commission. 1975. Great Lakes 1979. Temperature as an ecological resource. Basin Framework Study: Appendixes 5, 11, 12, 13, American Zoology 19:331-343. and 19. Ann Arbor, MI: Public Information Office. Great Lakes Basin Commission. Marchand, D., M. Sanderson, D. Howe, and C. Alpaugh. 1988. Climatic change and Great Lakes Hansen, J., I. Fung, A. Lacis, D. Rind, S. Lebedeff, levels: the impact on shipping. Climatic Change R. Ruedy, G. Russell, and P. Stone. 1988. Global 12:107-133. climate changes as forecast by the Goddard Institute for Space Studies three dimensional model. Journal Meisner, D.J., J.L. Goodier, H.A. Regier, B.J. of Geophysical Research 93:9341-9364. Shuter, and W.J. Christie. 1987. An assessment of the effects of climate warming on Great Lakes Horvath, F.J., M.R. Jannereth, and C.A. Shafer. Basin fishes. Journal of Great Lakes Research 1988. Impacts of Water Level Fluctuations. Water 13(3):340-352. Science and Technology Board Colloquium Series, Great Lake Water Levels: Shoreline Dilemmas. The New York Times. 1988. Out of mothballs, Chicago, IL. back on the lakes. p. 16. June 15. IJC. 1985. International Joint Commission. Great Quinn, F.H. 1978. Hydrologic response model of Lakes Diversions and Consumptive Uses. Report the North American Great Lakes. Journal of by International Great Lakes Diversion and Hydrology 37:295-307. Consumptive Uses Study Board. Washington, DC: International Joint Commission. Quinn, F.H. 1988. Fluctuations of Great Lakes Water Levels. Water Science and Technology Board Jones, C.A. and J.R. Kiniry, eds. 1986. CERES- Colloquium Series, Great Lakes Water Levels: Maize: A Simulation of Maize Growth and Shoreline Dilemmas. Chicago, IL. Ddevelopment. College Station, TX: Texas A&M Press. Ruttner, F. 1931. Hydrographische und hydrochemische Beobachtungen auf Java, Sumatra Kitchell, J.F., D.J. Stewart, and D. Weininger. 1977. und Vali. Archiv for Hydrobiologic Supplement Applications of a bioenergetics model to yellow 8:197-454. perch (Perca flavescens) and walleye (Stizostedion vitreum). Journal of Fish Research 34:1922-1935. 320 Great Lakes Sanderson, M. 1987. Implications of climatic U.S. Department of Commerce. 1983. U.S. change for navigation and power generation in the Department of Commerce, Bureau of the Census. Great Lakes. Climatic Change Digest. 1982 Census of Agriculture: Geographic Area Environment Canada. Downsview, Ontario. Series. Washington DC: Government Printing CCD87-03. Office. Solomon, A.M., and D. West. 1986. Atmospheric USDA. 1987. U.S. Department of Agriculture. carbon dioxide change: agent of future forest growth Agricultural Statistics. Washington, DC: U.S. or decline? In: Titus, J., ed. Effects of Changes in Government Printing Office. Stratospheric Ozone and Global Climate. Washington, DC: United Nations Environment USDA. 1982. U.S. Department of Agriculture, Program/U.S. Environmental Protection Agency. Forest Service. An analysis of the timber situation pp. 23-38. in the U.S. 1952-2030. Washington, DC: U.S. Government Printing Office. Forest Resource Southam, C., and S. Dumont. 1985. Status Report. Report No. 23. Impact of climate change on Great Lakes levels and outflows. Inland Waters Directorate. Ontario U.S. EPA and Environment Canada. 1987. The Region. Environment Canada. Unpublished report. Great Lakes: An Environmental Atlas and Resource Book. Chicago, IL. Tarlock, D.A. 1988. Multi-Jurisdictional Issues Presented to National Research Council, Water Wall, G. 1985. Climate Change and Its Impact on Science Technology Board Colloquium Series, Great Ontario Tourism and Recreation. Submitted to Lakes Water Levels: Shoreline Dilemma. Chicago, Environment Canada. IL. March. Wilkerson, G.G., J.W. Jones, K.J. Boote, K.J. U.S. Department of Commerce. 1987. U.S. Ingram, and J.W. Mishoe. 1983. Modeling soybean Department of Commerce, Bureau of the Census. growth for crop management. Transactions of the Statistical Abstract of the United States: 1987. American Society of Agricultural Engineers, St. (107th edition). Washington, DC: Government Joseph, Michigan, 26:1, 63-73. Printing Office. 321 CHAPTER 16 SOUTHEAST FINDINGS Considering various scenarios of climate change and CO2, the productivity of Global climate change could diminish the extent of southeastern agriculture could decline relative the region's forests, reduce agricultural productivity to northern areas, and 10 to 57% of the and increase the abandonment of farms, diminish region's farmland could be withdrawn from fish and shellfish populations, and increase cultivation. This analysis did not consider electricity demand. Approximately 90% of the whether new crops would be introduced. The national coastal wetland loss and two-thirds of the decline in cultivated acreage may tend to be national shoreline protection costs from sea level concentrated in areas where farming is only rise could occur in the Southeast. The impacts on marginally profitable today. A reduction in rivers and water supplies are uncertain. agriculture could hurt farm-related employment and the regional economy. Agriculture Forests Southeastern agriculture is generally more vulnerable to heat stress than to freezing, so There may be a significant dieback in southern the adverse impacts of more hot days would forests. Higher temperatures and drier soils more than offset the beneficial impact of a may make it impossible for most species to longer growing season. regenerate naturally and may cause forests to convert to shrub terrain or grassland. The As a result of climate change alone, yields of decline in the forests could be noticeable in 30 soybeans and corn would vary from no change to 80 years, depending on the site and scenario. in the cooler regions to up to a 91% decrease Southern noncoastal areas, such as Atlanta and in warmer areas, even if rainfall increases. Vicksburg, may have particularly large reductions. The moist coastal forests and the A preliminary assessment suggests that when relatively cool northern forests may survive, the direct effects of CO₂ are included, yields although with some losses. might increase in parts of the region if climate also becomes wetter. If climate becomes drier, The forest industry, which is structured around yields could decrease everywhere in the region. currently valuable tree species, would have to However, our understanding of the direct either relocate or modify its planting strategies. effects of CO₂ fertilization is less certain than our understanding of the impacts of climate Historically, abandoned farms have generally change. Increased CO₂ could also affect converted to forests. If large portions of the weeds, but these impacts were not analyzed. Southeast lose the ability to naturally generate forests, much of the region's landscape may If rainfall decreases, irrigation will become gradually come to resemble that of the Great necessary for farming to remain viable in much Plains. of the region. Water Supplies The range of such agricultural pests as potato leafhoppers, sunflower moths, and black Because the winter accumulation of snow plays cutworms could move north by a few hundred a negligible role in determining riverflow, our kilometers. This would most likely result in inability to predict whether rainfall will increase or increased use of pesticides. 323 Chapter 16 decrease makes it difficult to say whether riverflows finfish and shellfish, such as shrimp, flounder, and will increase or decrease. oysters. Oysters and other species could be threatened by the increased salinity that will The limited number of hydrologic studies accompany sea level rise. Some species, such as conducted in the Southeast further prevents us pink shrimp and rock lobster, could increase in from making any definitive statement about the abundance. regionwide implications for rivers. Electricity Demand Decreases in rainfall could disrupt navigation, drinking water availability, recreation, The annual demand for electricity in the hydropower, powerplant cooling, and dilution Southeast could rise by 14 to 22 billion of effluent, while increased rainfall could kilowatthours (kWh), or 2 to 3%, by 2010 and exacerbate the risk of flooding. by 100 to 197 billion kWh, or 7 to 11%, by 2055 as a result of increased temperature. For the scenarios used in this report, changes in operating rules for managed water systems By 2010, approximately 7 to 16 gigawatts (GW) would allow current water demands to be met could be needed to meet the increased in most instances. demand, and by 2055, 56 to 115 GW could be needed -- a 24 to 34% increase over baseline The Southeast generally has ample additions that may be needed without climate groundwater supplies. The potential change. The cumulative costs could be $77 to implications of increased irrigation on $110 billion by 2055. groundwater need to be examined. Policy Implications Sea Level Rise Federal laws constrain the U.S. Army Corps A 1-meter rise in sea level by the year 2100 of Engineers and other water resource would inundate 30 to 90% of the region's managers from rigorously considering tradeoffs coastal wetlands and flood 2,600 to 4,600 between many nonstatutory objectives of square miles of dryland, depending on the federal dams in the Southeast, including extent to which people erect levees to protect recreation, water supply, and environmental dryland from inundation. If current river quality. Increased flexibility would improve the management practices continue, Louisiana ability of these agencies to respond to and alone would account for 40% of national prepare for climate change. wetland loss, and developed areas could be threatened as soon as 2025. Given the potential withdrawals of acreage from agriculture, the potential for growing Holding back the sea by pumping sand or tropical crops needs to be examined. other measures to raise barrier islands, and protecting mainland areas with bulkheads and Strategies now being evaluated by the levees, would cost approximately $42 to $75 Louisiana Geological Survey and the U.S. billion through the year 2100 for a 1-meter Army Corps of Engineers to address coastal rise. wetland loss in Louisiana should consider a possible sea level rise of 0.5 to 2.0 meters. Marine Fisheries Measures that would enable this ecosystem to survive would require major public works and Gulf coast fisheries could be negatively affected changes in federal navigation and riverflow by climate change. A loss of coastal wetlands policies. Because of the decades required to due to sea level rise could eliminate critical implement necessary projects and the prospect habitats for shrimp, crab, and other that much of the ecosystem would be lost by commercially important species. Temperatures 2030 even without climate change, these in the gulf coast estuaries may exceed the programs need to proceed expeditiously. thermal tolerances for commercially important 324 Southeast Given the potentially important impacts on abundant rainfall, and generally flat terrain gave rise forests, private companies as well as agencies in the 17th century to a strong agricultural economy such as the U.S. Forest Service and state with a distinctive regional culture. The combination agencies may wish to assess the potential for of a benign climate and 60% of the nation's ocean large losses of southern forests and the beaches continues to attract both tourists and new implications for research and management residents to the southeastern coastal plain. Florida, strategies. for example, is the nation's fastest growing state and will be the third largest by the year 2000 (Meo et al., Volume J). CLIMATE AND THE SOUTHEAST The climate and the coastal zone of the CLIMATE-SENSITIVE Southeast are among the chief factors that RESOURCES OF THE distinguish the southeastern United States from the rest of the nation. 1 The warm temperatures, SOUTHEAST Water Resources 1 Except for the discussion of the economic implications for When statewide averages are considered, each of the seven states in the Southeast receives more agriculture, the term "Southeast" refers to the study area shown in Figure 16-1: North Carolina, South Carolina, Georgia, rainfall than any other state in the continental Florida, Alabama, Mississippi, Tennessee, and the coastal zones of Louisiana and Texas. NASHVILLE CHARLOTTE MEMPHIS ATLANTA BIRMINGHAM JACKSON JACKSONVILLE NEW ORLEANS MIAMI Figure 16-1. Southeast region. 325 Chapter 16 United States (although parts of some western but only 30% for public supplies (see Meo et al., states receive more). Moreover, the rivers of the Volume J). Southeast drain over 62% of the nation's lands; the Mississippi River alone drains 38% of the nation Atlanta and some other metropolitan areas (Geraghty et al., 1973). obtain their water supplies from federal reservoirs; however, even the many cities that do not still may The Southeast supports 50,000 square miles of benefit from federal and federal/state water bottomland hardwood forests (Mitch and Gosselink, management. For example, New Orleans obtains its 1986), 2 which are periodically flooded areas that water from the Mississippi River. Without the Old offer winter habitat for migratory birds such as River Control Structure in Simmesport, Louisiana, ducks, geese, and songbirds. Bass, catfish, and which prevents the river from changing its course to panfish are found in the slow-moving rivers, and the Atchafalaya River, the New Orleans water trout inhabit the fast-moving mountain streams. supply would be salty during droughts. Although Miami obtains its water from the Biscayne Aquifer, Dams have been constructed along most of the some coastal wells would be salty without the efforts region's major rivers. Although private parties have of the U.S. Army Corps of Engineers and the South built a few dams, most of the major projects were Florida Water Management District to recharge the built by the U.S. Army Corps of Engineers, the aquifer with supplemental freshwater from canals Tennessee Valley Authority, and other federal and Lake Okeechobee. agencies. In general, the statutory purposes of these reservoirs have been to ensure a sufficient flow of The various uses of water often conflict with water during droughts, to prevent floods, and to each other. Hydroelectric power generators, generate electricity. The nonstatutory objectives of lakefront residents, and boat owners benefit when environmental quality, recreation, and water supply water levels are maintained at high levels. However, also are considered in the operation of dams. high water levels make flood control more difficult, and municipal uses, navigation, hydropower, and Dam construction has created large lakes along environmental quality require that water be released which people have built houses, hotels, and marinas. during the dry season, which adversely affects These dams generate 22.2 billion kilowatthours recreation. (kWh) per year, approximately 7% of the region's power requirements (Edison Electric Institute, Estuaries 1985). In general, the reservoirs have sufficient capacity to retain flood surges and to maintain Over 43% of the fish and 70% of the shellfish navigation flows during the dry season. The one harvested in U.S. waters are caught in the Southeast notable exception is the Mississippi River: levees (NOAA, 1987). Commercially important fishes are and land-use regulations are the main tools for abundant largely because the region has over 85% preventing flood damages; although the Mississippi's of the nation's coastal wetlands; over 40% are in base flow usually is sufficient to support navigation, Louisiana alone. boats ran aground on many stretches of the river during the drought of 1988. Most of the wetlands in the Southeast are less than 1 meter above sea level. The wetlands in In Florida, which accounts for 45% of water Louisiana are already being lost to the sea at a rate consumption in the Southeast, groundwater supplies of 50 square miles per year because of the about half the water used by farms and 85% of the interaction of human activities and current rates of water used for residential and industrial purposes. relative sea level rise resulting from the delta's For the rest of the Southeast, groundwater supplies tendency to subside 1 centimeter per year. (This most water for agricultural and rural uses problem is discussed in greater detail below.) 2 This measure includes Mississippi, Arkansas, Louisiana, Texas Summer temperatures in many of the gulf and Virginia. coast estuaries are almost as warm as crabs, shrimp, oysters, and other commercially important fishes can tolerate (Livingston, Volume E). Winter 326 Southeast temperatures along the gulf coast are almost warm coasts, this region has wide low-lying coastal plains enough to support mangrove swamps, which and experiences several hurricanes annually. generally replace marshes once they are established; Florida, Texas, and Louisiana account for 62% of mangroves already dominate the Florida coast south the $144 billion of private property insured by the of Fort Lauderdale. Federal Flood Insurance Program (see Riebsame, Volume J). Beach Erosion and Coastal Flooding Agriculture The Southeast has 1,100 miles of sandy ocean beaches, many of which are found on low and In the last few years, droughts and heat waves narrow barrier islands. The Atlantic coast is heavily have caused crop failures in many parts of the developed, while much of the gulf coast is only now Southeast. Unlike much of the nation, cold weather being developed. In part because of their generally is not a major constraint to agricultural vulnerability to hurricanes, none of Mississippi's production, except for Florida's citrus industry. barrier islands has been developed, and only one of Louisiana's barrier islands is developed at present. Although cotton and tobacco were once the Because much of Florida's gulf coast is marsh, it is mainstays of the Southeast's economy, agriculture still largely undeveloped. now accounts for only 1% of the region's income (U.S. Department of Commerce, 1986). Since All eight coastal states are experiencing coastal World War II, substantial amounts of farmland have erosion. Along developed coasts, recreational been withdrawn from agriculture, and much of this beaches have narrowed, increasing the vulnerability land has been converted to forest. The cotton crop of shorefront structures to storms. In Louisiana, has been largely lost to the irrigated Southwest, and some undeveloped barrier islands are eroding and although tobacco remains profitable, it is grown on breaking up. Elsewhere, narrow barrier islands are only 500,000 acres. However, in the last few keeping pace with sea level rise by "overwashing" decades, southeastern farmers have found soybeans (i.e., rolling over like a rug) in a landward direction, to be profitable; this crop now accounts for 45% of while wide islands and mainland coasts have simply all cultivated land in the Southeast. Corn continues eroded. The coastal states of the Southeast are to account for 5% of southeastern agriculture (U.S. responding by holding back the sea in some areas Department of Commerce, 1982). Table 16-1 and by adapting to erosion in others. compares annual revenues by state for various crops. The two greatest natural disasters in U.S. history resulted from floods associated with Forests hurricanes in Galveston, Texas, and Lake Okeechobee, Florida, in which over 8,000 people The commercial viability of southeastern drowned. After the Mississippi River overflowed forests has increased greatly since World War II, its banks and inundated most of coastal Louisiana primarily as a result of the increased use of in the 1930s, Congress directed the U.S. Army softwoods, such as pines and firs, for plywood and Corps of Engineers to initiate a major federal for applications that once required hardwood. program of flood control centered around the Because this transition coincided with lower farm Southeast. Nevertheless, flood waters often remain prices and declining soils in the piedmont foothills over some low areas in Louisiana and Florida for of the Southeast, many mountain farms have been several days after a major rainstorm. converted to forests. However, in the last 10 years, 7 million acres of coastal plain forests have been Hurricanes continue to destroy recreational converted to agriculture (Healy, 1985). development in at least a few ocean beach communities almost every year in the Southeast. Approximately 45% of the nation's softwood The region presently experiences the majority of (mostly loblolly pine) and 50% of its hardwood are U.S. coastal flooding and probably would sustain grown in the region. Forests cover 60% of the the worst increases in flooding as a result of global Southeast, and 90% of forests are logged. warming. Unlike the Northeast and Pacific Oak-hickory covers 35%, and pine covers another 327 Chapter 16 Table 16-1. Annual Revenues by State for 33% of commercial forests. Only 9% of the Various Crops (thousands of 1986 southeastern forests are owned by federal and state dollars) governments, and 18% are owned by the forest industry. In contrast, 73% of the forests are owned by farmers and other private parties (Healy, 1985). Crop Value Indoor and Outdoor Comfort Corn for grain Alabama 856,550 The Southeast is one of the few areas that Florida 31,493 spends as much money on air-conditioning as on Georgia 203,931 heating. Figure 16-2 shows temperatures Mississippi 22,600 throughout the Southeast for the months of January North Carolina 324,789 and July. Even in January, about half the region South Carolina 104,333 experiences average temperatures above 50°F, and Tennessee 193,687 almost the entire region has a typical daily high above 50°F. Thus, with the possible exception of Cotton the cool mountains of Tennessee and North Alabama 145,540 Carolina, a global warming would increase the Florida 8,112 number of days during which outdoor temperatures Georgia 97,325 would be unpleasantly hot much more than it would Mississippi 449,630 reduce the number of unpleasantly cold days. North Carolina 30,944 Tennessee 109,610 PREVIOUS STUDIES OF THE Sugarcane for sugar and seed Florida 369,899 IMPACTS OF CLIMATE CHANGE ON THE SOUTHEAST Tobacco Florida NA Georgia NA Most studies examining the impact of global North Carolina warming on the Southeast have focused on sea level NA South Carolina NA rise. Recent efforts have addressed other topics. Tennessee NA Several dozen researchers presented papers on other global warming impacts on the Southeast at a Peanuts for nuts 1987 EPA conference held in New Orleans (Meo, Alabama 133,930 1987). Their papers suggested that agricultural Florida yields would decline, forest species would shift, and 48,600 that coastal and water supply officials should start to Georgia 472,645 North Carolina 122,941 plan for the consequences of global warming. South Carolina 5,882 Flooding Soybeans Alabama 140,719 Leatherman (1984) and Kana et al. (1984) Florida 31,036 applied flood-forecasting models to assess the Georgia 179,676 potential increases in flooding in Galveston, Texas, Mississippi 365,018 and Charleston, South Carolina. For the Galveston North Carolina 196,673 area, a 90-centimeter (3-foot) rise would increase South Carolina 125,214 the 100-year floodplain by 50%, while a 160- Tennessee 230,373 centimeter (5.2-foot) rise would enable the 100-year storm to overtop the seawall erected after the NA = Not available. disaster of 1900. For the Charleston area, a 160- Source: U.S. Department of Agriculture (1987). centimeter rise would increase the 10-year 328 Southeast A. B. 40 35 80 45 70 70 75 80 35 82 80 V 40 50 85 80 80 45 55 60 82 50 82 65 55 85 55 82 60 70 JANUARY JULY Figure 16-2. Typical temperatures in the Southeast: (A) January, (B) July. floodplain to the area currently covered by the 100- necessitate substantial changes in the ports and year floodplain. shipping lanes of the Mississippi River to prevent the loss of several thousand square miles of coastal Gibbs (1984) estimated that the economic wetlands. Titus et al. (1987) showed that a impact of a 90-centimeter rise by 2075 could be as reconstructed coastal drainage system in Charleston great as $500 million for Galveston and over $1 should be designed for a 1-foot rise in sea level if billion for Charleston. However, he also estimated the probability of such a rise is greater than 30%. that the adverse impacts of flooding and land loss Linder et al. (1988) found that warmer temperatures could be cut in half if the communities adopted would require an electric utility company to measures in anticipation of sea level rise. Titus substantially increase its generating capacity. (1984) focused on decisions facing Sullivans Island, South Carolina, in the aftermath of a storm. He concluded that rebuilding $15 million in oceanfront CLIMATE CHANGE STUDIES IN houses after a storm would not be economically THIS REPORT sound if future sea level rise is anticipated, unless the community is prepared to continuously nourish its beaches. Table 16-2 and Figure 16-3 illustrate the studies undertaken as part of this effort. Few Wetlands resources had previously been applied to examining the various impacts of climate change for the Kana et al. (1986) surveyed marsh transects Southeast. Models of coastal erosion, coastal and estimated that 90- and 160-centimeter (3.0- and wetland loss, agricultural yields, forest dynamics, 5.2-foot) rises in sea level would drown 50 and 90%, and electricity consumption were sufficiently refined, respectively, of the marsh around Charleston, South so that it was possible to inexpensively apply them Carolina. Armentano et al. (1988) estimated the to numerous sites and develop regional assessments. Southeast would lose 35 and 70% of its coastal Louisiana, which accounts for half of the region's wetlands for respective rises of 1.4 and 2.1 meters, wetlands, has been the subject of previous studies. assuming that developed areas are not protected. It is discussed following the studies for this report. Infrastructure By contrast, the impacts on water resources and ecosystems required more detailed site-specific The Louisiana Wetland Protection Panel studies, and it was not possible to undertake such (1987) concluded that a rise in sea level might case studies for a large number of watersheds or 329 Chapter 16 Table 16-2. Studies of the Southeast Regional Studies Impacts on Runoff in the Upper Chattahoochee River Basin - Hains, C.F. Haines, Hydrologist, Inc. (Volume A) Projected Changes in Estuarine Conditions Based on Models of Long-Term Atmospheric Alteration - Livingston, Florida State University (Volume E) Policy Implications of Global Climatic Change Impacts Upon the Tennessee Valley Authority Reservoir System, Apalachicola River, Estuary, and Bay and South Florida - Meo, Ballard, Deyle, James, Malysa, and Wilson, University of Oklahoma (Volume J) Potential Impacts on Climatic Change on the Tennessee Valley Authority Reservoir System - Miller and Brock, Tennessee Valley Authority (Volume A) Impact of Climate Change on Crop Yield in the Southeastern U.S.A. - Peart, Jones, and Curry, University of Florida (Volume C) Methods for Evaluating the Potential Impacts of Global Climate Change - Sheer and Randall, Water Resources Management, Inc. (Volume A) Forest Response to Climate Change: A Simulation Study for Southeastern Forests - Urban and Shugart, University of Virginia (Volume D) National Studies That Included Southeast Results The Economic Effects of Climate Change on U.S. Agriculture: A Preliminary Assessment - Adams, Glyer, and McCarl, Oregon State University (Volume C) National Assessment of Beach Nourishment Requirements Associated with Accelerated Sea Level Rise - Leatherman, University of Maryland (Volume B) The Potential Impacts of Climate Change on Electric Utilities: Regional and National Estimates - Linder and Inglis, ICF Inc. (Volume H) The Effects of Sea Level Rise on U.S. Coastal Wetlands -Park and Trehan, Butler University and Mausel and Howe, Indiana State University (Volume B) Potential Effects of Climatic Change on Plant-Pest Interactions - Stinner, Rodenhouse, Taylor, Hammond, Purrington, McCartney, and Barrett, Ohio Agricultural Research and Development Center (Volume C) Assessing the Responses of Vegetation to Future Climate Change: Ecological Response Surfaces and Paleological Model Validation - Overpeck and Bartlein, Lamont-Doherty Geological Observatory (Volume D) An Overview of the Nationwide Impacts of Rising Sea Level - Titus and Greene, U.S. Environmental Protection Agency (Volume B) The Cost of Defending Developed Shorelines Along Sheltered Waters of the United States from a Two Meter Rise in Mean Sea Level - Weggel, Brown, Escajadillo, Breen, and Doheny, Drexel University (Volume B) 330 Southeast Region and Coastal Studies of Wetland Loss and Cost of Holding Back the Sea Forest Study Sites Agriculture Study Sites TVA, Apalachicola, and South Florida Watersheds Figure 16-3. Overview of studies of the Southeast. ecosystems. Therefore, our analysis was limited to Table 16-3 illustrates how the frequency of representative case studies. For water resources, we mild days during the winter and the frequency of picked (1) the Tennessee Valley, because it is the very hot days during the summer might change largest managed watershed in the region; and (2) under the Goddard Institute for Space Studies Lake Lanier, because it serves Atlanta, the region's (GISS) doubled CO2 scenario. As explained in second largest city. In both cases, we were able to Chapter 4: Methodology, these estimates used identify researchers who were already familiar with average monthly changes in temperature and the area. The sole aquatic ecosystem studied in assumed no change in variability. Under this depth was Apalachicola Bay, picked because the scenario, the number of days per year in which the estuary had already been the subject of the most mercury would fall below freezing would decrease comprehensive data collection effort in the from 34 to 6 in Jackson, Mississippi; from 39 to 20 Southeast. in Atlanta; and from 41 to 8 in Memphis. The number of winter days above 70°F would increase from 15 to 44 in Jackson, from 4 to 14 in Atlanta, SOUTHEAST REGIONAL and from 5 to 24 in Memphis. CLIMATE CHANGE SCENARIOS Of the nine cities shown, only Nashville has summer temperatures that currently do not Figure 16-4 illustrates the scenarios of future regularly exceed 80°F. However, the number of climate change from general circulation models. days with highs below 80°F would decline from 60 Table 16-3 shows the more detailed seasonal to 34. Elsewhere, the heat would be worse. The changes. 331 Chapter 16 Table 16-3. The GISS Doubled CO₂ Scenario: Frequency of Hot and Cold Days (°F) Number of winter days with: Number of summer days with: Daily low <32 Daily high >70 Daily high <80 Daily high >90 Daily high >100 Location HISTᵃ 2xCO₂ HISTᵃ 2xCO₂ HISTᵃ 2xCO₂ HISTª 2xCO₂ HISTᵃ 2xCO₂ Atlanta, GA 38.3 20.5 4.2 13.6 10.0 2.2 17.1 53.3 0.6 4.2 Birmingham, AL 35.5 8.1 7.1 30.7 4.5 0.4 34.1 72.5 1.5 10.7 Charlotte, NC 42.1 23.8 3.4 9.9 11.9 3.7 23.1 56.5 0.1 5.9 Jackson, MS 33.5 5.9 15.3 43.5 0.8 0.2 55.1 83.1 2.0 19.5 Jacksonville, FL 9.3 1.7 34.6 49.6 2.3 0.3 46.4 81.3 0.6 14.1 Memphis, TN 41.2 8.1 5.2 23.6 4.9 0.7 50.5 74.8 2.6 19.1 Miami, FL 0.2 0.0 72.9 82.7 0.6 0.0 29.8 83.5 0.0 2.5 Nashville, TN 42.5 15.4 0.3 8.6 60.4 33.7 10.5 20.2 0.3 3.5 New Orleans, LA 14.9 3.5 24.9 39.5 0.9 0.1 55.4 84.9 0.3 13.5 ᵃHIST = Historic. Source: Kalkstein (Volume G). A. Temperature B. Precipitation 6 0.9 0.8 GISS 0.6 5 0.5 GFDL 0.4 4 OSU 0.3 CHANGE (°C) 3 CHANGE (mm/Day) 0.2 0.1 0 -0.1 2 -0.2 -0.3 1 -0.4 -0.5 0 0.6 Winter Spring Summer Fall Annual Winter Spring Summer Fall Annual Figure 16-4. 2xCO₂ less 1xCO₂ climate scenarios for the Southeast: (A) temperature, and (B) precipitation. 