vvEPA
United States
Environmental Protection
Agency
Multi-Model Framework for Quantitative
Sectoral Impacts Analysis
A Technical Report for the Fourth National Climate Assessment
May 2017
-------
FRONT MATTER
FRONT MATTER
CONTRIBUTORS
The Climate Change Impacts and Risk Analysis (CIRA) project is coordinated by the U.S. Environmental
Protection Agency's (EPA), with significant contributions from a number of Federal agencies, including
the Department of Energy (DOE), U.S. Forest Service (USFS), and Centers for Disease Control and
Prevention (CDC); academic experts; and consultants, including Industrial Economics, Inc., Abt
Associates Inc., and RTI International. Support for the report's production was provided by Industrial
Economics, Inc.
Individual Contributors and Modelers1 (listed in alphabetical order):
Ashley Allen (EPA)
Susan C. Anenberg (Env. Health Analytics)
Justin Baker (RTI International)
Chris Barker (U. of California Davis)
Matthew Baumann (Industrial Economics)
Robert Beach (RTI International)
Anna Belova (Abt Associates)
Charlotte Benishek (Industrial Economics)
Victor Bierman Jr. (LimnoTech)
Britta Bierwagen (EPA)
Margaret Black (Industrial Economics)
Brent Boehlert (Industrial Economics)
Alexandra Bothner (Industrial Economics)
Yongxia Cai (RTI International)
Karen Carney (Abt Associates)
Steven C. Chapra (Tufts University)
Paul S. Chinowsky (Resilient Analytics)
Stuart Cohen (National Renewable Energy Lab)
Jefferson Cole (EPA)
Joel Corona (EPA)
Jared Creason (EPA)
Allison Crimmins (EPA)
Pat Dolwick (EPA)
Michael Duckworth (Abt Associates)
Rebecca Eisen (CDC)
Xavier Espinet (Resilient Analytics)
Neal Fann (EPA)
Charles Fant (Industrial Economics)
Josh Graff-Zivin (U. of California San Diego)
Sahil Gulati (Industrial Economics)
Ethan Gutmann (Nat. Ctr. Atmos. Research)
Micah Flahn (CDC)
Ron Flail (Abt Associates)
Petr Havlik (Int'l Inst. Applied Systems Analysis)
Jacob Flelman (Resilient Analytics)
Jim Flenderson (Abt Associates)
Russell Florowitz (Pacific Northwest National Lab)
Fleather Flosterman (Abt Associates)
Gokul Iyer (Pacific Northwest National Lab)
Lesley Jantarasami (EPA)
Russ Jones (Abt Associates)
Michael Kolian (EPA)
John Kim (USFS Pacific Northwest Res. Station)
Patrick L. Kinney (Columbia University)
Peter Larsen (Independent Consultant)
Claire Lay (Abt Associates)
Jia Li (EPA)
Mark Lorie (Abt Associates)
Lindsay Ludwig (Industrial Economics)
Flardee Mahoney (Abt Associates)
Sergey S. Marchenko (U. of Alaska Fairbanks)
Jeremy Martinich (EPA)
Diane M.L. Mas (Fuss & O'Neill)
James McFarland (EPA)
April Melvin (American Assoc. for the Adv. Sci.)
Karen Metchis (EPA)
Dave Mills (Abt Associates)
Naoki Mizukami (Nat. Ctr. Atmos. Research)
1 Specific technical and modeling contributions by these individuals does not necessarily represent endorsement of material appearing in this
Technical Report.
i
-------
ACKNOWLEDGEMENTS
Christopher Moore (EPA)
Philip Morefield (EPA)
Matthew Neidell (Columbia University)
James E. Neumann (Industrial Economics)
Dmitry J. Nicolsky (U. of Alaska Fairbanks)
Christopher Nolte (EPA)
Sara Ohrel (EPA)
Hans W. Paerl (U. of North Carolina)
Stefani Penn (Industrial Economics)
Jason Price (Industrial Economics)
Lisa Rennels (Industrial Economics)
Matthew Rissing (Abt Associates)
Henry A. Roman (Industrial Economics)
Marcus Sarofim (EPA)
Rebecca Schultz (EPA)
Kate Shouse (EPA)
Eric Small (U. of Colorado Boulder)
Tanya Spero (EPA)
Alexis St. Juliana (Abt Associates)
Justin Stein (Abt Associates)
Ken Strzepek (Industrial Economics & MIT)
Nicole Thompson (Industrial Economics)
Hugo Valin (Int'l Inst. Applied Systems Analysis)
Stephanie Waldhoff (Pacific Northwest Nat. Lab)
Chris Weaver (EPA)
Kate R. Weinberger (Brown University)
Brittany Whited (EPA)
Jacqueline Willwerth (Industrial Economics)
Cameron Wobus (Abt Associates)
Raghavan Srinivasan (Texas A&M University)
Xuesong Zhang (Pacific Northwest National Lab)
Pearl Zheng (Abt Associates)
PEER REVIEW
The methods and results of the climate change impacts analyses described herein have been peer
reviewed in the scientific literature. In addition, this Technical Report was peer reviewed by seven
external and independent experts in a process independently coordinated by Eastern Research Group.
EPA gratefully acknowledges the following peer reviewers for their useful comments and suggestions:
Anna Alberini, Mikhail Chester, Helene Margolis, Michael Meyer, Timothy Randhir, Matthias Ruth, and
Susanna T.Y. Tong. The information and views expressed in this report do not necessarily represent
those of the peer reviewers, who also bear no responsibility for any remaining errors or omissions.
Details describing this review, and a comprehensive reference list for the CIRA peer-reviewed literature,
can be found in the Technical Appendix to this report.
RECOMMENDED CITATION
EPA. 2017. Multi-Model Framework for Quantitative Sectoral Impacts Analysis: A Technical Report for
the Fourth National Climate Assessment. U.S. Environmental Protection Agency, EPA 430-R-17-001.
DATA AVAILABILITY
Figures and metadata are being made available in the Global Change Information System
(https://data.globalchange.gov).
ii
-------
TABLE OF CONTENTS
TABLE OF CONTENTS
INTRODUCTION 1
About this Report 1
Interpreting the Results 2
EXECUTIVE SUMMARY 3
MODELING FRAMEWORK 6
1. SCENARIOS AND PROJECTIONS 6
1.1. Selection of Inputs 6
1.2. Projections of Future Climate 18
2. CIRA PROJECT BACKGROUND 22
2.1. Advancements in the CIRA Framework 22
2.2. Metrics 25
2.3. Sources of Uncertainty 27
2.4. Review of Related Literature 30
2.5.Terms and Acronyms Commonly Used in this Report 33
HEALTH 34
3. Air Quality 34
4. Aeroallergens 42
5. Extreme Temperature Mortality 48
6. Labor 54
7. West Nile Virus 60
8. Harmful Algal Blooms 66
9. Domestic Migration 74
INFRASTRUCTURE 79
10. Roads 79
11. Bridges 88
12. Rail 94
13. Alaska Infrastructure 100
14. Urban Drainage 100
15. Coastal Property 113
ELECTRICITY 120
16. Electricity Demand and Supply 120
WATER RESOURCES 128
17. Inland Flooding 128
18. Water Quality 135
19. Municipal and Industrial Water Supply 142
20. Winter Recreation 148
-------
TABLE OF CONTENTS
AGRICULTURE 156
21. Domestic Yield and Welfare Effects 156
22. U.S. and Global Agriculture Interactions 166
ECOSYSTEMS 171
23. Coral Reefs 171
24. Shellfish 176
25. Freshwater Fish 182
26. Wildfire 189
27. Carbon Storage 199
SYNTHESIS OF RESULTS 205
28. National Summary 205
29. Risk Reduction through Adaptation 211
30. Regional Summaries 218
30.1 Northeast 220
30.2 Southeast 227
30.3 Midwest 234
30.4 Northern Plains 241
30.5 Southern Plains 248
30.6 Southwest 254
30.7 Northwest 261
30.8 Alaska 268
30.9 Hawai'i and Puerto Rico 270
iv
-------
INTRODUCTION
ABOUT THIS REPORT
This Technical Report summarizes and communicates the results of the second phase of quantitative
sectoral impacts analysis under the Climate Change Impacts and Risk Analysis2 (CIRA) project (for
information on the first phase, see the CIRA Project Background section). This effort is intended to
inform the fourth National Climate Assessment3 (NCA4) of the U.S. Global Change Research Program
(USGCRP).4 The goal of this work is to estimate climate change impacts and economic damages to
multiple U.S. sectors (e.g., human health, infrastructure, and water resources) under different scenarios.
Though this report does not make policy recommendations, it is designed to inform strategies to
enhance resiliency and protect human health, investments, and livelihoods.
Each sectoral analysis is part of the CIRA multi-model framework which uses consistent inputs (e.g.,
socioeconomic and climate scenarios) to enable comparison of sectoral impacts across time and space.
In addition, the role of adaptation is modeled for some of the sectors to explore the potential for risk
reduction and, where applicable, to quantify the costs associated with adaptive actions.
The methods and results of the CIRA project have been peer-reviewed in the scientific literature,
including a special issue of Climatic Change entitled, "A Multi-Model Framework to Achieve Consistent
Evaluation of Climate Change Impacts in the United States."5The research papers underlying the
modeling and results presented herein are cited throughout this report and are listed in Section A.2 of
the Appendix to this Technical Report.
The Executive Summary follows this brief Introduction. The Modeling Framework (including climate
change projections, metrics modeled, and sources of uncertainty), further CIRA project background, and
commonly used terms and acronyms are described in the subsequent introductory chapters. The results
of 25 modeling analyses are reported in six chapters on Health, Infrastructure, Electricity, Water
Resources, Agriculture, and Ecosystems. The final chapter provides a synthesis of results at national and
regional (sub-national) scales as well as a summary of the analyses that modeled adaptation responses.
Additional resources can be found in the Technical Appendix or in the underlying research papers cited
throughout.
2 www.epa.gov/cira
3 www.globlachange.gov/NCA4
4 For a detailed assessment of the physical science basis for climate change, as well as a glossary of climate change terms, see the U.S. Global
Change Research Program (USGCRP)'s Climate Science Special Report (CSSR), which is similarly developed to inform the NCA4, and the fifth
assessment report of the Intergovernmental Panel on Climate Change. USGCRP, 2017. Climate Science Special Report: A Sustained Assessment
Activity of the U.S. Global Change Research Program [Wuebbles, D.J., D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, and T.K. Maycock
(eds.)]. U.S. Global Change Research Program, Washington, DC, USA. IPCC, 2014: Climate Change 2014: Synthesis Report. Contribution of
Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri
and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, 151 pp.
5 Martinich, J., J. Reilly, S. Waldhoff, M. Sarofim, and J. McFarland, Eds, 2015: Special Issue on "A Multi-Model Framework to Achieve Consistent
Evaluation of Climate Change Impacts in the United States." Climatic Change, 131,1-181.
1
-------
INTRODUCTION
INTERPRETING THE RESULTS
This Technical Report presents results from a large set of sectoral impact models that quantify and
monetize climate change impacts in the U.S., with a primary focus on the contiguous U.S., under
moderate and severe future climates. The CIRA analyses are intended to provide insights about the
potential direction and magnitude of climate change impacts. However, none of the estimates
presented in this report should be interpreted as definitive predictions of future impacts at a particular
place or time. Instead, the intention is to produce preliminary estimates of future effects using the best
available data and methods, which can then be revisited and updated over time as science and modeling
capabilities continue to advance.
The CIRA analyses do not evaluate or assume specific mitigation or adaptation policies in the U.S. or in
other world regions. Instead, they consider scenarios (Representative Concentration Pathways or RCPs6)
to illustrate potential impacts and damages of alternative future climates. The results should not be
interpreted as supporting any particular domestic or global mitigation policy or target. In addition, the
costs of reducing greenhouse gas (GHG) emissions, and the health benefits associated with co-
reductions in other air pollutants, are well-examined elsewhere in the literature and are beyond the
scope of this report. For this reason, the analysis presented in this Technical Report does not constitute
a cost-benefit assessment of climate policy.
Furthermore, only a small portion of the impacts of climate change are estimated, and therefore this
Technical Report captures just a fraction of the potential risks and damages that may be avoided or
reduced when comparing the alternative scenarios. To better estimate impacts, this ongoing project
continues to add new sectors, measures of economic damages, and adaptation scenarios, and to
improve methods and assumptions within existing sectoral modeling. Impacts that are not covered by
the modeling analyses and other important considerations or limitations are described in the discussion
sections of each individual sector chapter.
6 The RCPs are identified by their approximate total radiative forcing (not emissions) in the year 2100, relative to 1750: 2.6 W/m2 (RCP2.6), 4.5
W/m2 (RCP4.5), 6.0 W/m2 (RCP6.0), and 8.5 W/m2 (RCP8.5). RCP8.5 implies a future with continued high emissions growth with limited efforts
to reduce GHGs, whereas the other RCPs represent mitigation pathways of varying stringency; none of these scenarios represent any particular
national nor global policy.
2
-------
EXECUTIVE SUMMARY
This report quantifies potential physical and economic damages to multiple U.S. sectors (nationally and
within U.S. regions) using a consistent set of climate and socioeconomic scenarios and assumptions (see
following Modeling Framework section). Importantly, only a small portion of the impacts of climate
change are estimated. Looking across the large number of sectoral impacts described in this report, a
number of key findings emerge:
Under both atmospheric greenhouse gas (GHG) concentration scenarios modeled
(Representative Concentration Pathways or RCP4.5 and RCP8.5), climate change is
projected to significantly affect human health, the U.S. economy, and the environment.
These climate change impacts will not be uniform across the U.S., with most sectors
showing a complex pattern of regional-scale impacts.
For example, under RCP8.5, almost 1.9 billion labor hours across the national workforce are
projected to be lost annually by 2090 due to the effects of extreme temperature on suitable working
conditions, totaling over $160 billion in lost wages per year. More than a third of this national loss is
projected to occur in the Southeast ($47 billion lost annually by 2090).
In almost all sectors, projected physical and economic damages are significantly larger
under RCP8.5 than under RCP4.5. Lower global atmospheric GHG concentrations under
RCP4.5 substantially reduce damages associated with extreme weather, such as extreme
temperature, heavy precipitation, drought, and storm surge events.
For example, twice as many "100-year" flood events are projected across the contiguous U.S. under
RCP8.5 compared to RCP4.5 by the end of the century. By 2100, the difference between projected
damages from inland flooding under RCP8.5 and RCP4.5 is approximately $4 billion per year.
Avoided or reduced damages under the lower atmospheric GHG concentration scenario
(RCP4.5) are projected to increase over the course of the 21st century. Risks and damages
over the 21st century will not be avoided without significant reductions in GHG emissions.
For example, compared to RCP8.5, RCP4.5 avoids nearly 800 premature deaths from extreme
temperatures (both extreme heat and cold) per year by 2050, and more than 5,400 premature
deaths per year by 2090 in 49 U.S. cities —a reduction of temperature-related mortality of 24% by
2050, and nearly 60% from the 9,300 premature deaths projected to occur each year under RCP8.5
by 2090.
Adaptation actions, especially in the infrastructure sectors, are projected to substantially
reduce climate change impacts. Proactive adaptation measures implemented in
anticipation of future climate change risks are generally more cost-effective in reducing
damages than reactive adaptation responses implemented after impacts have already
3
-------
EXECUTIVE SUMMARY
occurred. For several infrastructure sectors, a combined portfolio of global mitigation and
regional adaptation strategies can eliminate a large portion of the economic impacts that
are otherwise projected to occur this century.
For example, average cumulative discounted reactive adaptation costs for roads are estimated at
$230 billion through 2100 under RCP8.5 and $150 billion under RCP4.5. Across all road types and
climate stressors, proactive adaptation to protect roads from climate change-related impacts is
projected to decrease costs over the century by more than 75% under both RCPs.
The rate and timing of climate change impacts, and actions to reduce them, are
important. While some impacts increase in frequency or magnitude in a gradual manner,
others exhibit threshold- type responses to climate change, as large changes manifest over
a short period of time.
For example, high-temperature bleaching events in coral (expulsion of symbiotic algae) occurred
infrequently in the past, but are projected to occur with greater frequency by 2030, resulting in
significant losses of U.S. coral reefs in Hawai'i and the Caribbean.
The figure on the next page provides an overview of the national-scale results presented throughout this
Technical Report, and shows that climate change is projected to cause economic impacts across most of
the sectors analyzed. Importantly, many of the sectoral methodologies do not estimate and monetize
the full extent of potential impacts from climate change on that sector, and as such, the results should
be treated as conservative. For example, the Air Quality analysis only estimates economic damages from
mortality caused by changes in ozone, omitting effects from changes in other air pollutants and
morbidity effects. In another example, the Extreme Temperature Mortality analysis only includes 49
major cities, covering about one third of the population. Although not available for all sectors,
cumulative impacts for the entire 21st century would very likely be much larger than the annual
estimates presented in the figure below. Please refer to the sectoral sections of this report describing
the individual modeling efforts for detailed information on the results, a summary of the methodologies
used, key sources of uncertainty, and references to the supporting peer-reviewed literature.
Annual damages are projected to increase over time and are generally larger under RCP8.5 compared to
RCP4.5. Projected impacts on Extreme Temperature Mortality, Labor, and Coastal Property are the most
economically significant under both RCPs in 2050 and 2090. Estimated impacts on Air Quality and Road
infrastructure are also large. For Wildfire, climate change is projected to generally have a beneficial
economic effect at a national level in the timeframes shown. It is important to note that while the
magnitude of estimated economic impacts for some of the sectors is relatively small, many of the
corresponding physical impacts have significant societal or iconic values that are generally not captured
in the estimates. For example, while damages associated with the loss of freshwater fishing days ($3.1
billion per year in 2090 under RCP8.5) is projected to be orders of magnitude less than coastal property
damages ($120 billion per year by 2090 under RCP8.5), this important recreational activity contributes
significantly to local economies, with more than 27 million people in the U.S. spending a total of $25
billion on over 365 million freshwater recreational fishing trips in 2011 alone.
4
-------
EXECUTIVE SUMMARY
Annual Economic Damages from Climate Change
Relative area represents projected economic damages under RCP8.5 in 2090. Sectors with estimated
damages larger than $5 billion per year are labeled (first number shown in $billions), and include the
percent decrease in damages under RCP4.5 compared to RCP8.5 in 2090 (second number shown in %).
Projected damages in 2050 and 2090 under both RCP8.5 and RCP4.5 for all sectors can be found in Table
28.2 of the National Summary section, including sectors with less than $5 billion annual damages in 2090
under RCP8.5, which are noted with asterisks in the legend but may not be visible at this figure scale.
Urban Drainage
$5.6 | -26%
Extreme
Temperature
Mortality
$140 | -58%
Rail
$5.5 | -36%
Labor
$160 | -48%
Coastal Property
$120 | -22%
Inland Flooding
$8.1 | -47%
Electricity Demand &
Supply
$9.2 | -63%
Roads
$20 | -59%
Air Quality
$26 | -31%
Water Quality
$4.6 | -35%
Air Quality
Aeroallergens*
Extreme Temperature Mortality |
Labor
West Nile Virus*
Harmful Algal Blooms*
Roads
Bridges*
Rail
Alaska Infrastructure*
Urban Drainage
Coastal Property
Electricity Demand and Supply
Inland Flooding
Water Quality*
Municipal and Industrial Water Supply*
Winter Recreation*
Agriculture*
Coral Reefs*
Shellfish*
Freshwater Fish*
Wildfire*
5
-------
MODELING FRAMEWORK
1. SCENARIOS AND PROJECTIONS
As this second phase of the CIRA project was undertaken to inform the development of NCA4, the
selection of scenarios and projections has been made consistent, to the maximum extent possible, with
the USGCRP-recommended inputs to the forthcoming assessment. These inputs have the benefits of
being well-known and commonly used by others in the climate change impacts modeling community.
1.1. SELECTION OF INPUTS
This section describes the selected scenarios and projections, as well as details regarding how they were
processed for use in the modeling framework.
Scenarios of GHG Emissions and Radiative Forcing
As described in the 2015 guidance from the USGCRP Scenarios and Interpretive Science Coordinating
Group/ the NCA4 will rely on climate scenarios generated for the Intergovernmental Panel on Climate
Change's (IPCC's) Fifth Assessment Report (AR5). The scenarios are based on four "representative
concentration pathways" (RCPs) that capture a range of plausible emission futures. The RCPs are
identified by their approximate total radiative forcing (not emissions) in the year 2100, relative to 1750:
2.6 W/m2 (RCP2.6), 4.5 W/m2 (RCP4.5), 6.0 W/m2 (RCP6.0), and 8.5 W/m2 (RCP8.5). RCP8.5 implies a
future with continued high emissions growth with limited efforts to reduce GHGs, whereas the other
RCPs represent mitigation pathways of varying stringency; none of these scenarios represent any
particular national or global policy.
As in many sectoral impact analyses, the analyses presented in this Technical Report use a subset of the
RCPs due to computational, time, and resource constraints. Based on USGCRP guidance, the analyses
utilize RCP8.5 as a high-end scenario and RCP4.5 as a low-end scenario.8 Comparing outcomes under
RCP8.5 and RCP4.5 captures a range of uncertainties and plausible futures and provides perspectives on
7 U.S. Global Change Research Program, 2015: U.S. Global Change Research Program General Decisions Regarding Climate-Related Scenarios for
Framing the Fourth National Climate Assessment. USGCRP Scenarios and Interpretive Science Coordinating Group. Available online at
https://scenarios.globalchange.gov/accouncement/1158
8 See the Third National Climate Assessment (2014) and Climate Impacts Group (2013) for useful descriptions of how the RCPs compare to other
common scenarios. References: Walsh, J., D. Wuebbles, K. Hayhoe, J. Kossinet al., 2014: Ch. 2: OurChanging Climate. Climate Change Impacts in
the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global Change
Research Program, 19-67. doi:10.7930/J0KW5CXT; Climate Impacts Group, 2013. Making sense of the new climate change scenarios. University
of Washington, available at: http://cses.washington.edu/db/pdf/snoveretalsok2013sec3.pdf.
6
-------
MODELING FRAMEWORK
Scenarios and Projections
the potential differences between scenarios.910 Figure 1.1 shows differences between RCP8.5 and
RCP4.5 in terms of global GHG emissions and atmospheric C02 concentration. Under RCP8.5, global
atmospheric C02 levels rise from current-day levels of approximately 400 up to 936 parts per million
(ppm) by 2100 relative to the 1986-2005 average. Under the RCP4.5, atmospheric C02 levels at the end
of the century remain below 550 ppm. For more information on RCP projections, see Appendix section
A.4 and the USGCRP Climate Science Special Report (CSSR).11
Figure 1.1. Global GHG Emissions and Atmospheric C02 Concentrations for RCP8.5 and RCP4.512
1000
900
800
700
600
500
400
300
Global CO, Concentration
VVVVV'lk'VJ'V''V
RCP 8.5 RCP 4.5
^ ^ ^ ^ ^ ^ ^ ^
RCP 8.5 RCP 4.5
Global C02 Emissions
Selection of Global Climate Model (GCM) Projections
To support development of the IPCC's AR5, over 20 climate modeling groups from around the world
agreed to a coordinated climate modeling experiment called the fifth phase of the Coupled Model
Intercomparison Project (CMIP5).13 More than 60 models from these groups were run with the RCPs
described above, and the resulting data archive has been made available for use by the scientific
community over the past several years.
Statistical Downscaling
The results of these global climate model simulations are displayed in coarse geographic grid cells
(roughly 2.5°x2.0°). This coarse spatial resolution can encompass disparate areas; for instance, a single
grid cell of a global climate model can cover the distance from San Francisco to Sacramento.14 To
provide more localized projections of climate changes—important for local impact assessment and
adaptation planning—and to provide more consistency with historical observations, downscaling
methodologies are typically employed. The approach used in this report is statistical downscaling, which
9 Ibid.
10 As described in the Coordinating Group's guidance, RCP4.5 is consistent with the SRES-B1 scenario used in NCA3, whereas RCP2.6 would
represent a significant departure.
11 USGCRP, 2017. Climate Science Special Report: A Sustained Assessment Activity of the U.S. Global Change Research Program [Wuebbles, D.J.,
D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, and T.K. Maycock (eds.)]. U.S. Global Change Research Program, Washington, DC, USA.
12 Data from: Meinshausen, M., S.J. Smith, K. Calvin, J. S. Daniel, M. L. T. Kainuma, J-F. Lamarque, K. Matsumoto, S.A. Montzka, S.C.B. Raper, K.
Riahi, A. Thomson, G.J.M. Velders, and D.P.P. van Vuuren, 2011: The RCP Greenhouse Gas Concentrations and their extension from 1765va to
2500. Climatic Change, 109, 213, doi: 10.1007/sl0584-011-0156-z. Data available at: http://www.pik-potsdam.de/~mmalte/rcps/.
13 Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485-498,
doi:10.1175/BAMS-D-ll-00094.1.
14 http://loca.ucsd.edu/what-is-loca/
7
-------
MODELING FRAMEWORK
Scenarios and Projections
develops statistical relationships between local climate variables (e.g., temperature or precipitation) and
large-scale predictors (e.g., pressure fields) and applies those relationships to the GCM output. While
many downscaled products using the CMIP515 archive are available, the modeling in this Technical
Report uses two of the highest-quality, publicly-available, and peer-reviewed downscaled primary
datasets:
Contiguous U.S.: A 2016 dataset of downscaled CMIP5 climate projections was commissioned by the U.S.