332 Southeast number of days per year above 90°F would increase from 30 to 84 in Miami, from 17 to 53 in Atlanta, and from 55 to 85 in New Orleans. Memphis, ALL DRYLAND PROTECTED Jackson, New Orleans, and Jacksonville, which currently experience 0 to 3 days per year above 100 SALT MARSH 100°F, would have 13 to 20 such days (Kalkstein, 80 BEACH/FLAT Volume G). RESULTS OF SOUTHEASTERN PERCENT OF 1986 WETLAND AREAS 60 MANGROVE FRESH MARSH 40 SWAMP 20 STUDIES 0 0.0 0.1 0.3 0.6 1.0 1.5 2.2 3.0 SEA LEVEL RISE (Meters) Coastal Impacts DEVELOPED AREAS PROTECTED A number of national studies for the report presented results for the effects of climate change 100 on the southeastern coast. Leatherman estimated 80 the cost of maintaining recreational beaches. Park et al. and Weggel et al. examined the impacts on PERCENT OF 1986 WETLAND AREAS 60 40 wetland loss and shoreline defense, and used their results to estimate the regionwide cost of raising 20 barrier islands. The projected rise in sea level 0 would cause shorelines to retreat, exacerbate coastal 0.0 0.1 0.3 0.6 1.0 1.5 2.2 3.0 flooding, and increase the salinity of estuaries, SEA LEVEL RISE (Meters) wetlands, and aquifers. (For a discussion of the NO PROTECTION rationale, methods, and nationwide results of these 100 studies, see Chapter 7: Sea Level Rise.) Coastal Wetlands PERCENT OF 1986 80 WETLAND AREAS 60 40 Park et al. (Volume B) examined 29 20 southeastern sites to estimate the regionwide loss 0 of coastal wetlands for a variety of scenarios of 0.0 0.1 0.3 0.6 1.0 1.5 2.2 3.0 future sea level rise. Their analyses included such SEA LEVEL RISE (Meters) societal responses as providing structural protection for all shorelines (total protection), protecting areas that are densely developed today (standard Figure 16-5. Wetlands loss in the Southeast for protection), and allowing shorelines to adjust three shoreline protection options (Park et al., naturally without coastal protection (no protection). Volume B). (NOTE: These numbers are different from those in Table 16-4 because they include Figure 16-5 illustrates their estimates for the nonvegetated wetlands, i.e., beaches and flats.) year 2100 for the various scenarios of sea level rise and coastal defense. Even if current sea level trends continue, 25% of the Southeast's coastal wetlands will be lost, mostly in Louisiana. a 100-centimeter rise could result in losses Excluding Louisiana: of 45 to 68%; and current trends imply a loss of 15%; a 200-centimeter rise implies losses of 63 to 80%. a 50-centimeter rise could result in a loss of 35 to 50%, depending on how shorelines Park et al. estimated losses of 50, 75, and 98% are managed; for Louisiana under the three scenarios. However, they did not consider the potential for mitigating the 333 Chapter 16 loss by restoring the flow of river water into these Cost of Protecting Recreational Beaches wetlands; no model exists that could do so (Louisiana Wetland Protection Panel, 1987). Titus In Volume B, Leatherman notes that the and Greene estimated statistical confidence intervals projected rise in sea level would threaten all illustrated in Table 16-4. developed recreational beaches. Even a 1-foot sea level rise would erode shorelines over 100 feet Total Coastal Land Loss throughout the Southeast. Along the coasts of North Carolina and Louisiana, the erosion would be Park et al. also estimated total land loss, considerably greater. Because the distance from the including both wetlands and dryland. Most of the high tide line to the first building is rarely more land loss from a rise in sea level would occur in than 100 feet, most recreational beaches would be Louisiana. A 50-centimeter (20-inch) sea level rise lost, unless either the buildings were removed or would result in the loss of 1,900 to 5,900 square coastal protection measures were undertaken. miles of land, while a 200-centimeter rise would inundate 10,000 to 11,000 square miles. Table 16-4 illustrates Leatherman's estimates of the cost of protecting recreational beaches by Table 16-4. Summary of Results of Sea Level Rise Studies for the Southeast (billions of dollars) Response Baseline 50-cm rise 100-cm rise 200-cm rise Developed areas are protected Land lost Dryland lost (mi²) 1,300-3,700 1,900-5,500 2,600-6,900 4,200-10,100 Wetlands lost (%)ᵃ 11-22 24-50 34-77 40-90 Cost of coastal defense 19-28 42-75 127-174 Open coast Sand 3 10-15 19-30 44-74 Elevated structures negligibleᵇ 5-9 10-40 60-75 Sheltered shores negligibleᵇ 2-5 5-13 9-41 All shores are protected Land lost Dryland lost (mi²) 0 0 0 0 Wetlands lost (%)ᵃ 0 38-61 47-90 68-93 No shores are protected Land lost Dryland lost (mi²) N/A 2,300-5,900 3,200-7,600 4,800-10,800 Wetlands lost (%)ᵃ N/A 22-48 30-75 37-88 "Wetlands" refers to vegetated wetlands only; it does not include beaches or tidal waves. b Costs due to sea level rise are negligible. Source: Titus and Greene (Volume B). 334 Southeast pumping sand from offshore locations. (See Table to over 7 million people and contains 675 miles of 7-3 for state-by-state results). A 1-meter rise in sea navigable waterways with annual commercial freight level could imply almost $20 billion in dredging of 28 million tons. The lakes created by the dams costs, with Texas spending $8.5 billion and Florida have over 10,000 miles of shorelines, which generate and Louisiana each spending over $3 billion. 75 million visits each year and along which people have invested $630 million, boosting the region's Using constant unit costs (except for Florida), annual economy by $400 million (Miller and Brock, Leatherman estimated that a 2-meter rise could only Volume A). double the total cost to $43 billion. Titus and Greene estimated that if the unit costs of sand To assess the potential impacts of climate increased, 1- and 2-meter rises could cost $30 and change, Miller and Brock conducted a modeling $74 billion, respectively. They also estimated that study of the water resource implications, and Meo the respective costs of rebuilding roads and utilities et al. examined the policy implications for the TVA. on barrier islands could be $5 to 9 billion, $10 to 40 billion, and $60 to 75 billion for the three scenarios. TVA Modeling Study Cost of Protecting Calm-Water Shorelines Methods While Leatherman focused only on the open Miller and Brock used the TVA's "Weekly ocean coast, Weggel et al. estimated the regionwide Scheduling Model," which the Agency currently uses costs of holding back the sea in developed sheltered in setting the guidelines for its operations, to assess and calm-water areas. Weggel et al. estimate that the impacts of climate change. This linear about $2 billion would be spent to raise roads and programming model selects a weekly schedule for to move structures, and $23 billion would be spent managing each reservoir in the TVA system by to erect the necessary levees and bulkheads for a 2- sequentially satisfying the objectives of flood control, meter rise. Table 16-4 shows confidence intervals navigation, water supply, power generation, water estimated by Titus and Greene, which imply a total quality, and recreation. Miller and Brock used this cost of $42 to 75 billion for a 1-meter rise. The model to simulate reservoir levels, riverflows, and combined cost is $68 to 83 billion. These estimates hydropower generation for wet and dry scenarios, do not include the costs of preventing flooding or of derived from the runoff estimates from the GISS protecting water supplies. doubled CO2 model run. Tennessee Valley Authority Studies TVA was unable to use a hydrologic model to estimate runoff for this study. Instead, they sought The Tennessee Valley Authority (TVA) was to use the runoff estimates from general circulation created in 1933 to spur economic growth in an area models. Unfortunately, the OSU and GFDL previously considered to be one of the nation's models estimate that there is no runoff today, which poorest. Geographically isolated by the would not permit derivation of a scenario. Appalachian Mountains, the region lacked electricity Therefore, the GISS runoff estimates were used as and roads, and the Tennessee River could not the "wet scenarios." Based on Rind (1988), the dry provide reliable transportation because it flooded in scenario simply assumed that the change in runoff the spring and dried to a trickle during the summer. would be the inverse of the change assumed in the By creating the TVA, Congress sought to remedy wet scenario. Therefore, a TVA study should be this situation by harnessing the river to provide viewed as an assessment of the system's sensitivity electricity, to prevent the flooding that had plagued to climate change, not as the literal implications of Chattanooga, and to ensure sufficiently stable particular general circulation models. riverflows that would permit maintenance of a 9- foot-deep navigation channel. Miller and Brock assessed the potential impacts of climate change on flood levels in The region administered by the TVA covers Chattanooga, Tennessee, using a model that had 40,000 square miles and includes parts of seven been developed to estimate the constraints on states. In the last half century, the TVA has weekly tributary releases. They also estimated the coordinated the construction of 43 major dams potential implications for water quality in the Upper along the river and its tributaries, many of which are Holston Basin of the valley, using a reservoir water shown in Figure 16-6. The system provides power quality model, a riverflow model, and a water 335 Chapter 16 A. NASHVILLE KNOXVILLE MEMPHIS CHATTANOOGA LEGEND: TVA POWER SERVICE AREA TENNESSEE RIVER WATERSHED B. BARKLEY RIVER OIHO CUMBERLAND RIVER NORMANDY COLUMBIA NORRIS TIMS FORD MISSISSIPPI RIVER KENTUCKY TENNESSEE RIVER MELTON HILL PICKWICK CHEROKEE WILSON WHEELER GUNTERSVILLE NICKAJACK : CHICKAMAUGA WATTS BAR FORT LOUDOUN DOUGLAS TELLICO ***** : Figure 16-6. (A) Map of the TVA region, and (B) schematic of the TVA reservoir system (Miller and Brock, Volume A). 336 Southeast quality model that TVA has used in the past to determine the environmental constraints affecting 90% riverflow. 1025 NORMAL 1020 1015 1010 Limitations 1005 (FEET) 1000 995 Because the riverflow scenarios were not based 990 on hydrologic analysis, conclusions cannot be drawn 985 980 regarding the sensitivity of riverflow to climate 975 BASE 970 change; a more thorough study should apply a DRY 965 WET basinwide hydrologic model to the region. A key 960 NORMAL MINIMUM 955 limitation for the flood analysis was that EPA NTM ; assumed that every storm in a given month would result in a change in riverflow proportional to the change in monthly runoff rather than incorporating 10% 1025 potential changes in flood frequency and intensity. NORMAL MAXIMUM 1020 (For climate change scenarios, see Chapter 4: 1015 1010 Methodology.) Finally, the study assumed that TVA 1005 1000 would not mitigate impacts by changing its ION (FEET) 995 operating rules for the reservoirs in response to 990 985 climate change. 980 R 975 BASE DRY 970 Results WET 965 NORMAL MINIMUM 960 955 Reservoir levels --- i isa Figure 16-7 shows the estimates of the changes in reservoir levels in the Norris Reservoir for the MEDIAN 1025 NORMAL wet and dry scenarios. Currently, water levels are 1020 1015 typically above 1,010 feet (NGVD) from early May 1010 1005 to early August. Under the wet scenario, the water (FEET) 1000 would generally be above this level from early April 995 TION 990 to early September; during the driest years (1%), 985 the water levels would be similar to the current 980 975 normal level between May and October. In the dry BASE 970 DRY 965 WET scenario, water levels would never exceed 1,005 feet NORMAL MINIMUM 960 in a typical year, and even during the wettest years 955 27M ; } in A : non : (1%) they would barely exceed the current normal condition between April and September. Figure 16-7. Water levels in Norris Reservoir under Changes in lake levels of this magnitude would climate scenarios: (A) 10% wettest years; (B) have important implications for recreation in the median; and (C) 10% driest years (adapted from Tennessee Valley, which is supported by facilities Miller and Brock, Volume A). worth over $600 million. Even today, recreation proponents are concerned with reservoir levels flows would reduce the dilution of municipal and dropping during some summers. Miller and Brock industrial effluents discharged into the river and its found that the wet scenario would largely eliminate tributaries. Moreover, because water would current problems with low lake levels; in contrast, generally remain at the bottom of reservoirs for a the dry scenario would make these problems the longer period of time, the amount of dissolved norm. oxygen could decline; this would directly harm fish and reduce the ability of streams to assimilate Water Quality wastes. Miller and Brock concluded that the water supplies from TVA would probably be sufficient, Miller and Brock found that a drier climate but that TVA could experience operational could also create environmental problems. Lower difficulties and customer dissatisfaction due to 337 Chapter 16 degraded water quality. During extended low-flow feasible operational change would be to cut back conditions, wastes would have increased power generation at fossil-fuel powerplants during opportunities to backflow upstream to water supply periods of low flow. However, hydropower intakes. production would also be reduced during periods of low flow, so cutting back production might not be Flooding acceptable. One alternative would be to construct cooling towers, which would eliminate discharges of Although a drier climate could exacerbate hot water, at a capital cost of approximately $75 many current problems facing TVA, a wetter million. climate could create difficulties, particularly the risk of flooding, in matters that are currently under Tennessee Valley Policy Study control. Miller and Brock found that in the wet scenario, during exceptionally wet years, storage Meo et al. (Volume J) analyzed the history, would be inadequate at the tributary reservoirs; this statutory authority, and institutional structure of the condition could result in uncontrolled spillage over TVA to assess the ability of the organization to dams. A high probability of flooding would also respond to climate change. Their analysis relied exist at Chattanooga. Miller and Brock examined both on the available literature and on interviews the levels of the five worst floods of the last 50 with a few dozen officials of TVA and states within years at Chattanooga, which did not overflow the the region. They divided the possible responses of banks of the Tennessee River or flood the city. TVA into two broad categories: (1) continuing the However, under the wet scenario, two of the floods current policy of maximizing the value of would overtop the banks. The worst flood could hydroelectric power, subject to the constraints of reach a level of 56.3 feet and cause over $1 billion flood control and navigation; and (2) modifying in damages; the second worst could reach a level of priorities so that power generation would be 46 feet and cause over $200 million in damages (see subordinated to other objectives if doing so would Figure 16-8). yield a greater benefit to the region. They concluded that if the climate became wetter, current Flooding could be reduced if operating rules policies would probably be adequate to address were modified to keep water levels lower in climate change because the only adverse effect reservoirs on tributaries (although this would would be the risk of additional flooding, which is diminish the hydropower benefits from a wetter already a top priority of the system. climate). However, changes in operating rules would not be sufficient to protect Chattanooga from If climate became drier, on the other hand, being flooded during a repeat of the worst storm, existing policies might be inadequate, because they because rainfall would be largely concentrated over require power generation to take precedence over the "mainstem" reservoirs, which do not have many of the resources that would be hardest hit. substantial flood-control storage. Although they expect that the TVA will be more successful at addressing future droughts, Meo et al. Power Generation found that during the 1985-86 drought, falling lake levels impaired recreation and reduced hydropower Miller and Brock calculated that the wet and generation, forcing the region to import power while dry scenarios imply, respectively, an annual increase five powerplants sat idle. of 3.2 megawatt-hours (16%, $54 million per year) and a decrease of 4.6 megawatt-hours (24%, $87 Meo et al. point out that groundwater tables million per year), given current capacity and are falling in parts of the region, in part because operating rules. numerous tributaries recharge the aquifers whenever water is flowing but are allowed to run Climate change could also have an impact on dry when water is not being released for fossil-fuel powerplants. If river temperatures hydropower. They suggest that even without climate become warmer, they will require additional dilution change, the deteriorating groundwater quality and water. Although sufficient water would be available availability are likely to lead a number of if the climate became wetter, meeting minimum communities to shift to surface water supplies in the flow requirements would be more difficult if climate coming decades, adding another use that must became drier. Miller suggested that the most compete for the water that is left over when the 338 Southeast Figure 16-8. Chattanooga was vulnerable to flooding until the TVA system of dams was constructed. The upper photo shows the 1867 Flood, with water levels similar to those projected by the Miller and Brock under the wet scenario (Miller and Brock, Volume A). demands for power have been met. Even with Studies of the Impacts on Lake Lanier current climate, they contend, the TVA should and Apalachicola Bay assess whether other uses of the region's water resources would benefit the economy more. If Figure 16-9 shows the boundaries of the climate becomes drier, the need for such a 19,800-square-mile Chattahoochee-Flint- reevaluation will be even more necessary. Apalachicola River Basin. The U.S. Army Corps of Engineers and others who manage the 339 Chapter 16 Lake Lanier Lake Lanier, located 30 miles northeast of Atlanta, is a source of water for the city and nearby jurisdictions. Federal statutes require the U.S. Army Corps of Engineers to manage Lake Lanier to BUFORD DAM provide flood control, navigation, and hydropower. TO PEACHTREE CREEK Nevertheless, the lake is also managed to meet nonstatutory objections such as recreation, minimum LAKE flows for environmental dilution, and water supply. LANIER Since Lake Lanier was dammed in 1957, the PEACHTREE CREEK statutory objectives of flooding and navigation have TO HEADWATERS OF WEST POINT LAKE been met; annual hydropower generation has been 134 MWH³, equal to 2% of today's power WEST POINT LAKE TO JIM WOODRUFF DAM requirements for Atlanta; and the releases of water have fulfilled the additional minimum flow needed to dilute the effluents from sewage treatment plants. During the last two decades, the lake's shoreline has been substantially developed with marinas, houses, and hotels. To a large degree, the residents have become accustomed to the higher water levels that prevailed from the 1970s through 1984. Droughts from 1985 to the present, however, have lowered lake levels, disrupting recreation. In APALACHICOLA RIVER the summer of 1986, navigation for recreational AND APALACHICOLA BAY boats located downstream of the lake was curtailed because of minimal releases from the lake. In 1988, Atlanta imposed water-use restrictions, with the objective of cutting consumption by 10 to 20%. A bill has been introduced to add recreation to the list of statutory purposes (HR-4257). Runoff in the Chattahoochee River Basin Figure 16-9. Drainage area of the Apalachicola- Chattahoochee-Flint River system. Study Design. Hains estimated runoff in the Chattahoochee River Basin and the flow of water into Lake Lanier for the three scenarios. He Chattahoochee River as it passes through Lake calibrated the Sacramento hydrology model Lanier on its way to the Apalachicola estuary and developed by the National Weather Service the Gulf of Mexico face many of the same issues as (Burnash et al., 1973) to the conditions found in those faced by the TVA. However, they also are the watershed of the upper Chattahoochee River. managing the water supply of Atlanta, the second He then generated scenarios of riverflow for the largest city in the Southeast, and the flow of water baseline climate and the GCM scenarios. into an estuary that supports the most productive fishery in Florida (U.S. Department of Commerce, Limitations. The Sacramento model was 1988). designed primarily for flood forecasting, not base flow. In addition, the model was calibrated using A number of researchers were involved in the data on evaporation of water from pans, which EPA's assessment of the potential implications of climate change for this watershed. A study of Lake Lanier and a study of the implications for the fish in 3 Personal communication from Harold Jones, Systems Apalachicola Bay are discussed in the following Engineer, Southeast Power Administration, Department of Energy, September 12, 1988. sections of this chapter. 340 Southeast is not perfectly correlated with evapotranspiration, modified a monthly water balance model/operations and these data came from a nearby watershed. model previously applied in southern California for the lake, based on current operating rules for the Since the analysis was based on scenarios of reservoir. For the first set of runs, the model average monthly change, it did not consider assumes that (1) minimum flows are maintained for potential changes in variability of events such as navigation and environmental dilution at all times, floods. The analysis did not incorporate changes in (2) lake levels are kept low enough to prevent vegetation, which could affect runoff. flooding, (3) historic rates of consumption continue, and (4) peak hydropower generation is maximized. Results. As with the Tennessee River, the To ensure that the assumptions adequately reflect major climate models disagree on whether the the actual decision rules used by water managers, Chattahoochee watershed would become wetter or Sheer and Randall reviewed the rules with local drier with an effective doubling of greenhouse gases. officials from the U.S. Army Corps of Engineers, Hains estimated that under the wetter GISS the Atlanta Regional Council, and others scenario, the average annual riverflow of the responsible for managing the water supply. In a Chattahoochee River would increase by 13%; the second set of runs, they examined the impacts of drier OSU and GFDL models imply declines of 19 climate change under alternative operating rules and 27%, respectively, as shown in Figure 16-10. that assume recreation is also a statutory objective. The GISS scenario implies slight decreases in winter flow and increases the rest of the year. Under the Limitations. Sheer and Randall did not GFDL scenario, these substantial decreases were consider changes in demand for water due to estimated throughout the year, with almost no flow climate change or population growth; thus, it in late summer. The OSU scenario also shows produces high estimates of future water availability reductions, but the reduction is greatest during the under all scenarios. Moreover, the results were not flood season (February to May) and negligible compared with historic lake levels. during the dry season (late summer/early fall). Results. Figure 16-11 shows the Sheer and Management of Lake Lanier Randall estimates of lake levels; Figure 16-12 shows quarterly hydropower production. Under the Study Design. Sheer and Randall (Volume A) relatively wet GISS scenario, annual power examined the implications for water management of production could increase by 9%. The higher the riverflow changes estimated by Hains. They streamflows in this scenario would still be well below those that occasionally occurred before Lake Lanier was closed; hence, no significant threat of flooding would exist for a repeat of the climate of SEASONAL FLOW RATIOS 1.5 1951-80. Under the relatively dry GFDL scenario, 1.4 GFDL however, power production could drop 47%, and OSU 1.3 lake levels would be likely to drop enough to GISS 1.2 substantially disrupt recreation. This scenario OBSERVED assumes that Atlanta would continue to take as 1.1 RATIO (SCENARIO/BASE) 1.0 much water as it does currently (allowing for growth would increase water supply problems). 0.9 0.8 Sheer and Randall also examined the 0.7 implications of making recreation a statutory 0.6 objective. Although it would be possible to 0.5 maintain lake levels, Atlanta's water supply would 0.4 be threatened. With the current climate, strict 0.3 OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP enforcement of such a policy would result in Lake MONTH Lanier supplying no water to metropolitan Atlanta for 8 months of every 30 years. Although under the Figure 16-10. Ratios of flow under doubled CO2 GISS scenario this would be reduced to 1 month, scenarios to base case in Upper Chattahoochee under the dry GFDL scenario, Atlanta would have River. to use an alternative source of water 1 to 3 months each summer. 341 Chapter 16 BASE GISS GFDL 1075 1065 TARGET ELEVATION (FEET) 1055 1045 50 55 60 65 70 75 80 YEAR Figure 16-11. Lake Lanier elevation (September) under doubled CO₂ scenarios (Sheer and Randall, Volume A). BASE GISS GFDL 20000 1 15000 MWH PER YEAR 10000 5000 0 50 55 60 65 70 75 80 YEAR Figure 16-12. Lake Lanier power generation under doubled CO2 scenarios (Sheer and Randall, Volume A). Implications. Climate change combined with would be unwise to assume that minimum flows population growth may require water managers to could be decreased because future growth may reexamine the tradeoffs between the various uses of increase the need for dilution of effluents, and the Chattahoochee River and Lake Lanier. A warmer temperatures would speed biological number of local water officials who met with Sheer activity. The likely impacts of climate change on suggested that an appropriate response to changing Apalachicola Bay may also increase the need to water availability might be to relax minimum flow maintain minimum flows. requirements for navigation and environmental quality. They reasoned that minimum flows for Apalachicola Bay environmental purposes are based on the assumption that sewage treatment plants are Apalachicola Bay supports hundreds of discharging at their maximum rates and that commercial fishermen; over 80% of Franklin temperatures are high, conditions that are usually County earns a livelihood from the bay (Meo et al. not met. They also argued that little is Volume J). The contribution of fishing to the area accomplished by maintaining minimum flows for was estimated at $20 million for 1980, representing navigation because ship traffic is light in the lower 90% of Florida's oyster harvest and 10% of its Chattahoochee. Others argued, however, that it shrimp harvest. This figure is projected to grow to $30 to $60 million by 2000. 342 Southeast Although the state has purchased most of the regression expressed the logarithm of riverflow as a land that is not part of a commercial forest, function of the logarithms of precipitation and economic pressures on forestry companies to sell evapotranspiration for a few weather stations land for coastal development are increasing. In located in the basin. 1979, the National Oceanic and Atmospheric Administration created the Apalachicola National Limitations. Hains' procedure greatly Estuarine Sanctuary to prevent development from oversimplified the relationships between the causal encroaching into this relatively pristine estuarine variables and riverflow, ignoring the impacts of environment. reservoir releases and the failure of the relationships to fit the simple log-linear form. These results The biology of the Apalachicola Bay estuary should be interpreted as an indication of the may be affected by higher temperatures, higher sea potential direction of change. levels, and different flows of water into the Apalachicola River. Hains estimated the flow of the Results. Figure 16-13 illustrates Hains' Apalachicola River, and Park et al. estimated estimates of average monthly flows for the wetland loss due to sea level rise. Livingston used Apalachicola estuary. Annual riverflow would both of these results and the temperature change scenarios to evaluate the potential impacts on the bay's fish populations. SEASONAL FLOW 1.2 Sea Level Rise GFDL 1.1 OSU 1.0 GISS The methods of Park et al. for estimating OBSERVED 0.9 wetland loss are described in Chapter 7: Sea Level 0.8 Rise. They estimated that a 1-meter rise in sea level would inundate approximately 60% of the salt marshes in Apalachicola Bay, and that mangrove DISCHARGE M³/SEC) (THOUSANDS) 0.7 0.6 swamps, which are rarely found outside southern 0.5 Florida today, would replace the remaining salt 0.4 marsh. Table 16-5 illustrates their estimates. 0.3 0.2 Apalachicola Riverflow 0.1 OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP MONTH Study Design. Hains estimated the impact of climate change on riverflow, using a regression model, which is simpler than the Sacramento model Figure 16-13. Doubled CO₂ flow into Apalachicola he used for the Chattahoochee River analysis. The Bay (Hains, Volume A). Table 16-5. Remaining Coastal Wetlands in Apalachicola Bay in the Year 2100 (hectares) Current sea 50-cm 100-cm 200-cm Area 1987 level rise rise rise rise Swamps 9.46 6.71 6.26 5.47 4.16 Fresh marsh 1.46 1.27 1.17 1.00 0.25 High marsh 1.19 0.37 0.04 0.04 0.02 Low marsh 3.42 2.33 0.39 0.06 0.03 Mangrove 0 0 3.06 2.13 1.80 Total wetlands 15.53 10.68 10.92 8.70 6.26 Source: Park et al. (Volume B). 343 Chapter 16 decrease under all scenarios, although it would unable to estimate the impact of wetland loss on increase in the summer and fall for the GISS and populations of finfish and shellfish. OSU scenarios, respectively. The limitations in Hains' estimates of riverflow Fish Populations in Apalachicola Bay do not significantly affect the results of Livingston's study because riverflow was only one of several Study Design. Using data from the literature variables to be considered. The uncertainties on the tolerance of various species to warmer surrounding changes in rainfall probably dwarf any temperatures, Livingston estimated the number of errors due to Hains' simplified hydrology, and months in a typical 30-year period during which the higher temperatures and sea level rise appear to be estuary would be too hot for these species and more important. extrapolated this information to estimate reductions in populations. Results. The results of this study suggest a dramatic transformation of the estuary from Hydrologic modeling was not used to estimate subtropical to tropical conditions. the combined impacts of sea level rise and changing riverflow on salinity. Instead Livingston used Warmer temperatures. Livingston concluded historic data to estimate regression equations that warmer temperatures would have a profound relating riverflow to salinity and salinity to effect on seafood species in the estuary because populations of some commercially important many species cannot tolerate temperatures much seafood species. above those that currently prevail. Figure 16-14 compares the number of months in a 6-year period Limitations. There is no historical record by (based on 1971-76) in which temperatures exceed a which to estimate the impact of warmer particular level for the current climate and the temperatures on the Apalachicola (or any other) GISS and GFDL scenarios, with known thresholds estuary. Livingston did not model the relationships for major commercial species. between various aquatic species or how they would change. He did not consider how finfish and Livingston concluded that crabs, shrimp, shellfish might adapt to climate change, and he was oysters, and flounder could not survive in the Base GISS 25 GFDL 20 Number of months 15 10 5 Blue Crab Blue Crab Oyster Larvae Croaker Redfish White Shrimp Larvae (30°) Juvenile (33°) Spotted Sea Trout (36°) (37.5°) (42°) Pinfish Flounder (35°) Figure 16-14. Months in a 6-year period during which temperatures (°C) would be too high for selected species under doubled CO2 scenarios (Livingston, Volume E). 344 Southeast estuary with the warming estimated in the GISS and impacts on the estuary. Table 16-6 shows GFDL scenarios, which imply close to 100% Livingston's estimates of losses in particulate mortality for blue crab larvae and juveniles. The organic carbon, the basic source of food for fish in GFDL scenario could cause over 90% mortality for the estuary. Sea level rise between 50 and 200 spotted sea trout, oyster larvae, panfish, and centimeters would reduce available food by 42 to flounder. The mortality under the milder GISS 78%. A proportionate loss in seafood populations scenario would be only 60%. would not necessarily occur, since organic carbon food supplies are not currently the constraining Although Livingston concludes that the oysters factor for estuarine populations. However, wetlands would probably be eliminated, he cautions that also are important to larvae and small shrimp, shrimp and other mobile species might adapt by crabs, and other species, serving as a refuge from fleeing the estuary for cooler gulf waters during the predators. A rise in sea level of a meter or more summer. However, such a flight would leave them could lead to a major loss of fisheries. vulnerable to predators. Despite the adverse impacts on shellfish and Increased salinity. Although sea level rise and flounder, a number of species might benefit from warmer temperatures seem likely to substantially global warming. For example, Livingston points out reduce the productivity of the estuary, the probable that pink shrimp could become more prevalent. impact of precipitation changes is less clear. If Moreover, some finfish spend their winters in riverflow in the Chattahoochee declines, it would Apalachicola Bay and occasionally find the estuary combine with sea level rise to increase salinity too cold. Other species such as rock lobster that concentrations in the estuary. Livingston concluded generally find the waters too cold at present may that oysters are the most vulnerable to increases in also be found in the estuary in the future. salinity because oyster drill and other predators, as well as the disease MSX, generally require high Implications. Based on Livingston's salinities. Livingston estimated losses of 10 to 35% projections, Meo et al. (Volume J) used current for oysters, blue crabs, finfish, and white shrimp retail prices of fish to estimate that the annual net under the GFDL scenario because of salinity economic loss to Franklin County could be $5 to increases alone. $15 million under the GFDL scenario, $1 to $4 million under GISS, and $4 to $12 million under the Sea level rise. Livingston also concluded that OSU scenario. the loss of wetland acreage would have important Table 16-6. Projected Changes of the Net Input of Organic Carbon (metric tons per year) to the Apalachicola Bay System for Various Scenarios of Sea Level Rise Fresh Sea- Salt Phyto- Factor wetlands grass marshes plankton Total Current scenario 30,000 27,200 46,905 233,280 337,385 for 2100 Baseline 26,100 28,700 23,500 144,640 222,940 sea level rise 0.5-meter rise 24,000 28,800 4,690 71,450 128,940 1.0-meter rise 21,300 30,100 940 58,790 111,130 2.0-meter rise 4,980 31,035 780 15,160 51,955 Source: Livingston (Volume E). 345 Chapter 16 Livingston's results should not be interpreted yields. The direct effects of CO2 in the crop to mean that fishing will be eliminated from modeling study results may be overestimated for Apalachicola Bay. The extent to which two reasons. First, experimental results from commercially viable tropical species could replace controlled environments may show more positive the species that are lost was not estimated. effects of CO2 than would actually occur in variable, windy, and pest-infested (weeds, insects, and Agriculture diseases) field conditions. Second, because other radiatively active trace gases, such as methane, also Agriculture in the Southeast will be affected are increasing, the equivalent warming of a doubled directly by changes in climate and indirectly by CO2 climate may occur somewhat before an actual changes in economic conditions and pests. This doubling of atmospheric CO2. A level of 660 ppm section presents results from a crop modeling study CO2 was assumed for the crop modeling of yield changes by Peart et al., and regional results experiments, while the CO₂ concentration in 2060 is from national studies of agricultural production estimated to be 555 ppm (Hansen et al., 1988) (see shifts by Adams et al. (Volume C) and of impacts of Chapter 6: Agriculture). changes in pest populations by Stinner et al. (Volume C). The study assumed that soils were relatively favorable for crops, with low salinity or compaction, Crop Modeling Study and assumed no limits on the supply of all nutrients, except nitrogen. The analysis considers neither Study Design change in technology nor adverse impacts due to changes in storm frequency, droughts, and pests and Peart et al. (Volume C) used the crop models pathogens. CERES-Maize (Jones and Kiniry, 1986) and SOYGRO (Wilkerson et al., 1985) to estimate the Results impacts of climate change on yields of corn and soybeans for 19 sites throughout the Southeast and Soybean Yields. Table 16-7 illustrates the adjacent states. Agricultural scientists have used results of the soybean model for 13 nonirrigated these models for several years to project the impacts sites in the study area, as well as Lynchburg, of short-term climatic variations. They incorporate Virginia, a colder site included for comparison the responses of crops to solar radiation, purposes. temperature, precipitation, and soil type, and they have been validated over a large range of climate The relatively wet GISS and relatively dry and soil conditions in the United States and other GFDL scenarios imply very different impacts on countries. yields. In the GISS scenario, the cooler sites in Georgia and the Carolinas mostly show declines in The major variable not considered by these soybeans yields of 3 to 25%, and the other sites and other existing agricultural models is the direct show declines of 20 to 39%, ignoring CO₂ "fertilization effect" of increased levels of fertilization. When the latter effect is included, the atmospheric carbon dioxide. Peart et al., therefore, Atlantic Coast States were estimated to experience modified their models to consider both the gains of 11 to 39%, and the other states could vary increased rate of photosynthesis and the increased from a 13% drop in Memphis to a 15% gain in water-use efficiency that corn and soybeans have Tallahassee. (Tennessee fares worse than the North exhibited in field experiments (see Chapter 6: Carolina sites at similar latitudes because its grid Agriculture). cell does not receive as favorable an increase in water availability.) Limitations By contrast, the dry GFDL scenario results in The analysis of combined effects is new very large drops in soybean productivity, with all but research and will need further development and one site experiencing declines greater than 50% and refinement. The model runs use simple parameters eight sites losing over 75%, considering only the for CO₂ effects, assume higher atmospheric impact of climate change. Even when CO2 concentration of CO2 than are predicted, and fertilization is considered, all but four sites probably overestimate the beneficial impact on crop experience losses greater than 50%. 346 Southeast Table 16-7. Impacts of Doubled CO2 Climate Change on Soybean Yields for Selected Southeastern Sites for Climate Change Alone and for Climate Change and CO2 Fertilization (percentage change in yield)ᵃ Climate change Climate change only and CO₂ fertilization Site GISS GFDL GISS GFDL Memphis, TN -38 -88 -13 -70 Nashville, TN -30 -52 +4 -81 Charlotte, NC -7 -92 +32 -88 Raleigh, NC -3 -87 + 39 -76 Columbia, SC -20 -78 + 18 -62 Wilmington, NC -11 -62 + 25 -41 Atlanta, GA -11 -78 + 27 -67 Macon, GA -25 -91 + 11 -82 Tallahassee, FL -20 -51 + 15 -17 Birmingham, AL -31 -54 0 -29 Mobile, AL -34 -43 -8 error Montgomery, AL -39 -84 -10 -68 Meridian, MS -37 -78 -9 -66 Lynchburg, vab +1 -74 +49 -55 The impacts of CO₂ fertilization cannot be quantified as accurately as climate change only. The climates shown here overstate the beneficial impact of CO₂ because Peart et al. assume that CO₂ has doubled. Because other gases contribute to the global warming, CO2 will have increased by a smaller fraction. ᵇPeart et al. investigated a number of sites in states adjacent to the Southeast. Lynchburg is included to permit comparison of results for the Southeast with a colder site. Source: Peart et al. (Volume C). Corn Yields. The two scenarios differ in a The results are mixed on whether currently dry similar fashion for nonirrigated corn. However, in land areas would be shifted to irrigation. Table 16- the case of irrigated corn, where the analysis 8 shows the percentage increases in yields that primarily reflects the impact of temperature would result from adding irrigation for particular increases, the two scenarios show more agreement. scenarios. All but four sites could increase yields When CO₂ fertilization was not considered, drops of today by 50 to 75% by irrigating. Under the wetter 13 to 20% were estimated in the GISS scenario, and GISS scenario, irrigation would increase yields only drops of 20 to 35% were calculated for the GFDL 7 to 53% (compared with not irrigating under the scenario. When CO₂ fertilization was included, the GISS scenario). However, under the dry GFDL GISS scenario implied declines of less than 8% for scenario, irrigation would increase yields by 50 to all sites, and the GFDL model showed similar 493% -- that is, it would mean the difference declines for two sites and respective declines of 17 between crop failure and a harvest slightly above and 27% for Charlotte, North Carolina, and Macon, today's levels in most years. Even without CO₂ Georgia. fertilization, 75% of the nonirrigated southeastern sites could gain more from irrigation than they Irrigation. The two scenarios show more would lose from the change in climate resulting agreement for agricultural fields that are already from the GFDL scenario. irrigated. Since the changes in water availability are irrelevant here, the impacts are dominated by the A farmer's decision to irrigate, shift to other increased frequency of very hot days. crops, or remove land from production would depend to a large degree on what happens to prices 347 Chapter 16 Table 16-8. Increases in Corn Yields from a Shift study are discussed in Chapter 6: Agriculture.) to Irrigation (percent, assuming no Their results suggest that the impact of climate CO2 fertilization)ᵃ change on southeastern agriculture would not be directly proportional to the impact on crop yields (Table 16-9). Current Site climate GISS GFDL Considering only the impact of climate change, Adams et al. found that the GISS and GFDL scenarios would reduce crop acreage by 10 and Memphis, TN 70 50 270 16%, respectively. When CO2 fertilization is Nashville, TN 65 49 205 considered, however, Adams et al. project respective Charlotte, NC 64 43 486 declines in farm acreage of 57 and 33% for the Raleigh, NC 51 28 444 GISS and GFDL scenarios. As yields increase, Columbia, SC 58 47 386 prices decline. Adams et al. estimate that most Wilmington, NC 16 8 50 areas of the nation would lose farm acreage. Atlanta, GA 15 7 79 However, they estimate that the Southeast would Macon, GA 61 33 489 experience the worst losses: while the Southeast has Birmingham, AL 6 9 61 only 13% of the cultivated acreage, it would account Mobile, AL 36 41 91 for 60 to 70% of the nationwide decline in farm Montgomery, AL 72 39 493 acreage. This result is driven by the increased Meridian, MS 62 53 323 yields that the rest of the nation would experience Lynchburg, vab 56 37 361 relative to the Southeast. a Estimates represent the change in yields, given When the CO2 fertilization effect is ignored, a particular scenario, from shifting to irrigation. the reductions in acreage would be much smaller, b Peart et al. investigated a number of sites in states although the Southeast would still account for 40 to adjacent to the Southeast. Lynchburg is included 75% of the nationwide loss. The general decline in to permit comparison of Southeast results with yields would boost prices, which could make it those for a colder site. economical for many farmers to irrigate and thereby Source: Column 1 from Peart et al. (Volume C); avoid the large losses associated with a warmer and Columns 2 and 3 derived from Peart et al. possibly drier climate. and Column 1. Agricultural Pests of both crops and water. Even though water is The modeling and economic studies of plentiful today, the capital costs of irrigation prevent agriculture do not consider the impact of pests on most farmers in the Southeast from taking crop yields. However, Stinner et al. (Volume C) advantage of the potential 50% increases in yields. suggest that global warming would increase the But if crop failures due to drought became as range of several agricultural pests that plague commonplace as Peart et al. project for the dry southeastern agriculture. (For details on the GFDL scenario, a major increase in irrigation methods of this nationwide study, see Chapter 6: probably would be necessary. Although Agriculture.) They point out that the northern groundwater is currently plentiful in the Southeast, ranges of potato leafhoppers, sunflower moths, no one has assessed whether there would still be black cutworms, and several other southeastern enough water if the climate became drier and pests are limited by their inability to survive a cold irrigation increased. Furthermore, climate change winter. Thus, milder winters would enable them to may increase the demand for water for move farther north, as illustrated in Figure 16-15. nonagricultural uses. Stinner et al. also note that increased drought frequency could increase the frequency of pest Shifts in Production infestations. Adams et al. (Volume C) examined the Implications of Agriculture Studies impacts of changes in crop yields on farm profitability and cultivated acreage in various Agriculture appears to be at least as vulnerable regions of the United States. (The methods for this to a potential change in climate in the Southeast 348 Southeast Table 16-9. Impact of Climate Change on Cultivated Acreage in the Southeastᵃ (figures in parentheses are percentage losses) With Direct CO, Without Direct CO₂ Region Baseline GISS GFDL GISS GFDL Acreage (millions) SE coast 12.5 8.7(30) 7.8(38) 11.5(8) 11.2(10) Appalachia 15.5 2.8(82) 7.4(52) 14.1(9) 12.9(17) Delta 19.9 9.3(53) 16.7(16) 17.7(11) 16.2(19) Total 47.9 20.8(57) 31.9(33) 43.3(10) 40.3(16) SE coast includes Florida, South Carolina, Georgia, Alabama. Appalachia includes North Carolina, Tennessee, Virginia, West Virginia, Kentucky. Delta includes Mississippi, Louisiana, Arkansas. Source: Adams et al. (Volume C). SUNFLOWER MOTH GREEN CLOVERWORM GISS PRESENT GFDL GISS PRESENT GFDL POTATO LEAFHOPPER BLACK CUTWORM GISS GFDL PRESENT PRESENT GISS GFDL Figure 16-15. Present and predicted northern ranges of various agricultural pests (Stinner et al., Volume C). as in any other section of the country. Unlike many Florida may present an important exception to of the colder regions, the benefits of a longer the generally unfavorable implications of climate growing season would not appreciably offset the change for crop yields. Although Florida is the adverse impacts of warmer temperatures in the warmest state in the Southeast, its agriculture Southeast, where cold weather generally is not a appears to be harmed by cold temperatures more major constraint to agricultural production. than the agriculture of other states in the region. In 349 Chapter 16 recent years, hard freezes have destroyed a large the stand simulation model did not include sub- fraction of the citrus harvest several times. As a tropical species, it was unable to simulate any result, the industry is moving south into areas near vegetation along the gulf coast under the very warm the Everglades, and sugarcane, which also thrives in doubled CO2 climate. The results for southern pine warm temperatures, is expanding into the were less conclusive but generally show the upper Everglades themselves. Global warming could border of the species range moving northward while enable the citrus and sugarcane areas to include the southern border remains stable. Growing most of the state. Warmer temperatures also would conditions along the gulf coastal region, however, help coffee and other tropical crops that are would also be favorable to subtropical species in a beginning to gain a foothold in the state. This doubled CO2 environment, but since the models study, however, did not examine how the frequency used in the study had no data on such species, it is of extreme events, such as the number of days unclear how southern pine might fare under below freezing in Florida, would change. competition with subtropical varieties. Although Florida's relative abundance of water Transitional Effects may make it the exception, the current situation there highlights an important aspect of climate Study Design change: Within the context of current prices and crop patterns, the impact of climate change appears Urban and Shugart (Volume D) applied a to be unfavorable. However, warmer temperatures forest simulation model to a bottomland hardwood may present farmers with opportunities to grow forest along the Chattahoochee River in Georgia different crops whose prices would justify irrigation and to upland sites near Knoxville, Tennessee, or whose seasonal cycles would conform more Macon, Georgia, Florence, South Carolina, and closely to future rainfall patterns. Vicksburg, Mississippi. Their study considered the OSU, GFDL, and GISS scenarios for doubled CO₂, Forests as well as the GISS transient A scenario through the year 2060. Potential Range Shifts The model these researchers used was derived Study Design from FORET, the "gap" model originally developed by Shugart and West (1977). The model simulates Overpeck and Bartlein (Volume D) used two forest dynamics by modeling the growth of each tree independent methods to study the potential shifts in in a representative plot of forest land. It keeps ranges of forest types over eastern North America. track of forest dynamics by assigning each of 45 tree These analyses suggest where trees are likely to species optimal growth rates, seeding rates, and grow in equilibrium doubled CO₂ climate conditions survival probabilities, and by subsequently adjusting after allowing for migration of tree species to fully these measures downward to account for less than catch up with climate change. The study only optimal light availability, temperature, soil moisture, indicates the approximate abundance of different and soil fertility. In the case of the bottomland species within a range, not what the transitional hardwood site, the model also considers changes in effects of climate on forests might be, or how fast river flooding, based on the flows in the lower trees will be able to migrate to the new ranges. Chattahoochee calculated in the Lake Lanier study. (For a discussion of the study's methodology and The researchers applied the model to both mature limitations, see Chapter 5: Forests.) forests and the formation of a new forest from bare ground. Results Limitations Three GCM scenarios and two vegetation models yielded similar results. The abundance of The results should not be taken literally owing deciduous hardwood populations (e.g., oak), which to a number of simplifying assumptions that Urban currently occupy the entire modeled eastern region and Shugart had to make. First, they assumed that from the Great Lakes region to the gulf coast, certain major species, such as loblolly pine, could would shift northward away from the gulf coast and not tolerate more than 6,000 (cooling) degree-days almost entirely out of the study region. Because per year. These species are not currently found in warmer areas, but the southern limits of their 350 Southeast range are also limited by factors other than be generated from bare ground, particularly if the temperature, such as the Gulf of Mexico and the climate becomes drier as well as warmer. For the dry climate of Texas and Mexico. Although the Knoxville site, the dry GFDL scenario implies that 6,000 degree-day line coincides with these species' a forest could not be started from bare ground, southern boundary across Florida, the peculiar while the GISS and OSU doubled CO2 scenarios environmental conditions of that state make it estimate reductions in biomass of 10 to 25%. For impossible to confidently attribute an estimate of the South Carolina site, only the GISS climate thermal tolerance to that observation alone. This would support a forest, albeit at less than 50% of caveat does not apply to most of the oaks, hickories, today's productivity. The Georgia and Mississippi and other species found in the cooler areas of the sites could not generate a forest from bare ground Southeast. for any of the scenarios. Thus, even with increased rainfall, some sites would have difficulty supporting Another important caveat is that the model regeneration. does not consider the potentially beneficial impact of CO2 fertilization on photosynthesis, changes in The transient analyses suggest that mature water-use efficiency, or leaf area. Nor did the forests could also be lost -- not merely converted to analysis consider introduction of new species into a different type -- if climate changes. Figure 16-16 the region. Thus, there is more confidence about shows that none of the forests would decline the fate of species currently in the region than about significantly within 50 years; however, all would what may replace those species. decline substantially before the end of the transient run in 80 years. The Mississippi forest would Results mostly die within 60 years, and the South Carolina and Georgia forests within 80 years. Only the The simulations by Urban and Shugart call relatively cool Tennessee site would remain into question the ability of southeastern forests to MISSISSIPPI TRANSIENT GEORGIA TRANSIENT Dynamics of Mature Forest Dynamics of Mature Upland Forest 160 180 140 160 140 120 Woody Biomass (T/ha) 120 100 80 Woody Biomass (T/ha) 100 80 60 60 Woody Biomass 40 40 Woody Biomass Control 20 Control 20 Transient Transient 0 0 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 Simulation Year Simulation Year SOUTH CAROLINA TRANSIENT EAST TENNESSEE TRANSIENT Dynamics of Mature Forest Dynamics of Mature Forest 180 180 160 160 140 140 Woody Biomass (T/ha) 120 Woody Biomass (T/ha) 120 100 100 80 80 60 60 40 Woody Biomass 40 Woody Biomass Control Control 20 Transient 20 Transient 0 0 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 Simulation Year Simulation Year Figure 16-16. Response of southeastern forests to GISS transient scenarios of climate change (Urban and Shugart, Volume D). 351 Chapter 16 somewhat healthy, although biomass would decline demand would rise more than annual demand. (This 35%. result is also sensitive to changes in variability.) Although the simulation results suggest that Linder and Inglis compared increases in southeastern forests are unlikely to benefit from the electric capacity required by climate change with global warming, the impact on forests may not be as those necessitated by economic growth. They bad as the model suggests, if new species move in estimated that through 2010, climate change could or if loblolly pine can tolerate more than 6,000 increase the expected capital costs of $137 billion by degree days per year. Nevertheless, major shifts in 6 to 9%; through 2055, it could increase expected forest types are almost certain to occur from the requirements of $350 to $500 billion by as much as warmer temperatures alone. 20%. Electric Utilities COASTAL LOUISIANA Linder and Inglis (Volume H) examined the impact of global warming on the demand for The sediment washing down the Mississippi electricity throughout the Southeast for the two River has formed the nation's largest delta at the GISS transient scenarios. (For additional details on river's mouth, almost all of which is in Louisiana. the methods and limitations of this study, see Composed mostly of marsh, cypress swamps, and Chapter 10: Electricity Demand.) Because their small "distributary" channels that carry water, study was limited to electricity, it did not consider sediment, and nutrients from the river to these the reduced consumption of oil and gas for space marshes and swamps, Louisiana's wetlands support heating that would result from warmer half of the nation's shellfish, one-fourth of its fishing temperatures. industry, and a large trapping industry. They also provide flood protection for metropolitan New Table 16-10 shows the percentage changes in Orleans and critical habitats for bald eagles and electric power requirements for various areas in the other migratory birds. Southeast. Along the gulf coast, annual power requirements could increase 3 to 4% by 2010 and 10 Water management and other human activities to 14% by 2055; elsewhere, the increases could be of the last 50 years are now causing this delta to somewhat less. Because peak demand for electricity disintegrate at a rate of about 100 square kilometers generally occurs during extremely hot weather, peak per year. Sediment that used to replenish the Table 16-10. Percentage Increases in Peak and Annual Demand for Electricity by 2010 and 2055 as a Result of Climate Change GISS A (2010) GISS B (2010) GISS A (2055) Area Annual Peak Annual Peak Annual Peak North Carolina, 1.6 7.3 1.3 2.4 5.9 24.4 South Carolina, Georgia Florida 2.7 4.9 2.7 3.6 9.3 20.0 Eastern Tennessee 1.6 3.7 1.3 1.2 5.9 12.2 Alabama, Western 1.9 3.8 2.2 5.7 6.8 13.5 Tennessee Mississippi 3.8 7.6 4.4 11.4 13.6 6.9 Louisiana 2.9 7.6 2.7 6.6 10.2 23.4 East Texas 3.1 7.9 2.8 6.6 11.3 25.3 Source: Linder and Inglis (Volume H). 352 Southeast delta now largely washes into the deep waters of the Strictly speaking, the entire loss of coastal gulf because flood-control and navigation guide Louisiana's estuaries should not be attributed to levees confine the flow of the river. Thus, the delta global warming because the ecosystem is already is gradually being submerged, and cypress swamps being lost. However, major efforts are being are converting to open-water lakes as saltwater initiated by the U.S. Army Corps of Engineers, the penetrates inland. If current trends continue, U.S. Fish and Wildlife Service, the Louisiana almost all the wetlands will be lost in the next Geological Survey, several local governments, and century. other federal and state agencies to curtail the loss, generally by erecting structures to provide A rise in sea level would further accelerate the freshwater and sediment to the wetlands. Technical rate of land loss in coastal Louisiana. As shown in staff responsible for developing these solutions Figure 16-17, even a 50-centimeter rise in sea level generally fear, however, that a 1-meter rise in sea (in combination with land subsidence) would level could overwhelm current efforts, and that if inundate almost all of the delta and would leave such a rise is ultimately going to take place, they New Orleans, most of which is below sea level and already should be planning and implementing a only protected with earthen levees, vulnerable to a much broader effort (Louisiana Wetland Protection hurricane. Panel, 1987). SISSIN N ISSIPPI Baton Rouge Lafayette Lake Charles New Orleans Morgan City Houma GULF OF MEXICO Projected Land Surface 0 20 40 miles 0 20 40 kilometers Figure 16-17. Projected future coastline of Louisiana for the year 2033, given a rise in sea level of 55 cm as predicted in the high scenario (Louisiana Wetland Protection Panel, 1987). 353 Chapter 16 POLICY IMPLICATIONS Impacts of Wetter Climate Agriculture and Forests Although most water resource problems have been associated with too little water, it does not Climate change could have a major impact on necessarily follow that a wetter climate would be land use in the Southeast. The estimated generally beneficial. The designs of water abandonment of 10 to 50% of the farmland in the management infrastructure and the location of Southeast and large declines in forests raise the an development along lakes and rivers have been based important question: How will this land be used? on current climate. Hence, shifts in either direction would create problems. In the past, forests have been cleared for agriculture, and when abandoned, they have been The chief problem from a wetter climate would converted to forest again. But the forest models be more flooding, particularly in southern Florida suggest that the impact of climate change on the and coastal Louisiana, where water often lingers for generation of new forests from bare ground would days and even weeks after severe rainstorms and be even more adverse than the impact on existing river surges. Inland communities, such as forests. If the forest simulations are correct, the Chattanooga, also might face flooding if wetter abandoned fields would become grasslands or would periods exceed the ability of dams to prevent become overgrown with weeds, and the Southeast flooding. could gradually come to resemble the scenery found today in the Great Plains. However, no one has Impacts of Drier Climate systematically investigated the extent to which human infrastructure might stabilize these changes. A drier climate, on the other hand, would Changes in crops might enable more farms to stay exacerbate current conflicts over water use during in business than Adams et al. project, and new dry periods. Hydropower would decline, increasing varieties of trees may find the region more the need to use fossil or nuclear power, both of hospitable. Because the commercial forests in the which would require more water for cooling. Southeast generally have short rotation cycles, it Conflicts between municipal water users and may be easier to respond to climate change there recreational interests also would intensify. Lake than in other regions. To a large degree, the ability levels could drop more during the summer, even if of human intervention to maintain the present municipal use of water did not grow. However, landscape would depend on international prices of warmer temperatures probably would increase agricultural and forest products, estimation of which municipal water demand for cooling buildings and is outside the scope of this report. watering lawns. These conflicts could be further exacerbated Water Resources if farmers increase the use of irrigation. Groundwater is available in reasonably shallow The water resource problems faced by the aquifers that drain into rivers. Any consumptive use Southeast are not likely to be as severe as the of water from these aquifers would reduce, and in problems faced by other regions of the country. some cases reverse, the base flow of water from Rainfall and runoff were estimated to increase in aquifers into these rivers. Water also could be the GISS scenario. Although most other drawn directly from rivers for irrigation in some assessments suggest that runoff would decline, the areas. magnitude of the decline does not appear to threaten the availability of water for municipal, A decline in riverflows could be important for industrial, or residential use. However, the both navigation and environmental quality. For the nonconsumptive uses for hydropower, navigation, Tennessee, as well as the Chattahoochee and other environmental quality, and recreation could be small rivers, adequate reservoir capacity exists to threatened. Although sufficient time exists to maintain flows for navigation, if this use continues develop rational strategies to implement the to take precedence over water supply and necessary tradeoffs, current federal statutes recreation. However, the 1988 drought has constrain the ability of water managers to do so. 354 Southeast graphically demonstrated that there are not enough development, long-term water supply sources, dams to guarantee navigation in the Mississippi. If powerplant construction, and other activities this situation became more commonplace, the sensitive to the availability of water would risk economic impact on New Orleans could be severe. basing their decisions on incorrect assumptions On the other hand, traffic on the Tennessee and regarding the future allocation of water. Ohio Rivers might use the Tennessee-Tombigbee Canal as an alternative, which would benefit the Estuaries Port of Mobile. Coastal plants and animals across the Lower flows also would reduce the dilution of Southeast may have difficulty surviving warmer municipal and industrial effluents discharged into temperatures. For example, along the northern rivers and would decrease the level of dissolved coast of the Gulf of Mexico, several types of fish oxygen. This would directly harm fish populations spend at least part of their lifetimes in estuaries that and would cause indirect harm by reducing the are already as hot as they can tolerate. If climate abilities of streams to assimilate wastes. Reduced became warmer, however, migrating north would flows also would threaten bottomland hardwood and not be feasible. While these species could escape estuarine ecosystems. To prevent these problems, the summer heat by fleeing to the cooler waters of factories and powerplants might have to erect the gulf, such a flight would make them vulnerable cooling towers or curtail their operations more to larger fish. frequently. In addition to the direct effect of climate Is Current Legislation Adequate? change on estuaries, human responses to climate change and sea level rise also could hurt coastal The same issues that face the TVA and Lake estuaries. Besides the impacts of flood control, Lanier would likely face decisionmakers in other increased reservoir construction would decrease the areas. Federal laws discourage water managers in amount of sediment flowing down the river and the Southeast from rigorously evaluating the nourishing the wetlands. If the climate becomes tradeoffs between the various uses of water. Most drier, irrigation could further reduce freshwater flow dams are more than sufficient to meet the statutory into estuaries. requirements for navigation and flood safety and to continue generating substantial hydropower on To a large extent, the policy implications for demand. Consequently, there has been little need wetland loss in the Southeast are similar to those to analyze the tradeoffs between these factors. For facing the rest of the U.S. coastal zone. Previous example, a literal application of the law would not studies have identified several measures to reduce allow the U.S. Army Corps of Engineers to cut the loss of coastal wetlands in response to sea level hydropower production or navigation releases to rise (e.g., Titus, 1988). These measures include the ensure a supply of water for Atlanta. Therefore, following: agencies have not analyzed the allocation of water that best serves the public for various levels of water increase the ability of wetlands to keep availability (although the TVA is beginning to do pace with sea level; so). remove impediments to landward creation At a practical level, federal water managers of new wetlands; and have shown flexibility, as in the case of cutting navigation along the Chattahoochee instead of dike the wetlands and artificially maintain further cutting Atlanta's water supply. If climate water levels. changes and more than a modest level of flexibility is necessary, water resource laws could be changed; All these measures are being employed or actively the physical infrastructure is largely in place to considered. address water problems of the Southeast. But until the laws are changed, the federal agencies in the Congress has authorized a number of Southeast often would be forced to allocate water freshwater and sediment diversion structures to inefficiently. Moreover, people making decisions assist the ability of Louisiana's wetlands to keep up concerning siting of recreational and industrial with relative sea level rise. These structures are 355 Chapter 16 engineered breaches in river levees that act as the implications for the mid-Atlantic and the spillways into the wetlands when water levels in the Northeast. If shore-protection measures are not river are high. Although decisions on where to taken, the majority of resorts will have no beach at build diversion structures are being based on high tide by 2025 under the midrange scenario of current climate and sea level, consideration of future sea level rise. The cost of undertaking the global warming would substantially change the necessary measures through 2025 probably would be assumptions on which current analyses are being economically justified for most resorts (see Chapter based and the relative merits of alternative options. 