Bureau of Reclamation and Army Corps of Engineers and developed by the Scripps Institution of
Oceanography with a number of collaborators.16This dataset, called LOCA (which stands for Localized
Constructed Analogs), is used in USGCRP's CSSR, which provides the physical climate science basis for
the upcoming NCA4. The LOCA dataset has many advantages; notably, the statistical approach produces
improved estimates of extremes, constructs a more realistic depiction of the spatial coherence of the
downscaled field, and reduces the problem of producing too many light-precipitation days.17 LOCA
projections have been developed for both RCP8.5 and RCP4.5 using 32 GCMs from the CMIP5 archive.18
However, as of the finalization date of this Technical Report, the LOCA dataset has only been
downscaled for the contiguous U.S. Finally, the LOCA dataset provides daily projections through 2100 at
a l/16th degree resolution for three variables: daily maximum temperature (tmax), daily minimum
temperature (tmin), and daily precipitation. Some of the sectoral models of this Technical Report
require additional variables, such as solar radiation, wind speed, and relative humidity. Appendix A.4
describes the historical binning approach used to develop internally-consistent projections for these
variables based on the LOCA projections.
Alaska: The Scenarios Network for Alaska + Arctic Planning (SNAP), a part of the International Arctic
Research Center at the University of Alaska Fairbanks, developed a downscaled climate dataset for
Alaska,19 which as described above, is not covered in the LOCA dataset. The commonly used SNAP
dataset focuses on five climate models from the CMIP5 ensemble that have the most skill for Alaska and
the Arctic.20 These five models, all of which were run using RCP8.5 and RCP4.5, are:
15 The CMIP5 climate data are widely available now, whereas products from the next phase of the project (CMIP6) are not envisioned to be
available in time to support the analyses of this project.
16 U.S. Bureau of Reclamation, Climate Analytics Group, Climate Central, Lawrence Livermore National Laboratory, Santa Clara University,
Scripps Institution of Oceanography, U.S. Army Corps of Engineers, and U.S. Geological Survey, 2016: Downscaled CMIP3 and CMIP5 Climate
Projections: Release of Downscaled CMIP5 Climate Projections, Comparison with Preceding Information, and Summary of User Needs. Available
online at http://gdo-dcp.ucllnl.org/downscaled cmip proiections/techmemo/downscaled climate.pdf. Data available at http://gdo-
dcp.ucllnl.org/downscaled cmip projections/.
17 University of California San Diego, cited 2017: LOCA statistical downscaling. Scripps Institution of Oceanography. Available online at
http://loca.ucsd.edu/
18 The LOCA projections were released in September 2016. For timing reasons, a pre-release version of the LOCA dataset was obtained and used
in the analyses of this Technical Report. This earlier version contained some very minor differences that have no significant impact on the
sectoral modeling results. Specifically, temporal discontinuities (jumps) in the time-series were found for a small number of grid cells on the
l/16th degree grid. Compared to the publicly available version, the earlier dataset contains a slight increase in noise of the precipitation field,
and slightly less agreement between the original model-predicted climate change signal and the change seen after downscaling with LOCA.
19 University of Alaska Fairbanks, cited 2017: SNAP: Scenarios Network for Alaska and Arctic Planning. International Arctic Research Center.
Available online at: https://www.snap.uaf.edu/
20 Model skill is a criterion to select a subset of GCMs that are able to best re-create historical climate in a region (e.g., see Rupp et al. 2013,
McSweeney et al. 2015, Vano et al. 2015). In the SNAP context, skill is measured based on the degree to which each GCM's output aligned with
observed climate data from 1958 to 2000 for precipitation, sea level pressure, and surface air temperature. Note that due to nonstationarity,
skill at matching historical observations may not always lead to superior representation of future climate conditions. References:
McSweeney, C.F., Jones, R.G., Lee, R.W. and Rowell, D.P., 2015. Selecting CMIP5 GCMs for downscaling over multiple regions. Climate
Dynamics, 44(11-12), pp.3237-3260.
Rupp, D.E., Abatzoglou, J.T., Hegewisch, K.C. and Mote, P.W., 2013. Evaluation of CMIP5 20th century climate simulations for the Pacific
Northwest USA. Journal of Geophysical Research: Atmospheres, 118(19).
Vano, J.A., Kim, J.B., Rupp, D.E. and Mote, P.W., 2015. Selecting climate change scenarios using impact-relevant sensitivities. Geophysical
Research Letters, 42(13), pp.5516-5525.
8
-------
MODELING FRAMEWORK
Scenarios and Projections
CCSM4
GFDL-CM3
GISS-E2-R
IPSL-CM5A-LR
MRI-CGCM3
National Center for Atmospheric Research
Geophysical Fluid Dynamics Laboratory (NOAA)
Goddard Institute for Space Studies (NASA)
Institut Pierre-Simon Laplace
Meteorological Research Institute
While it would be preferable to use one downscaled product to maximize consistency, neither dataset
covers all geographic areas needed for all analyses, and the strengths of each dataset offer significant
advantages over other available downscaled products.
Selection of GCMs
As in many sectoral impact analyses in the literature, the selection of a subset of GCMs is necessary due
to computational, time, and resource constraints. Table 1.1 presents the five GCMs that are used in the
sectoral analyses of this Technical Report. These GCMs were chosen based on: 1) the ability to leverage
existing dynamically-downscaled GCM data, 2) their availability in the SNAP and LOCA downscaled
datasets, 3) their ability to capture variability in temperature and precipitation outcomes, and 4) their
demonstrated independence and quality. A detailed description of the criteria used to select GCMs can
be found in the Technical Appendix (see section A.4).
9
-------
MODELING FRAMEWORK
Scenarios and Projections
Table 1.1. CMIP5 GCMs Used in the Analyses of this Technical Report
Center (Modeling Group)
Model
Acronym
Availability
LOCA SNAP
References
Canadian Centre for Climate Modeling and Analysis
CanESM2
X
Von Salzen et al.
201321
National Center for Atmospheric Research
CCSM4
X
X
Gent et al. 201122
Neale et al. 201323
NASA Goddard Institute for Space Studies
GISS-E2-R24
X
X
Schmidt et al. 200625
Met Office Hadley Centre
HadGEM2-ES
X
Collins et al., 201126
Davies et al. 200527
Atmosphere and Ocean Research Institute,
National Institute for Environmental Studies, and
Japan Agency for Marine-Earth Science and
Technology
MIROC5
X
Watanabe et al.
201028
Use of Eras and Control Scenarios
The LOCA and SNAP datasets, which are based on GCM projections from CMIP5, provide data for the
1950-2100 time period.29 In order to reduce the effects of inter-annual variability and obtain results that
are better representative of a particular point in the future, the analyses described in this Technical
Report use 20-year eras centered on specific years of interest. The selected timeframes provide
sufficient coverage of time periods across the century, including a reasonable estimation of late century
impacts (i.e., getting as close to 2100 as possible).
21 von Salzen, K., J.F. Scinocca, N.A. McFarlane, J. Li, J.N. Cole, D. Plummer, D. Verseghy, M.C. Reader, X. Ma, M. Lazare, and L. Solheim, 2013:
The Canadian fourth generation atmospheric global climate model (CanAM4). Part I: representation of physical processes. Atmosphere-Ocean,
51, 104-125.
22 Gent, P.R., G. Danabasoglu, L.J. Donner, M.M. Holland, E. Hunke, S. Jayne, D. Lawrence, R.B. Neale, P.J. Rasch, M. Vertenstein, and P.H.
Worley, 2011: The community climate system model version 4. Journal of Climate, 24, 4973-4991.
23 Neale, R.B., J. Richter, S. Park, P.H. Lauritzen, S.J. Vavrus, P. Rasch, and M. Zhang, 2013: The mean climate of the community Atmosphere
Model (CAM4) in forced SST and fully coupled experiments. Journal of Climate, 26, 5150-5168.
24 Some of the GCMs in the CMIP5 archive were run multiple times to develop individual initializations for each climate model. In general, the
LOCA dataset provides projections using the first initialization of each GCM. However, for the GISS-E2-R model, the LOCA dataset provided data
for RCP4.5 using run #r6ilpl and run #r2ilpl for RCP8.5. The main reasoning forthis difference is that the GCM initializations (raw data from
CMIP5) did not provide all of the climate data necessary for doing the LOCA constructed analog and bias correction technique. While the usage
of different initializations for the GISS-ER-R model could introduce inconsistency, the statistical differences across runs of the same GCM are
dramatically lower than across models, and those differences are further dampened by the LOCA bias correction. To evaluate the potential that
these alternative initializations could introduce inconsistencies, an analysis was completed comparing the raw #r6ilpl runs for both RCPs and
the raw #r2ilpl runs for both RCPs. The results of this comparative analysis confirmed that the differences are minimal and that it is reasonable
to use the LOCA projections.
25 Schmidt, G.A., R. Ruedy, J.E. Hansen, I. Aleinnov, N. Bell, M. Bauer, S. Bauer, B. Cairns, V. Canuto, Y Cheng, and A. Del Genio, 2006: Present-
day atmospheric simulations using GISS ModelE: Comparison to in situ, satellite, and reanalysis data. Journal of Climate, 19, 153-192.
26 Collins, W.J., N. Bellouin, M. Doutriaux-Boucher, N. Gedney, P. Halloran, T. Hinton, J. Hughes, C. D. Jones, M. Joshi, S. Liddicoat, G. Martin, F.
O'Connor, J. Rae, C. Senior, S. Sitch, I. Totterdell, A. Wiltshire, and S. Woodward, 2011: Development and evaluation of an Earth system model-
HadGEM2. Geoscience Model Development, 4, 1051-1075.
27 Davies, T., M.J. P. Cullen, A. J. Malcolm, M. H. Mawson, A. Staniforth, A. A. White, N. Wood, 2005: A new dynamical core for the Met Office's
global and regional modelling of the atmosphere. Quarterly Journal of the Royal Meteorological Society, 131, 1759-1782.
28 Watanabe, M., T. Suzuki, R. O'ishi, Y. Komuro, S. Watanabe, S. Emori, T. Takemura, M. Chikira, T. Ogura, M. Sekiguchi, and K. Takata, 2010:
Improved climate simulation by MIROC5: mean states, variability, and climate sensitivity. Journal of Climate, 23, 6312-6335.
29 Two of the five GCMs used in this project only provide data through the end of calendar year 2099. Therefore, the inclusion of the year 2100
data is not included in the eras used in the analyses of this Technical Report.
10
-------
MODELING FRAMEWORK
Scenarios and Projections
• Historic (1986-2005), typically referred to as the 'reference period'30
• 2030(2020-2039)
• 2050 (2040-2059)
• 2070(2060-2079)
• 2090 (2080-2099)
All the sectoral analyses in this Technical Report provide results for the 2050 and 2090 eras. Further,
results for some sectors are provided using a full time-series of results through 2100, as the outputs of
those models provide monthly or annual projections for the entire century.
For some sectors, it is necessary to compare projected climate change impacts with future 'control
scenarios' that do not include changes in climate. This is done to isolate the effects of climate change
from impacts occurring due to changes in non-climate effects, such as increasing population or
economic growth (see descriptions of these socioeconomic variable selections below). The Approach
sections for each sector describe where these 'control scenarios' are used and what they represent.
Sea Level Rise Scenarios
This Technical Report uses sea level rise scenarios described in the 2017 NOAA sea level rise technical
report for NCA431 and the CSSR of the USGCRP.32 The global mean sea level rise estimates underlying
these scenarios are based on the rates described in Kopp et al. (2014).33
To generate the global mean sea level rises estimates, the NOAA (2017) projections are stratified based
on rates in 2100, and the median34 for each subset of projections was identified to be consistent with
the 2100 global mean sea levels. These six values of global mean sea level change in 2100 are shown in
the first column of Table 1.2. After developing annual time series consistent with these 2100 sea levels,
these projections are used in the Coastal Property analysis described in this report. To account for the
differences in probabilities that each sea level trajectory could occur under each RCP, scenarios weights
based on NOAA (2017) are then applied to the Coastal Property sector results for each of the six levels
(Table 1.2).
30 While the 1986-2005 reference period was used across most sectors of this Technical Report, there are some where alternative periods were
defined due to model or data-specific needs and constraints. Each reference period used is defined in the sectors of the report.
31 National Oceanographic and Atmospheric Administration. 2017. Global and regional sea level rise scenarios for the United States. NOAA
Centerfor Operational Oceanographic Products and Services, Technical Report NOS CO-OPS 083.
32Sweet, W.V., R. Horton, R.E. Kopp, and A. Romanou, 2017. Sea level rise. In: Climate Science Special Report: A Sustained Assessment Activity
of the U.S. Global Change Research Program [Wuebbles, D.J., D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, and T.K. Maycock (eds.)]. U.S.
Global Change Research Program, Washington, DC, USA. pp. 333-363, doi: 10.7930/J0VM49F2.
33 Kopp, R.E., R.M. Horton, C.M. Little, J.X. Mitrovica, M. Oppenheimer, D.J. Rasmussen, B.H. Strauss and C. Tebaldi, 2014: Probabilistic 21st and
22nd century sea-level projections at a global network of tide-gauge sites. Earth's Future, 2, 383-406, doi:10.1002/2014EF000239.
34 These medians represent projections in which global mean sea level rise in 2100 is 28-32 cm, 48-52 cm, 98-102 cm, 145-155 cm, 195-205 cm,
or 245-255 cm.
11
-------
MODELING FRAMEWORK
Scenarios and Projections
Table 1.2. Global Mean Sea Level Rise in 2100 with Scenario Weights
RCP8.5
RCP4.5
Global Mean Sea Level
Rise in 2100 (cm)
Exceedance
Probability
Scenario Weight
Exceedance
Probability
Scenario Weight
30
0.9997
0.0069
0.9814
0.0948
50
0.9607
0.4064
0.7296
0.7197
100
0.1670
0.5464
0.0330
0.1746
150
0.0133
0.0352
0.0045
0.0087
200
0.0026
0.0038
0.0012
0.0014
250
0.0009
0.0013
0.0005
0.0008
Projections of location-specific differences in relative (or local) sea level change are also taken from
NOAA (2017), which account for land uplift or subsidence, oceanographic effects, and responses of the
geoid and the lithosphere to shrinking land ice. The probabilistic sea level rise estimates by location
from NOAA (2017) are not adopted; instead the mean values for each tide gauge location are used. A
distance weighting procedure for interpolating between tide gauge locations is used to attribute tide
gauge-level results to each coastal county. Figure 1.2 presents an example of the relative sea level rise
rates over the course of the century for Virginia Beach, VA.
Figure 1.2. Local Sea Level Rise Rates for Virginia Beach
Virginia Beach
350 i
300 - /
_ 250
30 cm
$ 200
100 cm
150
^—150 cm
200 cm
100
250 cm
50
0
2000
2020
2040
2060
2080
2100
Atmospheric Carbon Dioxide Concentrations
For the CMIP5 project, most climate model simulations used prescribed atmospheric C02
concentrations, and therefore did not interactively include the effect of carbon cycle feedbacks.35 The
analyses in this Technical Report therefore assume that C02 concentrations in each of the GCM
35 Friedlingstein, P., M. Meinshausen, V. K. Arora, C. D. Jones, A. Anav, S. K. Liddicoat, and R. Knutti, 2013: Uncertainties in CMIP5 climate
projections due to carbon cycle feedbacks. J. dim., 27, 511-526.
12
-------
MODELING FRAMEWORK
Scenarios and Projections
simulations are consistent, with reported values for each RCP taken from the Potsdam Institute for
Climate Impact Research (Meinshausen et al., 2011).36
Socioeconomic Projections
To capture the effects increasing population and income can have on impact estimates, the analyses in
this Technical Report use a single trajectory of socioeconomic change under each RCP. Using a single
control population projection isolates the differences in climate change impacts between the two RCPs,
such that the results will not be influenced by differing pathways of socioeconomic change.
The analyses in this Technical Report use the Median Variant Projection of the United Nation's (UN)
2015 World Population Prospects dataset to project future U.S. population for 2015-2100.37 This
scenario was chosen as it represents a reasonable, mid-range population projection, and allows for the
reasonable incorporation of future population growth.38 For historical U.S. population data for the
period 1986-2014, U.S. census data was used.39 As shown in Figure 1.3, the projected change in U.S.
population under the UN Median (shown in solid blue) lies between the U.S. Census High and Low
projections through 2060,40 and by 2100 lies in the middle of the range of scenarios from the UN and the
Shared Socioeconomic Pathways (SSPs).41The UN Median projection is also similar to the population
trajectory used in the first phase of the CIRA project (Paltsev et al., 2013).42 The impacts of climate
change on future U.S. population are not factored into the projections used throughout this Technical
Report, with the exception of the domestic migration sector. As such, the population projection does
not include important feedbacks that may lead to higher or lower populations at regional levels.
As the UN Median population projection is only available at a national scale, disaggregated population
projections were produced at the county-level using EPA's Integrated Climate and Land Use Scenarios
version 2 (ICLUSv2) model.43,44 The spatial pattern of population change in ICLUS is dependent upon
36 M. Meinshausen, S. Smith, S. J. Smith, K. Calvin, J. S. Daniel, M. L. T. Kainuma, J-F. Lamarque, K. Matsumoto, S. A. Montzka, S. C. B. Raper, K.
Riahi, A. Thomson, G. J. M. Velders, and D.P. P. van Vuuren, 2011: The RCP Greenhouse Gas Concentrations and their extension from 1765 to
2500, Climatic Change (Special Issue on RCPs), 109, 213, doi: 10.1007/sl0584-011-0156-z. Data available at: http://www.pik-
potsdam.de/~mmalte/rcps/
37 United Nations, 2015: World Population Prospects: The 2015 Revision. United Nations, Department of Economic and Social Affairs,
Population Division.
38 The choice of this particular population scenario versus another could have significant influence on the estimated impacts across sectors,
particularly those most influenced by changes in population and economic growth. The use of other population scenarios, such as SSP5 (713
million by 2100) orthe UN High Variant (647 million), were determined to be less ideal forthis analysis, as they would have a large effect in
inflating late-century impact estimates due to the large increase in national population. In addition, recent demographictrends in the U.S.
suggest that population growth lie closer to the mid-range scenarios presented in Figure 1.3. Given that the purpose of this analysis is to focus
on understanding the difference between the two RCPs, the exploration of uncertainty surrounding this particular population projection is
deferred to future work, and the robust literature exploring the differences amongst scenarios.
39 U.S. Census Bureau, cited 2017: Population Estimates Program. Available online at https://www.census.gov/programs-survevs/popest.html
40 Hollmann, F.W., T.J. Mulder, and J.E. Kalian, 2000: Methodology and Assumptions for the Population Projections of the United States: 1999 to
2100. U.S. Department of Commerce, Bureau of the Census, Population Division, Population Projections Branch. Population Division Working
Paper No. 38. Available online at https://www.census.gov/population/www/documentation/twps0038/twps0038.html
41 O'Neill, B. C., E. Kriegler, K. Riahi, K. L. Ebi, S. Hallegatte, T. R. Carter, R. Mathur, and D. P. v. Vuuren, 2014: A new scenario framework for
climate change research: the concept of shared socioeconomic pathways, Climatic Change, doi:10.1007/sl0584-013-0905-2.
42 Paltsev, S., E. Monier, J. Scott, A. Sokolov, and J. Reilly, 2013: Integrated economic and climate projections for impact assessment. Climatic
Change, doi:10.1007/sl0584-013-0892-3.
43 Bierwagen, B., D.M. Theobald, C.R. Pyke, A. Choate, A.P. Groth, J.V. Thomas, and P. Morefield, 2010: National housing and impervious surface
scenarios for integrated climate impact assessments. Proc Natl Acad Sci USA, 107, 20887-20892. See http://www.epa.gov/iclus for more
information.
44 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (Iclus) (Version
2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
13
-------
MODELING FRAMEWORK
Scenarios and Projections
underlying assumptions regarding fertility, migration rate, and international immigration. These
assumptions were parameterized using the storyline of SSP2,45 which suggests medium levels of fertility,
mortality, and international immigration.46 Figure 1.4 shows these county-scale population projections
(absolute and percent change from the reference period). The ICLUS model was also used to develop
county-scale demography projections (i.e., age, gender, and race),47 a developed-lands (municipal and
industrial development) map layer, and county-scale population projections driven by the future climate
patterns of the five LOCA GCMs described above.
Figure 1.3. Comparison of UN, U.S. Census, and the SSPs Population Projections
"i 400
UN Low
SSP3
Census Low
¦ UN Median
¦SSP2
Census Mid
2050
UN High
SSP5
— — • Census High
45 O'Neill, B. C., E. Kriegler, K. Riahi, K. L. Ebi, S. Hallegatte, T. R. Carter, R. Mathur, and D. P. v. Vuuren, 2014: A new scenario framework for
climate change research: the concept of shared socioeconomic pathways, Climatic Change, doi:10.1007/sl0584-013-0905-2.
46 The ICLUSv2 model, as well as the assumptions used in this particular run of the model, represent one framework for projecting population
and demographic change across the U.S. Alternative models or specifications could lead to different outcomes than those reported in this
Technical Report.
47 Projected changes in age distribution by county over time from ICLUSvl were scaled by the projected change in total county population using
the CIRA2.0 ICLUSv2 dataset.
14
-------
MODELING FRAMEWORK
Scenarios and Projections
Figure 1.4. Projected County-Scale Population Change
Total County Population
Percent Change from 2010
2010
Number of People
80 to 8,000
8,001 to 15,000
15,001 to 25,000
¦ 25,001 to 50,000
¦ 50,001 to 100,000
¦ 100,001 to 20,000,000
Percent Change
-99 to-50
-49 to -0
1 to 50
51 to 100
101 to 150
151 to 200
>200
2050
•vr* *.r-k
i . rh*-
A A v \\ /
.
2090
¦saw
Using the UN Median population projection for the U.S., the Emissions Predictions and Policy Analysis
(EPPA, version 6) model48 was run to generate a projection of economic growth (i.e., gross domestic
product, or GDP). This approach is similar to the one used in the first phase of CIRA modeling.49 The
projection of GDP growth through 2040 for the U.S. was taken from the 2016 Annual Energy Outlook
reference case,50 combined with EPPA-6 baseline assumptions for other world regions and time periods
(Figure 1.5). The impacts of climate change on economic activity (e.g., losses to labor supply or increased
capital expenditures for adaptation) are not accounted for in the macroeconomic input projections used
48 Chen, Y.-H. H., S. Paltsev, J. Reilly, J. Morris, and M, Babiker, 2015: The MIT EPPA6 Model: Economic Growth, Energy Use, and Food
Consumption. MIT Joint Program on the Science and Policy of Global Change, Report: 278, Cambridge, MA. Available online at
http://globalchange.mit.edu/research/publications/2892
49 Paltsev, S., E. Monier, J. Scott, A. Sokolov, and J. Reilly, 2013: Integrated economic and climate projections for impact assessment. Climatic
Change, doi:10.1007/sl0584~013-0892-3.
50 U.S. Energy Information Administration, 2016: Annual Energy Outlook.
15
-------
MODELING FRAMEWORK
Scenarios and Projections
throughout this Technical Report. As such, the economic growth projection may be overestimated when
considering the multi-sector damages, and the use of a single national-scale economic growth projection
that omits region-specific socioeconomic changes may lead to different results than those reported.
Figure 1.5. Projected Change in U.S. Gross Domestic Product
90
80
70
60
50
40
30
20
10
2010
2030
2050
2070
2090
Treatment of Discounting
When presenting annual economic results for future years (e.g., 2050 or 2090), results are not
discounted. When presenting cumulative economic results in present day value, defined as the year
2015, this report uses a 3% discount rate.51 In all cases, economic estimates are presented in $2015.52
Regional Summaries of Results
The sectoral sections of this report estimate impacts at various spatial scales (e.g., city-level, county, 8-
digit hydrologic unit code). To enable the development of regional (sub-national) comparisons and
summaries, results are aggregated to the scale of the NCA4 regions, which are shown in Figure 1.6. See
the Regional Summaries section of this Technical Report for estimated physical and economic impacts
across sectors.
51 In short, discounting provides an equal basis to compare the value of economic impacts that occur in different time periods. The discount rate
itself reflects the trade-off between consumption today and consumption tomorrow, meaning that with a positive discount rate, benefits that
occur today are worth more than they would be tomorrow. There are many ways to select a discount rate and little consensus about which
discount rate is most appropriate, particularly when assessing economic impacts that span generations. Therefore, this Technical Report uses
3%, a commonly employed rate in the climate impacts literature (e.g., see Goulder and Williams (2012)). This rate is also consistent with the
consumption rate of interest recommended by federal guidance for benefit cost analysis, known as OMB Circular A-4, to capture "the rate at
which 'society' discounts future consumption flows to their present value." OMB based this rate on the real rate of return on long-term
government debt averaged over a 30-year period prior to the issuance of Circular A-4 (2003). Goulder, Lawrence H. and Roberton C. Williams III,
"The Choice of Discount Rate for Climate Change Policy Evaluation," Climate Change Economics. Volume 4, Issue 3, November 2012,
http://dx.doi.org/10.1142/S201000781250Q248.
52 Dollar years are adjusted using the U.S. Bureau of Economic Affairs' Implicit Price Deflators for Gross Domestic Product, Table 1.1.9. See
"National Income and Product Accounts Tables" at https://bea.gov/national/index.htm
16
-------
MODELING FRAMEWORK
Scenarios and Projections
Figure 1.6. NCA4 Regional Aggregations53
NCA4 Regions
!¦! Northeast
Southeast •
H Midwest
I Northern Plains
Southern Plains
Southwest
Northwest
Alaska
Hawaii
I Puerto Rico
Esri, HERE. DeLcrme. Mspmylndia. ® OperStreetMspcortributcfs, ana theGIS user community
53 The NCA4 regions align with the NCA3 regions except that: 1) the Great: Plains is split into two regions, with the Northern Plains encompassing
North Dakota, South Dakota, Nebraska, Montana, and Wyoming, and the Southern Plains encompassing Kansas, Oklahoma, and Texas, and 2)
l-lawai'i and the Pacific Islands, as well as the Caribbean, are separate regions in NCA4,
17
-------
MODELING FRAMEWORK
Scenarios and Projections
1.2. PROJECTIONS OF FUTURE CLIMATE
This section provides a general overview of the characteristics of the climate projections driving the
sectoral analyses described in this Technical Report. Projections of secondary or indirect physical
impacts, such as runoff or water supply, are described in the relevant sector chapters.