7: Sea Level Rise). However, the cost of protecting More frequent or higher surges in the Mississippi all recreational beaches through 2100 would be $100 River would increase the amount of water delivered to $150 billion, which would probably lead some of to the wetlands. And if climate change resulted in the more vulnerable areas to accept a landward more soil erosion, more sediment might also reach migration much as areas on North Carolina's Outer the wetlands; lower flows could have the opposite Banks are facing today, particularly if warmer effect. Sea level rise might shorten the useful temperatures also lead to more hurricanes. lifetimes of these projects, but because the flood-protection benefits of protecting coastal The potential responses to global warming wetlands would be greater with a higher sea level should be viewed within the context of current (Louisiana Wetland Protection Panel, 1987). responses to erosion flooding. Florida has a trust fund to nourish its beaches and has received federal Artificially managing water levels also has been assistance for pumping sand onto the shores of proposed for Louisiana, particularly by Terrebonne Miami Beach. Mississippi has nourished the Parish, whose eastern wetlands are far removed beaches of Biloxi, Gulfport, and other resort from a potential source of sediment. Such an communities that lie on the mainland along the approach also might be possible for parts of Florida, protected waters behind the barriers. Louisiana is where wetlands already are confined by a system of rebuilding its undeveloped barrier islands because dikes and canals, and water levels already are they protect the mainland from storms. Most states managed. Although no one has yet devised a are moving toward "soft engineering" solutions, such practical means by which shrimp and other fish as beach nourishment, because of doubts about the could migrate between ocean and estuary, other effectiveness of hard structures in universal erosion species spend their entire lifetimes within the and their interference with recreational uses of the estuary, and freshwater species could remain in beach. artificially maintained freshwater wetlands. Land-use measures also have been employed A final response would be to accept the loss to adapt to erosion. Because of unusually high of existing wetlands, but to take measures to prevent erosion rates on the Outer Banks, houses along the development from blocking the landward creation of coast are regularly moved landward. North new wetlands. This approach has been enacted by Carolina requires houses, hotels, and condominiums the State of Maine (1987) and would be consistent to be set back from the shore by the distance of a with the proposals to discourage bulkheads that 100-year storm plus 30 years' worth of erosion on have been widely discussed by coastal zone the assumption that after 30 years, the house could managers and enacted by the State of South be moved back. Texas requires that any house left Carolina. Titus and Greene estimate that 1,800 standing in front of the vegetation line after the square miles of wetlands in the Southeast could be shore erodes must be torn down. created if developed areas were not protected. Although this area represents a small fraction of the If a global warming increases the frequency of potential loss, it would increase the remaining areas hurricanes, a number of southeastern communities of wetlands by 30 to 90%, and it would maintain will be devastated. However, the overall impact of and perhaps increase the proportion of shorelines increased hurricane frequency would be small on which at least some wetlands could be found. compared with the impact of sea level rise. While a doubling of hurricanes would convert 100-year Beach Erosion floodplains to 50-year floodplains throughout much of the Southeast, a 1-meter rise would convert them The implications of sea level rise for to 15-year floodplains. recreational beaches in the Southeast are similar to 356 Southeast Because the open-coast areas most vulnerable Gibbs, M. 1984. Economic analysis of sea level to sea level rise are generally recreational beach rise: methods and results. In: Barth, M.C., and resorts, the costs of erosion and flooding should be J.G. Titus, eds. Greenhouse Effect and Sea Level viewed within the larger context of why people go to Rise: A Challenge for This Generation. New York: the beach. People from the north visit southeastern Van Nostrand Reinhold Company. beaches to escape winter, and residents of the region go to escape the summer heat. As Hansen, J., I. Fung, A. Lacis, D. Rind, G. Russell, temperatures become warmer, Georgia and the S. Lebedeff, R. Ruedy, and P. Stone. 1988. Global Carolinas may be able to compete with Florida for climate change as forecast by the GISS 3-D model. northerners. Hotter temperatures also may increase Journal of Geophysical Research 93(D8):9341-9364. the desire of the region's residents to visit the beach. Healy, R.G. 1985. Competition for Land in the American South. Washington, DC: The Thus, it is possible that the cooler communities Conservation Foundation. will reap benefits from a longer and stronger tourist season that are greater than the increased costs for Jones, C.A., and J.R. Kiniry. 1986. CERES-Maize: erosion control. Areas that already have a A Simulation Model of Maize Growth and year-round season are less likely to benefit, and in Development. College Station, TX: Texas A&M a few areas like Miami Beach, the off-season may Press. be extended. Kana, T.W., J. Michel, M.O. Hayes, and J.R. Jensen. 1984. The physical impact of sea level rise REFERENCES in the area of Charleston, South Carolina. In: Barth, M.C., and J.G. Titus, eds. Greenhouse Armentano, T.V., R.A. Park, and C.L. Cloonan. Effect and Sea Level Rise: A Challenge for This Generation. New York: Van Nostrand Reinhold 1988. Impacts on coastal wetlands throughout the United States. In: Titus, J.G., ed. Greenhouse Company. Effect, Sea Level Rise, and Coastal Wetlands. Washington, DC: U.S. Environmental Protection Leatherman, S.P. 1984. Coastal geomorphic Agency. responses to sea level rise: Galveston Bay, Texas. In: Barth, M.C., and J.G. Titus, eds. Greenhouse Barth, M.C., and J.G. Titus. 1984. Greenhouse Effect and Sea Level Rise: A Challenge for This Generation. New York: Van Nostrand Reinhold Effect and Sea Level Rise: A Challenge for This Generation. New York: Van Nostrand Reinhold Company. Company. Linder, K.P., M.J. Gibbs, and M.R. Inglis. 1988. Burnash Robert J.C., R.L. Ferral, and R.A. Potential Impacts of Climate Change on Electric Mcguire. 1973. A Generalized Streamflow Utilities. New York: New York State Energy Simulation System, Conceptual Modeling for Digital Research and Development Authority. Computers. Sacramento, CA: National Weather Service and California Department of Water Louisiana Wetland Protection Panel. 1987. Saving Resources. Louisiana's Coastal Wetlands: The Need for a Long-Term Plan of Action. Washington, DC: U.S. Edison Electric Institute. 1985. Statistical Environmental Protection Agency/Louisiana Yearbook of the Electric Utility Industry. Geological Survey. EPA-230-02-87-026. Washington, DC: Edison Electric Institute. Meo, M. 1987. Proceedings of the Symposium on Geraghty, J., D. Miller, F. Van Der Leeden, and F. Climate Change in the Southern United States. Troise. 1973. Water Atlas of the United States. Norman, OK: University of Oklahoma. Port Washington, NY: Water Information Center. Mitch, W., and J. Gosselink. 1986. Wetlands. New York: Van Nostrand Reinhold Company. 357 Chapter 16 NOAA. 1987. National Oceanic and Atmospheric Titus, J.G. 1984. Planning for sea level rise before Administration. Fisheries Statistics Division, and after a coastal disaster. In: Barth, M.C., and National Marine Fisheries Service. 1987 Preliminary J.G. Titus, eds. Greenhouse Effect and Sea Level Statistics for United States Domestic Catch. Rise: A Challenge for This Generation. New York: Washington, DC: Unpublished data. Van Nostrand Reinhold Company. Rind, D. 1988. The doubled CO2 climate and the U.S. Department of Agriculture. 1987. Agricultural sensitivity of the modelled hydrologic cycle. Journal Statistics: 1987. Washington, DC: U.S. Department of Geophysical Research. of Agriculture. Shugart, H.H., and D.C. West. 1977. Development U.S. Department of Commerce. 1988. U.S. of an Appalachian deciduous forest succession Department of Commerce, Bureau of the Census. model and its application to assessment of the Statistical Abstract of the United States: 1988. impact of the chestnut blight. Journal of Washington, DC: Government Printing Office. Environmental Management 5:161-179. U.S. Department of Commerce. 1986. U.S. State of Maine. 1987. Dune Rule 355. Augusta, Department of Commerce, Bureau of Economic ME: Maine Department of Environmental Analysis. Protection. U.S. Department of Commerce. 1982. U.S. Titus, J.G. 1987. The greenhouse effect, rising sea Department of Commerce, Bureau of the Census. level, and society's response. In: Devoy, R.J.N., ed. Census of Agriculture. Vol. 1. Geographic Area Sea Surface Studies. New York: Croom Helm. Series. Washington, DC: Government Printing Office. Titus, J.G. 1988. Greenhouse Effect, Sea Level Rise, and Coastal Wetlands. Washington, DC: U.S. U.S. Department of Energy. 1988. U.S. Environmental Protection Agency. Department of Energy, Energy Information Administration. Electric Power Monthly; May. Titus, J.G. 1988. Sea level rise and wetland loss: an overview. In: Titus, J.G., ed. Greenhouse U.S. House of Representatives. Rep. Jenkins, Rep. Effect, Sea Level Rise, and Coastal Wetlands. Barnard, Rep. Darden. HR-4254. Georgia Washington, DC: U.S. Environmental Protection Reservoir Management Improvement Act of 1988. Agency. 100th Congress, 20th Session. Titus, J.G., C.Y. Kuo, M.J. Gibbs, T.B. LaRoche, Wilkerson, G.G., J.W. Jones, K.J. Boote, and J.W. M.K. Welts, and J.O. Waddell. 1987. Greenhouse Mishoe. 1985. SOYGRO V5.0: Soybean Crop effect, sea level rise, and coastal drainage systems. Growth and Yield Model. Technical Journal of Water Resource Planning and Documentation. Gainesville, FL: University of Management ASCE 113(2):216-227. Florida. 358 CHAPTER 17 GREAT PLAINS FINDINGS with intense groundwater use -- water depletion,so damage, altered farm and rural economics, and potential reversion to dryland Agriculture in the Great Plains (this study focused farming -- could be exacerbated by global on Nebraska, Kansas, Oklahoma, and Texas) is sensitive to climate fluctuations and would be at risk warming. from global warming. Although uncertainties remain regarding the rate and magnitude of global Water Quality climate change and the models used to estimate impacts, results indicate that climate change would It is not clear how climate change would affect cause reductions in regional agricultural production. water quality in the Great Plains. Groundwater Demand for irrigation is likely to increase, and quality may be less at risk than surface water quality of water may diminish. Regional electricity quality because of increased evaporation and use may increase. less leaching. These results are very sensitive to changes in the amounts and frequency of Agriculture rainfall, and groundwater impacts will be affected by total acres under production, by The effects of a warmer climate alone would application rates, by soil type under cultivation, and by changes in irrigated versus dryland generally reduce wheat and corn yields. Yield acres. changes range from + 15 to -90%. The direct effects of CO₂ on crop photosynthesis and water use may mitigate these effects, but the extent to Electricity Demand which the beneficial effects of CO2 on crop yields would be seen with climate change is Climate warming could cause the annual uncertain. demand for electricity in Kansas, Nebraska, Oklahoma, and West Texas to rise by an Crop yields in Texas and Oklahoma may decline additional 5 to 9 billion kilowatthours (kWh) (2 relative to northern areas of the United States. to 4%) by 2010, and by an additional 37 to 73 This change in productivity could lead to a 4 to billion kWh (10 to 14%) by 2055. Summertime 22% reduction of cultivated acreage in these use for air-conditioning and irrigation pumping states. could increase and outpace reductions in winter demand for space heating. Because of increased reliability of yields from irrigated lands relative to dryland yields, and Approximately 3 to 6 gigawatts (GW) of because of potentially higher crop prices, generating capacity would be needed by 2010 to demand for irrigation water on remaining farms meet the additional increased demand, and 22 would probably increase as global warming to 45 GW would be needed by 2055 -- a 27 to proceeds. The number of acres irrigated may 39% increase over baseline additions that may increase by 5 to 30%. be needed without climate change. The cumulative cost of these additions by 2055 Ogallala Aquifer would be $24 to $60 billion. Warming and/or drying in the Great Plains may Policy Implications place greater demand on regional groundwater resources. Many of the problems associated Agencies with responsibility for agricultural land use, such as the U.S. Department of 359 Chapter 17 Agriculture (USDA) Agricultural Stabilization and Conservation Service and the Soil Conservation Service, should begin to analyze how their missions may be affected by climate change and to consider development of flexible strategies to deal with Nebraska potential impacts. Water resource managers, such as those on river basin commissions and in state natural resource agencies, may wish to factor the OGALLALA potential effects of climate change into planning of AQUIFER Kansas land use, long-term water supply, irrigation, drainage, and water-transfer systems. Oklahoma CLIMATE-SENSITIVE RESOURCES IN THE GREAT PLAINS The Great Plains consists of a predominantly treeless region of relatively flat topography between the Rocky Mountains and the Mississippi lowlands Texas of central North America. Although very productive, the region (Figure 17-1) is sensitive to climate fluctuations, a fact that has been made apparent in several major droughts over the last few decades. Dryland Farming Area Despite this climate sensitivity, dryland agriculture provides the chief economic base for this thinly populated region with few cities. The region Figure 17-1. Boundaries of the Ogallala Aquifer was first settled by farmers in the late 1800s under and dryland wheat production in the Great Plains the Homestead Act, which created the family-farm (Science of Food and Agriculture, 1987, 1988). system in place today in the Plains (Bowden et al., 1981). The Great Plains, including portions of rice, and cotton has replaced dryland wheat Nebraska, Kansas, Oklahoma, and Texas, constitutes production, especially in western Kansas and the a vital part of the United States' agricultural base Texas Panhandle (Figure 17-1). However, the and is the focus of this report. Nearly 100,000 region's groundwater resources have been farms encompassing over 111 million acres produce overexploited in some areas, leading to some an important array of dryland and irrigated crops. reversion to dryland cropping. Major dryland crops include winter wheat and grain sorghum, and key irrigated grains include corn and Livestock constitute another important rice. In all, the four states have a combined agricultural commodity in the region. Almost 50% production of over 80, 30, and 25% of the nation's of all cattle fattened in the country are raised in the grain sorghum, wheat, and cotton, respectively four states, accounting for 40% of the total U.S. (Table 17-1). value of marketed livestock. Exploitation of water from the Ogallala Aquifer In addition to contributing substantially to has supported significant irrigated agricultural national food supplies, the four states are also major production in the Great Plains during the last two exporters of agricultural products. Foreign exports decades. In many areas, irrigated farming of corn, of grain and animal products are especially notable 360 Great Plains Table 17-1. U.S. Agricultural Ranking of Great Plains States and Percent of U.S. Total (for the four states combined) for Selected Products, 1982 U.S. total (all four Product Kansas Nebraska Oklahoma Texas states) (%) Sorghum harvested 2 3 5 1 80.5 Cattle fattened on 2 3 9 1 46.7 grain and concen- trates sold Value of cattle 2 3 7 1 40.7 and calves sold Wheat harvested 1 9 3 6 31.8 Cotton harvested -- -- 9 2 25.8 Hay harvested 9 2 16 7 15.9 Market value of 6 5 20 3 18.5 all agricultural products Source: USDA (1983). Table 17-2. Agricultural Exports from Selected Great Plains States, Fiscal Year 1984 (millions of dollars) U.S. total Exports U.S. Kansas Nebraska Oklahoma Texas (%) Feed grains and 7,585 372 903 - 385 22 byproducts Wheat and 4,526 797 150 353 276 35 byproducts Live animal and 1,161 130 134 18 161 38 meats All agricultural 31,187 1,719 1,762 1,471 2,031 19 products Source: USDA (1985). 361 Chapter 17 (Table 17-2). In total, these four states provide (Warrick and Bowden, 1981; Riebsame, 1983). approximately one-fifth of the dollar value of all These practices are designed to conserve moisture, U.S. agricultural exports. Yet, dependence on reduce energy input, and minimize erosion, and foreign markets puts Great Plains farmers at high thus, to increase yields and profits. Nevertheless, risk. While large historical fluctuations in grain and dryland crop yields still fluctuate widely with livestock production levels are partly related to temperature and precipitation variations between climatic variability, changing international demand, years. The coefficient of variation of wheat yields is and its effects on price, play an important role in close to 50% over much of the region, and the region's continuing economic and social approximately 30-40% of the planted acreage is instability. abandoned every year because of poor crops, especially on the western fringes of agriculture The Great Plains is also a major source of coal where the dominant crop is dryland wheat grown on and oil, though such extractive industries vary more summer fallow (Michaels, 1985). with international energy markets than with climate. Otherwise, the area exhibits little economic In addition to the developments in cropping diversity, a pattern that has led to a net systems, government policies and programs have outmigration, especially of younger segments of the also been devised to absorb or mitigate the impacts population. Regional population is growing slowly of climate stresses in the Great Plains and mostly in the fringe cities (e.g., Omaha), while rural elsewhere. These include federal programs for crop population and the total number of farms are slowly insurance, disaster grants and low-interest loans to decreasing. The region's economy remains farmers, and government-sponsored drought inexorably linked to the fortunes of agriculture and, research (Warrick, 1975). Such programs can be thus, to the climate. costly. For example, the projected cost of the 1988 Drought Relief is about $3.9 billion nationally Dryland Agriculture (Schneider, 1988). The dryland farming area of the Great Plains Despite the adoption of conservation tillage is one of the most marginally productive agricultural techniques, drought-resistant cultivars, and risk regions in the United States. Some observers have management programs, some analysts argue that stated that the southern Plains are simply too the region remains particularly vulnerable to sensitive to climate swings and that intensive dryland climate-induced reductions in crop yields and will be farming should be abandoned (Worster, 1979; one of the first U.S. agricultural regions to exhibit Popper and Popper, 1987). Yet in many years, the impacts of climate change (e.g., Lockeretz, 1978; Plains produce bumper crops of small grains that Warrick, 1984). Rapid acreage increases in the add significantly to the nation's export trade 1970s, destruction of windbreaks for larger fields to balance. accommodate bigger machinery, and speculative farm expansion all raise the possibility of renewed Dryland farmers in the Great Plains are land degradation and economic losses similar to particularly vulnerable to climate variability. The those of the Dust Bowl period, if climate change Great Plains States of Nebraska, Kansas, Oklahoma, creates an increased frequency of heat waves and and Texas were the hardest hit during the Dust droughts in the region. Most climate models Bowl of the 1930s (Worster, 1979; Hurt, 1981). indicate that the region would become drier as Yields of wheat and corn dropped as much as 50% global warming proceeds, suggesting potentially below normal, causing the failure of about 200,000 severe impacts on dryland farming. farms and migration of more than 300,000 people from the region. Irrigated Agriculture The Dust Bowl, other droughts, and the desire One response to the semiarid and highly for continued expansion and intensification of variable climate of the Great Plains has been dryland farming have led to numerous technological exploitation of surface and groundwater resources and social adjustments to climate and market for irrigation to replace dryland farming. In 1982, fluctuations. Especially critical, from a dryland 19 million acres, or 12% of all Great Plains farming perspective, has been the improvement of cropland, mostly in the southern Plains, were conservation tillage practices like summer fallowing irrigated. Groundwater provides most of the water 362 Great Plains for irrigation: 61 to 86% of the water used in cooling, etc.). Other types of energy are also Nebraska, Oklahoma, and Kansas as compared with sensitive to climate, but this study addresses only only 20% nationally. In this respect, irrigation electricity. farmers in the Great Plains are less sensitive to climate change relative to dryland farmers. However, the demand for irrigation water PREVIOUS CLIMATE IMPACT throughout the region is very sensitive to climate. STUDIES The improvement and application of well drilling and pumping technology after World War II Many studies of climate impacts on agriculture permitted the use of water from the immense in the Great Plains have been performed using a Ogallala Aquifer (Figure 17-1). Today, the aquifer variety of approaches and models. Dozens of supplies irrigation for approximately 14 million acres climate impact studies have focused specifically on in the Great Plains States of Colorado, Nebraska, the 1930s drought (e.g., Lockeretz, 1978; Bowden Kansas, Oklahoma, New Mexico, and Texas (High et al., 1981) and, more generally, on Great Plains Plains Associates, 1982). Use of the aquifer allows droughts (Warrick, 1975). Many recent studies have the irrigation of terrain too far from surface used crop-climate models to estimate impacts of supplies. The aquifer also provides water for climate on yields. Warrick (1984) analyzed the municipal and industrial purposes. vulnerability of the region to a possible recurrence of the 1930s drought by running a dryland crop yield Farmers in Nebraska recently began to use the model tuned to 1975 technology with 1934 and 1936 aquifer to irrigate corn, which is grown mostly for temperature and precipitation conditions. He found livestock feed. Corn, wheat, and some sugarbeets that recurrence of 1930s conditions in the region are irrigated farther south, while in Texas the would result in wheat yield reductions of over 50%. Ogallala is tapped chiefly for cotton. The aquifer Terjung et al. (1984) used a crop water demand and varies in depth from the land surface, in rate of yield model to investigate irrigated corn production natural discharge, and in saturated thickness across sensitivity to differing temperature, precipitation, the region. In Nebraska, the aquifer has a higher and solar radiation fluctuations. They found that in recharge rate (i.e., the rate at which the aquifer is the central Great Plains, evapotranspiration and replenished) than in the other Great Plains States, total water applied for irrigation were very sensitive and significant drawdown problems have not yet to climate variations. Liverman et al. (1986) occurred. In Texas and other states, high continued this modeling and found that the lowest withdrawal and low recharge rates of the aquifer irrigated yields occurred under cloudy, hot, and very have already resulted in "mining" of the resource dry climate scenarios. Under dryland cropping, (i.e., the rate of water withdrawal is greater than minimum yields occurred under sunny-hot and rate of recharge) and in the abandonment of sunny-warm scenarios with very dry conditions. thousands of irrigated acres (see Glantz et al., Volume J). Using an agroclimatic approach, Rosenzweig (1985) found that lack of cold winter temperatures Water Quality in the southern Great Plains may necessitate a change from winter to spring wheat cultivars with Nonpoint pollution (runoff and leaching) is the climate change projected for a doubling of CO2. main contributor to water quality problems in the Changes in temperature, precipitation, and solar Great Plains. Many of the groundwater supplies in radiation were considered. Decreased water the region contain elevated levels of fertilizer and availability may also increase demand for irrigation. pesticide-derived pollutants. In a later study, Rosenzweig (1987) showed that although the combined impact of doubled CO, Electricity Demand climate change (temperature, precipitation, and solar radiation changes) and the direct effects of Electricity use in the region is sensitive to elevated CO2 (increased photosynthesis and improved water use) compensated for the negative climate fluctuations in terms of space heating, effects of climate change in years with adequate cooling, and agricultural operations such as rainfall, this compensation did not reduce crop irrigation and livestock management (heating, failures in dry years. 363 Chapter 17 Robertson et al. (1987) estimated the combined Table 17-3. Great Plains Studies for EPA impact of temperature and precipitation changes Report to Congress on the Effects due to doubled CO₂ climate change and the direct of Global Climate Change effects of increased CO2 on rainfed corn and wheat yields and erosion using the Erosion Productivity Analyses Performed for This Case Study Impact Calculator (EPIC). Results showed that modeled wheat yields in Texas decreased and Potential Effects of Climate Change on modeled corn yields increased slightly. Such Agricultural Production in the Great Plains: A changes in productivity could result in long-term Simulation Study - Rosenzweig, Columbia changes in cropping patterns. University, NASA/Goddard Institute for Space Studies (Volume C) Glantz and Ausubel (1984) suggested that the Great Plains' mining of the Ogallala Aquifer and its Effects of Projected CO2-Induced Climatic susceptibility to future incidence of drought Changes on Irrigation Water Requirements in projected by global climate models be combined in the Great Plains States - Allen and Gichuki, analyses of the region, since both are critical to the Utah State University (Volume C) habitability of the area. National Studies That Included Great Plains Results GREAT PLAINS STUDIES IN THIS REPORT Economic Effects of Climate Change on U.S. Agriculture: A Preliminary Assessment - Adams, Oregon State University and Glyer and The studies for this report examine the McCarl, Texas A&M University (Volume C) implications of climate change for several important activities in the region: agricultural production and Impacts of Climate Change on the Movement economics, demand for irrigation water, and water of Agricultural Chemicals Across the U.S. quality. Climate change impact research on Great Plains and Central Prairie -Johnson, livestock, electricity use, and resource management Cooter, and Sladewski, Oklahoma policy relevant to the Great Plains is also described. Climatological Survey, University of Oklahoma The individual studies performed for this report are (Volume C) listed in Table 17-3. Changing Animal Disease Patterns Induced by The Great Plains studies explore the sensitivities the Greenhouse Effect - Stem, Mertz, Stryker, of regional activities to climate change scenarios. and Huppi, Tufts University (Volume C) The results are not meant to be predictions of what will happen; rather the studies aim to define the Effect of Climatic Warming on Populations of ranges and magnitudes of potential responses of the Horn Fly, with Associated Impact on critical regional systems to the predicted climate Weight Gain and Milk Production in Cattle - changes. Schmidtmann and Miller, U.S. Department of Agriculture, Agricultural Research Service (Volume C) GREAT PLAINS REGIONAL CLIMATE CHANGE SCENARIOS The Potential Impacts of Climate Change on Electric Utilities: Regional and National Estimates - Linder and Inglis, ICF Incorporated The estimated changes in seasonal and annual (Volume H) temperatures and precipitation for the scenarios are shown in Figure 17-2. For a description of the Climate Change and Natural Resources global climate models, climate scenarios, and a Management in the United States -Riebsame, discussion of the likelihood of these changes, see University of Colorado (Volume J) Chapter 2: Climate Change, and Chapter 4: Methodology. All three scenarios show large 364 Great Plains Average annual precipitation decreases by 0.26 millimeters per day (3.7 inches per year) in the GISS scenario, while GFDL and OSU have slight A. Temperature increases. However, these annual values mask a 6 pronounced reduction in rainfall in Nebraska and GISS Kansas in the GFDL scenario (see Figure 17-3). 