Temperature Change in the U.S.
Figure 1.7 presents the change in mean temperature across the lower 48 states (averaged across the
five LOCA GCMs described in the previous section), and Figure 1.8 shows the change in mean
temperature across Alaska (averaged across the two SNAP GCMs that are also included in the LOCA set -
CCSM4 and GISS-E2-R). As shown, the models project significant warming by the end of the century
under RCP8.5 compared to RCP4.5, particularly in parts of the Northeast, Midwest, and Northern Plains
of the contiguous U.S. In Alaska, the projected temperature increase is highest along the North Slope,
and the increase under RCP8.5 is significantly higher than that under RCP4.5. Section A.4 of the
Appendix to this Technical Report presents the results for each GCM.
Figure 1.7, Change in Mean Annual Temperature Relative to the Reference Period (1986-2005) across
the Contiguous U.S. (Average across the Five LOCA GCMs)
2030
2050
2070
2090
m
CO
Q_
O
in
Q_
O
cr
C F
Figure 1.8. Change in Mean Annual Temperature Relative to the Reference Period (1986-2005) across
Alaska (Average across Two SNAP GCMs)
2030
m
oo
Q_ X
O 1 /
or _
LO
"vf
Q_
O
^ fu
y*
2050
1/4
2070
-.4P
r
W V
2090
C F
18
-------
MODELING FRAMEWORK
Scenarios and Projections
In addition to increasing average temperatures, climate change is projected to result in an increase in
extreme temperatures. Figure 1.9 presents the projected number of days above 90°F across the
contiguous U.S., as well as the number of days above 90°F for the reference period (1986-2005). As
shown, the models project a significant increase in the number of days above 90°F by the end of the
century under RCP8.5 compared to RCP4.5, particularly in areas of the Southwest, Southern Plains, and
Southeast, where the number of days above 90°F reaches as high as 160 days per year. Figure 1.10
shows the projected number of days above 80°F across Alaska in future periods, as well as the
reference. The models project a significant increase in the number of days above 80°F, particularly in the
central part of the state, and the number of days above 80°F is greater under RCP8.5 than RCP4.5.
Section A.4 of the Appendix to this Technical Report presents the results for each GCM.
Figure 1.9. Number of Days above 90°F across the Contiguous U.S. (Average across the Five LOCA
GCMs)
Figure 1.10. Number of Days above 80°F across Alaska (Average across the Two SNAP GCMs)
2030 2050 2070 2090
Historical
(1986-2005) S
Historical
(1986-2005)
2030
—10
# of Days
2050 2070
Precipitation Change in the U.S.
Figure 1.11 presents the percent change in mean annual precipitation across the lower 48 states, and
Figure 1.12 shows the percent change in mean annual precipitation across Alaska. As shown, the models
project significant changes in precipitation by the end of the century under RCP8.5 compared to RCP4.5.
Some regions, particularly the Southwest, are projected to receive less annual rainfall while other
regions, particularly the Northwest, receive more annual rainfall. In Alaska, the projected increase in
precipitation is generally highest in the eastern parts of the state, but the projections under RCP8.5 for
the end of the century show extreme wetting across the majority of the state. Section A.4 of the
Appendix to this Technical Report presents the results for each GCM.
19
-------
MODELING FRAMEWORK
Scenarios and Projections
Figure 1.11. Percent Change from Reference Period in Mean Annual Precipitation across the
Contiguous U.S. (Average across the Five LOCA GCMs)
LO
oo
Q.
O
in
Q.
O
vl
2030
2070
-a m
m,
rj'2TV.
¦. i
wm*>.
frvt'^
% Change
In addition to changes in mean annual precipitation, climate change is projected to result in periods of
more intense rainfall. Figure 1.13 presents the percent change in the maximum daily precipitation across
the contiguous U.S, and Figure 1.14 presents the percent change in the maximum monthly precipitation
across Alaska. As shown, the models project more intense rainfall events by the end of the century
under RCP8.5 compared to RCP4.5. Across Alaska, the models project significant increases in the
maximum monthly precipitation by the end of the century in eastern areas of the state under RCP8.5,
while western and southwestern areas are projected to experience drying under RCP4.5. Section A.4 of
the Appendix to this Technical Report presents the results for each GCM.
Figure 1.12. Percent Change from Reference Period in Mean Annual Precipitation across Alaska
(Average across Two SNAP GCMs)
2070
2030
2050
2090
LO
CO
0-
O
a:
LO
•*s-
Q.
o
cr
^ s
vt
\
I 30
20
10
-10
-20
1 -30
% Change
20
-------
MODELING FRAMEWORK
Scenarios and Projections
Figure 1.13. Percent Change from Reference Period in Maximum Daily Precipitation across the
Contiguous U.S. (Average across the Five LOCA GCMs)
2030
2050
2070
2090
1 -30
% Change
Figure 1.14. Percent Change from Reference Period in Maximum Monthly Precipitation across Alaska
(Average across the Two SNAP GCMs)
2030
2050
2070
2090
™ -20
% Change
Lastly, climate change may affect the frequency of drought across the country. As shown in Figure 1.15,
areas of the Northwest and Southwest of the contiguous U.S. are projected to experience an increase in
the number of consecutive dry days by the end of the century, particularly under RCP8.5. Many other
parts of the country are projected to show decreases in the number of consecutive dry days. Section A.4
of the Appendix to this Technical Report presents the results for each GCM.
Period in Consecutive Dry Days across the Contiguous
2070
2090
¦ -10
I-20
-30
% Change
Figure 1.15. Percent Change from Reference
U.S. (Average across the Five LOCA GCMs)
2030 2050
21
-------
MODELING FRAMEWORK
CIRA Project Background
2. CIRA PROJECT BACKGROUND
2.1. ADVANCEMENTS IN THE CIRA FRAMEWORK
The results presented in this report build off the CIRA framework developed for the 2015 CIRA report
Climate Change in the United States: Benefits of Global Action.54 The CIRA modeling framework has been
updated to be consistent with USGCRP-recommended modeling guidelines such that these results may
serve as inputs to the forthcoming NCA4. Specifically, this Technical Report utilizes radiative forcing and
climate scenarios consistent with those being used in NCA4. For more information on the modeling
framework from the first phase of the CIRA modeling project, see pages 10-19 of the 2015 report.55 For
more information on the climate scenarios and projections used in this Technical Report, see the
Modeling Framework section.
Table 2.1 below provides a brief summary of differences between the two modeling frameworks. As a
result of using different scenario frameworks, and alternative or enhanced methodologies for some
sectors, differences exist in the magnitude of results reported in this document and those found in the
2015 report.56 Importantly, both frameworks enable the quantification of climate change impacts that
may have positive or negative effects on human health, the environment, and the economy.
In addition to updating the modeling framework, this report also includes multiple new sectoral impacts
not included in the 2015 report. These include: Aeroallergens; West Nile Virus; Harmful Algal Blooms;
Domestic Migration; Rail; Alaska Infrastructure; Winter Recreation; and the consideration of
international crop impacts affecting U.S. agricultural markets. Several sectoral models were expanded or
enhanced from the 2015 report, including: Bridges; Urban Drainage; Water Quality; Inland Flooding;
Shellfish; and the inclusion of hydropower availability in Electricity Demand and Supply. Two sectors
present in the 2015 report use different methods or models in this report, including: Air Quality and
Municipal and Industrial Water Supply (formerly referred to as Water Supply and Demand). Finally,
there are two sectors from the 2015 report that are not included in this report: Forestry (Timber Yields
and Market Effects) and Drought. In the case of forestry modeling, the underlying timber yield modeling
was not completed in time for inclusion in this report. For drought, the effects of climate change on
precipitation are captured in a number of other sectors which experience physical and economic
impacts associated with changes in drought. These include infrastructure sectors (e.g., Bridges), water
resource sectors, Agriculture, and Wildfires.
Several sectoral analyses in this Technical Report improve representation of adaptation and analysis of
the costs, benefits, and effectiveness of adaptation options, including Extreme Temperature Mortality,
Infrastructure (Roads, Bridges, Rail, and Coastal Property), and Winter Recreation. This not only allows
for improved estimates of climate change impacts and economic costs, but contributes to the
knowledge base to inform responses at national, regional, and local levels. Future CIRA work will likely
expand the coverage of sectoral impacts, leverage improved economic methods for valuing impacts,
explore the effectiveness of adaptation in more sectors, and expand the consideration of how impacts
are distributed across different socioeconomic populations.
54 EPA, 2015: Climate Change in the United States: Benefits of Global Action. United States Environmental Protection Agency, Office of
Atmospheric Programs, EPA 430-R-15-001.
"Ibid.
56 In addition, damages were presented in $2014 in the 2015 report and are presented in $2015 in this Technical Report.
22
-------
MODELING FRAMEWORK
CIRA Project Background
Table 2.1. Comparison of Modeling Frameworks
2015 Report (CIRA1.0)
2017 Report (CIRA2.0)
Climate Forcing
Scenarios
Emissions scenarios developed specifically
for benefits analysis57 using the Emissions
Predictions and Policy Analysis (EPPA-5)
model.
Representative Concentration Pathways
(RCPs), consistent with scenario selection
for NCA4.58
Business as usual with a GHG radiative
forcing of 8.6 W/m2.
Global GHG mitigation scenario at 3.2
W/m2, limiting the increase in global mean
temperature to 2°C by 2100.
Severe: RCP8.5
Moderate: RCP4.5
General Circulation
Models
Integrated Global System Model-
Community Atmospheric Model
Framework (IGSM-CAM), and the IGSM
pattern-scaling methodology.59
Coupled Model Intercomparison Project
(CMIP5)
Bias-Correction and
Downscaling
Bias-correction using delta method for
temperature and ratio method for
precipitation.
Externally-developed datasets: LOCA60for
the contiguous U.S. and SNAP for Alaska.61
Eras
Era length varies by sectors.
Reference period varies by sector.
• 2025
• 2050
• 2075
• 2100
20-year eras.
Reference period: (generally 1986-2005).
• 2030 (2020-2039)
• 2050 (2040-2059)
• 2070 (2060-2079)
• 2090 (2080-2099)
Sea Level Rise
Scenarios
Business as usual scenario: 56 in. by 2100.
Mitigation scenario: 37 in. by 2100.
Six 2100 sea levels (0.3, 0.5,1.0,1.5, 2.0,
and 2.5 m) are given RCP-specific
weights62 consistent with the USGCRP
CSSR.
57 Paltsev, S., E. Monier, J. Scott, A. Sokolov, and J. Reilly, 2013: Integrated economic and climate projections for impact assessment. Climatic
Change, doi:10.1007/sl0584-013-0892-3.
58 U.S. Global Change Research Program, 2015: U.S. Global Change Research Program General Decisions Regarding Climate-Related Scenarios
for Framing the Fourth National Climate Assessment. USGCRP Scenarios and Interpretive Science Coordinating Group. Available online at
https://scenarios.globalchange.gov/accouncement/1158
59 Monier, E., X. Gao, J.R. Scott, A.P. Sokolov, and C.A. Schlosser, 2014: Aframeworkfor modeling uncertainty in regional climate change.
Climatic Change, doi:10.1007/sl0584-014-1112-5.
60 U.S. Bureau of Reclamation, Climate Analytics Group, Climate Central, Lawrence Livermore National Laboratory, Santa Clara University,
Scripps Institution of Oceanography, U.S. Army Corps of Engineers, and U.S. Geological Survey, 2016: Downscaled CMIP3 and CMIP5 Climate
Projections: Release of Downscaled CMIP5 Climate Projections, Comparison with Preceding Information, and Summary of User Needs. Available
online at http://gdo-dcp.ucllnl.org/downscaled cmip proiections/techmemo/downscaled climate.pdf. Data available at http://gdo-
dcp.ucllnl.org/downscaled cmip projections/
61 University of Alaska Fairbanks, cited 2017: SNAP: Scenarios Network for Alaska and Arctic Planning. International Arctic Research Center.
Available online at: https://www.snap.uaf.edu/
62 Kopp, R.E., R.M. Horton, C.M. Little, J.X. Mitrovica, M. Oppenheimer, D.J. Rasmussen, B.H. Strauss and C. Tebaldi, 2014: Probabilistic 21st and
22nd century sea-level projections at a global network of tide-gauge sites. Earth's Future, 2, 383-406, doi:10.1002/2014EF000239.
23
-------
MODELING FRAMEWORK
CIRA Project Background
2015 Report (CIRA1.0)
2017 Report (CIRA2.0)
Population
County-scale projections from the ICLUS
model, driven by a single national
population projection from EPPA. U.S.
population reaches 514 million by 2100.
County-scale projections from the ICLUSv2
model, driven by the UN Median variant
scenario. U.S. population reaches 450
million by 2100.
Economic Growth
GDP projections under each emission
scenario from the EPPA-5 model.63
A single GDP projection from the EPPA-6
model64 consistent with the national
population projection.
Dollar Year of
Reported Damages
$2014
$2015
63 Paltsev, S., E. Monier, J. Scott, A. Sokolov, and J. Reilly, 2013: Integrated economic and climate projections for impact assessment. Climatic
Change, doi:10.1007/sl0584-013-0892-3.
64 Chen, Y.-H. H., S. Paltsev, J. Reilly, J. Morris, and M. Babiker, 2015: The MIT EPPA6 Model: Economic Growth, Energy Use, and Food
Consumption. MIT Joint Program on the Science and Policy of Global Change, Report 278, Cambridge, MA. Available online at
http://globalchange.mit.edu/research/publications/2892
24
-------
MODELING FRAMEWORK
CIRA Project Background
2.2. METRICS AND SCOPE OF ANALYSES
As noted previously, only a portion of the impacts of climate change are estimated, and therefore this
Technical Report captures just a fraction of the potential risks and damages (or benefits) that may be
avoided or reduced through GHG mitigation. The extent to which the methodologies encapsulate the
full range of impacts varies from sector to sector, and are also dependent on data availability and
modeling limitations. Table 2.2 below briefly summarizes the scope of physical and economic analysis
modeled in each sector. For more information on methods, see the individual sector sections and the
underlying research papers cited.
Table 2.2. Scope of Physical and Economic Analysis by Sector
Scope of Physical Analysis
Economic Valuation of the
Impact
HEALTH
Air Quality
Future ozone concentrations and resulting
number of premature deaths
Value of a statistical life (VSL)
Aeroallergens
Change in oak pollen season length and
concentrations, and resulting number of
emergency department visits for asthma
Emergency department cost-per-
visit
Extreme Temperature
Mortality
Number of premature deaths attributable to
extreme hot and cold temperatures in 49 cities
VSL
Labor
Lost labor supply hours due to changes in hot
and cold temperature, including extreme
temperatures
Lost Wages
West Nile Virus
Impact of temperature on number of West Nile
Neuroinvasive Disease cases
VSL and hospitalization costs
Harmful Algal Blooms
Change in occurrence of cyanobacterial
harmful algal blooms in 279 reservoirs
Lost consumer surplus from
reservoir recreation
Domestic Migration
Percent change in population
N/A
INFRASTRUCTURE
Roads
Vulnerability of paved, unpaved, and gravel
roads to changes in temperature, precipitation,
and freeze-thaw cycles
Reactive or proactive repair or
reconstruction costs to maintain
level of service
Bridges
Vulnerability of non-coastal bridges to changes
in peak water flow
Costs of proactive maintenance
and repairs to maintain level of
service
Rail
Vulnerability of the Class 1 rail network
(passenger and freight) to changes in
temperature
Costs of delays (reduced speed
and traffic) to railroad companies
and to public, and proactive
adaptation costs to install sensors
Alaska Infrastructure
Vulnerability of roads, buildings, airports,
railroads, and pipelines to changes in
permafrost thaw, freeze-thaw cycles,
precipitation, and precipitation-induced
flooding
Reactive and proactive adaptation
expenditures to maintain level of
service
Urban Drainage
Change in urban drainage volume from more
intense rainfall and increased runoff in 100
cities
Proactive adaptation costs to
implement stormwater best
management practices
Coastal Property
Vulnerability of on-shore property to sea level
rise and storm surge
Value of abandoned property and
costs of protection
25
-------
MODELING FRAMEWORK
CIRA Project Background
Scope of Physical Analysis
Economic Valuation of the
Impact
ELECTRICITY
Electricity Demand and
Supply
Changes in energy demand and supply and
hydropower generation in response to changes
in temperature and flow
Electric power system costs
(capital, O&M, fuel costs)
WATER RESOURCES
Inland Flooding
Changes in frequency of 100-year riverine
flooding events
Damages to assets located in
floodplains (e.g., buildings)
Water Quality
Changes in river, lake, and reservoir water
quality based on modeling of temperature,
dissolved oxygen, total nitrogen, total
phosphorus
Willingness to pay to offset
changes in water quality index
Municipal and
Industrial Water Supply
Changes in water supply to meet municipal
indoor, municipal outdoor, and industrial water
demands
Consumer welfare
Winter Recreation
Impact of snowpack on recreation visits for
downhill skiing and snowboarding, cross-
country skiing, and snowmobiling at 247
locations
Lost recreation (lift ticket and
entry prices)
AGRICULTURE
Domestic Yield and
Welfare Effects
Impacts of changing climate conditions on
yields of major U.S. crops (e.g., corn, soybean,
wheat, alfalfa hay, cotton), and the subsequent
decisions landowners may make regarding crop
mix, production practices, and land allocation
Producer and consumer welfare
U.S. and Global
Agriculture Interactions
Impact of global agricultural changes on U.S.
crop (corn, soybean, wheat) yields, production,
consumption, and price
N/A
ECOSYSTEMS
Coral Reefs
Percent change in shallow coral reef cover in
Hawaii, South Florida, and Puerto Rico
Lost recreational value
Shellfish
Effects of ocean acidification on growth rates of
oysters, scallops, geoducks, quahogs, and
clams, with subsequent effects on shellfish
supply
Consumer welfare
Freshwater Fish
Change in the spatial distribution of suitable
habitat for coldwater, warmwater, and rough
fish living in rivers and streams.
Lost recreational value
Wildfire
Change in terrestrial ecosystem vegetative
cover and acres burned on non-agricultural,
rural lands.
Response costs
Carbon Storage
Terrestrial carbon flux (storage and annual
flows) in metric tons
N/A
26
-------
MODELING FRAMEWORK
CIRA Project Background
2.3. SOURCES OF UNCERTAINTY
The modeling framework for analyses underlying this Technical Report enables the comparison of
physical and economic impacts of climate change across a large number of U.S. sectors in a consistent
fashion. As with any study, there are sources of uncertainty that are important to consider, several of
which are described below. Future work to address these will further strengthen confidence in the
estimates presented in this report. Limitations specific to the individual sectoral analyses are described
in those sections of this report, as well as in the peer-reviewed literature underlying the analyses.
Emissions and Climate Scenarios
With the goal of presenting a consistent and straightforward set of climate change impact analyses
across sectors, this Technical Report presents results using climate projections from five GCMs under
RCP8.5 and RCP4.5. These two RCPs, along with the use of climate projections from the LOCA
downscaled dataset, match those recommended for use in NCA4.65 Due to the level of effort necessary
to run each scenario through the large number of sectoral models of the project, the CIRA project uses
only five of the more than twenty GCMs with daily climate data from the CMIP5 ensemble. While these
five GCMs were chosen to capture a large range of the variability observed across the entire ensemble,
this subset is not a perfect representation. Analyzing results under the full set of CMIP GCMs would
better characterize the range and potential likelihood of future risks.
Even the full set of CMIP5 GCMs is not likely to span the entire range of potential physical responses of
the climate system to changes in the concentration of atmospheric GHGs. Previous literature has
demonstrated the importance of climate sensitivity assumptions in understanding a wide range of
potential changes to the climate system,66,67 as well as the effect of natural variability on timing and
magnitude of impacts.68,69The first phase of CIRA modeling investigated the relative importance of four
types of uncertainty inherent to projecting future climate: emissions scenarios, climate sensitivity,
natural variability, and climate model. For temperature, projected changes were most influenced by
decisions regarding whether to reduce GHG emissions and the value of climate sensitivity used (GHG
emissions scenario being the dominant contributor). Conversely, these same four sources of uncertainty
contribute in roughly equal measure to projected changes in precipitation over the U.S., with large
spatial differences.70
Coverage of Sectors and Impacts
The analyses presented in this Technical Report cover a broad range of potential climate change
damages or benefits in the U.S., but many important impacts are not included. Examples of these
omitted impacts include other health effects (e.g., mortality due to extreme events other than heat
waves; food safety and nutrition; mental health and behavioral outcomes), effects on ecosystems (e.g.,
changes in marine fisheries; impacts on specialty crops and livestock; species migration and
65 U.S. Global Change Research Program, 2015: U.S. Global Change Research Program General Decisions Regarding Climate-Related Scenarios
for Framing the Fourth National Climate Assessment. USGCRP Scenarios and Interpretive Science Coordinating Group. Available online at
https://scenarios.globalchange.gov/accouncement/1158
66 Paltsev, S., E. Monier, J. Scott, A. Sokolov, and J. Reilly, 2013: Integrated economic and climate projections for impact assessment. Climatic
Change, doi:10.1007/sl0584-013-0892-3.
67 Monier, E., X. Gao, J.R. Scott, A.P. Sokolov, and C.A. Schlosser, 2014: Aframeworkfor modeling uncertainty in regional climate change.
Climatic Change, doi:10.1007/sl0584-014-1112-5.
68 Monier, E., and X. Gao, 2014: Climate change impacts on extreme events in the United States: an uncertainty analysis. Climatic Change,
doi:10.1007/sl0584-013-1048-l.
69 Mills, D., R. Jones, K. Carney, A. St Juliana, R. Ready, A. Crimmins, J. Martinich, K. Shouse, B. DeAngelo, and E. Monier, 2014: Quantifying and
Monetizing Potential Climate Change Policy Impacts on Terrestrial Ecosystem Carbon Storage and Wildfires in the United States. Climatic
Change, doi:10.1007/sl0584-014-1118-z.
70 EPA. 2015. Climate Change in the United States: Benefits of Global Action. United States Environmental Protection Agency, Office of
Atmospheric Programs, EPA 430-R-15-001.
27
-------
MODELING FRAMEWORK
CIRA Project Background
distribution), and social impacts (e.g. national security; violence). Without information on these impacts,
this report provides only partial insight into the potential risks of climate change. Benefits of climate
change, for instance due to reductions in cold related deaths or from carbon fertilization, were
considered and are reported in the sectors.
In addition, it is important to note that impacts are only partially valued economically in many sectors.
For example, the Wildfire section presents estimated response costs, but not other damages (e.g.,
health effects from decreased air quality or water contamination, property damages, or loss of
recreational space). Therefore, the damages described in this report are likely an undervaluation of the
actual climate impacts that would occur under any given scenario. Furthermore, the methods used to
calculate economic impacts differ between sectors. For example, the Air Quality sector estimates the
economic value of premature death using a value of statistical life (VSL), a form of willingness-to-pay
(WTP) for reductions in the risk of death, while the Water Quality sector estimates economic
implications using WTP for improved water quality. Sectors like Aeroallergens and West Nile Virus use
mean costs of hospital visits. Some sectors, such as Freshwater Fish and Shellfish sectors, use estimates
of consumer surplus. Still other sectors measure adaptation or response costs. These varying methods of
estimating impacts may not capture all costs.
Finally, this Technical Report does not present results on the possibility of large-scale, abrupt changes
that have wide-ranging and possibly catastrophic consequences, such as the intensification of tropical
storms, or the rapid melting of the Greenland or West Antarctic ice sheets.71 In general, there are many
uncertainties regarding the timing, likelihood, and magnitude of the impacts resulting from these abrupt
changes, and data limitations have precluded their inclusion in the analyses presented in this report.
Their inclusion would assist in better understanding the totality of risks posed by climate change.
Sectoral Impacts Modeling
With the exception of several sectors of this report (e.g., Electricity Demand and Supply; Water Quality),
the impact estimates presented were developed using a single sectoral impact model. These models are
complex analytical tools, and choices regarding the structure and parameter values of the model can
create important assumptions that affect the estimation of impacts. Ongoing studies, such as the Inter-
sectoral Impact Model Intercomparison Project (ISI-MIP), are investigating the influence of structural
uncertainties across sectoral impact models.72 The use of additional models for each sector of this
Technical Report would help improve the understanding of potential impacts in the future.
The results presented in each sector were primarily developed independently of one another. As a
result, the estimated impacts may omit important interactive effects. For example, the Wildfire
projections presented in this report will likely generate meaningful increases in air pollution, a
potentially important linkage not included in the Air Quality analysis. Similarly, there are numerous
connections among the agriculture, water, and electricity sectors that affect impacts estimates.73
Although some of these interactions are captured within integrated assessment models, it is difficult for
these broader frameworks to capture all of the detail provided in the sectoral analyses. Although first
71 For more information on these types of impacts, see: National Research Council, 2013: Abrupt Impacts of Climate Change: Anticipating
Surprises. Washington, DC: The National Academies Press.
72 Huber, V., H.J. Schellnhuber, N.W. Arnell, K. Frieler, A.D. Friend, D. Gerten, I. Haddeland, P. Kabat, H. Lotze-Campen, W. Lucht, M. Parry, F.
Piontek, C. Rosenzweig, J. Schewe, and L. Warszawski, 2014: Climate impact research: beyond patchwork. Earth System Dynamics, 5, 399-408.
73 For a discussion of interactions among the energy, water, and land use sectors, see: Hibbard, K., T. Wilson, K. Averyt, R. Harriss, R. Newmark,
S. Rose, E. Shevliakova, and V. Tidwell, 2014: Ch. 10: Energy, Water, and Land Use. Climate Change Impacts in the United States: The Third
National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program.
28
-------
MODELING FRAMEWORK
CIRA Project Background
order connectivity was achieved in some cases (e.g., the water resource model informed irrigation water
availability in the agriculture analysis and availability for hydropower and thermo-electric cooling in the
Electricity Demand and Supply analysis), improved connectivity between sectoral models would aid in
gaining a more complete understanding of climate change impacts across sectors in the U.S.
Variability in Societal Characteristics
Though climate change will affect all Americans, it will not affect all Americans equally. In addition to
regional differences in impacts, socioeconomic factors (e.g., income, education) affect exposure,
sensitivity, or adaptive capacity, and can make some communities more or less vulnerable to impacts.74
In some cases, physiological differences place some individuals at greater risk, such as children, older
adults, persons with disabilities, or persons with preexisting or chronic medical conditions. An
individual's vulnerability to climate change impacts is also a function of their behavior, and is an
emerging area of research. In general, the results in this report do not separately report impacts for
vulnerable populations, nor analyze how individual behavior affects vulnerability. However, some
sectors explore these issues, such as the social vulnerability section of the Coastal Property sector, and
separation of impacts across different age groups in the Air Quality and Aeroallergen sectors.
Feedbacks
The modeling framework analyzes changes in socioeconomics and climatic drivers on impacts, with
consistent inputs across multiple models. The socioeconomic scenarios that drive the modeling analyses
do not incorporate potential feedbacks from climate change impacts to the socioeconomic system (e.g.,
changes in albedo from land use change or increased GHG emissions resulting from vegetative changes)
nor from sectoral damages to the economy (e.g., significant expenditures on protective adaptation
measures, such as seawalls, would likely reduce available financial capital to the economy for other
productive uses).
Geographic Coverage
In general, this Technical Report does not examine impacts and damages occurring outside of U.S.
borders. Aside from the inherent value of people and ecosystems around the world, these impacts could
also affect the U.S. through, for example, changes in migration and concerns for national security.
In addition, the primary geographic focus of this report is on the contiguous U.S., with most of the
sectoral analyses excluding Hawai'i, Alaska, and the U.S. territories.75 This omission is particularly
important given the unique climate change vulnerabilities of these high-latitude and/or island locales.
Finally, several sectoral analyses assess impacts in a limited set of major U.S. cities (e.g., Extreme
Temperature Mortality; Urban Drainage), and incorporation of additional locales would gain a more
comprehensive understanding of likely impacts.
74 Gamble, J.L., J. Balbus, M. Berger, K. Bouye, V. Campbell, K. Chief, K. Conlon, A. Crimmins, B. Flanagan, C. Gonzalez-Maddux, E. Hallisey, S.
Hutchins, L. Jantarasami, S. Khoury, M. Kiefer, J. Kolling, K. Lynn, A. Manangan, M. McDonald, R. Morello-Frosch, M.H. Redsteer, P. Sheffield, K.
Thigpen Tart, J. Watson, K.P. Whyte, and A.F. Wolkin, 2016: Ch. 9: Populations of Concern. The Impacts of Climate Change on Human Health in
the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 247-
286. http://dx.doi.org/10.7930/J0Q81B0T
75 Infrastructure and wildfire impacts were estimated in Alaska, and effects on coral reefs were estimated for Hawai'i and Puerto Rico.
29
-------
MODELING FRAMEWORK
CIRA Project Background
2.4. REVIEW OF RELATED LITERATURE
The third National Climate Assessment notes that valuation of the economic consequences of climate
change and estimation of how mitigation and adaptation can reduce risk have been lacking in recent
assessments, and were identified as priorities for future research.76,77 In recent years, research on
climate change risks and economic impacts has advanced considerably both in the U.S. and globally,
which has improved our understanding of the physical, environmental, social and economic impacts of
climate change. In the U.S., several large-scale, coordinated analytical efforts have been initiated to
evaluate the economic consequences of climate change to the U.S. Specifically, the CIRA project uses a
multi-model framework to systematically assess risks, impacts, and economic damages of climate
change on human health, key economic sectors, and ecosystems in the U.S.78 Two phases of CIRA
modeling have been completed to date: the first phase released in 2015,79 and the second phase, which
is presented in this Technical Report. Another study, the American Climate Prospectus (ACP), assesses
the economic risks of climate change to U.S. households, businesses, and the economy.80,81 At the state-
level, the California Climate Change Assessments provide region-specific projections of physical and
economic impacts across a number of sectors.82 These studies, among others, show that significant risks
and economic damages in the U.S. will occur in the 21st century under high emissions scenarios. Further,
climate change mitigation, and adaptation in some sectors, would substantially reduce the most
dangerous risks of climate change, thereby avoiding costly damages.
Both the CIRA and ACP analyses have developed internally consistent analytical frameworks (e.g.,
climate forcing, socioeconomic scenarios) to quantify climate change impacts and economic damages
across a range of impact categories (including human health, agriculture, coastal areas, and electricity)
at the national, regional and local scales. However, there are some key differences between the studies.
Both phases of CIRA modeling emphasize bottom-up, process-based understanding of climate change
impacts and rely on numerous sectoral simulation models to estimate the biophysical and economic
impacts of climate change, including a range of non-market, ecosystem impacts (e.g., wildfire, coral
reefs, and marine habitats). The ACP analysis focuses its quantitative analysis on a more limited number
of sectoral impacts and economic risks (i.e., in agriculture, labor, health, crime, energy and coastal
communities), and discusses important qualitative impacts and risks in other sectors (i.e., in water,
forests, tourism, and security). The ACP relies on empirically-derived econometric models of climate
impacts in four of the quantified sectors and on process-based models in two other sectors (energy and
coastal communities). Additionally, the ACP integrates these direct probabilistic impacts into a macro-
economic model to assess the indirect effects of climate impacts on the economy.
76 Corell, R. W., D. Liverman, K. Dow, K. L. Ebi, K. Kunkel, L. O. Mearns, and J. Melillo, 2014: Ch. 29: Research Needs for Climate and Global
Change Assessments. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.)
Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 707-718. doi:10.7930/J03R0QR3.
77 Liverman, D., 2015: U.S. national climate assessment gaps and research needs: overview, the economy and the international context.
Appearing in The U.S. National Climate Assessment, Springer doi: 10.1007/978-3-319-41802-5_13.
78 Waldhoff, S., Martinich, J., Sarofim, M., DeAngelo, B., McFarland, J., Jantarasami, L., Shouse, K., Crimmins, A., Ohrel, S. and Li, J., 2014:
Overview of the special issue: a multi-model framework to achieve consistent evaluation of climate change impacts in the United
States. Climatic Change, pp.1-20.
79 EPA, 2015: Climate Change in the United States: Benefits of Global Action. United States Environmental Protection Agency, Office of
Atmospheric Programs, EPA 430-R-15-001.
80 Houser, T., Kopp, R., Hsiang, S.M., Delgado, M., Jina, A., Larsen, K., Mastrandrea, M., Mohan, S., Muir-Wood, R., Rasmussen, D.J., Rising, J.,
and Wilson, P., 2015: Economic Risks of Climate Change: An American Prospectus. Columbia University Press.
81 Gordon, K., 2014: Risky Business: The Economic Risks of Climate Change in the United States. A product of the Risky Business Project.
82 State of California, cited 2017: California Climate Change Assessments. Available online at
http://climatechange.ca.gov/climate action team/reports/climate assessments.html
30
-------
MODELING FRAMEWORK
CIRA Project Background
While the CIRA analyses presented in this Technical Report use the deterministic projections based on
the IPCC Fifth Assessment Report (AR5) RCP and CMIP5 scenario framework, the ACP developed a new
dataset exploring the probabilistic distributions of AR5 projections to estimate and communicate the
risks of climate extremes. Both studies analyze and characterize many sources of uncertainty in the
estimates of climate change impacts and damages.
Internationally, there are significant ongoing multi-sector, multi-model analytical efforts that use
internally consistent scenarios to assess the impacts of climate change at sectoral, regional, and global
levels.83 These efforts have been developed with diverse purposes and exhibit different features
(Appendix A.5 provides a summary of the recent multi-sector, multi-model studies reviewed and their
key features). Of the studies reviewed, a majority of the projects evaluate climate change impacts and
the implications of climate change mitigation under alternative climate forcing scenarios (e.g., ISI-MIP84,
BRACE85, PESETA86); while a few focus on understanding climate change impacts under a business-as-
usual (i.e., no climate change mitigation) scenario (e.g., CIRCLE87), or impacts under high warming
scenarios (IMPRESSIONS88 and HELIX89). A majority of the studies examine multiple categories of climate
change impacts (e.g., agriculture, health, infrastructure, coastal areas, energy), while some focus on a
specific sector, such as agriculture (AgMIP90), ecosystems,91 and coastal areas (RISES-AM92). While most
of the studies estimate climate change impacts in individual sectors, several studies have begun to
investigate the effects of cross-sector interactions and the structural uncertainty across sectoral impact
models (e.g., ISI-MIP, IMPRESSIONS). These studies vary in their outputs: some studies have a stronger
focus on understanding the physical and biophysical impacts of climate change (e.g., BRACE) and a few
studies focus on understanding the economic impacts of climate change in key economic sectors (e.g.,
CIRCLE, PESETA). Most studies include both biophysical and economic outcomes. Geographic coverage
varies among these studies: some of the studies have a global focus (e.g., ISI-MIP), some focus on
Europe (e.g., PESETA), but a number of studies provide both a global perspective and investigation in
specific regions and/or countries (e.g., BRACE, IMPRESSIONS, HELIX). All of these studies have a strong
emphasis on understanding and characterizing uncertainties, such as those arising from climate and
socioeconomic scenarios and model structure. Some studies incorporate innovative approaches, such as
agent-based modeling to simulate adaptive responses and analysis of institutional and behavioral
constraints to adaptation (e.g., IMPRESSIONS). Several projects build in stakeholder engagement and
83 See discussion in Huber, V., H.J. Schellnhuber, N.W. Arnell, K. Frieler, A.D. Friend, D. Gerten, I. Haddeland, P. Kabat, H. Lotze-Campen, W.
Lucht, and M. Parry, 2014: Climate impact research: beyond patchwork. Earth System Dynamics, 5, 399.
84 Warszawski, L., K. Frieler, V. Huber, F. Piontek, O. Serdeczny, and J. Schewe, 2014: The inter-sectoral impact model intercomparison project
(ISI-MIP): project framework. Proceedings of the National Academy of Sciences, 111, 3228-3232.
85 O'Neill, B., and A. Gettelman, Eds., 2016: The Benefits of Reduced Anthropogenic Climate changE (BRACE) Project. University Corporation for
Atmospheric Research, National Center for Atmospheric Research. Climatic Change Special Issue. Available online at
https://chsp.ucar.edu/brace-climatic-change-special-issue
86 Ciscar, J.C., L. Feyen, A. Soria, C. Lavalle, F. Raes, M. Perry, F. Nemry, H. Demirel, M. Rozsai, A. Dosio, M. Donatelli, A. Srivastava, D. Fumagalli,
S. Niemeyer, S. Shrestha, P. Ciaian, et al., 2014: Climate Impacts in Europe. The JRC PESETA II Project. JRC Scientific and Policy Reports, EUR
26586EN.
87 OECD, 2015: The Economic Consequences of Climate Change. OECD Publishing, Paris, doi: 10.1787/9789264235410-en
88 Capela Lourengo, T., M.J. Cruz, H. Carlsen, A. Dzebo, J.D. Tabara, F. Cots, J. Haslett, and P. Harrison, 2015: Common Frame of Reference to
support the understanding of adaptation decisionmaking under high-end scenarios. EU FP7 IMPRESSIONS Project Deliverable Dl.l.
89 HELIX, cited 2017: High-End cLimate Impacts and extremes (HELIX). Available online at http://helixclimate.eu/home
90 Rosenzweig, C., J.W. Jones, J.L. Hatfield, A.C. Ruane, K.J. Boote, P.Thorburn, J.M. Antle, G.C. Nelson, C. Porter, S. Janssen, and S. Asseng,
2013: The agricultural model intercomparison and improvement project (AgMIP): protocols and pilot studies. Agricultural and Forest
Meteorology, 170, 166-182.
91 Scholze, M., Knorr, W., Arnell, N.W. and Prentice, I.e. 2006: A climate-change risk analysis for world ecosystems PNAS, 103 (35), 13116-13120
92 RISES-AM-, cited 2017: RISES-AM- EU Research Project. Available online at http://www.risesam.eu/
31
-------
MODELING FRAMEWORK
CIRA Project Background
decision support in the analytical frameworks to support adaptation and resilience decisions (RISES-AM,
IMPRESSIONS, HELIX).
In parallel to the advancements in climate impact assessments that utilize integrated assessment models
and sectoral or economic simulation models, there have been significant advancements in the use of
improved empirical methods and data that enable climate impact assessments based on empirically-
derived damage functions.93'94,95 A recent review discussed these advancements and synthesized the
state of knowledge on the social and economic impacts of climate change, including on human health
(mortality, morbidity, early life development), labor productivity, agriculture, energy, infrastructure,
trade, income, macroeconomy, social interactions, and demographic trend.96 The synthesis suggests that
both the current and projected future climate change has significant, negative impacts on human health,
welfare, society and the global economic growth. Overall, as suggested by various reviews,
representation of adaptation responses and pathways is limited in all the existing studies. Analysis of
ecosystem and non-market impacts of climate change varies considerably among the studies and is
missing in many studies.
To date, the CIRA and ACP analyses have produced the most detailed estimates of the economic impacts
of climate change across a range of sectors at both national and subnational levels in the U.S. The global
studies mentioned above mostly treat the U.S. as a region, or part of a larger region (e.g., North
America). For the global studies that have subnational (e.g., grid-level) estimates in the U.S.,
assumptions for analyses are mostly based on global projections or storylines that do not align closely
with nationally-derived socioeconomic scenarios and projections (e.g., for demographic change, land
use, economic growth). Such storylines are potentially less useful to inform the NCA4, as it would
require subnational estimates of climate change impacts for the major sectors that use those scenarios
to be consistent with the rest of the assessment.
93 Deschenes, O. and M. Greenstone, 2011: Climate change, mortality, and adaptation: evidence from annual fluctuations in weather in the US.
American Economic Journal: Applied Economics, 3, 152-185.
94 Lobell, D.B., W. Schlenker, and J. Costa-Roberts, 2011: Climate trends and global crop production since 1980. Science, 333, 616-620.
95 Schlenker, W. and M.J. Roberts, 2009: Nonlinear temperature effects indicate severe damages to US crop yields under climate change.
Proceedings of the National Academy of sciences, 106, 15594-15598.
96 Carleton, T.A. and S.M. Hsiang, 2016: Social and economic impacts of climate. Science, 353, aad9837.
32
-------
MODELING FRAMEWORK
CIRA Project Background
2.5. TERMS AND ACRONYMS COMMONLY USED IN THIS REPORT
Adaptation Costs: Economic damages incurred when implementing a wide range of adaptation options,
whether a reactive response (i.e., implemented in response to climate change impacts that have already
occurred) or proactive measure (i.e., implemented in anticipation of future climate change impacts).
Control Scenario: The control scenario, or the "no climate change scenario," refers to a modeled future
scenario that does not include climate change, but typically includes future changes in population
and/or economic growth. These control scenarios allow for the isolation of climate change impacts from
other non-climate drivers of change; specific sectoral application is described in each Approach section.
Damage: Economic damages are the monetized value of impacts attributed to climate change. Metrics
of damages differ by sector; these metrics are shown in Table 2.2 above and described in each sectoral
Approach section.
GCM: Global Climate Models are mathematical models that simulate the physics, chemistry, and biology
that influence the climate system. Related term: General Circulation Model.
5-GCM: In this report, the analyses generally use five global climate models: CanESM2, CCSM4, GISS-E2-
R, HadGEM2-ES, and MIROC5 (see Selection of GCMs above). In many instances, estimates are
presented as a "five-model average," which means the average of the results of these five GCMs.
Impact: Climate change impacts are the physical or economic effects (positive or negative) occurring in
response to climate stressors.
RCP: Representative Concentration Pathways are GHG concentration trajectories from the
Intergovernmental Panel on Climate Change's (IPCC's) Fifth Assessment Report (2014) that reflect
possible increases in radiative forcing associated with emissions over time.
RCP8.5: The Representative Concentration Pathway with an approximate total radiative forcing (not
emissions) in the year 2100, relative to 1750, of 8.5 W/m2. RCP8.5 implies a future with continued high
emissions growth under limited efforts to reduce GHGs. Thus, in this report, RCP8.5 represents the
"higher" or "more severe" scenario. Under RCP8.5, global atmospheric C02 levels rise from current-day
levels of approximately 400 up to 936 parts per million (ppm) by 2100.
RCP4.5: The Representative Concentration Pathway with an approximate total radiative forcing (not
emissions) in the year 2100, relative to 1750, of 4.5 W/m2. RCP4.5 implies a future with moderate
emissions growth under substantial global efforts to reduce GHGs. Thus, in this report, RCP4.5
represents the "lower" or "less severe" scenario. Under the RCP4.5, atmospheric C02 levels at the end of
the century remain below 550ppm.
Reference Period: The reference period or the historical reference period refers to the period of years
upon which future impacts are compared. The years included in these periods are described in each
sectoral Approach section.
Sector: This report is separated into impact sectors based primarily on the underlying model used in the
analysis and the category of climate change impacts estimated within that larger topic area.
33
-------
HEALTH
3. AIR QUALITY
3.1 KEY FINDINGS
• Climate change is expected to result in weather conditions that are increasingly conducive to high
concentrations of ground-level ozone over many parts of the U.S.
• Unless offset by additional domestic reductions in ozone precursor emissions, climate-driven
changes in ozone under RCP8.5 are projected to result in an additional 420-1,200 premature deaths
per year in 2050 and 920-2,500 premature deaths per year in 2090. Under RCP4.5, an increase of
300-810 premature deaths is projected annually in 2050, rising to 630-1,700 additional premature
deaths in 2090.
• Annual national costs of climate-driven, premature ozone-related deaths under RCP8.5 are
projected to be $9.8 billion in 2050 and $26 billion in 2090. Under RCP4.5, annual costs of
premature deaths are projected to be $6.9 billion in 2050 and $18 billion in 2090.
• Compared to RCP8.5, the level of summer season ozone projected under the RCP4.5 scenario
demonstrates a substantially lower burden of air pollution on U.S. respiratory health by reducing
ozone concentrations, thereby reducing the need for further emissions controls on domestic sources
to address ozone air pollution.
3.2 INTRODUCTION
As of 2015, more than 120 million Americans live in counties where air pollution levels exceed the
National Ambient Air Quality Standards (NAAQS). Ground-level ozone and fine particle pollution (PM2.5)
are the overwhelming contributors to these cases of poor air quality, and both pollutants have
significant adverse effects on human health through respiratory and cardiovascular impacts. Actions to
reduce the emissions that lead to high levels of ozone and fine particles have been highly successful over
the past 15 years; since 2000, ozone levels have been reduced by 17% and fine particle concentrations
have been reduced by 37% nationally.97 However, climate change has the potential to slow the
improvements to U.S. air quality by altering weather patterns and increasing the prevalence of
conditions that lead to episodes of poor air quality.98 Previous studies have suggested that the future
97 EPA, 2016: National Air Quality-Status and Trends of Key Air Pollutants. United States Environmental Protection Agency, Office of Air Quality
Planning and Standards. Available online at www.epa.gov/air-trends.
98 Fann, N., T. Brennan, P. Dolwick, J.L. Gamble, V. Ilacqua, L. Kolb, C.G. Nolte, T.L. Spero, and L. Ziska, 2016: Ch. 3: Air Quality Impacts. The
Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington,
DC, 69-98.
34
-------
HEALTH
Air Quality
health impacts of climate change resulting from degraded air quality could be significant, with an
economic burden ranging from hundreds of millions to hundreds of billions of dollars.99,100
3.3 APPROACH
This analysis projects impacts of climate change on future ground level ozone concentrations and
resulting health consequences within the contiguous U.S. for three periods representing the reference
(2000), mid-century (2050), and late-century (2090) conditions.101 A series of global and regional models
is used to assess future climate-driven changes in summer (May through September) ozone
concentrations over the contiguous U.S. Here, the regional climate across the contiguous U.S. is
represented by dynamically-downscaled data rather than statistically-downscaled data from LOCA. As a
result of this, there will be differences in the absolute magnitudes of climate variables (e.g., temperature
and precipitation) at local scales between the dynamically downscaled Weather Research and
Forecasting (WRF) data and the LOCA dataset used elsewhere in this Technical Report and described in
the Scenarios and Projections section. See Appendix A.6 for the WRF regional climate summaries. Air
pollutant levels can be strongly influenced by meteorological conditions other than temperature and
precipitation (e.g., mixing layer depths, wind speed and direction, among others). The dynamical
downscaling is used to create the temporal evolution of those three-dimensional meteorological data
that are required as inputs for the simulations of air quality. See Appendix A.6 for a discussion of climate
impacts on PM2.5.
Climate impacts are simulated with the CCSM4 GCM102 under RCP8.5 and RCP4.5 and dynamically
downscaled over North America using the WRF model103,104 to provide climate-influenced
meteorological inputs for 2050 and 2090 relative to 2000.105 These meteorological data, along with
emissions that are projected to represent mid-century conditions,106 are used to drive the Community
Multiscale Air Quality model (CMAQ)107 over the contiguous U.S. using a 36-kilometer grid resolution.
The 2000, 2050, and 2090 estimates are each represented by 11 years of simulations centered on those
years to consider interannual variability in meteorology. The reference period simulation is driven by the
99 Fann, N., C.G. Nolte, P. Dolwick, T.L. Spero, A. Curry Brown, S. Phillips, and S. Anenberg, 2015: The geographic
distribution and economic value of climate change-related ozone health impacts in the United States in 2030. Journal of the Air & Waste
Management Association, 65, 570-580. http://dx.doi.org/10.1080/10962247.2014.996270.
100 Garcia-Menendez, F., R.K. Saari, E. Monier, and N.E. Selin, 2015: U.S. air quality and health benefits from avoided climate change under
greenhouse gas mitigation. Environmental Science and Technology, 49, 7580-7588, doi:10.1021/acs.est.5b01324
101 The estimates for 2050 and 2090 represent the average for the periods 2045-2055 and 2085-2095, respectively. The reference period for the
analysis is 1995-2005.
102 Gent, P.R., G. Danabasoglu, L.J. Donner, M.M. Holland, E.C. Hunke, S.R. Jayne, D.M. Lawrence, R.B. Neale, P.J. Rasch, M. Vertenstein, P.H.
Worley, Z.-L. Yang, and M. Zhang, 2011: The Community Climate System Model Version A.J. Climate, 24, 4973-4991,
doi:10.1175/2011JCLI4083.1
103 Skamarock, W.C., and J.B. Klemp, 2008: A time-split nonhydrostatic atmospheric model for weather research and forecasting applications. J.
Comput. Phys., 227, 3465-3485.
104 Following the methods of: Spero, T.L., C.G. Nolte, J.H. Bowden, M.S. Mallard, and J.A. Herwehe, 2016: The impact of incongruous lake
temperatures on regional climate extremes downscaled from the CMIP5 archive using the WRF model. J. Climate, 29, 839-853.
105 Due to the computational demands in dynamically downscaling GCMs, only one of the five GCMs being used throughout this Technical
Report is applied in this air quality analysis. As a result, this approach does not capture uncertainty across climate models, which may be
important for characterizing air quality impacts. As described in the Modeling Framework section of this Technical Report, CCSM4 lies close to
the ensemble mean in terms of variability in national and regional projections of annual average and seasonal temperature and precipitation
change.
106 EPA, 2016: Emissions Inventory for Air Quality Modeling Technical Support Document: Heavy-Duty Vehicle Greenhouse Gas Phase 2 Final
Rule, EPA-420-R-16-008, 210 pp.
107 Byun, D., and K. L. Schere, 2006: Review of the governing equations, computational algorithms, and other components of the Models-3
Community Multiscale Air Quality (CMAQ) modeling system. Appl. Mech. Rev., 59, 51-77.
35
-------
HEALTH
Air Quality
CCSM4 representation of meteorological conditions for that period rather than using observed
meteorological conditions. The CMAQ model is used to simulate the transformation of air pollutant
emissions to ozone in the reference period and each of the multi-year scenarios of climate change-
driven changes in meteorology. While natural emissions from vegetation respond to changes in
meteorology, domestic non-GHG emissions are held constant through the future period so that the
effects of climate change can be isolated from changes in emissions policies.108
The potential impacts of the climate-driven changes in air pollutant concentrations for specific health
endpoints are estimated using the Benefits Mapping and Analysis Program - Community Edition
(BenMAP-CE).109 Estimates of the number of individuals exposed to future levels of ozone
concentrations are projected using population counts from the Integrated Climate and Land Use
Scenarios version 2 (ICLUSv2).110 BenMAP-CE quantifies ozone-attributable deaths and illnesses using
the same suite of concentration-response functions documented in Fann et al. (2015).111 The economic
value of these premature deaths is estimated with a value of statistical life (VSL) adjusted to future years
based on a projection of economic growth.112 The results presented in this section represent the
additional deaths due to climate change compared to the observed mortality in the reference period.
Additional information on the approach is provided in Fann et al. (2015).113
3.4 RESULTS
Ozone, which is regulated by NAAQS, is known to adversely affect human health through respiratory and
cardiovascular pathways, resulting in additional reported acute respiratory symptoms, missed school
and work days, hospital admissions, and premature deaths. The analysis is focused on the summer
months when ozone concentrations tend to be higher in response to higher temperatures. The changes
presented here are for each future period relative to the reference period. The simulations represent a
single member of a large ensemble of potential future climate outcomes, and this analysis should be
viewed in that context.