5 The large temperature increase and pronounced GFDL summer drying combine to make the GFDL 4 OSU scenario severe in these states, and the most severe CHANGE (°C) case among the climate change scenarios. 3 The magnitudes of climate changes in the 2 spring and summer from the GFDL scenario and the climate of the 1930s drought in Nebraska and 1 Kansas are compared in Figure 17-3. While the scenario decreases in growing season precipitation 0 are about the same as those during the most severe Winter Spring Summer Fall Annual drought years (1934 and 1936) in the area, the B. Precipitation climate change scenario temperatures are about 0.5 3°C higher than the Dust Bowl temperatures. 0.4 0.3 0.2 A. Temperature CHANGE (mm/Day) 0.1 8 0 7 -0.1 6 -0.2 -0.3 Change (°C) 5 -0.4 4 -0.5 3 -0.6 Winter Spring Summer Fall Annual 2 1 0 Figure 17-2. Average change in (A) temperature, 1934 1936 GFDL and (B) precipitation over Great Plains gridpoints in Spring GISS, GFDL, and OSU global climate models B. Precipitation / Summer (2XCO₂ run less 1XCO₂ run). 0.4 0.2 0.0 increases in temperature for the Great Plains States -0.2 under a doubled CO₂ climate. The GISS scenario has an annual warming of 4.5°C, the GFDL scenario Change (mm/day) -0.4 has an annual warming of 5.0°C, and OSU has an -0.6 annual warming of 3.3°C. In general, winter -0.8 temperatures increase more than summer -1.0 temperatures in the GISS model, and summer -1.2 temperature changes are greater than winter -1.4 temperature changes in the GFDL and OSU 1934 1936 GFDL scenarios. The differences between the models Figure 17-3. Comparison of observed drought (1943 range from 0.2 to 1.5°C. The impact studies used and 1936) and GFDL climate change in Nebraska only the GISS and GFDL climate change scenarios and Kansas for (A) temperature, and (B) because of time limitations. precipitation (Rosenzweig, Volume C). 365 Chapter 17 RESULTS OF THE GREAT Limitations PLAINS STUDIES This work does not consider changes in frequencies of extreme events, even though Crop Production extremes of climatic variables, particularly runs of extremes, are critical to crop productivity (see To better understand the potential physical Chapter 3: Variability). Development of the impact of climate change on crops, Rosenzweig CERES models was based on current climate; the modeled changes in corn and wheat yields in the relationships in the models may or may not hold Great Plains using crop growth models. under differing climate conditions, particularly the high temperatures predicted for greenhouse Study Design warming. Two crop growth models, CERES-Wheat The direct effects of CO2 are only (Ritchie and Otter, 1985) and CERES-Maize (Jones approximated in the crop modeling study, because and Kiniry, 1986) were used to test the sensitivity of the models do not include a detailed simulation of crop yields to the GISS and GFDL climate change photosynthesis. Also, experimental results from scenarios. These models are designed for large-area controlled environments may show more positive yield prediction and for farm decisionmaking and effects of CO2 than would actually occur in variable, have been validated for a wide range of conditions windy, and pest-infested (e.g., weeds, insects, and (Otter-Nacke et al., 1986). The CERES models diseases) field conditions; thus, this study probably simulate crop responses to the major factors that overestimated the beneficial effects of increased affect crop yields: climate, soils, and management. CO2. The models employ simplified functions to predict Results crop growth stages; development of vegetative and reproductive structures; growth of leaves and stems; dieback of leaves; biomass production and use; root Climate change scenarios cause simulated system dynamics; and the effects of soil-water deficit wheat (Figure 17-4) and corn (Figure 17-5) yields to decrease in the southern and central Great Plains. on photosynthesis and biomass use in the plant. Results shown are means of modeled yields at study At each of 14 locations, the crop models were sites grouped by latitude for 30 years of baseline run with three soils present in the region and climate change scenarios. With climate change representing low, medium, and high productive alone, decreases in modeled yields appear to be capacity. Model results were generated for changes caused primarily by increases in temperature, which in yield, water used for irrigation (if crop is would shorten the duration of crop life cycle (the irrigated), crop evapotranspiration, and planting and period during which a crop grows to maturity). This maturity dates for both dryland and irrigated cases. results in reduced yields. When the direct effects The direct effects of CO2 (i.e., increased of CO₂ on crop photosynthesis and transpiration are photosynthesis and decreased transpiration per unit- included in the climate change simulations, modeled leaf area) were simulated with the climate change crop yields overcome the negative effects of climate scenarios in another set of runs. A method for change in some cases, but not in others. In general, approximating the direct effects in the CERES the more severe the climate change scenario, the models was developed by computing ratios of daily less compensation provided by direct effects of CO2. photosynthesis and evapotranspiration rates for a canopy exposed to elevated (660 ppm) CO₂ to those Corn and wheat yields were estimated to rates for the same canopy exposed to current (330 respond differently to dryland and irrigated climate ppm) CO2 conditions (see Peart et al., Volume C). change conditions and to the direct effects of CO2. Daily photosynthesis rates of wheat and corn Dryland corn yield decreases were very high in the canopies were increased 25 and 10%, respectively, hotter and drier GFDL scenario, particularly at based on published results of controlled higher latitudes. These decreases were caused by environmental experiments with crops growing in air the combined effects of high temperatures with increased CO2 levels. 366 Great Plains (A) DRYLAND 5 4 YIELD (thousands kg/Ha) 3 2 1 0 40-42 N 38-40 N 36-38 N 34-36 N <34N LATITUDE (B) IRRIGATED 8 6 YIELD (thousands kg/Ha) 4 2 0 40-42 N 38-40 N 36-38 N 34-36 N < 34 N LATITUDE BASE GISS GISS DE GFDL GFDL DE DE = Direct Effects of CO2 Figure 17-4. CERES-Wheat yields in the Great Plains with GISS and GFDL climate change scenarios with and without the direct effects of CO2: (A) dryland, (B) irrigated (Rosenzweig, Volume C). shortening the grain-filling period and increased yields. These results suggest an increased demand moisture stress. The GFDL scenario has for irrigation in the region. pronounced reductions in summer precipitation (decreases of about 30 mm per month) in the two Adjusting the planting date of wheat to later in northern gridboxes of the study area, which occur the fall, one potential farmer adjustment to a during critical growth stages of corn. Irrigated corn warmer climate, was not estimated to significantly was more negatively affected than irrigated wheat in ameliorate the effects of the GISS climate change the combined climate and direct effects runs scenario on CERES-Wheat yields. Changing to because of the lower photosynthetic response of varieties with lower vernalization requirements corn to CO2. (need for a period of cold weather for reproduction) and lower photoperiod sensitivity (sensitivity to In general, the amount of water needed for daylength), in addition to delaying planting dates, irrigation in the crop models is estimated to overcomes yield decreases at some sites but not at increase in the areas where precipitation decreases others. and irrigation reduces interannual variability in 367 Chapter 17 (A) DRYLAND 8 6 YIELD (thousands kg/Ha) 4 2 0 40-42 N 38-40 N 36-38 N 34-36 N < 34 N LATITUDE (B) IRRIGATED 14 12 10 YIELD (thousands kg/Ha) 8 6 4 2 0 40-42 N 38-40 N 36-38 N 34-36 N < 34 N LATITUDE BASE GISS GISS DE GFDL GFDL DE DE = Direct Effects of CO2 Figure 17-5. CERES-Maize yields in the Great Plains with GISS and GFDL climate change scenarios with and without the direct effects of CO2: (A) dryland, (B) irrigated (Rosenzweig, Volume C). Implications acreage would be irrigated as high temperatures increase the risk of crop failures. Increased There is potential for climate change to cause irrigation would be needed to ensure acceptable and decreased crop yields in the southern Great Plains. stable yield levels. However, availability of and Farmers would need varieties of corn and wheat competition for water supplies also may change with that are better acclimated to hotter and possibly climate change, and defining the extent to which drier conditions to substitute for present varieties, irrigation can provide an economic buffer against and adjustment strategies tailored for each crop and climate change requires further study. location. Agricultural Economics Pressure for increased irrigation may grow in the region, particularly with more severe climate Many economic consequences are likely to changes. This would occur for two reasons: first, result from the physical changes in crop yields and crops currently irrigated would require more water water availability caused by climate change. where precipitation decreases; and second, more Decreased yields will further stress farmers already 368 Great Plains affected by marginal productivity and economic Results fluctuations. Additional irrigation needs could place greater demand on the Ogallala Aquifer and other The estimates of Adams et al. (see Volume C) water resources in the region. To examine the for total agricultural and irrigated acreage changes agricultural implications of climate change more in the southern Great Plains States (Oklahoma and closely, Adams et al. introduced yield changes from Texas only) are shown in Table 17-4. Agricultural the Great Plains and other regional crop modeling land is estimated to decrease in the southern Great studies, and changes in crop water use and water Plains in all scenarios, with and without the direct availability from the GISS and GFDL scenarios into effects of CO2. Decreases range from 4 to 22%. an economic model to translate the physical effects Irrigated acreage, on the other hand, increases in all of climate change into economic consequences. (For scenarios, from 9 to 30%. This is because of study design and limitations, see Chapter 6: increased stability of irrigated yields relative to Agriculture.) Analyses were done both for climate dryland yields, and because of a rise in commodity change alone and for the combined effects of prices that makes expansion of irrigation production climate change and enhanced CO₂ concentrations to economically feasible. explore the sensitivity of the agricultural system to the projected changes. The economic study did not Implications address the issues of whether the physical and institutional changes required to accommodate The results of the agricultural economics study increased demand for irrigated acreage are feasible imply that wheat and corn production may shift or whether new crops would be introduced. The away from the southern Great Plains. This may study did not consider changes in global agriculture. weaken the economic base of many rural Table 17-4. Estimated Changes in Agricultural Land Usage in Oklahoma and Texas (millions of acres) Base GISS GFDL Usage acreage Acreage Change % Change Acreage Change % Change Agricultural land Without direct effects 54.7 42.6 -12.1 -22.1 52.0 -2.7 -4.9 With direct effects 54.7 48.8 -10.9 -19.9 52.7 -2.0 -3.8 Irrigated acreage Without direct effects 5.3 6.9 1.6 29.6 5.6 0.3 4.9 With direct effects 5.3 5.8 0.5 9.4 6.1 0.8 15.3 Source: Adams et al. (Volume C). 369 Chapter 17 communities in the region and cause dislocations of similarly to increased CO2 (which may reduce rural populations. Uncertainties exist about transpiration), although published reports of adaptation in the region, such as substitution of experimental results show different responses more heat- and drought-tolerant varieties and crops. among crops (see Rose, Volume C). The majority If irrigated acreage expands as predicted in the of results presented in this study assumed that crop economic analysis, changes in capital requirements varieties would not change, even though farmers for agriculture would also occur. may shift to crops more adapted to the changed climate. If irrigated acreage does increase in the area, groundwater overdrafts also would be likely, along Results with associated increases in surface and groundwater pollution and other forms of In general, modeled results showed that environmental degradation. The current analysis seasonal irrigation requirements for an area growing did not address the issue of whether the physical alfalfa, corn, and winter wheat in the Great Plains and institutional changes required to accommodate would increase by about 15% under the doubled such an increase in irrigated acreage are feasible. CO2 scenario. These results are based on averages of the two GCM doubled CO₂ scenarios and the Irrigation likely occurrence of only moderate CO2-induced decreases in transpiration. Higher air temperatures cause increased evaporative demands, which largely govern crop Irrigation requirements were estimated to vary water use and irrigation water requirements. The depending on the type of crop, changes in climatic climate and crop production changes that might be factors, and variations in response to CO₂. The associated with global warming in the southern perennial crop alfalfa showed persistent increases in Great Plains are likely to heighten farmer interest seasonal net irrigation water requirements (see in irrigation, both because evapotranspiration may Figure 17-6). These increases are driven primarily increase and because irrigated crops might obtain a by higher temperatures, with less influence from larger economic advantage in a less favorable stronger winds, greater solar radiation, and a longer climate. Therefore, climate change impacts on growing season. irrigation water requirements were analyzed in more detail. Study Design 120 Allen and Gichuki (see Volume C) evaluated 100 the effects of climate change and reduced transpiration due to enhanced CO₂ on crop irrigation water requirements in the Great Plains. Percent Change From Baseline Value 80 They used an irrigation water requirement model to 60 calculate daily soil moisture balances, 40 evapotranspiration, and irrigation water requirements for corn, wheat, and alfalfa. The 20 model employed the Penman-Monteith combination 0 method to estimate crop evapotranspiration Nebraska Kansas Oklahoma Texas (Monteith, 1965). Four levels of potential direct GISS effects of CO2 on transpiration were simulated. GFDL Limitations Some uncertainty is embedded in the evapotranspiration and irrigation water requirement Figure 17-6. Seasonal irrigation water requirement estimates owing to mismatching of weather profiles for alfalfa for GISS and GFDL climate change and crop characteristics. Also, this study assumed scenarios and a moderate CO2-induced decrease in that alfalfa, corn, and wheat all would respond transpiration (Allen and Gichuki, Volume C). 370 Great Plains On the other hand, decreases in seasonal net 100 irrigation requirements were estimated for the (A) ALFALFA region's two most important crops, winter wheat 80 and corn, in most areas, depending on the projected 60 direct effects of CO2 on transpiration. These water need decreases would be generally due to shorter 40 crop growing periods caused by higher temperatures, which accelerate crop maturity. 20 When crop varieties appropriate to the longer 0 growing season were modeled, irrigation requirements for winter wheat were estimated to 100 increase. Water requirements during peak irrigation (B) CORN 80 periods (when plant growth and temperatures are greatest) increased in almost all cases (Figure 17- PERCENT CHANGE FROM BASELINE VALUE 60 7). These results are consistent with results from 40 the crop modeling study. 20 Plant canopy (leaf) temperatures were estimated to increase above current baseline values 0 for all crops and sites studied. Increases in leaf 50 temperatures may reduce photosynthetic activity and (C) WINTER WHEAT crop yields. They also would make crops more 40 sensitive to moisture stress. (See discussion on 30 direct effects of CO2 in Chapter 6: Agriculture.) 20 Implications 10 0 Any reduction in irrigation requirements for corn and winter wheat would be beneficial in the -10 Great Plains because less water and energy would -20 GISS GFDL GISS GFDL GISS GFDL GISS GFDL be required to produce the crops. However, the NEBRASKA KANSAS OKLAHOMA TEXAS shortened crop growth periods might allow for 0% double-cropping (planting two crops in one season), 20% / 40% thus increasing total irrigation requirements. 60% Farmers may shift to longer-season varieties, which 80% would also increase water needs. Figure 17-7. Percent change in net peak monthly irrigation requirement from baseline values for Expanded farm irrigation systems will require alfalfa, corn, and winter wheat for GISS and GFDL increased capital investments and larger peak drafts climate change scenarios and five levels of CO2- on groundwater systems and on energy supplies. induced decreases in transpiration (Allen and Increased groundwater extraction could pose Gichuki, Volume C). environmental and economic problems, especially where "water mining" is currently a major problem. Any action of irrigators to increase irrigation Water Quality efficiency as an attempt to cope with projected water shortages, while economically beneficial, may Agricultural pesticides are a high-priority lead to increased salinity problems if sufficient water pollution problem in at least half of the states within is not applied to meet soil leaching requirements. the U.S. Great Plains and Central Prairie. 371 Chapter 17 Potentially toxic agricultural chemicals can be removed from farmers' fields through degradation, A. PESTICIDE RUNOFF LOSSES surface runoff, sediment transport, and downward 200 percolation. An understanding of potential climate BASE change effects on the movements of agricultural 150 GISS chemicals is needed to identify potential changes in drinking water quality. PERCENTAGE OF BASE GFDL 100 Study Design 50 Johnson et al. used the Pesticide Root Zone 0 Model (PRZM) (Carsel et al., 1984) to simulate the WINTER WHEAT COTTON CROP REGION partitioning of pesticides between plant uptake, chemical degradation, surface runoff, surface B. PESTICIDE EROSION LOSSES erosion, and soil leaching in the Great Plains under baseline climate and climate change scenarios. The 160 locations modeled were representative of cropping 140 practices for winter wheat and cotton in the region. 120 The interactions among soil, tillage, management were studied. (For further discussion of the study's PERCENTAGE OF BASE 100 80 systems, pesticide transport, and climate change 60 40 design and limitations, see Chapter 6: Agriculture.) 20 0 Results WINTER WHEAT COTTON CROP REGION As Figure 17-8 shows, surface runoff and surface erosion of agricultural pesticides increased C. PESTICIDE LEACHING under the GISS scenario for the winter wheat 100 regions of the Great Plains. In the southern Great Plains cotton simulations, both the GISS and GFDL 80 scenarios produced increases in surface pesticide losses with runoff and eroded soils. The quantity of pesticides leached below the PERCENTAGE OF BASE 60 40 crop root zone is estimated to decrease everywhere 20 except on silty soils in the cotton region. This 0 overall decline most likely results from higher WINTER WHEAT COTTON evaporative demands in response to temperature CROP REGION increases and to less available moisture for Figure 17-8. Regional summary of surface and infiltration and deep percolation. subsurface pesticide loss as a percentage of the base climate scenario losses (Johnson et al., Volume C). Implications Results of the modeling imply that water quality groundwater impacts will depend on total acres in the southern Great Plains may be affected by under production, application rates, soil type under climate change. However, because these results are cultivation, and changes in irrigated versus dryland highly dependent on the frequency and intensity of acres. precipitation events, directions of change are uncertain. Surface water appears to be vulnerable From a water quality perspective, decreased to deterioration under climate change conditions, pesticide leaching may be advantageous. From a although the result does not hold for all cases. water quantity perspective, these results could be Groundwater quality in some areas appears to be cause for concern. Less leaching can imply less less at risk than surface water quality. However, water movement through soil profiles and less water 372 Great Plains availability for aquifer recharge. If water demands warmer environments permit longer seasonality of were to increase (as suggested by the crop diseases currently present. Stem et al. calculated production, economic, and irrigation analyses) at the that the ranges of bluetongue and Rift Valley fever same time that recharge rate decreased, competition (both serious or potentially serious diseases of for scarce water resources could increase cattle) could be extended northward from Texas to dramatically in the region. Kansas and Nebraska with climate warming. Climate change thus has the potential to cause Livestock increased incidence of animal disease and to increase stress on livestock production in the Great Livestock production is a critical agricultural Plains. activity in the Great Plains and may be sensitive to climate fluctuations in several ways. The warming Electricity Demand in the climate change scenarios may alleviate cold stress conditions in the winter but would exacerbate Linder and Inglis (see Volume H) estimated heat stress in the summer. Warmer summers are the changes in demand for electricity for the years likely to necessitate more hours of indoor cooling. 2010 and 2055. (For a description of the study's Reproductive capabilities have been shown to design and methodology, see Chapter 10: Electricity decline as a result of higher temperatures. Higher Demand.) In each case, they first estimated the temperatures also may enable tropical diseases and change in electricity demand due to projected pests to extend their ranges northward into the regional economic and population growth, and then southern Great Plains. High temperatures also may factored in changes in demand based on the GISS reduce insect pest activities in some locations and transient climate change scenarios A and B. The increase them in others. (For a discussion of results for the southern and central Great Plains are livestock issues, see Chapter 6: Agriculture.) discussed here. Schmidtmann and Miller (see Volume C) Results modeled the effect of climate warming on the horn fly, a common pest of pastured cattle that causes Estimates of changes in peak demand, capacity reductions in weight gain and milk production. (For requirements, and cumulative and annual costs a description of study design and limitations, see projected for the climate change scenarios in the Chapter 6: Agriculture.) This study used only the Great Plains are shown in Table 17-5. The results GFDL scenario; since it had the highest are driven by seasonal changes in weather-sensitive temperatures, results should be considered as an demands for electricity: summertime use for air- extreme case. In Texas, horn fly populations were conditioning and irrigation-pumping increases and estimated to become lower in summer than they are outpaces reductions in demand for space heating in currently because high temperatures are lethal to the winter. Electricity demand grows by 2 to 4% by the insects when they are immature. Thus, weight 2010, and new capacity requirements are estimated gains of calves and feeder/stocker cattle could to increase by 15 to 28% by 2010 for the climate increase relative to current rates in Texas. In change scenarios as compared with the base case Nebraska, however, temperatures in the GFDL (i.e., economic growth without climate change). By scenario would not reach lethal levels, and increases 2010, additional cumulative capital costs induced by of 225 to 250 horn flies per head were estimated. climate change may be $3.7 to $6.7 billion, and This would result in greater weight reductions than annual costs of generating power may rise by 3 to those currently observed. These results suggest that 6%. greater stress may occur in livestock production in the northern part of the Great Plains, and that In 2055, new capacity generating requirements stress may be alleviated in Texas. are estimated to increase by 22 to 45 gigawatts or 27 to 39%. Annual electricity demand in the region Stem et al. (see Volume C) studied the effects increased an additional 10 to 14% by 2055 under of climate change on animal disease patterns. (For the climate change scenarios. New capacity study design and limitations, see Chapter 6: requirements without climate change are estimated Agriculture.) The ranges of some diseases may be to be 20 GW by 2010 and 112 to 134 GW by 2055. extended as habitats of disease vectors enlarge or as 373 Chapter 17 Table 17-5. Estimated Change in Peak Demand and Annual Energy Requirements Induced by Climate Change (%) 2010 2055 GISS A GISS B GISS A Utility area Ann. Peak Ann. Peak Ann. Peak Kansas/Nebraska 1.7 6.8 1.3 5.2 5.7 22.1 Oklahoma 3.0 7.9 2.8 6.6 11.3 25.3 Texas, east 3.0 7.9 2.8 6.6 11.3 25.3 Texas, south 3.3 10.0 1.7 5.1 10.6 24.6 Texas, west 3.1 8.6 2.4 6.1 11.1 25.1 Source: Linder and Inglis (Volume C). Linder and Inglis calculated that cumulative resources. Although the Ogallala Aquifer has come capital costs for electricity in the region would under close scrutiny in the past, it is important to increase from $20 to $53 billion by 2055 with note that previous studies have not addressed climate change. The estimated changes in annual potential climate change impacts on this resource. costs induced by climate change range from $5 to Many of the problems associated with intense $10 billion. groundwater use (water depletion, soil damage, altered rural and farm economics, and potential Implications reversion to dryland farming) could be exacerbated by global warming. This study shows that irrigated Increased electrical capacity requirements and acreage in the Great Plains could increase and that the need to maintain the reliability of utility systems the demand on the aquifer could rise by up to 15%. could place additional stress on the Great Plains. These potential adjustments to climate change This is especially important if climate change should be studied to understand their implications increases the demand for irrigation, which is an for land use, resource conservation, regional important consumer of electricity in the region. economics, and community issues in the Ogallala Also, the potential exists for conflicts between area. power production and agriculture over the use of scarce resources such as water. Powerplants may take the cooling water they need from rivers or POLICY IMPLICATIONS from the already overused Ogallala Aquifer, and increased coal and oil production in the region The policy options for responding, either in would utilize land that might be farmed. However, anticipation or in reaction, to climate change in the energy production may provide alternative income Great Plains range from noninterference, in which sources in an area whose economy is poorly agricultural, water, and other resource systems are diversified. left to adjust without assistance, to a more active approach in which federal, state, and local government agencies plan for and assist in the CLIMATE CHANGE AND THE process of adaptation. OGALLALA AQUIFER Given the historical government involvement in Warming and/or drying in the Great Plains may agriculture, especially in this marginal region where place greater demand on regional groundwater support programs may mean the difference between 374 Great Plains farm survival and failure, it is likely that an active flexible repertoire of anticipatory strategies; new adjustment process will be called for. Policymakers institutional arrangements may be needed. in the Great Plains may have to respond to decreased agricultural production in the area, Some programs already in place can help to increased demand for water and electricity, poorer lessen the negative effects of climate change on the water quality, and changes in livestock production. Great Plains. Federal legislation such as the "Sod- The major issues that policymakers should address Buster Bill" and programs such as the Conservation include land-use management, water resource Reserve Program are examples of new policies management, and agricultural risk management (see designed to reduce the use of marginal lands for Riebsame, Volume J). Regional utility planners and agriculture. The basic goals of these laws are to policymakers should also begin to consider climate protect the most erodible farmlands by removing change as a factor -- along with other uncertainties them from crop production, and to use conservation -- affecting their resource availability analyses and as a tool for reducing overproduction. Such planning decisions. programs are prudent now for reducing erosion and may become even more important for protecting Of course, uncertain and limited impact soil and water quality in a changing climate. assessments such as those described above cannot However, protection of marginal lands may have to be used to create and implement detailed policy. be weighed against the need for greater crop Rather, they should be viewed as scenarios that production if climate change lowers yields. For suggest the types of policies and the range of policy example, the government's response to the 1988 mechanisms and flexibilities that could alleviate drought was to release some conservation land for potentially disruptive impacts from climate change. cropping in 1989. This would help replenish food The eventual problem for the policymaker, of stocks but also would place a greater amount of course, is deciding when to switch from scenario marginal land at risk of erosion. analysis to actual policy formulation and implementation. The last few sections of this Water Resource Management chapter suggest some of the policy implications raised by the impacts described earlier. If GCM projections of climate change are qualitatively correct, parts of the Great Plains are Land-Use Management likely to suffer increasing aridity. Farmers may demand more water for irrigation, although Land managers should analyze how their groundwater sources are already taxed. missions and holdings may be affected by climate Competition for water resources between change and should develop flexible strategies to deal agricultural and nonagricultural demands may be with potential impacts. Federal agencies, such as exacerbated. Water managers need to factor the the Department of Agriculture, the Forest Service, potential effects of climate change into their the Fish and Wildlife Service, and the Department decisions on irrigation, drainage, and water transfer of Interior, should work with state agriculture, systems, and they should consider potential climate forest, and park agencies on such plans. change as they formulate supply allocation rules, reservoir operating criteria, safety protocols, and Climate change may cause agriculture and other plans for long-term water development. Water land uses to become more environmentally and conservation techniques, water reallocation between economically marginal in the Great Plains. competing uses, water transfers and marketing, and Consequently, land uses may shift in intensity, type, land-use adjustments should be evaluated for their and location. Indeed, locational shifts may involve ability to absorb the effects of a range of future several states or multiple regions. This adjustment climate changes. The goal at this point may not be process can be made more efficient and less to formulate detailed policy, but rather to test the disruptive if individual jurisdictions, such as climate sensitivity and feasibility of alternative water municipalities, states, and federal regions, respond management policies and practices. in a coordinated manner. Decisions made by managers of agriculture will affect forests, wildlife, Decisionmakers should also consider the and water resources. Decisionmakers should begin potential effects of climate change on water quality now to work together to develop a sound and and the use of pesticides. They should examine 375 Chapter 17 alternative pest control strategies, such as Integrated Jones, C.A., and J.R. Kiniry. 