108 Domestic non-GHG emissions in the future periods were based on EPA projections for the mid-century, as described in technical support
documents for the 2040 Heavy Duty Greenhouse Gas Phase 2 Rule, and did not change across climate scenarios or time periods so that the
effects of climate change could be isolated. These technical support documents are available from www.regulations.gov: EPA-HQ-OAR-2014-
0827-2301 and EPA-HQ-OAR-2014-0827-2303.
109 EPA, 2014: Environmental Benefits Mapping and Analysis Program -Community Edition (BenMAP-CE). United States Environmental
Protection Agency, Office of Air Quality Planning and Standards, Health and Environmental Impacts Division. Available at www.epa.gov/benmap
110 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (ICLUS)
(Version 2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
111 Fann, N., C.G. Nolte, P. Dolwick, T.L. Spero, A. Curry Brown, S. Phillips, and S. Anenberg, 2015: The geographic distribution and economic
value of climate change-related ozone health impacts in the United States in 2030.Journal of the Air & Waste Management Association, 65,
570-580. http://dx.doi.org/10.1080/10962247.2014.996270.
112 At the time of this analysis, the EPA's Guidelines for Preparing Economic Analyses recommends a VSL of $7.9 million ($2008) based on 1990
incomes. To create a VSL using $2015 and based on 2015 incomes, the standard value was adjusted for inflation and for income growth
adjustment based on the approach described in EPA's BenMAP-CE model and its documentation. The resulting value, $10.0 million for 2015
($2015), was adjusted to future years by assuming an elasticity of VSL to GDP per capita of 0.4. Projections of U.S. GDP and population
described in the Modeling Framework section of this Technical Report were employed. Using this approach, the VSL is estimated at $12.4
million in 2050 and $15.2 million in 2090. Sources: 1) EPA, 2014: Guidelines for Preparing Economic Analyses. National Center for
Environmental Economics. Available online at http://vosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-0568-52.pdf/$file/EE-0568-52.pdf: and 2)
EPA, cited 2017: Benefits Mapping and Analysis Program (BenMAP): Manual and Appendices for BenMAP-CE. Available online at
https://www.epa.gov/benmap/manual-and-appendices-benmap-ce.
113 Fann, N., C.G. Nolte, P. Dolwick, T.L. Spero, A. Curry Brown, S. Phillips, and S. Anenberg, 2015: The geographic distribution and economic
value of climate change-related ozone health impacts in the United States in 2030.Journal of the Air & Waste Management Association, 65,
570-580. http://dx.doi.org/10.1080/10962247.2014.996270.
36
-------
HEALTH
Air Quality
Under both RCPs, daily maximum summer temperatures are projected to increase across the contiguous
U.S., with summer temperatures generally intensifying through the end of the century (see Figure A.6.1
in Appendix A.6). In 2090, daily maximum summer temperatures are projected to rise by 3-7°C under
RCPS.5. Although warmer than the reference period conditions, the summer daily maximum
temperatures in 2090 under RCP4.5 are lower than those projected in 2050 under RCP8.5. Annual
precipitation is projected to increase in the Northwest, Southeast, and Northeast, but decrease in the
Southwest and Southern Plains (see Figure A.6.2).
Figure 3.1 shows the projected change in summer-average maximum daily 8-hour ozone concentrations
over the U.S. in 2050 and 2090 relative to the reference period (2000). The results are consistent with
other published studies in that they show that a warming climate is generally expected to lead to
increased ozone concentrations in parts of the U.S. Seasonal-average ozone increases of up to 5 parts
per billion (ppb) are simulated at 2090 under RCPS.5 over parts of the U.S. This effect is often referred to
as the "climate penalty,"114 where attaining national air quality standards will become more difficult
because climate change will counteract improvements that may be expected from emissions
reductions.115 In parts of the Southeast and Southern Plains, climate-driven meteorological changes
result in conditions slightly less conducive to ozone formation, potentially due to an increase in
precipitation during summer months (not shown in Figure A.6.2) and wind trajectories that transport
cleaner marine air into these regions.
Figure 3.1. Change in Summer-Average Maximum Daily Ozone
Maps show the change in summer-average maximum daily 8-hour ozone concentrations (ppb) in 2050
(2045-2055) and 2090 (2085-2095) compared to 2000 (1995-2005).
>5
4 to 5
3 to 4
2 to 3
1 to 2
0.5 to 1
-0.5 to 0.5
-1 to -0.5
-2 to -1
-3 to -2
-4 to -3
114 Wu, S., L. J. Mickiey, E. M. Leibensperger, D. J. Jacob, D. Rind, arid D. G. Streets, 2008: Effects of 2000-2050 global change on ozone air quality
in the United States. Journal of Geophysical Research, 113, D06302. doi:10.1029/2007JD008917.
115 Climate-driven changes in meteorological patterns will also impact PM2.5 concentrations throughout: the U.S. However, unlike ozone, there is
no current consensus as to whether these changes will result: in increasing or decreasing PM2.5 levels. The PM2.5 results from this modeling are
provided in Appendix A.6.
RCP8.5
2050
RCP4.5
2090
37
-------
HEALTH
Air Quality
At the national scale, these projected climate-attributable increases in summer ozone have a
quantifiable adverse effect on human health as shown in Figure 3.2 and summarized by region in
Table 3.1. The analysis estimates that an additional 790 premature ozone-related deaths occur annually
(between May and September) in 2050 under RCP8.5 relative to the reference period. As climate
warming persists through 2090, this value increases to 1,700 additional premature deaths each year.
Compared to RCP8.5, RCP4.5 reduces mortality impacts by avoiding 240 deaths by 2050 and 500 deaths
by 2090. Projected increases in premature deaths are largest in the Midwest and Northeast, while
decreases in ozone-related deaths are projected in the Southeast and Southern Plains under some
scenario and time period combinations. Table 3.2 provides the monetized values associated with the
estimated changes in premature mortality. RCP4.5 is also projected to significantly reduce the number
of incidents of ozone-related, acute respiratory symptoms leading to hospital visits and school absences.
See Appendix A.6 for more details on ozone morbidity effects.
Figure 3.2. Change in Ozone-Related Premature Deaths
Maps show county-level estimates for the average change in ozone-related premature deaths over the
summer months in 2050 (2045-2055) and 2090 (2085-2095) compared to 2000 (1995-2005).
RCP4.5
Ozone-Related Premature Deaths
50.1 -193
10.1 to 50
5.1 to 10
1.1 to 5
1 | -0.9 to 1
-4.9 to-1
-9.9 to -5
¦ -20 to -10
RCP8.5
2090
2050
38
-------
HEALTH
Air Quality
Table 3.1. Excess (or Avoided) Ozone-Related Premature Deaths
The table presents estimates for 2050 (2045-2055) and 2090 (2085-2095) under RCP8.5 and RCP4.5
compared to 2000 (1995-2005). The 95th percentile confidence intervals are provided in parentheses116.
Values may not sum due to rounding.
2050 2090
Region
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
230
200
670
310
(120 to 340)
(110 to 300)
(360 to 980)
(160 to 450)
Southeast
69
-40
-72
88
(37 to 100)
(-59 to -21)
(-100 to -38)
(47 to 130)
Midwest
380
300
910
580
(200 to 550)
(160 to 440)
(490 to 1,300)
(310 to 840)
Northern
23
20
42
29
Plains
(12 to 33)
(11 to 30)
(22 to 61)
(16 to 43)
Southern
3.2
-4.2
-37
89
Plains
(1.7 to 4.7)
(-6.2 to -2.3)
(-54 to -20)
(48 to 130)
Southwest
62
71
110
57
(33 to 91)
(38 to 100)
(59 to 160)
(30 to 83)
Northwest
20
(10 to 29)
5.2
(2.8 to 7.6)
93
(50 to 140)
29
(16 to 43)
National Total
790
550
1,700
1,200
(420 to 1,200)
(300 to 810)
(920 to 2,500)
(630 to 1,700)
116 The confidence intervals are calculated using a Monte Carlo technique, in which random draws are taken from a distribution of standard
errors reported in epidemiological and economic value studies. This distribution reflects sampling error alone, and does not account for
uncertainty introduced in other "upstream" elements of the analysis—including the projection of emissions, meteorology, air quality modeling,
etc.
39
-------
HEALTH
Air Quality
Table 3.2. Cost of Excess (or Avoided) Ozone-Related Premature Deaths
The table presents estimates for 2050 (2045-2055) and 2090 (2085-2095) under RCP8.5 and RCP4.5
compared to 2000 (1995-2005). Units are millions of $2015. The 95th percentile confidence intervals are
provided in parentheses. Values may not sum due to rounding.
Region
2050
RCP8.5 RCP4.5
2090
RCP8.5 RCP4.5
Northeast
$2,900
($260 to $8,200)
$2,500
($230 to $7,200)
$10,000
($910 to $29,000)
$4,700
($420 to $13,000)
Southeast
$850
($77 to $2,400)
-$500
(-$1,400 to -$45)
-$1,100
(-$3,100 to -$98)
$1,300
($120 to $3,800)
Midwest
$4,700
($420 to $13,000)
$3,700
($330 to $11,000)
$14,000
($1,200 to $39,000)
$8,800
($790 to $25,000)
Northern
Plains
$280
($25 to $810)
$250
($23 to $720)
$630
($57 to $1,800)
$440
($40 to $1,300)
Southern
Plains
$40
($3.6 to $110)
-$53
(-$150 to -$4.7)
-$560
(-$1,600 to -$50)
$1,400
($120 to $3,800)
Southwest
$770
($69 to $2,200)
$880
($79 to $2,500)
$1,700
($150 to $4,800)
$860
($77 to $2,500)
Northwest
$240
($22 to $690)
$65
($5.8 to $180)
$1,400
($130 to $4,000)
$450
($40 to $1,300)
National
Total
$9,800
($880 to $28,000)
$6,900
(-$900 to $21,000)
$26,000
(-$2,200 to $78,000)
$18,000
($1,600 to $51,000)
3.5 DISCUSSION
The findings shown here indicate that increasing temperatures will non-uniformly alter summer
concentrations of ozone across the contiguous U.S. in 2050 and 2090 relative to 2000. Although ozone is
strongly influenced by temperature, other meteorological factors (such as wind speed, cloud cover, wind
trajectories, and precipitation amounts) will also affect those ozone concentrations. Adverse health
outcomes projected from increased ozone are aligned with population centers, such that densely
populated areas that experience increased ozone will see a greater impact on health. The largest
increases in ozone are projected to occur from the Northern Plains through the eastern Great Lakes,
which are generally the regions with the largest temperature increases (see Figure A.6.1). The
simulations held pollutant emissions at 2040 levels in both the reference period and the future periods,
so any ozone increases are driven by changes in meteorology or in biogenic emissions. Additionally,
average relative humidity values during the summer are projected to be lower in this region in the
future due to climate change, which is also conducive to higher ozone concentrations. Other
meteorological parameters that could potentially impact ozone (e.g., depth of the mixed layer, wind
speed and direction, precipitation frequency) appear to have smaller impacts in this region than the
higher temperatures and drier conditions projected.
Projections under RCP4.5 demonstrate a substantially decreased burden of air pollution on U.S.
respiratory health by reducing ozone concentrations relative to RCP8.5. In other words, RCP8.5 will lead
to meteorological conditions that are generally more favorable for ozone production than RCP4.5.
Future ozone levels will depend not only on the meteorological conditions in which ozone is formed, but
also on future trends in ozone precursor emissions. Thus, the need for further emissions controls on
domestic sources to meet certain air quality goals will likely be greater under RCP8.5. RCP4.5 could
40
-------
HEALTH
Air Quality
substantially decrease the future impacts of air pollution on U.S. respiratory health by reducing ozone
concentrations relative to unconstrained climate change, thereby reducing the need for additional
emissions controls on domestic sources of ozone air pollution. These findings are consistent with others
described in the assessment literature,117 and the results of a previous CIRA analysis using different
methods.118 This analysis does not quantify the additional benefits to air quality and health that would
stem from simultaneous reductions in traditional air pollutants along with GHG emissions.
Wildfires are strong local sources of air pollutants, including compounds that form PM and ozone, and
their occurrence is often linked to a confluence of extreme meteorological conditions (e.g., drought,
strong winds, lightning) combined with a natural fuel load in forested areas. Wildfires are localized, rare
events that can have severe impacts on air quality and human health. However, the natural initiation of
wildfires, the extreme meteorological conditions that increase susceptibility to wildfires, and the
trajectories of the plumes from wildfire smoke are all difficult to project with confidence (even with
observation-driven meteorological models), particularly in areas of rapidly changing terrain elevation.
Furthermore, there are important but highly uncertain human and economic components to whether a
wildfire occurs, how people respond, and what impacts are associated (e.g., wildfires resulting from
arson or negligence, accessibility of wildfire locations, availability of resources and trained response
personnel, evacuation efforts, size of the population directly impacted by or downwind of the wildfire,
and personal behaviors to reduce exposure to wildfire smoke). Future research is needed to link the
methods applied in this section to wildfire modeling so that these effects can be investigated.
Health outcomes from climate-driven impacts on PM2.5, even excluding expected increases in wildfire
emissions, may also be signficant but remain uncertain. The changes in PM2.5 are driven by complex
meteorological processes related to the distributions of cloud cover, radiation, and precipitation at any
given location. However, there is considerable uncertainty in the geographic distribution of these
changes to PM2.5 and the overall trend, not just in this modeling analysis but in the literature overall.119
The interactions of some sources of PM2.5 with sunlight, clouds, and solar radiation are also uncertain.120
See Appendix A.6 for further discussion.
117 Fann, N., T. Brennan, P. Dolwick, J.L. Gamble, V. Ilacqua, L. Kolb, C.G. Nolte, T.L. Spero, and L. Ziska, 2016: Ch. 3: Air Quality Impacts. The
Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington,
DC, 69-98.
118 Garcia-Menendez, F., R.K. Saari, E. Monier, and N.E. Selin, 2015: U.S. air quality and health benefits from avoided climate change under
greenhouse gas mitigation. Environmental Science and Technology, 49, 7580-7588, doi:10.1021/acs.est.5b01324.
119 Fiore, A. M., V. Naik, and E. M. Leibensperger, 2015: Air quality and climate connections. Journal of the Air & Waste Management
Association, 65, 645-685. doi:10.1080/10962247.2015.1040526.
120 Bond, T.C., S.J. Doherty, D.W. Fahey, P.M. Forster, T. Bernsten, B.J. DeAngelo, M.G. Flanner, S.Ghan, B. Karcher, D. Koch, S. Kinne, Y. Kondo,
P.K. Quinn, M.C. Sarofim, M.G. Schultz, M. Shulz, C. Venkataraman, H. Zhang, S. Zhang, N. Bellouin, S.K. Guttikunda, P.K. Hopke, M.Z. Jacobson,
J.W. Kaiser, Z. Klimont, U. Lohmann, J.P. Swartz, D. Shindell, T. Storelvmo, S.G. Warren, and C.S. Zender, 2013: Bounding the role of black
carbon in the climate system: a scientific assessment. Journal of Geophysical Research - Atmospheres, 118, 5380-5552.
41
-------
HEALTH
Aeroallergens
4. AEROALLERGENS
4.1 KEY FINDINGS
• Aeroallergens currently pose a substantial U.S. public health burden, including emergency
department visits for the most severe reactions; climate change is projected to increase U.S. allergic
disease incidence from oak pollen. People living in the Northeast and children less than 18 years old
are at higher risk.
• Though all three regions analyzed (Northeast, Southeast, and Midwest) show projected increases in
pollen season length under all time periods and scenarios, changes are particularly large under
RCP8.5 in 2090, when season length is increased by 3.8 days in the Midwest, 3.5 days in the
Northeast, and 1.7 days in the Southeast.
• By 2090, total oak pollen-related asthma emergency department (ED) visits in the Northeast,
Southeast, and Midwest are projected to increase by 3.9% and 9.6% compared to the reference
period under RCP4.5 and RCP8.5, respectively. Increases in ED visits in 2090 under RCP8.5 are
particularly high in the Midwest (13%) and Northeast (12%).
• Across the three regions analyzed, costs from oak pollen-related asthma ED visits from climate
change increase in all scenarios, time periods, and regions, particularly in the Northeast and
Midwest. The increase in annual costs of ED visits in all three regions in 2090 is $1.2 million under
RCP8.5 and $0.52 million under RCP4.5. Inclusion of additional pollen types would likely increase
these damages.
4.2 INTRODUCTION
Rising C02 concentrations and associated climate change is expected to lengthen and intensify pollen
season in parts of the U.S.,121122 potentially leading to additional cases of allergic rhinitis (commonly
known as "hay fever") and allergic asthma episodes.123,124 For example, the duration of pollen release for
common ragweed, the aeroallergen that most commonly affects persons in the United States, has been
increasing as a function of latitude in recent decades in the Midwest region.125 Among individuals with
allergic asthma, exposure to allergens can result in exacerbation of symptoms, including asthma
episodes, sinusitis, or anaphylaxis.126 Asthma is widespread in the U.S., affecting approximately 7% of
121 Zhang, Y., L. Bielory, T. Cai, Z. Mi, and P. Georgopoulos, 2015: Predicting onset and duration of airborne allergenic pollen season in the
United States. Atmospheric Environment, 103, 297-306.
122 Fann, N., T. Brennan, P. Dolwick, J.L. Gamble, V. Ilacqua, L. Kolb, C.G. Nolte, T.L. Spero and L. Ziska, 2016: Ch. 3: Air Quality Impacts. In: The
Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington,
DC, 69-98, doi: 10.7930/J0GQ6VP6.
123 Reid, C.E., and J.L. Gamble, 2009: Aeroallergens, allergic disease, and climate change: Impacts and adaptation. EcoHealth, 6, 458-470.
124 Sheffield, P.E., K.R. Weinberger, and P.L. Kinney, 2011: Climate change, aeroallergens and pediatric allergic disease. Mount Sinai Journal of
Medicine, 78, 78-84.
125 Fann, N., T. Brennan, P. Dolwick, J.L. Gamble, V. Ilacqua, L. Kolb, C.G. Nolte, T.L. Spero and L. Ziska, 2016: Ch. 3: Air Quality Impacts. In: The
Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington,
DC, 69-98, doi: 10.7930/J0GQ6VP6.
126 Nielsen, G.D., J.S. Hansen, R.M. Lund, M. Bergqvist, S.T. Larsen, S.K. Clausen, P. Thygensen, and O.M. Poulsen, 2002: IgE-mediated asthma
and rhinitis I: A role of allergen exposure? Pharmacology & Toxicology, 90, 231-242.
42
-------
HEALTH
Aeroallergens
adults and 9% of children,127 and resulting in $56 billion in medical expenditures, missed work and
school days, and early deaths in 2007.128
Rising C02 concentrations and climate-induced changes in temperature and precipitation may impact
pollen season timing and length, the amount of pollen produced throughout the season, the allergen
content of pollen grains, and the spatial distribution of species producing allergenic pollen.129 For tree
pollen specifically, warmer temperatures both year-round and in the months preceding the pollen
season have been linked to increased season intensity and length.130
4.3 APPROACH
This analysis examines the health consequences of present-day and climate-induced changes in pollen
exposure in the Northeast, Southeast, and Midwest regions of the contiguous U.S., focusing on oak
pollen- and asthma-related ED visits. Oak pollen season length and seasonal average pollen
concentrations are collected from observations for the reference period (1994-2010) at 59 National
Allergy Bureau monitoring stations and simulated for future years (2030, 2050, 2070, and 2090) using
published relationships between temperature, precipitation, and oak pollen season length. These
relationships are applied to climate projections under RCP8.5 and RCP4.5 using the five GCMs.
Epidemiologically-derived health impact functions are combined with demographic projections to
estimate oak pollen-associated asthma ED visits for all days in the recent past and simulated future oak
pollen season using the Environmental Benefits Mapping and Analysis Program (BenMAP-CE). A
monetary value of pollen-related ED visits is determined by applying the mean of two cost-per-visit
estimates.131,132 Adjusted for inflation, the mean per-visit cost in 2015 dollars is $490/visit. Importantly,
this analysis only considers climate impacts on oak pollen in three regions of the contiguous U.S., and
therefore does not estimate the total potential national health effects of climate-driven changes to
aeroallergens. For more information regarding the approach used in this section to estimate climate
change impacts on aeroallergens, please refer to Anenberg et al. (2017).133
4.4 RESULTS
As shown in Figure 4.1, oak pollen season is projected to lengthen in the Northeast, Southeast, and
Midwest under both RCPs and in both time periods (except under the GISS-E2-R model in the
Southeast). Impacts are projected to be greater under RCP8.5 than RCP4.5, particularly in 2090. The
projected increase in season length is highest in the Midwest (2.1 days in 2050 and 3.8 days in 2090
under RCP8.5 for the five-GCM average), followed closely by the Northeast (1.9 days in 2050 and 3.5
127 USGCRP, 2016: The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research
Program. Available online at http://www.globalchange.gov/health-assessment
128 CDC, 2011: Vital Signs, May 2011. United States Department of Health and Human Services, Centers for Disease Control and Prevention.
Available online at http://www.cdc.gov/vitalsigns/asthma/
129 Fann, N., T. Brennan, P. Dolwick, J.L. Gamble, V. Ilacqua, L. Kolb, C.G. Nolte, T.L. Spero and L. Ziska, 2016: Ch. 3: Air Quality Impacts. In: The
Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington,
DC, 69-98, doi: 10.7930/J0GQ6VP6.
130 Ariano, R., G.W. Canonica, and G. Passalacqua, 2010: Possible role of climate changes in variations in pollen seasons and allergic
sensitizations during 27 years. Annals of Allergy, Asthma & Immunology, 104, 215-222.
131 Smith, D.H., D.C. Malone, K.A. Lawson, L.J. Okamoto, C. Battista, and W.B. Saunders, 1997: A national estimate of the economic costs of
asthma. Am J Resp Crit Care Med, 156, 787-793.
132 Stanford, R., T. McLaughlin and L.J. Okamoto, 1999: The cost of asthma in the emergency department and hospital. Am J Resp Crit Care Med,
160, 211-215.
133 Anenberg, S. C., K. R. Weinberger, H. Roman, J. E. Neumann, A. Crimmins, N. Fann, J. Martinich, and P. L. Kinney (2017), Impacts of oak pollen
on allergic asthma in the United States and potential influence of future climate change, GeoHealth, 1, doi:10.1002/2017GH000055.
43
-------
HEALTH
Aeroallergens
days in 2090 under RCP8.5 for the five-GCM average).134 Two models (HadGEM2-ES and MIROC5)
projected increases in season length of more than 5 days in the Midwest by 2090 under RCP8.5. In the
Southeast, the projected change in season length is smaller (0.96 days in 2050 and 1.7 days in 2090
under RCP8.5 for the five-GCM average).
Figure 4.1. Change in Oak Pollen Season Length
Graph shows the projected change in oak pollen season length (days) in 2050 and 2090 relative to the
1994-2010 reference period for the Northeast, Southeast, and Midwest regions for each GCM. The
results represent the average of the results for all monitoring locations in each region (15 locations in the
Northeast, ten in the Southeast, and seven in the Midwest). The season length in the reference period is
26 days in the Northeast, 29 days in the Southeast, and 26 days in the Midwest, based on the average
season length across each region's monitoring locations.
C
o
•
•
• •
•
•
•
m
•
• n
•
§ . § •
•
o
•
° o
• °
• 8
0
o
•
0 ®
y
•
• • •
o •
o •
o
•
° o
• •
O
•OC
o o«
•
o
o
A
-1
w
cc
CO
LO
CtO
-------
HEALTH
Aeroallergens
respectively (five-GCM average). In the Midwest, the 2090 projected changes are 490 and 270 visits for
the 0-17 age group and 230 and 120 visits for the 18+ age group under RCP8.5 and RCP4.5, respectively
(five-GCM average).
Figure 4.2. Change in Asthma-Related Emergency Department Visits
The graphs show change from the reference period (1994-2010) by age groups for the three regions
studied under RCP8.5 and RCP4.5. Results represent averages of the five GCMs.
1,300
NORTHEAST
1-1/
(100)
4.5
8.5 4.5
8.5 4.5
8.5 4.5
2030
2050
2070
2090
800
700
£ 600
S
Q 500
Lbl
"O
% 400 -
5
cc
| 300
.£
to
< 200
100
SOUTHEAST
~ 18+
¦ 0-17
li
J.5 | 4.5
2030
ill
8.5 | 4.5 8.5
2050 ;
111
5.5 | 4.5
2090
800
MIDWEST
~ 18+
700
¦ 0-17
600
5 500
¦o 400
300
200
100
(100)
8.5
4.5
4.5
4.5
4.5
2030
2050
2070
2090
45
-------
HEALTH
Aeroallergens
Costs from oak pollen-related asthma ED visits are generally projected to increase due to climate change
across the three regions, particularly in the later part of the century. The projected annual costs in 2090
are highest in the Northeast under both RCPs (Table 4.1).
Table 4.1. Change in Annual Costs of Emergency Department Visits
Values reported in thousands of $2015, and represent averages of the five GCMs. The 90% confidence
interval results (5th and 95th percentiles) are presented in parentheses following the mean estimate.