1986. CERES-Maize: Pest Management, which use biological control, A Simulation Model of Maize Growth and genetic resistance, and innovative cropping systems Development. College Station, TX: Texas A&M to reduce pesticide applications. Press. Risk Management Liverman, D.M., W.H. Terjung, J.T. Hayes, and L.O. Mearns. 1986. Climatic change and grain corn Several government, semiprivate, and private yields in the North American Great Plains. Climatic institutions have a large financial stake in Great Change 9:327-347. Plains agriculture through land credit, commodity and equipment loans, and insurance. Additionally, Lockeretz, W. 1978. The lessons of the Dust Bowl. the federal government provides disaster relief for American Scientist 66:560-569. climate extremes affecting regional agriculture. Climate warming poses a potential long-term risk to Michaels, P.J. 1985. Economic and climatic factors the financial institutions supporting agriculture, to in "acreage abandonment" over marginal cropland. the resources available for emergency relief, and to Climatic Change 7:185-202. individual farmers. This possibility should be carefully assessed, and plans should be made now to Monteith, J.L. 1965. Radiation and crops. monitor risk as climate changes. Experimental Agriculture Review 1(4):241-251. Otter-Nacke, S., D.C. Goodwin, and J.T. Ritchie. REFERENCES 1986. Testing and Validating the CERES-Wheat Model in Diverse Environments. Houston, TX: Bowden, M.J., R.W. Kates, P.A. Kay, W.E. Lyndon B. Johnson Space Center. AgRISTARS YM-15-00407. JSC 20244. Riebsame, D. Johnson, H. Gould, and D. Weiner. 1981. The effect of climate fluctuations on human populations: two hypotheses. In: Wigley, T.M.L., Popper, D.E., and F.J. Popper. 1987. The Great M.J. Ingram, and G. Farmer, eds. Climate and Plains: from dust to dust. Planning (December):1 18. History. Cambridge, United Kingdom: Cambridge University Press, pp. 479-513. Powers, W.L. 1987. The Ogallala's bounty Carsel, R.F., C.N. Smith, L.A. Mulkey, J.D. Dean, evaporates. Science of Food and Agriculture 5(3):2- 5. and P. Jowise. 1984. Users' Manual for Pesticide Root Zone Model: PRZM. USEPA/ERL Report EPA-600-3-84-109. Washington, DC: U.S. Riebsame, W.E. 1983. Managing agricultural Government Printing Office. drought: the Great Plains experience. In: Platt, R., and G. Macinko, eds. Beyond the Urban Fringe: Dregne, H.E., W.O. Willis, and M.K. Adams. 1988. Land Use Issues in Non-Metropolitan America. The metamorphosis of the "Great American Minneapolis: University of Minnesota Press, pp. 257-270. Desert." Science of Food and Agriculture 6(11):2- 7. Riebsame, W.E. 1987. Human Transformation of the United States Great Plains: Patterns and Glantz, M.H., and Ausubel, J.H. 1984. The Ogallala Causes. Proc. Symp. on the Earth as Transformed Aquifer and carbon dioxide: comparison and by Human Action, Clark University. convergence. Environ. Conser. 11(2):123-31. Ritchie, J.T., and S. Otter. 1985. Description and High Plains Associates. 1982. Six-State High Plains- performance of CERES-Wheat: user-oriented wheat Ogallala Aquifer Regional Resources Study. Austin, yield model. In: Willis, W.O., ed. ARS Wheat Yield TX. Project. USDA-ARS. ARS-38. pp. 159-175. Hurt, R.D. 1981. The Dust Bowl. Chicago: Nelson- Hall. 376 Great Plains Robertson, T., V.W. Benson, J.R. Williams, C.A. USDA. 1983. U.S. Department of Agriculture, U.S. Jones, and J.R. Kiniry. 1987. Impacts of climate Bureau of the Census. 1982 Census of Agriculture. change on yields and erosion for selected crops in Washington, DC: U.S. Government Printing Office. the southern United States. In: Meo, M., ed. Proc. Symp. on Climate Change in the Southern U.S.: USDA. 1985. U.S. Department of Agriculture, Impacts and Present Policy Issues, Science and Economic Research Service. Foreign Agricultural Public Prog., Univ. of Oklahoma, Norman, OK. Trade of the United States. Washington, DC: U.S. Government Printing Office. Rosenzweig, C. 1985. Potential CO2-induced climate effects on North American wheat-producing regions. Warrick, R.A. 1984. The possible impacts on wheat Climatic Change 7:367-389. production of a recurrence of the 1930s drought in the U.S. Great Plains. Climatic Change 6:5-26. Rosenzweig, C. 1987. Climate change impact on wheat: the case of the High Plains. In: Meo, M., ed. Warrick, R.A. 1975. Assessment of Research on the Proc. Symp. on Climate Change in the Southern Drought Hazard in the United States. Monograph U.S.: Impacts and Present Policy Issues, Science and No. 4. Boulder, CO: Natural Hazards Research and Public Prog., Univ. of Oklahoma, Norman, OK. Applications Information Center, University of Colorado. Schneider, K. 1988. Drought cutting U.S. grain crop 31% this year. The New York Times August Warrick, R.A., and M.J. Bowden. 1981. The 12:A1. changing impacts of drought in the Great Plains. In: Lawson, M.P., and M.E. Baker, eds. The Great Terjung, W.H., D.M. Liverman, and J.T. Hayes. Plains: Perspectives and Prospects. Lincoln: 1984. Climatic change and water requirements for University of Nebraska Press, pp. 111-137. grain corn in the North American Great Plains. Climatic Change 6:193-220. Worster, D. 1979. Dust Bowl: The Southern Great Plains in the 1930s. New York: Oxford University Press. 377 CHAPTER 18 RESEARCH NEEDS This report has suggested that concerns over Research in the natural and social sciences the adaptability and fate of both natural and must have an important role in developing well- managed ecosystems in a changed climate are well reasoned adaptation strategies because it will founded. Natural forested ecosystems, aquatic and provide the data and understanding of processes marine biota, wildlife in refuges, water quality in necessary to design efficient responses to a new small lakes, and other resources may be vulnerable climate, and better management techniques for the to rapid climate change. Strategies for mitigating resources that must be conserved. changes in these systems are likely to be complex and difficult to implement. While it may be difficult The needs of U.S. and international to quantify the consequences, climate change may policymakers for information on the possible have large effects on biodiversity, primary environmental effects of climate change and the productivity, and cycling of nutrients, and it may be processes that control them should not be difficult, if not impossible, to reverse these impacts. underestimated, especially since the task of attempting to mitigate emissions of greenhouse This report has also shown that while gases is so large and complex. This chapter intensively managed ecosystems, especially identifies some of the major topics for research in agroecosystems, may also be affected by a climate the natural and social sciences that should be change, there seem to be more opportunities for pursued to help policy analysis and development in human intervention to mitigate or adapt to their this area. responses. Thus, the critically important question is whether the capacity for human intervention can The scope of this chapter is necessarily broad. keep pace with the rate of change induced by It addresses both the research proposed by EPA changing climate. Areas of major concern are the and the research recommendations of the scientific interactive effects of climate change and carbon research community from a perspective that the dioxide increases on crop yields, and the adaptation development of sound environmental policy, both rate of management practices. for mitigation and adaptation, depends on the capability of the scientific research community to Although it is clearly not possible to study all respond to increasingly specific demands for the potential effects of a change in the climate information from policymakers. system, or to consider all the possible social or political ramifications of responding to climate change, there will be a continuing need to RELATIONSHIP BETWEEN understand better the possible consequences of climate change because adaptation to different POLICY AND SCIENCE climates will be a necessary part of any complete societal strategies to cope with the greenhouse Secretary of State James Baker and EPA effect. Therefore, it is important to have in place a Administrator William Reilly recently set forward research framework for both the natural and the four principles to guide policy development: social sciences that will provide the information required to allow societies to respond to the The first is that we can probably not challenge of large-scale, rapid changes in the afford to wait until all of the uncertainties climate system. This research should be undertaken have been resolved before we do act. simultaneously and in coordination with programs Time will not make the problem go away. directed at establishing a broad consensus for governmental actions, both domestic and The second is that while scientists refine international, that address energy, land use, and the state of our knowledge, we should other social policies that might lead to reduced focus immediately on prudent steps that emissions of greenhouse gases. are already justified on grounds other 379 Chapter 18 than climate change. These include international scientific and policy consensus on reducing CFC emissions, greater energy greenhouse issues. efficiency, and reforestation. EPA's domestic responsibilities, and the The third is that whatever global solutions research reported on in this document, have led us to global climate change are considered, to formulate several important questions that should they should be as specific and cost-effective be thought of as overriding themes, rather than as as they can possibly be. a list of all the potential issues: The fourth is that those solutions will be How rapidly might climate change as a most effective if they transcend the great result of future manmade emissions? fault line of our times, the need to reconcile the transcendent requirements for What are the likely regional atmospheric both economic development and a safe manifestations of such global atmospheric environment. changes? These four principles establish a framework What are the likely extent and magnitude of within which both domestic and international ecological, environmental, and societal programs will develop. They balance the needs for changes associated with a given change in both scientific research and policy development, regional atmospheres? while clearly recognizing the international scope of the issue. In doing so, these four principles will act What technologies and policy options exist as the basis for U.S. participation in international to reduce the rate of growth in greenhouse assessment activities, as well as for domestic policy gas emissions, and how much would they development. cost? The Global Climate Protection Act of 1987 What are the cultural and institutional directs EPA and the State Department to barriers that might limit the implementation coordinate the development of national policy for of such options? global climate change. This coordination involves many other agencies with essential policy roles, such What are the likely consequences of as the Department of Energy. proposed mitigative or adaptive policies? In addition, the Global Climate Protection Act These questions are viewed as the foundation directs EPA, in cooperation with other agencies, to for analyzing possible environmental changes due to prepare a scientific assessment of climate change. climate change, and eventually for analyzing possible This assessment is now being coordinated through approaches to managing risks. They begin to match the Intergovernmental Panel on Climate Change, an needs for policy development with scientific needs organization created under the joint auspices of the for understanding the functioning of the Earth as an United Nations Environment Programme and the integrated system. By doing so, they define the World Meteorological Organization (WMO). It will specific areas in which scientific research is be developed by a work group with extensive U.S. necessary: biogeochemical dynamics, physical participation coordinated through the Federal climate and the hydrologic cycle, ecosystem Coordinating Committee on Science and dynamics, Earth system history, and human Engineering Technology Committee on Earth interactions with the geosphere-biosphere. Indeed, Sciences. A second work group will analyze climate they justify an overall program of research, with one change impacts, and a third work group is of the main goals being to "establish the scientific responsible for examining response strategies. Each basis for national and international policymaking work group has approximately 18 months to develop related to natural and human-induced changes in an interim report. Reports from these three work the global earth system" (Federal Coordinating groups will be critical to the development of Committee on Science and Engineering Technology). 380 Research Needs RESEARCH AND ASSESSMENT Water resource managers face similar problems. In California, all the scenarios indicated NEEDS IN THE SOCIAL that large changes in the management of water SCIENCES might need to be considered if the snowpack were smaller and melted earlier. In the Great Lakes, This report has identified many important issues lower water levels may necessitate changes in that policy analysts and decisionmakers must begin management. While changes in precipitation or have begun to address. It is apparent that even remain the most uncertain of the outputs from for the heavily managed environmental resources GCMs, the lessons for research in water such as agriculture and water supply, an existing management are relatively clear. We need to range of concerns makes the response of resource understand the degree to which there is flexibility in managers to climate change difficult to predict. water allocation decisions, and to develop the Even current climate variability is not always information needed by water managers to evaluate accounted for in resource management. Yet it is possible changes in allocation under climate change. the response of resource managers and environmental policymakers to climate change that In each of these cases, both the institutional will ultimately determine how society responds to a and historical factors that affect the decisionmaking changed climate both for managed and natural process must be analyzed and understood, as must resources. The inadequacy of our current local, regional, and national political influences. In knowledge regarding how their decisions are made particular, the problems of designing resource demands closer attention from the social science management systems for flexible response need to be addressed as institutional and investment research community. questions. While the need for flexible resource Institutional Response to Climate management is clear, the reality of maintaining flexibility while still making decisions regarding Variability and Climate Change large capital expenditures, such as building powerplants and dams, may be quite difficult. One of the major issues identified in this report There will be a continued need to conduct targeted is how institutions respond to current variability in case studies of how resource managers currently climate. It is well known that current climate consider climate variability and to address potential variability, represented by such episodes as the future changes in variability (see Chapter 19: recurrence of the El Niño and periodic droughts, Preparing for Climate Change). can have catastrophic effects on major regional industries, that in turn have larger, sometimes global In addition, while climate change may consequences on supply and processing of resources. ultimately be one of the most important variables It is also well known that in both the relatively that managers must consider in the decisionmaking distant and relatively recent past, variability in prócess, it may not be the most immediate. climate has led to severe regional economic Research is necessary to show how devoting dislocation and subsequent migration of large attention and resources to a developing issue such numbers of people, even in industrialized societies as climate change makes sense from a management such as the United States. What is not as well and policy standpoint. Research is also necessary to known is how the U.S. institutions responsible for examine the differences in how a wealthy, highly managing agriculture, forestry, and water resources industrialized society, such as the United States, will be able to respond to future climate variability, makes decisions about responding to climate especially if that variability increases. The drought variability and change and how other societies, of the summer of 1988 clearly illustrates that U.S. especially lesser developed countries, make such farms are still susceptible to severe weather decisions. Since climate change is intrinsically a conditions; it does not, however, answer the global issue, such studies will be necessary to form question of whether a succession of such droughts, a consensus regarding the need for coordinated as might be expected in future scenarios of a responses and management strategies. warmer, drier Grain Belt, could be accommodated by the existing government programs. 381 Chapter 18 RESEARCH AND ASSESSMENT regional consequences of global atmospheric change. The transition from traditional disciplinary NEEDS IN THE NATURAL investigations of processes to interdisciplinary SCIENCES investigations of the links between processes on such large spatial scales will demand new As reviewed by the National Academy of approaches from the scientific research community. Sciences Committee on Global Change (NRC, 1988), in order to be responsive to policy concerns, Figure 18-1 represents in schematic fashion the the primary scientific research needs are in those information flow that must occur among scientific phenomena and processes that occur on global disciplines while explicitly taking into account the scales, or that occur on regional scales but will have transitions between spatial scales. It indicates that global consequences over the next few decades to a the purpose of conducting research in emissions of few centuries. Therefore, research and assessment trace gases, inventorying and evaluating the emission activities must examine global scale questions of factors of anthropogenic and biogenic sources of emissions and atmospheric chemistry as well as the trace gases, evaluating possible technological Global Emissions Global Concentrations CO, CO2 CH Global Global CH, N.O Emissions Atmospheric N,O CFC's Scenarios Chemistry CFC's CO & Scenarios O & Models NOx CO NMHC NOx NMHC Physical Climate Models & Climatology Regional Climatology R SOx T NOx Regional NMHC Regional CO, Emissions SO, Environmental Ecosystem Air NO, Resources Response CO Scenarios Part.'s Quality Models H+ at Risk & Models Models & Scenarios CH, etc. etc. Atmospheric Regional Emissions Scenarios Productivity Yield Distribution Water Quality etc. Impacts Figure 18-1. Relationship between global and regional information flow. 382 Research Needs controls, investigating the possibility of positive Socioeconomic Impacts feedbacks, and attempting to realistically simulate the emissions of trace gases is to provide The final major link is between the ecological information for understanding the composition of and environmental consequences of climate change the atmosphere. Models can then be used to create and emissions of greenhouse gases. This link must estimates of atmospheric composition on include the interaction between societal impacts, approximately the same temporal and spatial scales. such as changes in energy demand and end-use, and changes in emissions. It will be critical to establish Climate System interdisciplinary communication because of feedbacks between the biosphere and the The scientific research community should fully atmosphere. Clearly, changes in the growth and investigate the dynamic consequences of different distribution of major terrestrial vegetation types, as compositions of the atmosphere, including the well as changes in ocean chemistry and biology, will dynamics of the ocean as it influences both alter biogenic emissions of trace gases. Of critical atmospheric composition and heat transfer. The importance is the possibility that these biogenic derivation of regional climate scenarios from either emission changes may lead to even greater modeling output or analog methods and scientific temperature changes (positive feedbacks), as has understanding are then necessary to link the been hypothesized for methane. How climate processes on global scales with environmental and change will affect anthropogenic emissions, and ecological research questions on regional and local whether changes would be positive or negative scales. The climate system modeling community, as feedbacks, is largely unexplored. well as the statistical climatology community, must devote significant effort to improving the ability of Data the atmospheric sciences to make predictions on relatively small regional scales, so that policymakers Underlying all these concerns for the can begin to have some quantitative confidence in interaction among processes in the natural world is the results from environmental and ecological a critical need for long time-series of data on Earth modeling. system processes, and the information systems necessary to manage the data. No amount of Research Scales modeling or experimentation of processes will replace actual observations of how the Earth system A further critical link identified in Figure 18-1 responds to changes in climate forcing and the is that estimates of environmental changes will be degree and characteristics of its natural variability. needed on spatial scales that are larger than ecologists and environmental scientists have Objectives of Federal Global Change traditionally used in their research (e.g., ecoregions Program to biomes). While initially qualitative, as in much of this report, these estimates will be used both as Both the NAS (NRC, 1988) and the Federal input for assessments and as a way to formulate Global Change Program (CES, 1989) have identified series of testable hypotheses concerning the the scientific elements intrinsic to understanding the processes that control projected ecological changes. Earth's behavior as an integrated system, and especially its response to global atmospheric change. The ecological and environmental research The section below summarizes the scientific community must, therefore, define those elements and their rationale, and presents the broad atmospheric variables that control the growth and scientific objectives of the research to be sponsored distribution of major vegetation types, including in the Federal Global Change Program. These crops, and must explore the physical and biological scientific elements refer directly back to the needs processes that control the distribution of water and for information identified in Figure 18-1, as shown nutrients in natural and managed landscapes. These in Figure 18-2. definitions and processes must be those that affect the characteristics and dynamics of ecosystems on Biogeochemical dynamics include (1) the spatial scales commensurate with the atmospheric sources, sinks, fluxes, and interactions scales defined above. 383 Chapter 18 Scenario Assessment Global Emissions Atmospheric Atmospheric Chemistry Dynamics Scenario Assessment Figure 18-2. Two-stage scenario approach to integration. between biogeochemical constituents within cryosphere (i.e., glaciers, snow), land the Earth system; (2) the cycling of surface, and biosphere. biogeochemical elements in the atmosphere, oceans, terrestrial regions, biota, and These are clearly central to the description, sediments over Earth's history; and (3) the understanding, and prediction of global climate influence of biogeochemical elements on change, particularly in terms of impacts on global the regulation of ecological systems and climate conditions and the hydrologic system. contribution to potential greenhouse constituents (CO₂, CH4, N₂O, CFCs) that Human interactions has been defined as the have a direct influence on climate. study of the impacts of changing global conditions on human activities. The global Ecological systems and dynamics would environment is a crucial determinant of involve the responses of ecological systems, humanity's capacity for continued and both aquatic and terrestrial, to changes in sustained development. Research should global environmental conditions and of the focus on the interface between human influence of biological systems on the activities and natural processes. atmospheric, climatic, and oceanic systems. This includes studies of plant succession, Earth system history is the study of the terrestrial and aquatic biodiversity, natural record of environmental change that extinctions, and relationships with is contained in the rocks, terrestrial and geological substrate. Monitoring and marine sediments, glaciers and ground ice, specific ecosystem experiments can provide tree rings, eumorphic features (including information on stresses influencing the the record of eustatic changes in sea level), biota and on the biotic response to natural and other direct or proxy documentation of and societal environmental stresses. Such past environmental conditions. These information is needed to achieve the basic archive the Earth's history and document understanding required for the development the evolution of life, past ecosystems, and of models. Identification and study of human societies. Past ecological epochs particularly sensitive ecosystems will be with warmer or cooler climates relative to especially informative. the present climate are of particular scientific interest. Climatic and hydrologic systems would involve the study of the physical processes Solid-earth processes include the study of that govern the atmosphere, hydrosphere certain processes that affect the life- (oceans, surface and groundwaters, etc.), supporting characteristics of the global 384 Research Needs environment, and especially the processes 3. Develop integrated conceptual and that take place at the interfaces between predictive Earth system models. the Earth's surface and the atmosphere, hydrosphere, cryosphere, and biosphere. Each of these objectives simultaneously leads Solid-earth processes that directly affect the toward improving the monitoring, understanding, environment are of primary interest; and predicting of global change. They aim to processes that have only indirect effects are provide, by the year 2000, detailed assessments of excluded. the state of the knowledge of natural and human- induced changes in the global Earth system and The solar influence is the study of the appropriate predictions on time scales 20 to 40 years variability in solar radiation and its impact into the future. Assessments of uncertainties in on atmospheric density, chemistry, model outputs will be an integral part of these dynamics, ionization, and climate. predictions. Research on the effects of solar variability on biogeochemical cycles as well as the impact of ultraviolet light on biology and THE ROLE OF EPA IN POLICY chemistry would be particularly important AND SCIENTIFIC RESEARCH here. Of these scientific elements, studies of EPA's own activities have been structured to biogeochemical dynamics, climate and hydrologic provide leadership in both policy analysis and systems, ecosystem dynamics, Earth system history, development, as required by the Global Climate human interactions, and to a lesser extent, solar Protection Act, and in scientific research, especially influences, are the most important from the on the consequences of changes in the climate standpoint of developing a policy-oriented research system. The development of a broad-based, program. The degree to which the solid-earth interdisciplinary scientific research program that processes are important depends entirely on their responds to the policy-oriented questions identified contribution to global change over the time-scale of earlier in this chapter has depended strongly on a few decades to a few centuries. A better concurrent scientific planning efforts by the National understanding of these processes remains an Academy of Sciences and the Federal Global important scientific aspect of a Federal Global Climate Change Program. Change Program but can be anticipated to have less value from a public policy perspective. Specifically, the goals and objectives of the EPA Global Climate Change Research Program Three Major Scientific Objectives have been structured to respond both to the policy- oriented questions, and to the scientific needs The scientific elements relevant to the identified by NAS in the U.S. proposal for the development of well-informed public policy must be International Geosphere Biosphere Program and as structured in a way that permits the overall adopted by the Federal Global Change Program. objectives of the U.S. program to contribute to both The program is designed to provide information on scientific and policy communities. To accomplish the biosphere and its response to climate change this, the Federal Global Change Program has and technical information to develop policy options outlined three major objectives in its Strategy to limit and adapt to climate change. EPA's Document (CES, 1989). proposed research has two goals: 1. Establish an integrated, comprehensive 1. To assess the probability and magnitude of program for Earth system measurements on changes in the composition of the global a global scale. atmosphere, the anthropogenic contributions to those changes, and the 2. Conduct a program of focused studies to magnitude of subsequent impacts on the improve our understanding of the physical, environment and society. chemical, and biological processes that influence Earth system changes and trends 2. To assess the likely extent, magnitude, and on global and regional scales. rate of regional environmental effects as a 385 Chapter 18 function of changes and variability in policy responsibilities in issues of global climate climate, for the purpose of evaluating the change are going to be able to take advantage of risks associated with changes in the climate developments in all areas previously discussed. system. Many of the developmental needs in the atmospheric and space sciences, and many of the Eight associated scientific and institutional global monitoring needs, will be beyond the objectives have been identified: capability of any one federal agency and will require the cooperation of all. 1. To develop improved estimates for both anthropogenic and natural sources of The goals and objectives of proposed policy radiatively important trace gases, and to research and activities in EPA closely follow the investigate the feedback processes by which previously listed recommendations. The main foci climate variability influences the sources of will be on the development and coordination of a these gases. national policy, as called for in the Global Climate Protection Act, and the coordination and 2. To develop techniques for estimating implementation of the International Response current and future emissions of radiatively Strategies Assessment of the IPCC. Both mitigation important trace gases. and adaptation policies will be investigated, as outlined in the following chapter. 3. To improve understanding of global atmospheric chemistry in order to project future concentrations of trace gases, IMPACT ASSESSMENT including tropospheric ozone. METHODOLOGY 4. To relate global changes in climate to regional changes by constructing a series of Continued efforts at assessing the causes and regional atmospheric scenarios. consequences of climate change are clearly needed. This report has illustrated one potentially valuable 5. To predict ecosystems' responses to climate method for conducting such an assessment. change and to test the processes that However, because the need will continue, there is a control those responses. corresponding need to consider how best to do assessments in a way that preserves both the 6. To document the spatial covariation of understanding of what may happen and the certainty regional climate change with regional with which we know it. This section outlines the ecological change in order to establish approach that will be taken in future impact comprehensive ecological monitoring in assessment efforts led by EPA. selected locations, cooperatively with EPA and other federal programs. Integrated modeling of large-scale environmental issues has been attempted many 7. To develop information on technologies and times before and may be useful for policy analysis practices that could limit greenhouse gases or for heuristic purposes. However, there is general and to adapt to climate change. agreement within the scientific community that a model adequate to simulate the dynamics of 8. To produce periodic scientific assessments geophysical, chemical, and biological processes on in conjunction with other federal agencies global scales will be developed only after decades of and international research organizations, research (ESSC, 1988). and to perform research to evaluate the consequences of adaptation and mitigation Although achieving such a goal lies so far in policies. the future, the question of how to deal with integrating diverse aspects of science in global While defining the framework for EPA's own climate change and its potential effects in the nearer scientific research, these goals and objectives also term remains. One promising approach for assume that all federal agencies with significant integrating research results is to treat the entire 386 Research Needs Scenario Human & Scenario Assessment Industrial Effects Assessment Atmospheric Regional Emissions Dynamics Atmospheres Scenario Environmental & Ecological Effects Assessment Figure 18-3. Three-stage approach to integration. cycle of information flow (Figure 18-1) as a series of results and data, in both qualitative and potentially two-stage processes (Figures 18-2 and 18-3). quantitative fashion. Within each two-stage process, research results Each pair of scenario-response steps is should be treated as follows: The first part of the explicitly decoupled from other pairs, while process is the creation of a set of scenarios, where remaining consistent with them. Thus, such an a scenario is defined as a plausible combination of approach can indicate both ranges and sensitivities variables derived from a set of internally consistent of responses in potentially verifiable fashion within assumptions. The second part of the process will each pair, but does not attempt the premature task evaluate the range of changes that are potentially of modeling uncertainty all the way through the attributable to each scenario and will evaluate the global system. sensitivity of the underlying systems to different aspects of the scenarios. Thus, scenarios of changes The use of scenarios as assessment and in land use could be used to evaluate possible integrative tools is not part of the traditional changes in emissions; scenarios of emissions could scientific approach toward prediction and validation. be used to evaluate the possible changes in Nevertheless, it is important from three standpoints: atmospheric composition; scenarios of atmospheric composition could be used to evaluate changes in 1. For scientific information to be of use to climate; climate scenarios could be used to evaluate policymakers, a continued iterative process the possible changes in ecosystems; and scenarios of of evaluating the state of knowledge in the ecosystem and land-use changes can in turn be used suite of sciences relevant to global change to evaluate possible changes in emissions. must be maintained. An iterative process of using and analyzing scenario-based The use of a scenario-assessment approach for assessments can provide such information impact assessments has several advantages. It could in a usable and informative way. provide clear priorities for research on the sensitivities of important environmental processes in 2. To achieve the multidisciplinary syntheses each scientific area. It maintains a realistically needed to make scientific advances in holistic view of the problems of global change, and problems of global climate change, it preserves information on the uncertainty of model evaluation of the methods by which 387 Chapter 18 predictions are made and by which REFERENCES scenarios of change can be composed, and evaluation of the sensitivities of affected CES. 1989. Committee on Earth Sciences. Our processes must continue. The scenario- Changing Planet: The FY 1990 Research Plan. based assessment approach provides a The U.S. Global Change Research Program. A ready-made integrating framework for such continual evaluations. Report by the Committee on Earth Sciences. July 1989. Washington, DC: Executive Office of the 3. Because of the importance of this proposed President, Office of Science and Technology Policy. July. research in public policy arenas, it is critical not to lose sight of what is and is not ESSC. 1988. Earth Systems Science Committee. predictable. By distinguishing between a set of scenarios and actual verifiable A Program for Global Change: Earth Systems Science a Closer View. Report of the Earth predictions, the scenario-based approach can best illustrate the difference without Systems Science Committee. Washington, DC: NASA Advisory Council. January. becoming a morass of hedged bets. IGBP. 1988. International Geosphere-Biosphere Program. Toward an Understanding of Global Change: Initial Priorities for the U.S. Contribution to the International Geosphere-Biosphere Program. Washington, DC: National Research Council, Committee on Global Change, National Academy Press. 388 CHAPTER 19 PREPARING FOR CLIMATE CHANGE The preceding chapters suggest that a global require thousands, perhaps millions, of warming could have significant impacts on farms decisionmakers to consciously consider global and forests, rivers and lakes, fish and wildlife, and warming as they plan their activities. many practical aspects of everyday life. This issue is very different from other environmental problems. These differences need not thwart the process It is global in scope: all nations emit greenhouse of preparing for global warming. First, many types gases and all will experience the impacts. of institutions already cope with equally long-term Moreover, the changes are likely to last for and uncertain trends; transportation planners, for centuries and could shape the very nature of society. example, routinely consider economic growth over Although many of the possible consequences may 30- to 50-year periods when picking routes for not occur for decades, it is important that we begin highways and urban rail systems. Second, reaching now to examine how we might respond. a consensus on what is fair would be easiest when no one feels immediately threatened. Finally, the The potential responses fall broadly into two decentralized nature of adaptation would enable the categories: (1) limiting the change in climate; and communities and corporations most sensitive to (2) adapting to it. These two responses are climate change to respond quickly, rather than complementary, not mutually exclusive. Because having to await a national consensus on the most past emissions of greenhouse gases may eventually appropriate response. warm the Earth one degree Celsius, some adaptation will be necessary, and efforts to prepare Because a companion report ("Policy Options for global warming can contribute information to for Stabilizing Global Climate") examines options the process of deciding whether, when, and how to for limiting future global warming, this chapter limit it. On the other hand, slowing the rate of focuses on adaptation strategies. We briefly discuss global warming would make it easier for humans the process of choosing such strategies, then present and other species to adapt. several examples. Although limiting climate change would require worldwide cooperation, responding to its WHEN IS A RESPONSE consequences would not. Private citizens and companies can relocate or modify their operations. WARRANTED? Communities and states can undertake public works or enact planning measures. Charitable foundations Strategic Assessments and profit-making corporations can support research to develop better response strategies. One of the most fundamental issues facing National governments can support all of these activities. decisionmakers is whether to implement responses today or to defer preparation until the timing and magnitude of future climate change are more Preparing for global warming raises three certain and the potential impacts are more challenges. First, the uncertainties make it difficult imminent. Although global warming might to be sure that we are employing the correct eventually require particular actions, such actions response: the climate may change more (or less) need not necessarily be taken today. On the other than anticipated; in the case of precipitation, we do not even know the direction of change. Second, the hand, the likelihood of at least some global warming long-term nature increases the difficulty of is sufficiently well established and the time required forecasting the impacts and gaining the attention of to develop a response sufficiently long that deferring all preparation could lead us to miss opportunities decisionmakers more accustomed to focusing on to substantially reduce the eventual economic and near-term problems. Finally, adaptation would environmental costs of the greenhouse effect. 389 Chapter 19 Individual organizations must decide for potential implications of climate change may require themselves whether or not to prepare for the little more than a few additional computer greenhouse effect. The first question is whether simulations. global warming is likely to alter the success of current activities or projects now being planned. If The Council on Environmental Quality has held not, preparing for the impacts of climate change public meetings on the possibility of requiring usually would be unnecessary; if so, the next federal agencies to consider climate change in question is whether doing something today would be environmental impact statements. The rationale is worthwhile. that (1) if climate changes, the environmental impact of some federal projects may be different We use the term "strategic assessment" to than the impact if the climate does not change; and refer to the process by which people and (2) these assessments are an inexpensive way to organizations examine whether, when, and how to increase our understanding of the potential respond to global warming, based on what people implications of global warming. The Corps of know today. In some cases, these assessments Engineers has recently announced that it intends to formally consider the costs and benefits of estimate the impacts of sea level rise in future alternative responses; in others, a qualitative analysis feasibility studies and environmental impact is sufficient to reach a conclusion. statements for coastal projects. (Baldwin, Volume J, discusses including climate change as a Strategic assessments would be good consideration in environmental impact statements.) investments for almost any organization whose activities are sensitive to climate or sea level and Program-Oriented Assessments whose decisions have outcomes stretching over periods of 30 years or longer. In many cases, these Agencies with many potentially vulnerable studies can use existing analytical tools and activities may need programwide assessments. In consequently be relatively inexpensive. If they some cases, the combined impact of climate change reveal that action today is worthwhile, the savings can be summarized by a single variable, such as from such action may be orders of magnitude flood insurance claims. On the other hand, many greater than the cost of the studies. Even if they agencies, such as the TVA, the Corps of Engineers, show that no action is necessary, many organizations and EPA, have programs that face several impacts, will find it useful to know that their projects are not each of which must be examined separately. vulnerable, and the studies would contribute to society's understanding of the impacts of global Problem-Oriented Assessments warming. These studies are sometimes necessary because These assessments can be conducted as project-oriented studies lack a mandate to examine decision-oriented analyses (e.g., supplements to broader implications. Utility companies, for ongoing evaluations of proposed projects) or as example, may want to consider the implications of special studies focusing on particular programs or increased demand due to warmer temperatures. particular problems; Table 19-1 lists examples of Moreover, problems that are explicitly the each type. responsibility of no one while implicitly the responsibility of several different groups could be Decision-Oriented Assessments beyond the scope of program-oriented assessments. For example, the combined impact of farm closures The most cost-effective strategic assessments and forest dieback raises land-use questions that are those conducted as a routine part of the would be outside the responsibility of any single evaluation of ongoing projects. Because they are organization. oriented toward a specific near-term decision, they are not likely to be ignored. Moreover, their cost is Criteria for Choosing a Strategy often minimal because they supplement existing studies and therefore have little overhead. For Strategic assessments can objectively identify example, once a consultant has developed a the implications of climate change and possible hydrologic model for a levee or dam, examining the responses, but picking the "best" response will 390 Preparing for Climate Change Table 19-1. Examples of Strategic Assessments Decisionmaker Question Decision-Oriented Home buyers Is the buyer willing to accept long-term risk of erosion and flooding? Forestry companies Are appropriate species being planted? If so, when would a shift be necessary? Farmers Would a new well be even more useful if climate changed? Utility companies Is the size of a proposed powerplant optimal given projected climate change? City engineers Should new drainage facilities be designed with extra margin for sea level rise and possibly increased rainfall? Water resource agencies Is the dam designed properly? Would its benefits be different? Federal agencies Would sea level rise or climate change significantly alter the environmental developing environmental impacts of a project? impact statements Program-Oriented Research directors For which impacts can we develop a solution? What would be the costs of the research and the potential benefits of anticipated solutions? Utility companies Does system capacity need to be expanded? If not, when would expansion be necessary? Flood insurance programs By how much would insurance claims increase? Does expanding the program to include erosion increase the impact of climate change? Agricultural planners Do current farm programs help or hinder the adjustments climate change might require? Public health agencies Would climate change increase the incidence of malaria and other tropical diseases in the United States? Air pollution regulatory Should current regulatory approaches be supplemented with incentive agencies systems, new chemicals, or relocation policies? 391 Chapter 19 Table 19-1. Examples of Strategic Assessments (continued) Decisionmaker Question Problem-Oriented Natural resource Do we need a program to aid the survival of forests and other terrestrial agencies ecosystems? Federal and state Which options would ensure long-term survival of Louisiana's coastal agencies wetlands? Wetland protection How do we ensure that wetlands can migrate as sea level rises? agencies Canada and the How do we manage changes in levels of the Mississippi River and Great United States Lakes? State coastal zone Would the state provide necessary funds to hold back the sea on barrier agencies and barrier islands? If not, would the town bear the cost of retreat? Are current island communities erosion and flood programs consistent with long-term response? Water resource What should be done to address increased salinity in estuaries? agencies Air pollution Will climate change alter the results of current air-pollution strategies? agencies Publicutility Should power companies be building extra capacity for increasing demand? commissions sometimes be a subjective decision based on a Economic Efficiency: Are the benefits greater number of criteria: than the costs? Flexibility: Is the strategy reasonable for the Profitability: Does the investment provide a entire range of possible changes (including no return greater than alternative investments, change) in temperature, precipitation, and sea i.e., greater than the "discount rate"? level? Political Feasibility: Is the strategy acceptable Urgency: Would the strategy be successful if to the public? implemented today but fail if implementation were delayed 10 or 20 years? Health and Safety: Would the proposed strategy increase or decrease the risk of Low Cost: Can the strategy be implemented disease or injury? with a negligible investment today? Legal and Administrative Feasibility: Can Irreversibility: Would failure to adopt a existing organizations implement the strategy strategy result in irreversible loss of a under existing law? resource? Equity: Would implementing (or failing to Consistency: Does the policy support other implement) the strategy impose unfair costs on national, state, community, or private goals? some regions or on a future generation? 392 Preparing for Climate Change Environmental Quality: Would the strategy in monetary terms. Many decisionmakers do not maintain clean air and water or help natural feel comfortable with economic estimates of the systems survive? value of a lost human life, unique cultural resource, or endangered species. Although economic theory Private versus Public Sector: Does the provides a procedure (discounting) for comparing strategy minimize governmental interference present and future costs, it provides less guidance with decisions best made by the private on how much wealth and how many unsolved sector? problems one generation should pass along to future generations. Although it provides tools for assessing Unique or Critical Resources: Would the risk and uncertainty, economic theory does not strategy protect against the risk of losing specify the extent to which society should be risk- unique environmental or cultural resources? averse. Because there is no objective formula for addressing these types of issues, responses are more The highest priorities would generally be likely to be based on intuitive judgment and on what actions that meet the criteria of flexibility, urgency, is broadly acceptable to the public. irreversibility, and low cost, because they inherently address the major obstacles encountered in preparing for global warming: (1) flexible policies EXAMPLE RESPONSES FOR meet the challenge of uncertainty because they are appropriate regardless of how the climate eventually ADAPTING TO GLOBAL changes; (2) although analytical techniques WARMING substantially discount the benefits of taking action sooner rather than later, delaying action is not a This chapter presents a variety of example viable option when the urgency criterion is met; (3) responses rather than a single integrated strategy irreversible losses can be avoided only by because the process of adapting to climate change anticipating a problem; and (4) low-cost options are would be relatively decentralized. Although the always easiest to implement. various impacts would not be completely independent of each other, responses to one type of Nevertheless, these responses would not impact in one region generally could be always be sufficient to address the implications of implemented regardless of whether strategies are climate change. More comprehensive solutions implemented to address other types of impacts in would often involve measures with more significant other regions. The need to protect California's costs that might prove, in retrospect, to have been water supplies, for example, would be largely unnecessary if climate does not change as projected. independent of the impact of global warming on The costs of not acting may still be great enough to southeastern forests, midwestern agriculture, justify such actions, but decisionmakers would have mid-Atlantic barrier islands, and the level of the to carefully weigh the various tradeoffs. Great Lakes. To a large degree, the procedures for doing For purposes of this discussion, approaches for so have already been developed and applied. Most adapting to global warming can be broadly divided corporations and many government agencies into four categories, three of which require a conduct profitability or cost-benefit analyses. If the response before the climate changes: principal costs and benefits of a strategy can be quantified in monetary terms, economic theory No immediate action is necessary if least-cost provides a rigorous procedure for making tradeoffs solutions could be implemented using existing between present and future costs, and for technology and institutions as the problem considering uncertainty, profitability, and most of emerges. the other criteria. Anticipatory action is appropriate where taking Nevertheless, subjective assessments are concrete actions today would avert irreversible necessary when the impacts cannot be readily valued and expensive costs. 393 Chapter 19 Planning is appropriate where we do not need in temperature and water availability. If the climate to physically change what we are doing of one state gradually comes to resemble the immediately, but where we need to change the climate currently experienced in another state, "rules of the game" now, so that people can farmers in the former state may gradually begin to respond to new information in a way that plant the crops currently grown in the latter. But furthers social goals. there is no advantage in switching crops today. Research and education are appropriate in Anticipatory Action cases where decades would® be required to develop solutions and to train people to Although many responses will not be necessary implement them, or where uncertainties must for a few decades, studies have identified a number be reduced before the appropriate action can of instances in which physical responses to global be identified. warming are appropriate even today. These circumstances fall broadly into two categories: (1) We discuss each of these categories in turn. incorporating awareness of global warming into long-term projects that are already under way, No Immediate Action where climate change must be addressed either now or not at all; and (2) taking actions today that, The urgency of responding to climate change without climate change, would not be necessary depends not only on the severity of a potential until later, if at all. impact but also on the extent to which taking action today would diminish the ultimate cost of adaptation Modifying Ongoing Projects to Consider Climate or allow us to avoid problems that would be Change unavoidable if we waited before taking action. Even where the impacts of climate change would be The rationale for incorporating global warming severe, if the solution to a problem is well defined into current decisions is that the outcome of and can be implemented quickly, or if no known projects initiated today will be altered by changes solution would substantially mitigate the problem, in temperature, rainfall, sea level, or other impacts immediate action is not necessary (although in the of global warming. For many long-term projects, latter case, research may be appropriate). Two factoring climate change into the initial design is examples follow. economically efficient because the failure to do so would risk premature failure of the project, while Reservoir Operating Rules the cost of doing so would be only a few percent of the total project cost. Because consideration of The decision rules that govern the timing and global warming would also ensure that projects are magnitudes of water releases are generally based adequate to address current climate variability and on historic climate variability. For example, if the trends in sea level, such modifications may prove to flood season is March to May and droughts are be worthwhile investments even if the anticipated from July to September, reservoir managers climate change does not occur, as described in the typically lower the water levels by the end of following examples. Thus, these actions can satisfy February to ensure adequate flood control capacity, the criteria of flexibility, urgency, irreversibility, and and they allow the levels to rise in June to ensure low cost. adequate water in case of a drought. If global warming advanced the flood season by one month, Street Drains managers could eventually shift the schedule of water releases. But since such modifications could Consider the replacement of a century-old be implemented quickly, there is no advantage in street drain. If designed for the current 5-year modifying the schedule until the climate changes. storm, such a system might be insufficient with a 10% increase in the severity of the design storm or Choice of Crops a 1-foot rise in sea level, necessitating a completely new system long before the end of the project's The differences among crops grown in various useful life. On the other hand, installing slightly regions of the country result largely from differences larger pipes to accommodate climate change might 394 Preparing for Climate Change cost only an additional 5%. In such a case, the technical advisory panel that the barrier would designing for changes in climate might prove to be become necessary; once that eventuality was worthwhile if these changes occurred; even if they generally recognized, the consensus was that the did not occur, benefits would be realized because project should go forward. the system would provide additional protection during the more severe 10-year storm. (For Constructing a project today solely because of additional examples, see Chapter 7: Sea Level Rise, the greenhouse effect requires more certainty than and Chapter 13: Urban Infrastructure.) incorporating climate change into the design of a project that would be undertaken anyway, primarily Commercial Forests for two reasons: (1) undertaking a new project requires the legislature or the board of directors to Because some commercial tree species live as initiate major appropriations rather than approve long as 70 years before being harvested, small increases in the cost of a project already consideration should be given to modifying the approved; and (2) because it is not motivated by the locations of commercial forests and types of species need to address current problems, the project can planted to account for global warming. For be delayed until there is more certainty. Even if example, some types of Douglas-firs need at least a decisionmakers are sufficiently certain of future few weeks of cold winter temperatures to produce impacts, they do not have to initiate the project seeds. Forestry companies currently concentrate today unless the time expected to pass before the planting efforts at the mountain bases, from which impacts occur is not much greater than the time logs can be most readily transported. However, if required to design, approve, and build the project temperatures rise, the forests there may no longer intended to prevent those impacts. Thus, only near- produce young firs to replace the old. Thus, it term impacts of global warming and those whose might be reasonable to begin planting farther up the solution would take several decades to implement mountain or in a colder region of the country. require remedial action today. Two examples follow. A shift from long-lived species vulnerable to climate change to species having less vulnerability or River Deltas shorter growing cycles may also be appropriate. If two species are equally profitable today but one The loss of wet and dry land in the Mississippi would fare much better if climate changed, shifting River Delta in coastal Louisiana is one example of to the latter species would involve little risk and how global warming could alter the timing of a might substantially help long-term profits. Shifting project (see Chapter 16: Southeast). If current to a species whose life cycle is only 20 years would trends continue, most of the delta will be lost by enable harvests to take place before the climate 2100. But if sea level rise accelerates, this can occur changes enough to adversely affect growth, and as soon as 2050. The immediacy of the problem is would make it easier to respond to climate change greater than these years suggest, because the loss of as it occurs (see Chapter 5: Forests). land is steady. Assuming the additional loss of land to be proportional to sea level rise, half the delta Undertaking New Projects Primarily Because of could be lost by 2030, with some population centers Future Climate Change threatened before then. In a few cases, where authorities are already Whether or not sea level rise accelerates, the contemplating public works for which the economic majority of the delta can survive in the long run justification is marginal, the prospect of climate only if society restores the natural process by which change might encourage decisionmakers to proceed the Mississippi River once deposited almost all of its today with such projects. For example, a storm sediment in the wetlands. Because billions of surge that almost flooded London during the 1950s dollars have been invested over the last 50 years in led the Greater London Council to develop plans to flood-control and navigation-maintenance projects build a movable barrier across the Thames River. that could be rendered ineffective, restoring natural Although many questioned whether the barrier was sedimentation would cost billions of dollars and worth building, steadily rising flood levels (1 foot could take 20 years or longer. Because of the wide every 50 years for the past 5 centuries) convinced variety of interests that would be affected and the 395 Chapter 19 large number of options from which to choose, Land purchases for protecting ecosystems have another 10 to 20 years could pass from the time the two important limitations. First, they would almost project was authorized until construction began. certainly be inadequate to address all the species migration that might be required by climate change: Thus, if sea level rise accelerates according to protecting coastal wetlands would require current projections and a project were initiated purchasing most of the nation's coastal lowlands, today, about half of the delta would remain when and many types of terrestrial species would have to the project was complete; however, if the project shift by hundreds of miles. Second, land purchases were authorized in the year 2000, 60 to 70% might do not handle uncertainty well: if temperatures, be lost before it was complete. By contrast, if sea rainfall, or sea level change more than anticipated, level rise does not accelerate, the two the land purchased will eventually prove to be implementation dates might imply 25% and 35% insufficient. losses of coastal wetlands. Planning: Changing the Rules of the Undertaking a project today satisfies the Game flexibility criterion, because even current trends imply that something eventually must be done. Because a failure to act soon could result in an Although concrete action in response to irreversible loss of much of the delta, it also satisfies climate change is necessary today for only a few the urgency criterion. types of problems, defining the "rules of the game" may now be appropriate for a much wider class of Purchase of Land problems. Doing so increases flexibility: if climate changes, we are better prepared; if it does not change, preparation has cost us nothing. Another Purchasing land could keep options open for advantage of this type of long-range planning is water resource management and wetlands that reaching a consensus on what is fair is easier protection. In regions where climate becomes when no one is immediately threatened. Moreover, drier, additional reservoirs may become necessary. However, because accurate forecasts of regional such planning reduces risk to investors: although climate change are not yet possible, water managers they still face uncertainty regarding the timing and in most areas cannot yet be certain that they will magnitude of climate change, planning can prevent need more dams. Even in areas such as California that uncertainty from being compounded by where dams will probably be required, these will not uncertainty regarding how the government will have to be built for decades. Nevertheless, it may respond. Two examples in which changing the rules of the game might be appropriate follow. make sense to