2030 2050 2070 2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
-$15
$140
$240
$240
$440
$220
$520
$230
(-$90 to
($15 to
($19 to
($16 to
($57 to
($32 to
($60 to
($25 to
$74)
$290)
$480)
$470)
$870)
$430)
$1,000)
$470)
Southeast
$60
$52
$180
$140
$260
$96
$360
$100
($17 to
($15 to
($31 to
($30 to
($31 to
($37 to
($43 to
($31 to
$110)
$98)
$350)
$270)
$520)
$190)
$720)
$200)
Midwest
$110
$84
$170
$64
$300
$180
$350
$190
(-$3.4 to
(-$1.2 to
(-$2.5 to
($0 to
(-$5.7 to
(-$5.4 to
($3.3 to
($5.7 to
$24)
$190)
$380)
$120)
$670)
$400)
$810)
$440)
3 Region Total
$150
$280
$590
$440
$1,000
$490
$1,200
$520
(-$76 to
($29 to
($48 to
($46 to
($82 to
($64 to
($110 to
($62 to
$420)
$580)
$1,200)
$870)
$2,100)
$1,000)
$2,600)
$1,100)
4.5 DISCUSSION
This analysis isolates the impact of climate change on oak pollen season length in the Northeast,
Midwest, and Southeast regions of the U.S. By 2090, climate change driven changes in oak pollen season
length are projected to increase asthma ED visits by 3.9% (1,100) and 9.6% (2,500) compared to the
reference period for RCP4.5 and RCP8.5, respectively, with the largest changes in the Midwest (13%),
followed by the Northeast (12%) and Southeast (6.5%). The study finds that moderate versus severe
climate change could avoid more than half (1,400) of the additional oak pollen-related asthma ED visits
projected in 2090. Although these results were limited to one pollen type (oak) and one health outcome
(asthma ED visits), they suggest that aeroallergens pose a substantial burden on U.S. public health and
that future climate change is likely to increase allergic disease incidence in the U.S.
Compared with nationwide impacts of ambient air pollution, estimated oak pollen asthma ED visits in
the Northeast, Southeast, and Midwest are approximately 10% of estimated asthma ED visits associated
with fine particulate matter (PM2.5) among children age <18 years (110,000) and approximately 90% of
those associated with ozone among all ages in 2005 (19,000).135 Since this analysis included only one
pollen type, these comparisons suggest that the burden of aeroallergens on asthma ED visits in the U.S.
could be of a similar magnitude to that of ambient air pollution. However, oak pollen exposure-response
relationships may already capture some portion of health effects from other pollen types since some are
temporally correlated.136 This analysis focused on regions containing the highest prevalence of oak trees
135 Fann, N., A.D. Lamson, S.C. Anenberg, K. Wesson, D. Risley, and B.J. Hubbell, 2011: Estimating the national public health burden associated
with exposure to ambient PM2.5 and ozone. Risk Analysis, 32, 81-95.
136 Ito, K., K.R. Weinberger, G.S. Robinson, P.E. Sheffield, R. Lall, R. Mathes , Z. Ross, P.L. Kinney, and R.D. Matte, 2015: The associations between
daily spring pollen counts, over-the-counter allergy medication sales, and asthma emergency department visits syndrome in New York City,
2002-2012. Environ Health, 14, 71.
46
-------
HEALTH
Aeroallergens
and approximately 50% of the population in the U.S., but additional health impacts from oak pollen
exposure would be expected elsewhere. Climate change consequences could be underestimated
because some of the largest increases in oak pollen season length occur in the West. This analysis also
excludes climate impacts on seasonal average pollen concentrations, changes in pollen allergenicity, and
the geographic range of oak trees. Finally, the health-response functions used in this analysis are not
adjusted for future physiological changes in immunity or changes in behavior (e.g., increases in self-
protection), both of which possess large uncertainties when projecting into the future.
47
-------
HEALTH
Extreme Temperature Mortality
5. EXTREME TEMPERATURE MORTALITY
5.1 KEY FINDINGS
• Changes in extreme temperatures are projected to result in a net average increase of approximately
9,300 premature deaths per year under RCP8.5 by 2090 in the 49 modeled cities. Under RCP4.5,
more than 5,000 deaths are avoided each year by 2090.
• The projected reduction in deaths from extremely cold days is far less than the projected increase in
deaths from extremely hot days in all climate models, scenarios, and time frames.
• Annual damages associated with additional extreme temperature related deaths are estimated at
$140 billion under RCP8.5 and $60 billion under RCP4.5 by the end of the century.
• Mortality from extremely hot days decreased more than 50% under both RCP8.5 and RCP4.5 in 2050
and 2090 when the human health response to extreme temperatures was evaluated using Dallas'
threshold for extreme heat (in all cities with thresholds initially cooler than Dallas), as a sensitivity
analysis to consider the effect of adaptation.
5.2 INTRODUCTION
Climate change will alter the weather conditions to which we are accustomed. Extreme temperatures
are projected to rise in many areas across the U.S., bringing more frequent and intense heat waves and
increasing the number of heat-related illnesses and deaths.137 138 Exposure to extreme heat can
compromise the body's ability to regulate its temperature, resulting in heat exhaustion and/or heat
stroke, and can also exacerbate existing medical problems, such as heart and lung diseases.139 By one
measure, heat waves are already the largest cause of fatalities from extreme weather in the U.S.140 For
instance, during a 1995 heat wave in Chicago, an estimated 700 individuals died as a result of the
extreme heat.141 Increases in both average and extreme temperatures are also projected to result in
fewer extremely cold days, and this is expected to reduce deaths associated with extreme cold.
5.3 APPROACH
This analysis estimates the number of deaths over the course of the 21st century attributable to extreme
temperatures in 49 cities in the contiguous U.S., which account for approximately one third of the
national population. City-specific relationships between daily deaths (from all causes) and extreme
137 USGCRP, 2016: The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. Crimmins, A., J. Balbus, J.L.
Gamble, C.B. Beard, J.E. Bell, D. Dodgen, R.J. Eisen, N. Fann, M.D. Hawkins, S.C. Herring, L. Jantarasami, D.M. Mills, S. Saha, M.C. Sarofim, J.
Trtanj, and L. Ziska, Eds. U.S. Global Change Research Program, Washington, DC, 312 pp., doi: 10.7930/J0R49NQX.
138 Luber, G., K. Knowlton, J. Balbus, H. Frumkin, M. Hayden, J. Hess, M. McGeehin, N. Sheats, L. Backer, C.B. Beard, K.L. Ebi, E. Maibach, R. S.
Ostfeld, C. Wiedinmyer, E. Zielinski-Gutierrez, and L. Ziska, 2014: Ch. 9: Human Health. Climate Change Impacts in the United States: The Third
National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G.W. Yohe, Eds., U.S. Global Change Research Program, 220-256.
doi:10.7930/J0PN93H5.
139 Sarofim, M.C., S. Saha, M.D. Hawkins, D.M. Mills, J. Hess, R. Horton, P. Kinney, J. Schwartz, and A. St. Juliana, 2016: Ch. 2: Temperature-
Related Death and Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change
Research Program, Washington, DC, 43-68, doi: 10.7930/J0MG7MDX.
140 Bell, J.E., S.C. Herring, L. Jantarasami, C. Adrianopoli, K. Benedict, K. Conlon, V. Escobar, J. Hess, J. Luvall, C.P. Garcia-Pando, D. Quattrochi, J.
Runkle, and C.J. Schreck, III, 2016: Ch. 4: Impacts of Extreme Events on Human Health. The Impacts of Climate Change on Human Health in the
United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 99-128, doi: 10.7930/J0BZ63ZV.
141 EPA, 2016. Climate Change Indicators in the United States, 2016. Fourth edition. EPA 430-R-16-004. Available online at
https://www.epa.gov/climate-indicators
48
-------
HEALTH
Extreme Temperature Mortality
temperatures were estimated based on historical observations, and are combined with the projections
of extremely hot and cold days (average of three years centered on 2050 and 2090) using city-specific
extreme temperature thresholds to project future deaths from extreme heat and cold under RCP8.5 and
RCP4.5 in five GCMs. Extremely hot days are defined as those with a daily minimum temperature
warmer than 99 percent of the days in the reference period (1989-2000) and that are at least 20°C
(68°F). Extremely cold days are defined as those with a daily maximum temperature colder than 99
percent of the days in the reference period (1989-2000) and do not exceed 10°C (50°F). As a result, the
study explicitly addresses the question of the net mortality impact of changes in extreme temperature
days in the future due to climate change.
The potential impact of future population change is accounted for using the ICLUSv2 population
projections described in Modeling Framework section of this Technical Report. To monetize the effects
of changing mortality, the analysis uses a baseline VSL adjusted to future years based on a projected
change in economic growth.142 The results presented in this section have been updated since Mills et al.
(2014) to include additional cities and more recent mortality rate data.143 This analysis does not estimate
impacts across ages or socioeconomic status. As these demographics change, they could impact the
results. Finally, this analysis does not estimate extreme temperature effects on morbidity, which could
result in a higher number of cases compared to mortality.144 For more information on the approach and
results for the extreme temperature mortality sector, please refer to Mills et al. (2014).145
5.4 RESULTS
Under RCP8.5, the average number of extremely hot days in the U.S. is projected to nearly double from
2050 to 2100. The projected increase in deaths due to more frequent extremely hot days is much larger
than the projected decrease in deaths due to fewer extremely cold days (Table 5.1), a finding that is
consistent with the conclusions of the assessment literature.146 Under RCP8.5, the net increase in
projected deaths from more extremely hot days and fewer extremely cold days in 49 cities is
142 At the time of this analysis, the EPA's Guidelines for Preparing Economic Analyses recommends a VSL of $7.9 million (2008$), based on 1990
incomes. To create a VSL using $2015 and based on 2015 incomes, the standard value was adjusted for inflation and for income growth
adjustment based on the approach described in EPA's BenMAP-CE model and its documentation. The resulting value, $10.0 million for 2015
($2015), was adjusted to future years by assuming an elasticity of VSL to GDP per capita of 0.4. Projections of U.S. GDP and population
described the Modeling Framework section of this Technical Report were employed. Using this approach, the VSL in 2050 is estimated at $12.4
million and $15.2 million in 2090. Sources: 1) EPA, 2014: Guidelines for Preparing Economic Analyses. National Centerfor Environmental
Economics. Available online at http://vosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-0568-52.pdf/$file/EE-0568-52.pdf: and 2) EPA, cited 2017:
Benefits Mapping and Analysis Program (BenMAP): Manual and Appendices for BenMAP-CE. Available online at
https://www.epa.gov/benmap/manual-and-appendices-benmap-ce
143 The approach described in Mills et al. (2014) was updated in several ways. First, the analysis was expanded from 33 cities to encompass a
total of 49 out of 50 of the cities (excluding Honolulu) analyzed by Medina-Ramon and Schwartz (2007). Medina-Ramon and Schwartz did not
calculate heat mortality response functions for cities where the minimum temperature for the 99 percentile hottest day was equal to or below
20°C (8 cities), or cold mortality response functions where the maximum temperature for the 1 percentile coldest day was greater than or equal
to 10°C (7 cities). In a warming climate, cities that were too warm to meet the criteria for the cold threshold will continue to be too warm,
making a cold mortality response function insignificant. Most of the cities that were too cool to meet the criteria for the hot threshold are
expected to warm enough that their 99 percentile hottest days will exceed 20°C in the future. Therefore, inclusion of cities without a heat
mortality response function will lead to an underestimate of the change in future mortality in those cities, and therefore an underestimate of
avoided deaths. However, inclusion of a wider range of cities gives a more complete picture of impacts in the U.S. Additionally this study was
updated to limit the analysis to the actual counties corresponding to the cities specified in Medina-Ramon and Schwartz (2007), ratherthan the
MSAs used in Mills et al. (2014). This reduces the total population considered within the original 33 cities. Furthermore, Ben MAP data for the
all-age mortality rates in the cities is used, resulting in some small differences in calculations.
144 Eisenman, D., Wilhalme, H., Tseng, C., English, P., Chester, M., Fraser, A., Pincetl, S, 2016. Heat death associations with the built
environment, social vulnerability and their interactions with rising temperature. Health Place, doi: 10.1016/j.healthplace.2016.08.007.
145 Mills, D., J. Schwartz, M. Lee, M. Sarofim, R. Jones, M. Lawson, M. Duckworth, and L. Deck, 2014: Climate Change Impacts on Extreme
Temperature Mortality in Select Metropolitan Areas in the United States. Climatic Change, doi: 10.1007/sl0584-014-1154-8.
146 Sarofim, M.C., S. Saha, M.D. Hawkins, D.M. Mills, J. Hess, R. Horton, P. Kinney, J. Schwartz, and A. St. Juliana, 2016: Ch. 2: Temperature-
Related Death and Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change
Research Program, Washington, DC, 43-68, doi: 10.7930/J0MG7MDX.
49
-------
HEALTH
Extreme Temperature Mortality
approximately 3,400 deaths per year in 2050, and 9,300 deaths per year in 2090. In comparison, RCP4.5
avoids nearly 800 deaths each year by 2050, and more than 5,400 deaths each year by 2090, a mortality
reduction of 24% and 58% respectively.
Table 5.1. Changes in Annual Mortality
The results represent the change in annual mortality from the 1989-2000 reference period due to
extreme heat, extreme cold, and combined stressors. Values represent results for the 49 cities, and
assume increased population growth. Estimates may not sum due to rounding.
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Heat
CanESM2
4,200
2,800
12,000
3,700
CCSM4
2,300
1,700
7,200
2,500
GISS-E2-R
2,500
2,300
5,500
2,700
HadGEM2-ES
5,900
3,900
13,000
7,400
MIROC5
2,500
2,300
8,900
3,400
5-Model Average
3,500
2,600
9,300
3,900
Cold
CanESM2
-49
-41
-51
-36
CCSM4
-34
-47
-48
-35
GISS-E2-R
-10
-16
-52
-6
HadGEM2-ES
-32
-49
-58
-37
MIROC5
-40
-10
-55
-51
5-Model Average
-33
-33
-53
-33
Total (Change in Combined Heat and Cold Deaths)
CanESM2
4,100
2,700
12,000
3,700
CCSM4
2,300
1,700
7,100
2,400
GISS-E2-R
2,500
2,300
5,400
2,700
HadGEM2-ES
5,900
3,900
13,000
7,400
MIROC5
2,400
2,300
8,900
3,300
5-Model Average
3,400
2,600
9,300
3,900
Mortality rates from extreme hot and cold temperatures by city are greatest under RCP8.5 in 2090
(Figure 5.1). In this time period and climate scenario, nearly all cities experience net mortality rates
greater than eight deaths per 100,000 residents from extreme hot and cold, with the exception of the
Northwest. However, several cities, notably Kansas City, have high mortality rates under RCP4.5 and in
2050.
50
-------
HEALTH
Extreme Temperature Mortality
Figure 5.1. Projected Extreme Temperature Mortality in Select Cities
Estimated net mortality rate from extremely hot and cold days (average of five GCMs; number of deaths
per 100,000 residents). Cities without circles should not be interpreted as having no extreme
temperature impact.
RCP8.5
RCP4.5
Mortality rate (deaths/100k) Reference
Impacts vary regionally; however, it is important to note that the number of cities and the number of
people included within each region of this study are not homogenous. In the reference period, most of
the cities that had the highest increase in mortality from exceeding heat thresholds were located in the
Northeast and the Midwest - therefore, for a similar amount of warming, these cities will see a greater
increase in mortality than some of the more southern cities (see Figure 5.1). The finding that some hot
cities in the Southwest and southern Florida have lower increases in mortality during extreme heat
events is consistent with a number of studies147148 and is thought to be a result of adaptation measures
(physiological, behavioral, and infrastructure based).149 In the reference period, cities had insufficient
exceedances of extreme heat thresholds in the Northwest or parts of the Southwest region (Colorado
and northern California), and therefore this analysis projects no increase in mortality in those cities. The
147 Kalkstein, L. S., S. Greene, D. M, Mills, and J. Samenow, 2011: An evaluation of the progress in reducing heat-related human mortality in
major U.S. cities. Natural Hazards, 56, 113-129. doi:10.1007/sll069-010-9552-3
148 Anderson, G. B., and M. L. Bell, 2011: Heat waves in the United States: Mortality risk during heat: waves and effect modification by heat wave
characteristics in 43 U.S. communities. Environmental Health Perspectives, 119, 210-218. doi:10.1289/ehp.l002313
149 Sarofim, M.C., S. Saha, M.D. Hawkins, D.M. Mills, J. Hess, R. Horton, P. Kinney, J. Schwartz, and A. St. Juliana, 2016: Ch„ 2: Temperature-
Related Death and Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change
Research Program, Washington, DC, 43-68, doi: 10.7930/J0MG7MDX
51
-------
HEALTH
Extreme Temperature Mortality
empirical dataset did not include cities located in the Northern Plains, therefore no results are reported
for that region. Annual damages from increases in extreme heat related deaths are $140 billion under
RCP8.5 and $60 billion under RCP4.5 by the end of the century.
As a sensitivity study, this analysis was also conducted using assumptions that approximated higher
physiological adaptation and increased availability of air conditioning. The "with adaptation" values
(Table 5.2) evaluate the potential impacts of temperature adaptation by setting the threshold
temperature for extreme heat days equal to the values for Dallas, Texas, the second warmest city in the
analysis (unless the city had a higher threshold). Assuming all cities have the adaptive capacity of
residents of Dallas results in substantially lower increases in premature deaths, more than halving
mortality under all years and climate scenarios.
Table 5.2. Change in Annual Premature Mortality from Extreme Heat and Resulting Damages
Changes in annual mortality by region are five-GCM means of premature deaths from the 1989-2000
reference period, due to extreme heat only (not including change in cold-related mortality),. Totals based
on the 49 modeled cities are also shown for combined heat and cold (noted as '*with cold' below) and
for the heat-only mortality results of the adaptation sensitivity study (noted as '*with adaptation'
below). No cities in the Northern Plains region were modeled. Values may not sum due to rounding.
2050
RCP8.5
RCP4.5
Premature Deaths
(Annual)
Valuation
(millions of $2015)
Premature Deaths
(Annual)
Valuation
(millions of $2015)
Northeast
660
$8,200
660
$8,200
Southeast
610
$7,600
470
$5,800
Midwest
800
$10,000
700
$8,700
Southern Plains
550
$6,900
360
$4,400
Southwest
850
$11,000
400
$5,000
Northwest
0
$0
0
$0
Total
3,500
$43,000
2,600
$32,000
*With Cold
3,400
$42,000
2,600
$32,000
*With Adaptation
1,000
$13,000
650
$8,100
2090
RCP8.5
RCP4.5
Premature Deaths
(Annual)
Valuation
(millions of $2015)
Premature Deaths
(Annual)
Valuation
(millions of $2015)
Northeast
2,300
$35,000
970
$15,000
Southeast
1,600
$25,000
670
$10,000
Midwest
2,000
$31,000
840
$13,000
Southern Plains
1,300
$19,000
620
$9,400
Southwest
2,000
$31,000
830
$13,000
Northwest
0
$0
0
$0
Total
9,300
$140,000
3,900
$60,000
*With Cold
9,300
$140,000
3,900
$60,000
*With Adaptation
4,300
$65,000
1,300
$19,000
52
-------
HEALTH
Extreme Temperature Mortality
5.5 DISCUSSION
The result that climate change will lead to an annual increase of thousands of deaths in the U.S. is
consistent in direction and magnitude with the majority of studies in this field.150 The main estimates
presented in this section account for the difference in sensitivity due to geography (e.g., a 100-degree
day in Texas generally leads to fewer health impacts than a 100-degree day in Vermont), but do not
account for changes in sensitivity over time as humans adapt to a changing climate, whether due to
increased availability of air conditioning or how the human body can become accustomed to high
temperatures over time. The sensitivity analysis considered a future in which the human health
response to extreme temperatures in all 49 cities in the future was equal to that of Dallas today, and in
this case the results showed that mortality dropped more than 50% compared to the main estimates.
This provides one estimate of the number of lives adaptation might save, but even in this scenario,
deaths increased significantly compared to present-day conditions.
Beyond this one sensitivity analysis, the overall study only covers 49 cities, or approximately 1/3 of the
U.S. population, and so these values are likely underestimates. Furthermore, this study does not
consider the interactive effects of extreme heat and air quality, which could have compounding effects.
It is also important to recognize that some populations are at greater risk to extreme heat conditions,
particularly older adults, children, people working outdoors, the socially isolated and economically
disadvantaged, those with chronic illnesses, and some communities of color.151
The study also only considers deaths related to extreme temperatures, though extreme heat will very
likely lead to an increase in morbidity as well. Loss of internal temperature control can result in a
cascade of illnesses, including heat cramps, heat exhaustion, heatstroke, and hyperthermia in the
presence of extreme heat, and hypothermia and frostbite in the presence of extreme cold. Temperature
extremes can also worsen chronic conditions such as cardiovascular disease, respiratory disease,
cerebrovascular disease, and diabetes-related conditions. Prolonged exposure to high temperatures is
associated with increased hospital admissions for cardiovascular, kidney, and respiratory disorders.
Exposures to high minimum temperatures may also reduce the ability of the human body to recover
from high daily maximum temperatures.152 None of these health impacts, nor any associated mental
health impacts, are included in this analysis.
150 Ibid.
151 Ibid.
152 Ibid.
53
-------
HEALTH
Labor
6. LABOR
6.1 KEY FINDINGS
• Under RCP8.5, labor hours in the U.S. are projected to decrease due to increases in extreme
temperatures, especially for outdoor industries whose workers are exposed to the elements.
Considering changes in both extreme heat and cold, approximately 1.9 billion labor hours are
projected to be lost in 2090, costing an estimated $160 billion in lost wages.
• RCP4.5 avoids the loss of more than 900 million labor hours and nearly $75 billion in wages in 2090
compared to RCP8.5.
6.2 INTRODUCTION
Climate change affects labor in a number of ways, but projections of hotter summer temperatures raise
a particular concern. Extreme summer temperatures will be more frequent and intense in the future.153
Exposure to higher average temperatures and temperature extremes affect workers' health, safety, and
productivity.154 When exposed to high temperatures, workers are at risk for heat-related illnesses (e.g.
heat stroke and heat exhaustion) and fatigue155 and therefore may take more frequent breaks, or may
have to stop work entirely, resulting in lower overall labor capacity. This is especially true for high-risk
industries where workers are doing physical labor and have a direct exposure to outdoor temperatures
(e.g., agriculture, construction, utilities, and manufacturing).156
6.3 APPROACH
This analysis focuses on the impact of changes in extreme temperatures on labor supply157 across the
contiguous U.S. Specifically, the analysis estimates the number of labor hours lost due to changes in
extreme temperatures using dose-response functions for the relationship between temperature and
labor from Graff Zivin and Neidell (2014).158 Mean maximum temperatures from the five GCMs are
projected for four future periods (five-year averages centered on 2030, 2050, 2070, and 2090) at the
county level for RCP8.5 and RCP4.5. The analysis estimates the total labor hours lost in all categories of
the labor force and also for workers in high-risk industries (most likely to be strongly exposed to
extreme temperature), taking into account county-level population projections from the ICLUSv2
153 Walsh, J., D. Wuebbles, K. Hayhoe, J. Kossin, K. Kunkel, G. Stephens, P. Thome, R. Vose, M. Wehner, J. Willis, D. Anderson, S. Doney, R. Feely,
P. Hennon, V. Kharin, T. Knutson, F. Landerer, T. Lenton, J. Kennedy, and R. Somerville, 2014: Ch. 2: Our Changing Climate. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 19-67. doi:10.7930/J0KW5CXT.
154 USGCRP, 2016: The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. Crimmins, A., J. Balbus, J.L.
Gamble, C.B. Beard, J.E. Bell, D. Dodgen, R.J. Eisen, N. Fann, M.D. Hawkins, S.C. Herring, L. Jantarasami, D.M. Mills, S. Saha, M.C. Sarofim, J.
Trtanj, and L. Ziska, Eds. U.S. Global Change Research Program, Washington, DC, 312 pp. doi: 10.7930/J0R49NQX.
155 Gamble, J.L., J. Balbus, M. Berger, K. Bouye, V. Campbell, K. Chief, K. Conlon, A. Crimmins, B. Flanagan, C. Gonzalez-Maddux, E. Hallisey, S.
Hutchins, L. Jantarasami, S. Khoury, M. Kiefer, J. Kolling, K. Lynn, A. Manangan, M. McDonald, R. Morello-Frosch, M.H. Redsteer, P. Sheffield, K.
Thigpen Tart, J. Watson, K.P. Whyte, and A.F. Wolkin, 2016: Ch. 9: Populations of Concern. The Impacts of Climate Change on Human Health in
the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 247-286. doi: 10.7930/J0Q81B0T.
156 Graff Zivin, J. and M. Neidell, 2014: Temperature and the allocation of time: implications for climate change. Journal of Labor Economics, 32,
1-26, doi:10.1086/671766.
157 This analysis uses the term labor supply to refer to hours worked, but cannot determine whether that choice is driven by employees or
employers.
158 Graff Zivin, J. and M. Neidell, 2014: Temperature and the allocation of time: implications for climate change. Journal of Labor Economics, 32,
1-26, doi:10.1086/671766.
54
-------
HEALTH
Labor
model.159 The fraction of workers in high-risk industries is calculated using Bureau of Labor Statistics
data from 2003-2007 and is assumed to remain fixed over time for each county.160 The dose-response
functions are estimates of short-run responses to changes in weather, and as such do not account for
longer-term possibilities, such as acclimation of workers, relocation of industries, technological
advancements to reduce exposure, or broader changes in the labor force.161 The analysis estimates the
cost of the projected losses in labor hours based on the Bureau of Labor Statistics' estimated average
wage in 2005, adjusted to future years based on the projected change in GDP per capita.162 163 For more
information on the approach for the labor sector, please refer to Graff Zivin and Neidell (2014)164 and
EPA (2015).165
6.4 RESULTS
Without global GHG mitigation, an increase in extreme heat is projected to have a large negative impact
on U.S. labor hours, especially for outdoor labor industries. Under RCP8.5, almost 1.9 billion labor hours
across the national workforce are projected to be lost annually by 2090 due to unsuitable working
conditions (individual GCM results range from 1.0 to 2.7 billion hours lost). Losses are particularly large
in the Southeast (0.57 billion labor hours annually) and Midwest (0.40 billion labor hours annually). Loss
of labor hours across the U.S. is projected to be very costly, totaling over $160 billion in lost wages per
year by 2090 (range from $87- $220 billion). More than a third of this national loss is projected to occur
in the Southeast ($47 billion annually).