purchase the necessary land today. Otherwise, the most suitable sites may be Land Use developed, making future reservoir construction more expensive and perhaps infeasible. A number of potential reservoir sites have been protected by The potential consequences of global warming creation of parks and recreation areas, such as suggest that it may already be appropriate to guide development away from areas where it could Tocks Island National Park on the Delaware River. conflict with future environmental quality or public Federal, state, and local governments often safety. This can be done through master plans, laws purchase land to prevent development from and regulations, and revisions of ownership rights. Land use is generally regulated by local encroaching on important ecosystems. Particularly governments and planning commissions, with state in cases where ecosystem shifts are predictable, such governments also playing a role in some areas. as the landward migration of coastal wetlands, it may be worthwhile to purchase today the land onto A primary rationale for most local land-use which threatened ecosystems are likely to migrate. planning is that by themselves, real-estate markets Even where the shifts are not predictable, expanding do not always produce economically efficient or the size of refuges could limit their vulnerability socially desirable outcomes, because people do not (see Chapter 8: Biodiversity). bear all the costs or reap all the benefits from their actions. The uses to which people put their 396 Preparing for Climate Change property often can have significant impacts on other need to migrate; if temperature, rainfall, or sea level property owners and the environment. Because change more than expected, zoning provides only zoning and other land-use restrictions are usually temporary protection. implemented long before anyone would want to undertake the prohibited activities, people have time Flexible Planning: Allowing the Market to Decide to plan their activities around the constraints. If people know the rules of the game well in advance, The rationale for these mechanisms is that those who want the option of subdividing their preventing development is inefficient; in some cases property or clearing a forest buy land where these developing a property might be worthwhile even if activities are permissible, and those who want it would subsequently have to be abandoned. property in an area where such activities will not Flexible planning has the desirable feature of take place buy land where the activities are minimizing governmental interference with private prohibited. Thus, in the long run, planning helps decisions. For example, the overall constraint of maintain environmental quality while imposing few keeping natural shorelines is set by the government, costs that individuals could not avoid by buying but the market decides whether nearby property is property elsewhere. still worth developing given that constraint. If the effects of climate change do not materialize, the The institutional capabilities of planning are government has not unnecessarily prevented well suited for addressing environmental impacts of development (satisfying the low-cost criterion). climate change when the direction of the impact is Most importantly, these measures do not require a known. The example of coastal wetland loss precise determination of how much climate will (outside Louisiana) has been extensively examined change, and thus satisfy the flexibility criterion. in the literature; many of the same principles would also apply to shifts in forests, interior wetlands, With this situation in mind, the State of Maine changing water levels in the Great Lakes, and has recently issued regulations stating that structures keeping land vacant for reservoirs. would have to be removed to allow wetlands to migrate inland in response to sea level rise. South A possible goal of land-use planning would be Carolina has recently enacted legislation to to ensure that development does not block substantially curtail construction of bulkheads. migration of ecosystems or preclude construction of Because these rules do not interfere with the use of a dam. Without planning, the land could be vacated property for the next several decades, they have a only by requiring abandonment with relatively little minimal impact on property values, and thus do not advance notice, which would require compensation deprive people of their property. The major (except for the case of coastal wetlands in states limitation of this approach is that it may be too where property owners do not currently have the flexible: if sea level rise begins to require a right to erect shore-protection structures). Planning large-scale abandonment, a state or local measures can either prevent development through government may find it difficult to resist pressure zoning (or purchase of land, discussed above), or set to repeal the rule. the basic social constraint that ecosystems will be able to migrate, while allowing the market to decide An alternative that avoids the risk of whether or not development should proceed given backsliding is to modify conventions of property this constraint. ownership. One example would be long-term leases that expire 50 to 100 years hence or when high tide Preventing Development: Zoning rises above a property's elevation. This approach, which has been applied to Long Island, allows the The most common tools for directing land use market to explicitly incorporate its assessment of sea are master plans and the zoning that results from level rise into its valuation of the leases. Although them. Zoning to ensure that land is available for a leaseholders have requested no-cost extensions on dam would be similar to zoning to keep land their leases when they expire, local governments available for a freeway. For protecting ecosystems, generally have found enforcing the provisions of however, zoning has the same problem as land leases easier than enforcing regulations requiring purchases: it has to be based on a particular people to abandon property. Moreover, this assumption regarding how far the ecosystem will approach can be implemented by the private sector; 397 Chapter 19 for example, a conservancy willing to lease the land Research and Development back to developers for 99 years might be able to buy lowlands inexpensively (see Chapter 7: Sea Level Research and development expenditures can Rise). often be economically justified in cases where other responses cannot. Most of the impacts of climate Water Allocation change at least theoretically could be mitigated, but in many cases, effective solutions have not yet been Particularly in the Southwest, the nation's developed. Like strategic assessments, research is water supply infrastructure is guided by policies as valuable as the savings it makes possible. embedded in contracts and laws that prescribe who gets how much water. Many of these rules are not Research is also one of the major vehicles by economically efficient; water is wasted because of which one generation improves life for succeeding rules that do not allow people with too much water generations. Even if the economic efficiency of to sell it to people with too little. The equity of taking action to mitigate impacts of climate change these formulas is often sensitive to climate; during cannot be demonstrated, some policymakers might wet periods, everyone may receive plenty, but in dry find it equitable for this generation to provide periods some get enough while others get none. solutions to accompany the problems we pass on to the next generation. To a large degree, the means by which the impact of climate change might be reduced are Table 19-2 lists a number of research questions already being advocated to address current climate and applications that would assist adaptation. variability and potential supply shortages due to However, for the most part, strategic assessments population growth. These measures include have not been undertaken to determine the cost and legalizing water markets; curtailing federal subsidies, probability of developing solutions or the magnitude which lead to waste by keeping prices artificially of potential savings that might result, so it is difficult low; and modifying allocation formulas (see Chapter to be certain that the research would benefit society. 9: Water Resources). The most notable exception is improvement in estimates of future climate change; for virtually Nevertheless, the changes required by global every impact examined in this report, the relevant warming may be different in one crucial aspect: the decisionmakers have told EPA that improved effective date of any rule changes. Because the climate projections are critical for developing most severe changes in rainfall from the greenhouse responses. (For more details on necessary research, effect may still be decades in the future, the see Chapter 18: Research Needs.) problem can be addressed even if the effective date is not until 2020. This situation, however, may Education enhance the political feasibility of instituting a rational response today, since no one need be Efforts to prepare for climate change can be immediately threatened. By contrast, if planning is only as enlightened as the people who must carry deferred another 20 years, the impacts of climate them out. Education will be a critical component change may become too imminent for potential of any effort to address the greenhouse effect losers to agree to the necessary changes. because (1) decisionmakers in various professions will need to routinely consider the implications of Research and Education: Increasing Our global warming; and (2) an informed citizenry will Understanding be necessary for the public to support the public policy and institutional changes that may be Although a particular problem may not required. Governments will almost certainly have a require solutions for a few decades, society should major role. begin preparing now. In some cases, we are decades away from having viable solutions or the To factor global warming into their decision public awareness necessary to reach a consensus. processes, people will need information about We now examine two vehicles for expanding our changes in climate variables, the resulting effects, knowledge: research and education. and techniques for mitigating the impacts. Federal 398 Preparing for Climate Change Table 19-2. Example Research Problems and Applications Research problem Application Synergistic impacts of CO2, climate change, Shifts in mix of trees and crops, drought-tolerant crops and air pollution on plants Shifts in habitats of birds, fish, and Restoration ecology: rebuilding ecosystems that are lost land animals Ability of wetlands and coral reefs to Mechanisms to accelerate vertical growth keep up with sea level changes Erosion of beaches due to climatology and More efficient placement of sand when beaches are restored sea level changes Ability of alternative plant strains to Development of heat- and drought-resistant crops tolerate harsh climate Magnitude of changes in global sea level All responses to global warming and regional climate Shifts in pests due to climate change Development of integrated pest management programs and better background data for groundwater protection policies Shifts in microorganisms that currently Long-term water supply planning diminish water quality in tropical areas and state agencies have already sponsored large the issue to the public at large so that the various conferences on sea level rise each year since 1983; options can be fully considered. To a large degree, coastal engineers and policymakers are increasingly the news media and school systems will be considering accelerated sea level rise in land-use responsible for explaining the issue to people. decisions and the design of public works. This Nevertheless, governments can support these process is now beginning to unfold in the fields of institutions by sponsoring public meetings, issuing utility planning and water-resource management, press releases, and perhaps most important, and may emerge in other areas. translating the results of its technical studies into brochures and reports that are accessible to Because climate change could require major reporters, teachers, and the general public. public policy initiatives, governments must explain 399 AUTHORS Joseph J. Bufalini USEPA - Atmospheric Research and Exposure Assessment Laboratory - Research Triangle Park, North Carolina Lauretta M. Burke The Bruce Company Margaret M. Daniel The Bruce Company Robert L. DeVelice USEPA - Environmental Research Laboratory - Corvallis, Oregon Eugene C. Durman USEPA Office of Policy, Planning and Evaluation Peter L. Finkelstein USEPA - Atmospheric Research and Exposure Assessment Laboratory - Research Triangle Park, North Carolina Anthony Janetos USEPA - Office of Research and Development Roy Jenne National Center for Atmospheric Research Ross A. Kiester USEPA - Environmental Research Laboratory - Corvallis, Oregon George A. King USEPA - Environmental Research Laboratory - Corvallis, Oregon Kenneth P. Linder ICF, Inc. Janice A. Longstreth ICF, Inc. Linda O. Mearns National Center for Atmospheric Research Ted R. Miller The Urban Institute Mark W. Mugler Apogee Research, Inc. Ronald P. Neilson USEPA - Environmental Research Laboratory - Corvallis, Oregon Alan Robock University of Maryland Cynthia Rosenzweig Columbia University/Goddard Institute for Space Studies William E. Riebsame University of Colorado Michael C. Rubino Apogee Research, Inc. Joel B. Smith USEPA Office of Policy, Planning and Evaluation 401 Dennis A. Tirpak USEPA Office of Policy, Planning and Evaluation James G. Titus USEPA Office of Policy, Planning and Evaluation Jack K. Winjum USEPA - Environmental Research Laboratory - Corvallis, Oregon Robert C. Worrest USEPA - Environmental Research Laboratory - Corvallis, Oregon 402 CONTRIBUTING INVESTIGATORS AND PROJECTS Authors: Adams, Richard M., J. David Glyer, and Bruce A. McCarl Institution: Oregon State University and Texas A & M University Title: The Economic Effects of Climate Change on U.S. Agriculture: A Preliminary Assessment. Appendix: Volume C - Agriculture Authors: Allen, Richard C., and Francis N. Gichuki Institution: Utah State University Title: Effects of Projected CO2-Induced Climate Changes on Irrigation Water Requirements in the Great Plains States (Texas, Oklahoma, Kansas, and Nebraska). Appendix: Volume C - Agriculture Author: Assel, Raymond, A. Institution: Great Lakes Environment Research Laboratory Title: Impact of Global Warming on Great Lakes Ice Cycles. Appendix: Volume A - Water Resources Author: Baldwin, Malcolm F. Institution: Environmental Management Support, Inc. Title: Applicability of Federal Long-Range Planning and Environmental Impact Statement Processes to Global Climate Change Issues. Appendix: Volume J - Policy Authors: Blumberg, Alan F., and Dominic M. DiToro Institution: HydroQual, Inc. Title: The Effects of Climate Warming on Lake Erie Water Quality. Appendix: Volume A - Water Resources Authors: Botkin, Daniel B., Robert A. Nisbet, and Tad E. Reynales Institution: University of California, Santa Barbara Title: Effects of Climate Change on Forests of the Great Lake States. Appendix: Volume D - Forests Authors: Byron, Earl R., Alan Jassby, and Charles R. Goldman Institution: University of California, Davis Title: The Effects of Global Climate Change on the Water Quality of Mountain Lakes and Streams. Appendix: Volume E - Aquatic Resources Authors: Changnon, Stanley A., Jr., Steven Leffler, and Robin Shealy Institution: Illinois State Water Survey and University of Illinois Title: Impacts of Extremes in Lake Michigan Levels Along Illinois Shorelines: Low Levels. Appendix: Volume H - Infrastructure 403 Authors: Croley, Thomas E., II, and Holly C. Hartmann Institution: Great Lakes Environment Research Laboratory Title: Effects of Climate Changes on the Laurentian Great Lakes Levels. Appendix: Volume A - Water Resources Author: Davis, Owen K. Institution: University of Arizona Title: Ancient Analogs for Greenhouse Warming of Central California. Appendix: Volume D - Forests Author: Dudek, Daniel J. Institution: Environmental Defense Fund Title: Climate Change Impacts upon Agriculture and Resources: A Case Study of California. Appendix: Volume C - Agriculture Author: Easterling, William E. Institution: Resources for the Future Title: Farm-Level Adjustments by Illinois Corn Producers to Climate Change. Appendix: Volume C - Agriculture Authors: Glantz, Michael H., Barbara G. Brown, and Maria E. Krenz Institution: National Center for Atmospheric Research Title: Societal Responses to Regional Climate Change: Forecasting by Analogy. Appendix: Volume J - Policy Author: Haile, Daniel G. Institution: U.S. Department of Agriculture, Agriculture Research Service - Gainesville Title: Computer Simulation of the Effects of Changes in Weather Patterns on Vector-Borne Disease Transmission. Appendix: Volume G - Health Author: Hains, David K. Institution: C.F. Hains, Hydrologist, Inc. Title: Impacts of Global Warming on Runoff in the Upper Chattahoochee River Basin. Appendix: Volume A - Water Resources Authors: Johnson, Howard L., Ellen J. Cooter, and Robert J. Sladewski Institution: University of Oklahoma Title: Impacts of Climate Change on the Transport of Agricultural Chemicals Across the USA Great Plains and Central Prairie. Appendix: Volume C - Agriculture Authors: Josselyn, Michael, and John Callaway Institution: San Francisco State University Title: Ecological Effects of Global Climate Change: Wetland Resources of San Francisco Bay. Appendix: Volume E - Aquatic Resources 404 Author: Kalkstein, Laurence S. Institution: University of Delaware Title: The Impact of CO₂ and Trace Gas-Induced Climate Changes upon Human Mortality. Appendix: Volume G - Health Authors: Keith, Virgil F., J. Carlos DeAvila, and Richard M. Willis Institution: Engineering Computer Optecnomics, Inc. Title: Effect of Climatic Change on Shipping within Lake Superior and Lake Erie. Appendix: Volume H - Infrastructure Author: Leatherman, Stephen P. Institution: University of Maryland Title: National Assessment of Beach Nourishment Requirements Associated with Accelerated Sea Level Rise. Appendix: Volume B - Sea Level Rise Authors: Lettenmaier, Dennis P., Thian Yew Gan, and David R. Dawdy Institution: University of Washington Title: Interpretation of Hydrologic Effects of Climate Change in the Sacramento-San Joaquin River Basin, California. Appendix: Volume A - Water Resources Authors: Linder, Kenneth P., and Mark R. Inglis Institution: ICF, Inc. Title: The Potential Impacts of Climate Change on Regional and National Demands for Electricity. Appendix: Volume H - Infrastructure Author: Livingston, Robert J. Institution: Florida State University Title: Projected Changes in Estuarine Conditions Based on Models of Long-Term Atmospheric Alteration. Appendix: Volume E - Aquatic Resources Authors: Longstreth, Janice, and Joseph Wiseman Institution: ICF/Clement Associates, Inc. Title: The Potential Impact of Climate Change on Patterns of Infectious Disease in the United States. Appendix: Volume G - Health Authors: Magnuson, John J., David K. Hill, Henry A. Regier, John A. Holmes, J. Donald Meisner, and Brian J. Shuter Institution: University of Wisconsin, University of Toronto, and Ontario Ministry of Natural Resources Title: Potential Responses of Great Lakes Fishes and their Habitat to Global Climate Warming. Appendix: Volume E - Aquatic Resources Author: McCormick, Michael J. Institution: Great Lakes Environment Research Laboratory Title: Potential Climate Changes to the Lake Michigan Thermal Structure. Appendix: Volume A - Water Resources 405 Authors: Mearns, Linda O., S.H. Schneider, S.L. Thompson, and L.R. McDaniel Institution: National Center for Atmospheric Research Title: Analysis of Climate Variability in General Circulation Models: Comparison with Observations and Changes in Variability in 2xCO2 Experiments. Appendix: Volume I - Variability Authors: Meo, Mark, Thomas E. James, Jr., Steve Ballard, Lani L. Malysa, Robert E. Deyle, and Laura A. Wilson Institution: University of Oklahoma Title: Policy Implications of Global Climate Change Impacts upon the Tennessee Valley Authority Reservoir System, Apalachicola River, Estuary, and Bay, and South Florida. Appendix: Volume J - Policy Authors: Miller, Barbara A., and W. Gary Brock Institution: Tennessee Valley Authority Title: Potential Impacts of Climate Change on the Tennessee Valley Authority Reservoir System. Appendix: Volume A - Water Resources Authors: Morris, Ralph E., Mike W. Gery, Mei-Kao Liu, Gary E. Moore, Christopher Daly, and Stanley M. Greenfield Institution: Systems Applications, Inc. Title: Sensitivity of a Regional Oxidant Model to Variations in Climate Parameters. Appendix: Volume F - Air Quality Authors: Overpeck, Jonathan T., and Patrick J. Bartlein Institution: Lamont-Doherty Geological Observatory and University of Oregon Title: Assessing the Response of Vegetation to Future Climate Change: Ecological Response Surfaces and Paleoecological Model Validation. Appendix: Volume D - Forests Authors: Park, Richard A., Manjit S. Trehan, Paul W. Mausel, and Robert C. Howe Institution: Butler University and Indiana State University Title: The Effects of Sea Level Rise on U.S. Coastal Wetlands. Appendix: Volume B - Sea Level Rise Authors: Peart, Robert M., James W. Jones, R. Bruce Curry, Ken Boote, and L. Hartwell Allen Jr. Institution: University of Florida Title: Impact of Climate Change on Crop Yield in the Southeastern USA: A Simulation Study. Appendix: Volume C - Agriculture Authors: Penner, Joyce E., Peter S. Connell, Donald J. Wuebbles, and Curtis C. Covey Institution: Lawrence Livermore National Laboratory Title: Climate Change and Its Interactions with Air Chemistry: Perspective and Research Needs. Appendix: Volume F - Air Quality 406 Authors: Ray, Daniel K., Kurt N. Lindland, and William J. Brah Institution: The Center for the Great Lakes Title: Effects of Global Warming on the Great Lakes: The Implications for Policies and Institutions. Appendix: Volume J - Policy Author: Riebsame, William E. Institution: University of Colorado Title: Climate Change Perceptions Among Natural Resource Decision-Makers: The Case of Water Supply Managers. Appendix: Volume J - Policy Authors: Riebsame, William E., and Jeffrey W. Jacobs Institution: University of Colorado Title: Climate Change and Water Resources in the Sacramento-San Joaquin Region of California: Policy Adjustment Options. Appendix: Volume J - Policy Authors: Rind, David, R. Goldberg, and R. Ruedy Institution: Goddard Institute for Space Studies, Columbia University, and Sigma Data Service Corporation Title: Change in Climate Variability in the 21st Century. Appendix: Volume I - Variability Authors: Ritchie, Joe T., B.D. Baer, and T.Y. Chou Institution: Michigan State University Title: Effect of Global Climate Change on Agriculture: Great Lakes Region. Appendix: Volume C - Agriculture Author: Rose, Elise Institution: Consultant Title: Direct (Physiological) Effects of Increasing CO2 on Crop Plants and Their Interactions with Indirect (Climatic) Effects. Appendix: Volume C - Agriculture Author: Rosenzweig, Cynthia Institution: Columbia University/Goddard Institute for Space Studies Title: Potential Effects of Climate Change on Agricultural Production in the Great Plains: A Simulation Study. Appendix: Volume C - Agriculture Authors: Schmidtmann, Edward T., and J.A. Miller Institution: U.S. Department of Agriculture, Agriculture Research Service - Beltsville, Maryland Title: Effect of Climatic Warming on Populations of the Horn Fly, with Associated Impact on Weight Gain and Milk Production in Cattle. Appendix: Volume C - Agriculture 407 Author: Schuh, G. Edward Institution: University of Minnesota Title: Agricultural Policies for Climate Changes Induced by Greenhouse Gases. Appendix: Volume C - Agriculture Authors: Sheer, Daniel P., and Dean Randall Institution: Water Resources Management Inc. Title: Methods for Evaluating the Potential Impacts of Global Climate Change: Case Studies of the State of California and Atlanta, Georgia. Appendix: Volume A - Water Resources Authors: Stem, Edgar, Gregory A. Mertz, J. Dirck Strycker, and Monica Huppi Institution: Tufts University Title: Changing Animal Disease Patterns Induced by the Greenhouse Effect. Appendix: Volume C - Agriculture Authors: Stinner, Benjamin R., Robin A.J. Taylor, Ronald B. Hammond, Foster F. Purrington, David A. McCartney, Nick Rodenhouse, and Gary Barrett Institution: Ohio Agricultural Research and Development Center and Ohio State University Title: Potential Effects of Climate Change on Plant-Pest Interactions. Appendix: Volume C - Agriculture Authors: Titus, James G., and Michael S. Greene Institution: U.S. Environmental Protection Agency Title: An Overview of the Nationwide Impacts of Sea Level Rise. Appendix: Volume B - Sea Level Rise Authors: Urban, Dean L., and Herman H. Sheer Institution: University of Virginia Title: Forest Response to Climate Change: A Simulation Study for Southeastern Forests. Appendix: Volume D - Forests Authors: Walker, Christopher J., Ted R. Miller, G. Thomas Kingsley, and William A. Hyman Institution: The Urban Institute Title: Impact of Global Climate Change on Urban Infrastructure. Appendix: Volume H - Infrastructure Authors: Weggel, J. Richard, Scott Brown, Juan Carlos Escajadillo, Patrick Breen, and Edward L. Doheny Institution: Drexel University Title: The Cost of Defending Developed Shorelines Along Sheltered Waters of the United States from a Two Meter Rise in Mean Sea Level. Appendix: Volume B - Sea Level Rise Author: Williams, Philip B. Institution: Philip Williams & Associates Title: The Impacts of Climate Change on the Salinity of San Francisco Bay. Appendix: Volume A - Water Resources 408 Authors: Woodman, James N., and Cari L. Sasser Institution: North Carolina State University Title: Potential Effects of Climate Change on U.S. Forests: Case Studies of California and the Southeast. Appendix: Volume D - Forests Author: Yohe, Gary W. Institution: Wesleyan University Title: The Cost of Not Holding Back the Sea: Phase 1, Economic Vulnerability. Appendix: Volume B - Sea Level Rise Authors: Zabinski, Catherine and Margaret B. Davis Institution: University of Minnesota Title: Hard Times Ahead for Great Lakes Forests: A Climate Threshold Model Predicts Responses to CO2-Induced Climate Change. Appendix: Volume D - Forests 409 ROBERT T STAFFORD. VERMONT. CHAIRMAN OHN H. CHAFEE. RHODE ISLAND LLOYD BENTSEN, TEXAS LAN K. SIMPSON. WYOMING QUENTIN N. BURDICK, NORTH DAKOTA AMES ABDNOR. SOUTH DAKOTA GARY HART, COLORADO TEVE SYMMS, ЮАНО DANIEL PATRICK MOYNIHAN, NEW YORK GORDON HUMPHREY. NEW HAMPSHIRE GEORGE J. MITCHELL. MAINE ETE V DOMENICI. NEW MEXICO MAX BAUCUS. MONTANA DAVE DURENBERGER, MINNESOTA FRANK R. LAUTENBERG. NEW JERSEY United States Senate BAILEY GUARD. STAFF DIRECTOR LEE O. FULLER, MINORITY STAFF DIRECTOR COMMITTEE ON ENVIRONMENT AND PUBLIC WORKS WASHINGTON, DC 20510 September 12, 1986 Mr. Lee Thomas Administrator Environmental Protection Agency Washington, D.C. 20460 Dear Mr. Thomas: The purpose of this letter is to formally request that EPA undertake two studies on climate change due to the greenhouse effect and submit them to Congress no later than March 31, 1988. At the outset, we want to thank you for appearing before the Subcommittee on Environmental Pollution at hearings last June on the problems of global climate change and stratospheric ozone depletion. Your testimony showed a refreshing appreciation for the magnitude of the environmental risks presented by these problems and the need to be exploring incremental actions that can be taken to reduce these risks. As summarized at those hearings and elsewhere, the scientific community appears to have reached agreement that substantial ozone depletion may result from continued use of chlorofluorcarbons (CFC's) and that increases in CFC's and other greenhouse gases are like to produce global climate changes greater than any in man's history. There is a very real possibility that man - through ignorance or indifference, or both - is irreversibly altering the ability of our atmosphere to perform basic life support functions. What is urgently needed now is for us to begin to deal with these issues. They can no longer be treated solely as important scientific questions. First, some actions including limits on CFC's appear warranted in the near term. Second, we need to expand efforts to more fully understand the effects that atmospheric pollution has on the environment and to develop an extensive range of policy options for dealing with the serious global problem of climate change due to the greenhouse effect. This second need has led to our request for two EPA studies. One of the studies we are requesting should examine the health and environmental effects of climate change. This study should include, but not be limited to, the potential impacts on agriculture, forests, wetlands, human health, rivers, lakes and estuaries as well as other ecosystems and societal impacts. This 411 Mr. Lee Thomas September 9, 1986 Page 2 study should be designed to include original analyses to identify and fill in where important research gaps exist, and to solicit the opinions of knowledgeable people throughout the country through a process of public hearings and meetings. The other study should include an examination of the policy options that, if implemented, would stabilize current levels of atmospheric greenhouse gas emissions. This study should address: the need for and implications of significant changes in energy policy, including energy efficiency and development of alternatives to fossil fuel; reductions in the use of CFC's; ways to reduce other greenhouse gases such as methane and nitrous oxides; as well as the potential for and effects of reducing deforestation and increasing reforestation efforts. It should include a series of policy options and recommendations for concrete steps to be taken along with a discussion of the potential effectiveness of each for limiting climate change. Since the United States must take a leadership role in addressing these global problems, the policy options that you develop should include a specific plan for what the United States can do to stabilize its share of greenhouse gas emissions as well as a plan for helping other nations to achieve comparable levels of control. We realize that undertaking this project will be a significant challenge and will require substantial resources. We therefore urge you to immediately direct the necessary funds in both FY-87 and FY-88 to assure that you can comply with our request to promptly conduct these studies. Many of us believe that these are among the most important environmental problems of the next decade. The sooner you can provide recommendations to Congress, the sooner we will be able to provide leadership throughout the world to prevent a pending environmental disaster. Your personal attention and prompt reply to this request will be greatly appreciated. We look forward to working with you on these important environmental problems. Please do not hesitate to contact us for additional guidance and assistance. Sincerely, Grage Natiball George J. Mitchell John John H. Chafee Chefee 412 Mr. Lee Thomas September 9, 1986 Page 3 allent Are f. Albert Gore Robert T. Stafford Max Max Baucus Baucus The Dave Durenberger Diverture Retucl Lealy Patrick J. Leahy Gozdon Humpht 413 EPA United States Environmental Protection Agency (PM-221) Washington, DC 20460 Official Business Penalty for Private Use $300

nara / coll_gb-pod / doc

EPA - Global Climate Change [The Potential for Global Climate Change on the United States 12/89]

These records pertain to Global Climate Change.

ID: 285411055
Type: document

Citation Image

Metadata

id	285411055
sourceUrl	https://catalog.archives.gov/id/285411055
contentType	document
identifierLocal	06683-004
collections	Records of the White House Office of Policy Development (George H. W. Bush Administration) Ed Goldstein Subject Files
iiifBase	https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf
thumbnailUrl	https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf
largeImageUrl	https://s3.us-east-1.amazonaws.com/NARAprodstorage/lz/presidential-libraries/bush/foia/2017/41-bpr-poldev-goldstein-06683-004.pdf
imageCount	1
hasImages	yes
source	import

Additional Source Fields

url	https://catalog.archives.gov/id/285411055
naId	285411055
levelOfDescription	fileUnit
recordType	description
ocrSource	nara-archive

Pages

Page 1 OCR

Survos folio item chat

Ask the Scholar

Source Description

Scholar Source Context