As shown in Figure 6.1, the majority of the country is projected to experience decreases in labor hours
due to extreme temperature effects. In 2090, losses of high-risk labor hours up to 6.5% are estimated
for counties in the Southwest, Texas, and Florida under RCP8.5. Only a limited number of counties are
projected to experience increases in labor hours, shown by the green-shaded areas Figure 6.1, which
occur in areas that do not frequently experience extreme heat but whose extremely cold temperatures
will be warmed in the future. See Section A.7 of the Appendix for GCM-specific maps of percent change
in labor hours.
159 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (Iclus) (Version
2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
160 Bureau of Labor Statistics, cited 2017: Quarterly Census of Employment and Wages. [Available online at http://www.bls.gov/cew/]. High-risk
workers were defined as those employed in agriculture, forestry, and fishing; hunting, mining, and construction; and manufacturing,
transportation, and utilities industries. It is important to note the national distribution of workers remains fixed relative to the average
distribution from 2003-2007, and that air conditioning and othertime-allocation behaviors remain fixed.
161 The underlying method described in Graff Zivin and Neidell (2014) found no temporal displacement of labor across days in the sample
dataset, indicating decreased working hours caused by high temperatures do not cause workers to supply additional labor at othertimes (i.e.,
to make up for lost work).
162 Bureau of Labor Statistics, cited 2017: Quarterly Census of Employment and Wages. [Available online at http://www.bls.gov/cew/]. Average
wage ($23.02 per hour in a 35-hour work week) calculated using high-risk labor categories only, as most extreme temperature impacts on labor
hours occur in these industries.
163 Cost of projected losses in labor hours were estimated by adjusting the Bureau of Labor Statistics' estimated average wage in 2005 of $23.02
to $27.47 and calculating to 2100 using an index reflecting projected changes in GDP per capita given real $2015 GDP and population.
164 Graff Zivin, J. and M. Neidell, 2014: Temperature and the allocation of time: implications for climate change. Journal of Labor Economics, 32,
1-26, doi:10.1086/671766.
165 EPA, 2015: Technical Appendix for Report: Climate Change in the United States: Benefits of Global Action. Section G: Technical Details
Related to Labor Analysis. U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA 430-R-15-001. Available online at
https://www.epa.gov/cira/downloads-cira-report
55
-------
HEALTH
Labor
Figure 6.1. Percent Change in Hours Worked
Estimates represent change in hours worked from the 2003-2007 reference period at the county level for
high-risk industries only, and are normalized by the high-risk working population in each county. Values
represent five-year averaged results across the five GCMs in 2050 and 2090 under RCP4.5 and RCP8.5.
2090
Percent Change in
Hours Worked
-6.5% to -5%
-4.9% to -4%
-3.9% to -3%
-2.9% to -2%
-1.9% to -1 %
-0.9% to 0%
0.1% to 1.5%
2050
RCP4.5
56
-------
HEALTH
Labor
Table 6.1. Change in Annual Hours Worked and Wages Lost for High Risk Industries
Estimates are five-year averaged changes from the 2003-2007 reference period in 2050 and 2090 under
RCP8.5 and RCP4.5. Values may not sum due to rounding.
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Change in Hours (millions) Compared to 2005
CanESM2
-920
-640
-2100
-840
CCSM4
-810
-590
-1700
-920
GISS_E2_R
-500
-380
-1000
-620
HadGEM2_ES
-1400
-1100
-2700
-1500
MIROC5
-760
-780
-1800
-960
5-Model Average
-880
-700
-1900
-970
Change in Wages (millions of $2015) Compared to 2005
CanESM2
-$46,000
-$32,000
-$180,000
-$70,000
CCSM4
-$41,000
-$30,000
-$140,000
-$76,000
GISS_E2_R
-$25,000
-$19,000
-$87,000
-$52,000
HadGEM2_ES
-$70,000
-$56,000
-$220,000
-$120,000
MIROC5
-$38,000
-$39,000
-$150,000
-$79,000
5-Model Average
-$44,000
-$35,000
-$160,000
-$80,000
At the national level, impacts to hours worked (Figure 6.1) and to labor costs (Figure 6.2) are
substantially smaller under RCP4.5 than RCP8.5, particularly in 2090. The difference between RCP8.5
and RCP4.5 is approximately 180 million labor hours per year across the workforce by 2050,
representing a savings of nearly $9.0 billion in annual wages (Table 6.1). In 2090, RCP4.5 would prevent
the loss of more than 910 million labor hours annually and nearly $75 billion in wages compared to
RCP8.5166. The avoided loss of labor hours under RCP4.5 compared to RCP8.5 is more than five times
higher in 2090 than in 2050.
166 Differences are based on the actual values and not the rounded numbers reported in Table 6.1.
57
-------
HEALTH
Labor
Figure 6.2. Wages Lost for All Labor Categories
Estimates represent the change in wages compared to the reference period (2003-2007) for the
contiguous U.S. in 2030, 2050, 2070, and 2090 under RCP4.5 and RCP8.5
$180
$160
In $140
& $120
to
¦2 $ioo
ifa
r $80
(D
Ctf)
ro
$60
$40
$20
$0
I RCP8.5 ¦ RCP4.5
2030
2050
2070
2090
6.5 DISCUSSION
Rising temperatures and changes in extreme heat events are projected to lead to significant losses in
hours worked in high-risk industries resulting in significant losses in wages across the contiguous U.S.,
particularly by the end of the century. The results presented show a significant difference between
RCP8.5 and RCP4.5 impacts on both changes in hours worked and wages for high-risk industries by the
end of the century. The Southeast and the Midwest are particularly vulnerable to future labor
productivity losses. A similar study of climate change impacts on labor also found that increasingly
extreme heat across the nation—especially in the Southwest, Southeast, and Upper Midwest-
threatens productivity and human health.167
This analysis only partially captures the effects that changes in humidity could have on worker's health
and productivity. Studies have found that the Southeast is particularly vulnerable to high wet-bulb
temperatures.168 169 For example, one study found that labor productivity by the end of the century is
projected to decrease up to 3.1% in the Southeast, the area with the highest wet-bulb temperatures, for
high-risk job sectors like construction, mining, utilities, transportation, agriculture, and
manufacturing.170 In addition to changes in temperature and humidity, climate change also affects the
167 Gordon, K., 2014: Risky Business: The Economic Risks of Climate Change in the United States: a Climate Risk Assessment for the United
States. M. Lewis and J. Rogers, Eds. Available online at
http://riskvbusiness.org/site/assets/uploads/2015/09/RiskvBusiness Report WEB 09 08 14.pdf
168 Dunne, J. P., R.J. Stouffer, and J.G. John, 2013: Reductions in labour capacity from heat stress under climate warming. Nature Climate
Change, 3, 563-566, doi:10.1038/nclimatel827.
169 Rhodium Group, 2014: American Climate Prospectus: Economic Risks in the United States. Input to the Risky Business Project. Available
online at http://climateprospectus.org/assets/publications/AmericanClimateProspectus vl.2.pdf
170 Gordon, K., 2014: Risky Business: The Economic Risks of Climate Change in the United States: a Climate Risk Assessment forthe United
States. M. Lewis and J. Rogers, Eds. Available online at
http://riskvbusiness.org/site/assets/uploads/2015/09/RiskvBusiness Report WEB 09 08 14.pdf
58
-------
HEALTH
Labor
health of outdoor workers by worsening air quality and increasing pollen exposure, increasing the
frequency or severity of other forms of extreme weather, and altering exposure to vector-borne
diseases.171 Some of these impacts are quantified in this Technical Report (see the Air Quality,
Aeroallergen, and West Nile Virus sections), but others are not assessed, including the potential
compounding effects from multiple climate-induced stressors.
As noted in the Approach section, this study does not consider the influence of potential adaptation
measures either by workers (e.g. physiological acclimation), employers (e.g. policies or infrastructure to
reduce exposure), or industries (e.g. technological advancements). Adaptation measures would likely
improve the health and safety of workers, but resulting impacts on worker productivity remain
uncertain.
171 Gamble, J.L., J. Balbus, M. Berger, K. Bouye, V. Campbell, K. Chief, K. Conlon, A. Crimmins, B. Flanagan,
C. Gonzalez-Maddux, E. Hallisey, S. Hutchins, L. Jantarasami, S. Khoury, M. Kiefer, J. Kolling, K. Lynn, A. Manangan, M. McDonald, R. Morello-
Frosch, M.H. Redsteer, P. Sheffield, K. Thigpen Tart, J. Watson, K.P. Whyte, and A.F. Wolkin, 2016: Ch. 9: Populations of Concern. The Impacts of
Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 247-
286. doi: 10.7930/J0Q81B0T.
59
-------
HEALTH
West Nile Virus
7. WEST NILE VIRUS
7.1 KEY FINDINGS
• Increases in annual average temperature are expected to continue to increase incidence of West
Nile virus in multiple regions of the U.S.
• Annual national cases of West Nile neuroinvasive disease (WNND) are projected to more than
double by 2050, increasing by 1,300 and 1,000 cases above the reference period (970 cases in 1995)
under RCP8.5 and RCP4.5, respectively. In 2090, an additional 3,300 and 1,700 annual cases
compared to the reference period are projected under RCP8.5 and RCP4.5, respectively.
• The associated monetized impact of these additional climate-attributable cases is estimated at $1.1
billion and $0.87 billion in 2050 under RCP8.5 and RCP4.5, respectively, and $3.3 billion and $1.8
billion in 2090 under RCP8.5 and RCP4.5, respectively.
7.2 INTRODUCTION
West Nile virus (WNV) is the most widely distributed arthropod-borne virus in the world, and the leading
cause of arthropod-borne viral disease in the U.S.172,173 It was first detected in the Western Hemisphere
in 1999 and rapidly spread across much of the Americas by 2005.174,175 The virus is now endemic
throughout most of the contiguous U.S., and is transmitted between birds and several species of
mosquitoes, which can cause infection in humans.176 Climate change has the potential to alter the
geographic distributions of WNV and its vectors.177 Above-normal temperatures have been among the
most consistent predictors of outbreaks, due in part to the acceleration of viral incubation in mosquitoes
and increased mosquito reproduction rates at higher temperatures.
WNV is classified as a nationally notifiable health outcome; accordingly, state health agencies are
responsible for reporting West Nile virus cases to the Centers for Disease Control and Prevention
(CDC).178 WNV cases can be distinguished by severity of the patient's symptoms; milder cases may
produce symptoms (e.g., fever, headache, rash, vomiting) that are indistinguishable from other
illnesses.179 As a result, there are questions about the accuracy of reported case counts for these milder
expressions of WNV because of potential under-reporting and incorrect attribution. In contrast, West
Nile neuroinvasive disease (WNND) cases, which occur for less than 1% of people infected with the
disease, can affect the patient's brain or cause neurologic dysfunction and these cases typically result in
172 Kramer, L.D., L.M. Styer, and G.D. Ebel, 2008: A Global Perspective on the Epidemiology of West Nile Virus. Annu Rev Entomol, 53, 61-81.
173 Lindsey, N.P., J.A. Lehman, E. Staples, and M. Fischer, 2015: West Nile Virus and Other Nationally Notifiable Arboviral Diseases - United
States, 2014. Morbidity and Mortality Weekly Report, 64, 929-934.
174 Gubler, D.J., 2007: The Continuing Spread of West Nile Virus in the Western Hemisphere. Clin Infect Dis, 45, 1039-1046.
175 Hayes, E.B. and D.J. Gubler, 2006: West Nile Virus: Epidemiology and Clinical Features of an Emerging Epidemic in the United States. Annu
Rev Med, 57, 181-194.
176 Reisen, W.K., 2013: Ecology of West Nile Virus in North America. Viruses, 5, 2079-2105.
177 Beard, C.B., R.J. Eisen, C.M. Barker, J.F. Garofalo, M. Hahn, M. Hayden, A.J. Monaghan, N.H. Ogden, and P.J. Schramm, 2016: Chapter 5:
Vector-Borne Diseases. In: The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment, U.S. Global Change
Research Program, Washington, DC, 129-156. http://dx.doi.org/10.7930/J0765C7V
178 CDC, 2015: West Nile Virus: Surveillance Resources. Centers for Disease Control and Prevention. Available online at
https://www.cdc.gov/westnile/resourcepages/survresources.html
179 CDC, 2016: National Notifiable Diseases Surveillance System (NNDSS): Arboviral Diseases, Neuroinvasive and Non-Neuroinvasive 2015 Case
Definition. Centers for Disease Control and Prevention. Available online at https://wwwn.cdc.gov/nndss/conditions/arboviral-diseases-
neuroinvasive-and-non-neuroinvasive/case-definition/2015/
60
-------
HEALTH
West Nile Virus
a patient's hospitalization. Because it is unlikely that these WNND patients could or would avoid
hospitalization given the severity of their syndromes, there is more certainty with these reported WNND
cases.180
7.3 APPROACH
This analysis explores the relationship between temperature and WNND in the contiguous U.S. using
WNND cases, population, and temperature data from 2004-2012 to develop county-specific health
impact functions. Specifically, the study estimates regional associations between temperatures and the
probability of above-average WNND incidence. Based on these health impact functions, county-level
expected WNND incidence rates are then estimated for a 1995 reference period (1986-2005) and two
future years (2050: 2040-2059 and 2090: 2080-2099) using temperature data from the five GCMs under
two RCPs. These results are combined with projections of county-level populations to calculate the
potential number of cases for 2050 and 2090. All-age, county-level population projections are from the
ICLUSv2181 for 2010, 2050, and 2090. The 2010 population estimates were used with the reference
period for 1986-2005 to provide a more recent representation of the population.
Cases were allocated to fatal and nonfatal outcomes based on a 6.5% mortality rate reported in the
national summary of 2014 WNND cases.182 For nonfatal outcomes, a cost of $41,391 ($2015) was used,
which reflects incurred hospital charges for patients with associated meningitis, encephalitis, and acute
flaccid paralysis syndrome. Hospitalization costs do not account for lost productivity during
hospitalization, related outpatient costs, or associated pain and suffering. Fatal WNND case costs are
estimated using a baseline VSL adjusted to future years based on changes in economic growth.183 For
more information on the approach to estimating climate change impacts on WNV cases, please refer to
Belova et al. (2017).184
7.4 RESULTS
Approximately half of all U.S. counties reported a WNND case from 2004 to 2012. Because high
temperatures are associated with a higher probability of above-average WNND incidence, this study
projects the number of years within the future 20-year time periods that are substantially warmer than
temperatures between 2004 and 2012, and therefore more likely to result in WNND outbreaks. In 2050
(under both RCPs) and under RCP4.5 in 2090, few counties are projected to have more than four years
(out of the possible 20 year eras) during which projected temperatures are substantially warmer than
180 Hahn, M.B., A.J. Monaghan, M.H. Hayden, R.J. Eisen, M.J. Delorey, N.P. Lindsey, R.S. Nasci, and M. Fischer, 2015: Meteorological Conditions
Associated with Increased Incidence of West Nile Virus Disease in the United States, 2004-2012. Am J Trop Med Hyg, doi:10.4269/ajtmh.l4-
0737.
181 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (ICLUS)
(Version 2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
182 Lindsey, N.P., J.A. Lehman, E. Staples, and M. Fischer, 2015: West Nile Virus and Other Nationally Notifiable Arboviral Diseases - United
States, 2014. Morbidity and Mortality Weekly Report, 64, 929-934.
183 At the time of this analysis, the EPA's Guidelines for Preparing Economic Analyses recommends a VSL of $7.9 million (2008$), based on 1990
incomes. To create a VSL using $2015 and based on 2015 incomes, the standard value was adjusted for inflation and for income growth
adjustment based on the approach described in EPA's BenMAP-CE model and its documentation. The resulting value, $10.0 million for 2015
($2015), was adjusted to future years by assuming an elasticity of VSL to GDP per capita of 0.4. Projections of U.S. GDP and population
described in the Modeling Framework section of this Technical Report were employed. Using this approach, the VSL in 2050 is estimated at
$12.4 million and $15.2 million in 2090. Sources: 1) EPA, 2014: Guidelines for Preparing Economic Analyses. National Centerfor Environmental
Economics. Available online at http://vosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-0568-52.pdf/$file/EE-0568-52.pdf: and 2) EPA, cited 2017:
Benefits Mapping and Analysis Program (BenMAP): Manual and Appendices for BenMAP-CE. Available online at
https://www.epa.gov/benmap/manual-and-appendices-benmap-ce
184 Belova, A., Mills, D., Hall, R., Juliana, A.S., Crimmins, A., Barker, C. and Jones, R. (2017) Impacts of Increasing Temperature on the Future
Incidence of West Nile Neuroinvasive Disease in the United States. American Journal of Climate Change, 6, 166-216.
https://doi.org/10.4236/aicc.2017.61010.
61
-------
HEALTH
West Nile Virus
the 2004 to 2012 observed annual mean (based on the five GCM average). However, many counties are
projected to have multiple future years (e.g. more than 10 years out of the possible 20 year eras) with
annual mean temperatures that are substantially different compared to the observed 2004-2012 period
under RCP8.5 in 2090.
Estimates of the expected change in the annual number of WNND cases in 2090 under RCP8.5 compared
to the reference period (970 cases nationally for the 1986-2005 period) differ considerably from the
2050 and RCP4.5 estimates, as seen in Figure 7.1 and Table 7.1. While all regions show increases in the
future expected annual number of cases, the results for the Southeast are the most striking: the number
of total cases grows from approximately 100 in the reference period to more than 1,200 in 2090 under
RCP8.5. Variability in GCM results are shown in Figure 7.2, using the Southeast as an example.
Figure 7.1. Projected Regional WNND Cases
Maps show annual cases by region in the 1995 reference period (1986-2005), 2050 (2040-2059), and
2090 (2080-2099) under RCP8.5 and RCP4.5. Populations are consistent with the representative year,
and results are averaged over the GCMs and modeled climate periods.
Projected WNND Cases
20 -100
101 - 250
m 251 - 400
I I 401 - 550
l~~l 551 - 700
701 - 950
951 - 1,230
|. . . | Results not statistically significant
RCP8.5
2050
Reference
2090
RCP4.5
62
-------
HEALTH
West Nile Virus
Table 7.1. Projected Increases in WNND Cases
Future cases by region are estimated under a dynamic population (populations are modeled for 2050
(2040-2059) and 2090 (2080-2099)) and holding population constant at 2010 levels to demonstrate the
portion of additional cases attributable to climate change. Cases are reported as increases from the 1995
reference period (1986-2005). Values may not sum due to rounding.
Dynamic Population
2010 Population
2050
2090
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
170
130
490
210
140
99
420
160
Southeast
370
270
1100
440
300
210
960
330
Midwest
170
130
450
210
110
81
350
130
Northern Plains
100
79
330
150
71
51
250
85
Southern Plains
220
200
450
330
70
51
160
71
Southwest
240
230
420
380
28
21
56
29
Northwest
6.5
6.4
11
11
0.44
0.35
0.84
0.47
Total
1,300
1,000
3,300
1,700
720
510
2200
800
There are often considerable differences across the estimates of the expected annual number of WNND
cases for the U.S. in a specific a reporting period. These differences are largely a reflection of year-to-
year variability in the projected annual average temperatures across the calendar years selected to
represent the reporting period, as well as model and scenario uncertainty. In the example of the
Southeast shown in Figure 7.2, across RCPs and reporting years (i.e., comparing results within the
panels), the largest modeled mean value is approximately -three-to-four times the smallest value,
highlighting this inter-model uncertainty. There is also increased year-to-year variability under RCP8.5
compared to the RCP4.5 scenario.
The costs of temperature-related increases in WNND cases (Table 7.2) in 2050 are approximately $1.1
billion and $0.87 billion for RCP8.5 and RCP4.5, respectively. Damages increase in 2090 to $3.3 billion
and $1.8 billion under RCP8.5 and RCP4.5, respectively. Removing cases from regions where the
temperature- incidence functions were not statistically significant in the underlying literature185 reduces
these values to $0.87 billion and $0.68 billion in 2050, and to $2.9 billion and $1.4 billion in 2090, under
RCP8.5 and RCP4.5, respectively.
185 Hahn, M.B., A.J. Monaghan, M.H. Hayden, R.J. Eisen, M.J. Delorey, N.P. Lindsey, R.S. Nasci, and M. Fischer, 2015: Meteorological Conditions
Associated with Increased Incidence of West Nile Virus Disease in the United States, 2004-2012. Am J Trop Med Hyg, doi:10.4269/ajtmh.l4-
0737.
63
-------
HEALTH
West Nile Virus
Figure 7.2. Projected Change in WNND Cases in the Southeast Region
The graphs present the mean estimated WNND cases in 2050 (2040-2059) and 2090 (2080-2099) by
GCM and RCP in addition to the cases in the 1986-2005 reference period under a dynamic population
(populations are modeled for 2050 and 2090). For the five GCMs, 2.5% and 97.5% confidence intervals
are also provided. Data on the confidence intervals for the five-GCM average are not available.
2050
2090
2,200
2,000
1,800
1,600
1,400
1,200
1,000
800
600
400
200
2,200
2,000
1,800
1,600
1,400
1,200
1,000
800
600
400
200
RCP8.5
u t ' i
-1 r-
-------
HEALTH
West Nile Virus
7.5 DISCUSSION
The results presented in this section indicate that climate change will increase the risk of WNND
infection rates across the contiguous U.S. Placing these results in a broader context is challenging
because to date, most of the work projecting the impact of climate change on vectorborne disease has
focused on evaluating potential changes in season lengths and range expansion rather than developing
incidence projections.
This study only accounts for projected changes in temperature, one of a number of factors that can
influence WNND disease occurrence and case totals. Future research could be improved by accounting
for changes in precipitation (an important factor for mosquito breeding) and changes in land use
characteristics that may affect bird, mosquito, and human distributions. Furthermore, it is difficult to
predict how human behavior may respond to a changing climate. This analysis assumes that the effects
of interventions (e.g., mosquito control and public outreach regarding personal protection from
mosquito biting) are captured by the original temperature-incidence relationships and that the nature of
those relationships will not change over time.
The results presented in this analysis, which only account for neuroinvasive WNV cases, are likely
conservative, as projections do not capture the potential for increases in WNV cases in the roughly half
of U.S. counties where no WNV cases were reported from 2004 to 2012. In addition, the approach for
monetizing hospitalization costs does not account for lost productivity during hospitalization, related
outpatient costs, or associated pain and suffering. Still, the projected increases in WNND cases are
noteworthy considering the absolute number of cases involved, the severity of associated health
impacts, and the magnitude of these increases relative to the projected reference.
65
-------
HEALTH
Harmful Algal Blooms
8. HARMFUL ALGAL BLOOMS
8.1 KEY FINDINGS
• Warming temperatures and changes in precipitation, which will drive alterations in river flow and
nutrient availability, are projected to increase the occurrence of harmful algal blooms in many
watersheds of the contiguous U.S.
• In 2090, a warming climate under RCP8.5 results in an additional full month with harmful algal
concentrations above a recommended public health threshold, while RCP4.5 results in an additional
half month. The resulting losses to reservoir recreation in the contiguous U.S. are estimated to be on
the order of hundreds of millions of dollars per year by 2090.
8.2 INTRODUCTION
Harmful algal blooms (HABs) can affect human health and welfare by degrading the quality of drinking
water supplies and forcing activity restrictions at recreational waterbodies. Today, lakes and reservoirs
that serve as sources of drinking water for between 30 and 48 million Americans are susceptible to
periodic contamination by algal or cyanobacterial toxins (also called cyanotoxins). For example, in
August 2014 nearly 500,000 residents of Toledo, Ohio lost access to their drinking water after tests from
a cyanobacterial harmful algal bloom (CyanoHAB) in Lake Erie revealed the presence of cyanotoxins near
the water plant's intake.186 Certain drinking water treatment processes can address algal toxins;
however, removal efficiency can be as low as 60 percent.187 Recreational exposure to cyanotoxins in
lakes and other freshwater bodies can also adversely affect human health. In 2009 and 2010,
cyanotoxins were responsible for nearly half of all disease outbreaks linked to recreational freshwater
exposure.188
More freshwater CyanoHABs and increased human exposure risk are expected in the future as climate
change contributes to more suitable environmental conditions for bloom formation.189 Such conditions
are primarily due to warming surface water temperatures, but are also affected by changing
precipitation patterns. More frequent extreme precipitation events can increase freshwater runoff
carrying excess nitrogen, phosphorus, other nutrients, and sediments into water bodies. In addition,
intense precipitation followed by periods of drought increases residence time in reservoirs and supports
bloom formation.
186 Toledo, City of, 2014: Microcystin Event Preliminary Summary. 73 pp. City of Toledo Department of Public Utilities. [Available online at:
http://toledo.oh.gov/media/132055/Microcystin-Test-Results.pdf]
187 EPA, 2015: Recommendations for Public Water Systems to Manage Cyanotoxins in Drinking Water. United States Environmental Protection
Agency, Office of Water. EPA 815-R-15-010. Available online at: http://www2.epa.gov/sites/production/files/2015-06/documents/cvanotoxin-
management-drinking-water.pdf
188 Hilborn, E.D., V.A. Roberts, L. Backer, E. DeConno, J.S. Egan, J.B. Hyde, D.C. Nicholas, E.J. Wiegert, L.M. Billing, M. DiOrio, M.C. Mohr, F.J.
Hardy, T.J. Wade, J.S. Yoder, and M.C. Hlavsa, 2014: Algal bloom-associated disease outbreaks among users of freshwater lakes — United
States, 2009-2010. MMWR. Morbidity and Mortality Weekly Report, 63, 11-15. Available online at
http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6301a3.htm
189Trtanj, J., L. Jantarasami, J. Brunkard, T. Collier, J. Jacobs, E. Lipp, S. McLellan, S. Moore, H. Paerl, J. Ravenscroft, M. Sengco, and J. Thurston,
2016: Ch. 6: Climate Impacts on Water-Related Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific
Assessment. U.S. Global Change Research Program, Washington, DC, 157-188. http://dx.doi.org/10.7930/J03F4MH
66
-------
HEALTH
Harmful Algal Blooms
8.3 APPROACH
Building on the water resource framework described in the Water Quality section of this report,190 this
analysis uses a series of linked models to evaluate the biophysical impacts of climate change on
CyanoHAB occurrence in the contiguous U.S. The linked model chain starts with climate projections from
five GCMs under RCP8.5 and RCP4.5 that are input into a rainfall-runoff model (CLIRUN-II), which
simulates monthly runoff in each of the 2,119 eight-digit hydrologic unit codes (HUCs) of the contiguous
U.S. A water demand model then projects water requirements of the municipal and industrial (M&l) and
agriculture sectors under a scenario assuming population growth consistent with the other sectors of
this Technical Report. With these runoff and demand projections, a water resources systems model
produces a time series of reservoir storage, release, and allocation to the various demands in the system
(e.g., M&l, agriculture, environmental flows, transboundary flows, hydropower).
A modified version of the QUALIDAD water quality model191 incorporates this information on managed
flows and reservoir parameters while simulating a number of water quality metrics in waterbodies,
including cyanobacteria concentrations. To address uncertainty in the relationship between
cyanobacteria growth and water temperature, the analysis uses two algal growth scenarios: a) a "low-
growth" scenario that plateaus as temperatures reach 26°C, and b) a "high-growth" scenario that
assumes a linear growth rate with changes in temperature.
Finally, the analysis evaluates cyanobacteria concentrations for their potential recreational impacts in
terms of potential days of restricted recreational activity at specific sites. The approach quantifies lost
recreation at 279 reservoirs to estimate the partial economic effects associated with changes in algal
bloom risk. To isolate the effects of climate change from the effects of population growth, the projected
changes are all expressed relative to the a control scenario, which uses historical climate from 1986-
2005, in addition to population growth effects. For more information on the approach to estimating
climate change impacts on harmful algal blooms, please refer to Chapra et al. (2017).192
8.4 RESULTS
Climate change is projected to increase the risk of CyanoHAB occurrence in the future. Figure 8.1 shows
the changes in cyanobacteria concentrations for the surface layer of all waterbodies at the four-digit
HUC level. Across most watersheds of the country, concentrations increase overtime and are higher
under RCP8.5 compared to RCP4.5. The projected change in cyanobacteria concentrations is generally
larger under the high-growth scenario compared to the low-growth scenario. The largest increases are
projected for the Midwest, Northern Plains, and Southern Plains.
Under the low-growth scenario, areas of large decreases in annual average cell counts are observed in
parts of the West, especially under RCP8.5. This counterintuitive finding occurs when the growth of
cyanobacteria plateaus under the low-growth scenario (at water temperatures exceeding 30°C) while
the respiration rate continues to increase with temperature, eventually causing decreases in
cyanobacteria concentrations.193 This effect is not observed in the high-growth scenario because growth
190 As described in: Boehlert, B., K.M. Strzepek, S.C. Chapra, C. Fant, Y. Gebretsadik, M. Lickley, R. Swanson, A. McCluskey, J.E. Neumann, and J.
Martinich, 2015: Climate change impacts and greenhouse gas mitigation effects on U.S. water quality. Journal of Advances in Modeling Earth
Systems, 7, 1326-1338, doi:10.1002/2014MS000400.
191 Chapra, S.C., 2014: QUALIDAD: A parsimonious modeling framework for simulating river basin water quality, Version 1.1, Documentation and
users manual, Civil and Environmental Engineering Dept., Tufts University, Medford, MA.
192 Chapra, S.C., B. Boehlert, C. Fant, J. Henderson, D. Mills, D.M.L. Mas, L. Rennels, L. Jantarasami, J. Martinich, K.M. Strzepek, V.J. Jr. Bierman,
and H.W. Paerl, 2017: Climate change impacts on harmful algal blooms in U.S. freshwaters: a screening-level assessment. Environmental Science
and Technology, doi: 10.1021/acs.est.7b01498.
193 This phenomenon of decreasing cyanobacteria concentrations at higher temperatures occurs when light and nutrient limitations restrict
growth to about 15% of maximum.
67
-------
HEALTH
Harmful Algal Blooms
does not plateau at these higher temperatures, but rather continues to increase along with respiration
as temperature increases.
Figure 8.1. Projected Change in Cyanobacteria Concentrations
Changes in average annual waterbody surface cyanobacteria (thousands of cells per ml) for the low-
growth and high-growth scenarios in 2050 (2040-2059) and 2090 (2080-2099) relative to the control
scenario. Values for each RCP represent the average results for the five GCMs, and are aggregated to the
four-digit HUC level, weighted by waterbody surface area.
Low-Growth Scenario
RCP8.5
RCP4.5
2050
A
fi
2090
i \
T\rn
At &
High-Growth Scenario
RCP8.5 RCP4.5
2050
> A
2090
< -8.0 -6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 >
Thousands of cells per ml
Figure 8.2 shows the population-weighted aggregated cyanobacteria concentration across the year for
the results from the 2090 era. In the Control scenario, the aggregate concentrations are slightly lower
than 20,000 cells/ml at the peak in August and September. In a changing climate, these peaks rise above
38,000 cells/ml on the low-end and above 70,000 on the high-end. Projected changes in cyanobacteria
concentrations vary between the two growth scenarios, especially during the late summer, and for
scenarios with higher projected temperature changes. In the low-growth scenario, the aggregate
68
-------
HEALTH
Harmful Algal Blooms
concentrations show less of a spread across the climate scenarios than in the high-growth scenario,
where increasing temperatures above 30°C continue to increase the growth rate. Under both growth
scenarios, projected concentrations are largest under the hottest GCMs (CanESM2 and HadGEM2-ES),
and the lowest under the GISS-E2-R GCM that projects a smaller amount of average warming.
69
-------
HEALTH
Harmful Algal Blooms
Figure 8.2. Seasonal Profile of Aggregate Cyanobacteria Concentration
Population-weighted aggregate cyanobacteria concentration (in thousands of cells per ml) in 2090.
Low-Growth Scenario
so
Q.
40
30
T3
20
10
0
High-Growth Scenario
70
60
50
40
30
20
10
0
GCM
RCP8.5
RCP4.5
CanESM2
CCSM4
— — — —
GISS-E2-R
HadGEM2-E$
MIROC5
Control
— — — —
70
-------
HEALTH
Harmful Algal Blooms
Table 8.1 shows the change in the mean number of days per year above two levels, 20,000 cells/ml and
100,000 cells/ml, recommended by the World Health Organization as important for managing human
health risk.194 The 20,000 cells/ml level is associated with moderate risk of short-term adverse health
effects. The 100,000 cells/ml level represents a very high risk of short or long-term adverse health
effects. The lowest increase in number of days is produced from the GISS-E2-R GCM, and the highest is
produced by the HadGEM2-ES GCM due, primarily, to the respective projected changes in air
temperature (which affects water temperature). For the 20,000 cells/ml threshold, the mean number of
additional risk days ranges from about 4 to 44 in 2090 across all five GCMs, both climate forcing
scenarios, and both growth scenarios. In 2090, a warming climate under RCP8.5 results in an additional
full month with counts above 20,000 cells/ml, while RCP4.5 results in an additional half month.
Table 8.1. Projected Change in Bloom Occurrence
Change in the mean number of days above the 20,000 cells/ml and 100,000 cells/ml thresholds per year
per waterbody, aggregated by population, relative to the control scenario. Results are shown for the low-
and high-growth scenarios for each RCP/GCM combination. Darker green colors represent greater
change in days per year with bloom occurrence.
20,000 cells/ml 100,000 cells/ml
Low High Low High
2050 2090 2050 2090 2050 2090 2050 2090
CanESM2
11
23
15
34
7
14
10
20
CCSM4
9
20
14
30
8
10
11
17
RCP8.5
GISS-E2-R
6
11
9
19
5
6
9
11
HadGEM2-ES
14
23
25
44
8
13
15
24
MIROC5
10
22
14
32
7
12
11
19
5-Model Average
10
20
16
32
7
11
11
18
CanESM2
9
9
12
13
6
6
9
9
CCSM4
8
11
12
17
7
8
11
12
RCP4.5
GISS-E2-R
4
4
7
8
4
3
7
5
HadGEM2-ES
12
17
19
27
8
9
13
15
MIROC5
10
12
14
19
8
9
12
14
5-Model Average
9
11
13
17
7
7
10
11
Changes in the risk of HAB occurrence will have wide-ranging economic impacts. This analysis estimates
the effects on recreation at 279 freshwater reservoirs across the contiguous U.S. As shown in Table 8.2,
climate change under both RCPs is estimated to result in economic losses to recreation in nearly all
regions of the country, with estimated damages being greater under the high-growth scenario.
Projected losses are generally larger under RCP8.5 compared to RCP4.5 in both 2050 and 2090, with the
Northeast, Southeast, and Southern Plains experiencing the largest adverse effects. Importantly, these
economic damage estimates represent only part of the total potential impact to human health and the
environment. In addition, this analysis does not evaluate the economic effects of climate change on
194 World Health Organization, 1999: Toxic Cyanobacteria in Water: A guide to their public health consequences, monitoring and management.
London: E & FN Spon. Available online at http://www.who.int/water sanitation health/resourcesqualitv/toxcvanobacteria.pdf?ua=l
71
-------
HEALTH
Harmful Algal Blooms
HABs in marine or estuarine waters or other freshwater bodies outside of the 279 reservoirs modeled.
Economic damages associated with cyanotoxin exposure in humans, increased water treatment,
ecological effects, and other impacts associated with HABs are not included in this analysis.
Table 8.2. Economic Impacts on Reservoir Recreation from HABs
Results represent the average of the five GCMs, and are shown in millions of undiscounted $2015.
Positive dollar values indicate losses in recreation value due to HABs. Negative values indicate that for a
given period there were less projected days of recreation loss under the climate change projection than
under the control scenario that includes population growth (but no climate change).
2050
2090
Region
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Low Growth
Northeast
$11
$6.8
$28
$16
Southeast
$48
$36
$82
$57
Midwest
$3.1
$1.7
$18
$3.5
Northern Plains
-$3.4
-$3.1
-$3.7
-$4
Southern Plains
$10
$10
$21
$16
Southwest
$7.1
$5
$5.1
$6.8
Northwest
$0.17
$0.21
$1.6
$0.12
Total
$77
$57
$150
$95
High Growth
Northeast
$5.5
$4.9
$27
$13
Southeast
$47
$41
$110
$70
Midwest
$5.3
$3.6
$35
$5.4
Northern Plains
$1.7
$1.2
$4.3
$2.9
Southern Plains
$14
$16
$56
$25
Southwest
$8.1
$4.8
$8.2
$8.9
Northwest
$0.19
$0.22
$5.4
$0.18
Total
$82
$71
$250
$130
8.5 DISCUSSION
This analysis demonstrates that climate change is likely to increase the risk of HAB occurrence across
many parts of the contiguous U.S. These results are consistent with the findings of the assessment
literature, which describe that CyanoHABs are strongly influenced by rising temperatures and altered
precipitation patterns, and that the seasonal and geographic range of suitable habitat for cyanobacterial
species is projected to expand.195
It is important to keep in mind several caveats when interpreting the results presented above. All
waterbodies are modeled as well mixed systems, although these are split into two vertical layers during
195Trtanj, J., L. Jantarasami, J. Brunkard, T. Collier, J. Jacobs, E. Lipp, S. McLellan, S. Moore, H. Paerl, J. Ravenscroft, M. Sengco, and J. Thurston,
2016: Ch. 6: Climate Impacts on Water-Related Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific
Assessment. U.S. Global Change Research Program, Washington, DC, 157-188. http://dx.doi.org/10.7930/J03F4MH4
72
-------
HEALTH
Harmful Algal Blooms
stratification. Given that cyanobacteria may concentrate vertically over specific depth ranges or
horizontally due to lateral winds, potentially compounding concentrations by orders of magnitude,196
the estimates presented here will be less than the potential maxima. This approach is not able to model
cyanobacteria toxicity, which varies by many factors and would be a more direct indicator of human
health risk. Since cyanobacteria range in shape and size, there is also uncertainty in the calculations to
convert from biomass to cell count. In addition, the effects of very high water temperatures (>30°C) on
the growth and respiration rate of cyanobacteria are not well understood. Although the results reported
here use low and high growth rate scenarios and assumptions, these may not fully capture the range of
potential outcomes. Finally, this analysis only partially quantifies the economic impacts of changes in
HAB risk. Valuation of the broader effects of HABs on human health and ecosystems, as well as those
that may occur outside of the reservoir-focus of this analysis, would increase the damages associated
with these impacts.
196 World Health Organization, 1999: Toxic Cyanobacteria in Water: A guide to their public health consequences, monitoring and management.
London: E & FN Spon. Available online at http://www.who.int/water sanitation health/resourcesqualitv/toxcvanobacteria.pdf?ua=l
73
-------
HEALTH
Domestic Migration
9. DOMESTIC MIGRATION
9.1 KEY FINDINGS
• Warmer weather and altered precipitation patterns due to climate change are projected to
influence human migration patterns. Compared to population projections assuming no climate
change, changes in regional climate are projected to increase populations along the coasts and in
the Midwest, and decrease populations in the Great Plains and the Gulf states of the Southeast.
• Differences in migration patterns between RCPs and climate model projections are modest.
However, the overall direction and magnitude of change at regional levels are consistent across all
scenarios analyzed.
• While not a dominant driver, climate is one of many factors that influence human migration in the
U.S. Some of these non-climate related factors are incorporated into the migration model, while
others are too complex and variable to predict.
9.2 INTRODUCTION
Domestic migration in the U.S. most often occurs when people are young adults, as they finish college,
make initial career decisions, serve in the military, and form families.197 Many factors can influence
migration, including economic opportunities or recessions, availability and affordability of housing, and
climate conditions. For example, people often prefer to move to places with warmer winters or cooler,
less-humid summers. Recent research has found a substantial increase in U.S. households' valuation of
moderate or "nice" weather's contribution to quality of life, and migration to areas with such climates is
likely to continue.198,199 Anticipating areas in the U.S. that may have population growth or decline in the
near future based on changes in climate variables can help communities plan for rising demands on
housing, transportation, energy infrastructure, and the many other needs of a growing population.
9.3 APPROACH
This analysis examines the effects of climate change on migration within the U.S. (at a county-scale)
using the ICLUSv2 (Integrated Climate and Land Use Scenarios, version 2) model.200 The ICLUSv2 model
uses 2010 U.S. Census county-level population data and shared socioeconomic pathways201 to project
population change using assumptions about fertility, mortality, and immigration through the end of the
century. ICLUSv2 includes a climate amenity value by using historic climate data and Internal Revenue
Service migration data covering the 1980-1999 period. Future projections of climate change are used to
update these amenity values at each time step of the migration model, enabling ICLUSv2 to reflect the
current and future amenity value of climate.
197 Cromartie, J., and P. Nelson, 2009: Baby Boom Migration and Its Impact on Rural America, ERR-79, U.S. Dept. of Agri., Econ. Res. Serv.
198 Rappaport, J., 2007: Moving to nice weather. Regional Science and Urban Economics, 37, 375-398, doi: 10.1016/j.regsciurbeco.2006.11.004.
199 Rappaport, J., 2008: Consumption amenities and city population density. Regional Science and Urban Economics, 38, 533-552, doi:
10.1016/j. regsciu rbeco .2008.02.001.
200 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (Iclus) (Version
2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
201 O'Neill, B. C., E. Kriegler, K. Riahi, K. L. Ebi, S. Hallegatte, T. R. Carter, R. Mathur, and D. P. v. Vuuren, 2014: A new scenario framework for
climate change research: the concept of shared socioeconomic pathways. Climatic Change, 122, 387-400, doi:10.1007/sl0584-013-0905-2.
74
-------
HEALTH
Domestic Migration
For this analysis, climate factors from the five GCMs are incorporated into the model for both RCP8.5
and RCP4.5. ICLUSv2 uses a spatial interaction model to simulate the migration of people within the U.S.
(net migration into and out of a county, but not into or out of the country), and this model incorporates
measures of climate as predictive variables. Specifically, the model uses mean summer (July) apparent
temperature (10-year running average), mean summer (June, July, August) precipitation (10-year
running average), mean winter (January) apparent temperature (10-year running average), and mean
winter (December, January, February) precipitation (10-year running average). Other factors included in
the model include population density, population growth rate, and developable land area. To determine
the role of climate change in future migration, the results presented in this section compare differences
in migration between a "no-climate change" reference projection202 and the migration model that
includes climate factors. Unlike other sectoral analyses in this Technical Report, this study does not
monetize the economic effects of the projected changes in climate-driven migration. Additional
information about ICLUSv2 can be found in EPA (2017).203
9.4 RESULTS
The results presented in this section focus on the effects of climate change on domestic migration, and
do not represent broader migration trends at the international level due to non-climate factors. While
changes in climate have a relatively small influence on migration in the short term, the cumulative effect
of this influence has a meaningful impact overtime (Figures 9.1 and 9.2). However, results show only
small differences between RCP8.5 and RCP4.5 over the course of the century (Figure 9.1), and
differences across the GCMs are modest204 (see Figure A8-2 in the Appendix for GCM-specific maps).
Clusters of county-scale, climate-driven emigration (negative percent changes) are more obvious in the
Southern Plains, for example, on the western edges of Texas, Kansas, and Nebraska; and throughout
Montana and the Southeast. Positive percent changes are clear throughout the Midwest and Northeast.
Some western states like Colorado, Oregon, and Washington are projected to have overall positive
percentage change trends, though select counties have negative areas. On average, the total population
of Midwest, Southwest, Northeast, and Northwest was higher relative to the no-climate change model,
while a decline in population is projected in the Northern Plains, Southeast, and Southern Plains by 2090
relative to the no-climate change reference (Figure 9.2).
Across all scenarios analyzed, the largest percent change for any NCA region was in the Northeast
(Figures 9.1 and 9.2). By 2090 and under RCP8.5, results under the HadGEM2-ES model show the
greatest increase compared to the no-climate change scenario, approximately 7%, equaling a population
gain of approximately 6 million individuals. The largest population decrease compared to the no-climate
change scenario was an 8.7% loss of people projected in the Southern Plains, equivalent to 5 million
individuals by 2090 under RCP8.5 of the HadGEM2-ES GCM. However, it is important to note that both
of these climate change-driven changes are occurring on top of underlying increases in population
growth in both regions, as other non-climate factors are dominant drivers of regional population
change.
202 This is the population projection used throughout the CIRA2.0 modeling framework to estimate county-level population changes. In this
projection, county-scale changes in population are driven by underlying shifts in demographic trends (birth rates, mortality, migration) and
preferences for urban, suburban, and rural surrounding. Please see the ICLUS 2017 documentation report for additional information regarding
the structure and assumptions of the ICLUSv2 model.
203 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (Iclus) (Version
2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
204 Nationally, the choice of climate model has only minimal influence over population changes. The largest difference is in the Northern Plains,
which ranges from a -4.8% change (HadGEM2-ES, RCP4.5) to a -8.0% (GISS-E2-R, RCP8.5) by 2090 (Figure 9.2). Seethe Appendix A.8 for
additional results.
75
-------
HEALTH
Domestic Migration
Figure 9.1. Climate Change-Induced Domestic Migration
Relative net differences in county-level population projections by RCP and year. Values represent the
average percentage change across the five GCMs compared to a "no climate change" control scenario.
RCP8.5
RCP4.5
2050
.
2090
Percent Change
< -9.o
IH-8.9to -6.0
-5.9 to -3.0
-2.9 to 0.0
0.1 to 3.0
3.1 to 6.0
M 6.1 to 9.0
>9.0
76
-------
HEALTH
Domestic Migration
Figure 9.2. Regional Climate Change-Induced Migration
Relative net differences in projected regional population by forcing scenario and climate model. Values
are expressed as the percentage change from a "no climate change" control scenario.
9
8
7
6
I &
0
§ 4
3
2
1
0
21
9
8
7
6
1 &
u
H
S 4
£
3
2
1
0
21
0
-1
-2
-3
| -4
§ -5
£ -6
-7
-8
-9
-10
Northeast
2050
Midwest
Southern Plains
Southeast
0
-l
•2
-3
o>
f-4
u
c
S -5
-6
-7
-8
-9
9
8
7
6
I 5
"5
C
ii 4
jr
13
a
1
o
-l
Northern Plains
2050
Southwest
2050
Northwest
GCM
Ml ROCS
HadGEM2-ES
GISS-E2-R
CanESM2
CCSM4
RCP8.5
2015 2025
77
-------
HEALTH
Domestic Migration
9.5 DISCUSSION
The findings described above suggest that climate change will influence human migration within the U.S.
in the coming decades. This finding is consistent with results from other studies that have documented a
link between climate and domestic migration.205,206 However, the actual distribution is contingent upon
many variables, and the ICLUSv2 model used in this analysis represents one configuration of how those
variables may affect future migration. For example, land area and population density are included in the
ICLUSv2 model, and have a greater influence on migration than climate change. In addition, this analysis
assumes a single scenario regarding overall birth, death, and migration rates, which can be challenging
to project over long time scales. Other factors not explicitly included in the model, like economic
opportunity, can also influence migration patterns. Finally, this analysis also assumes that every
individual will have the same response to climate change, when in reality people have unique responses
to temperature and precipitation patterns based on preferences overtime.
Climate change will also have other impacts beyond changing precipitation and temperature patterns
which may influence population distribution. Shifting areas of vectors that can carry diseases, land loss
due to sea level rise, or declining air quality207 could increase or decrease the likelihood that people
move, as well as their destination. Finally, migration is likely to change the geography of demands on
housing, transportation, energy infrastructure, and the many other needs of a growing population.
Using a migration model that evaluates the effect of climate change on these parameters would provide
a more robust characterization of these affects.
205 Sin ha, P., and M.L. Cropper, 2013: The value of climate amenities: evidence from US migration decisions. Resources for the Future,
Discussion paper: RFF DP 13-01.
206 Cragg, M., and M. Kahn, 1997: New estimates of climate demand: evidence from location choice. Journal of Urban Economics, 42, 261-284.
207 USGCRP, 2016: The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. Crimmins, A., J. Balbus, J.L.
Gamble, C.B. Beard, J.E. Bell, D. Dodgen, R.J. Eisen, N. Fann, M.D. Hawkins, S.C. Herring, L. Jantarasami, D.M. Mills, S. Saha, M.C. Sarofim, J.
Trtanj, and L. Ziska, Eds. U.S. Global Change Research Program, Washington, DC, 312 pp. http://dx.doi.org/10.7930/J0R49NQX
78
-------
INFRASTRUCTURE
10. ROADS
10.1 KEY FINDINGS
• Climate change-driven changes in temperature and precipitation are projected to result in
significant impacts to U.S. roads. Discounted, reactive adaptation costs (rehabilitation measures) are
estimated at $230 billion through 2100 under RCP8.5 and $150 billion under RCP4.5, on average.
• The highest per-lane-mile reactive adaptation costs are associated with impacts on paved roads due
to changes in temperature and precipitation. Changes in the freeze-thaw cycle are projected to lead
to a cost savings relative to the reference period.
• Across all road types and climate stressors, proactive adaptation to protect roads against climate
change-related impacts is projected to decrease costs over the century by 98% under RCP8.5 and
83% under RCP4.5.
10.2 BACKGROUND
The U.S. road network is one of the nation's most important capital assets. Roads are susceptible to
damage from various climate stressors, including temperature, precipitation, and flooding. Increased
temperatures can cause accelerated aging of binder material and rutting of asphalt; precipitation can
cause cracking and erosion; and flooding can lead to washouts and overtopping of roads. As these
climate change stressors continue to change, damages to roads and costs of maintenance and repair will
vary across the U.S.208 For example, roads may experience more frequent buckling due to increased
temperatures, more frequent washouts of unpaved surfaces from increased flooding, and changes in
freeze-thaw cycles that cause cracking.209
10.3 APPROACH
The analysis estimates the costs of reactive adaptation measures resulting from climate change impacts
to roads in the contiguous U.S. and evaluates the ability of proactive adaptation measures (i.e.,
modification of roads prior to the occurrence of climate change-related damages) to improve resiliency
and reduce costs. To develop these estimates, the analysis relies on the Infrastructure Planning Support
System (IPSS), a software tool that integrates stressor-response algorithms, engineering data on the U.S.
208 Schwartz, H. G., M. Meyer, C. J. Burbank, M. Kuby, C. Oster, J. Posey, E. J. Russo, and A. Rypinski, 2014: Ch. 5: Transportation. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 130-149. doi:10.7930/J06QlV53.
209 Transportation Research Board, 2008: Potential Impacts of Climate Change on U.S. Transportation. Special Report 290, Committee on
Climate Change and U.S. Transportation, National Research Council of the National Academies.
79
-------
INFRASTRUCTURE
Roads
road network, and the climate projections described in the Modeling Framework section of this
Technical Report.210,211 The IPSS tool estimates the potential impacts related to three climate stressors
(temperature, precipitation,212 and timing of freeze-thaw cycles213) for three road types (paved,
unpaved, and gravel), as summarized in Table 10.1, and quantifies the costs of reactive adaptation in the
form of maintenance activities required to ensure current levels of service.214 These costs represent the
incremental change in expenditures associated with projected climate change relative to the reference
period (1950-2013) as modeled by the five GCMs under RCP8.5 and RCP4.5. In addition, many parts of
the U.S. road network are under-maintained today, which can increase their vulnerability to climate
change. This analysis focuses on the additional impacts due to climate change independent of this
underlying vulnerability.
The IPSS tool also quantifies the costs of proactive adaptation measures to protect and rehabilitate
roads against impacts caused by climate stressors, where applicable. The differences between the costs
of proactive adaptation measures and the costs of reactive adaptation measures to address climate
change-related impacts represent the effects of proactive adaptation for the roads sector.215 For more
information on the approach, please refer to Chinowsky and Arndt (2012), Espinet et al. (2016), and
Neumann et al. (2014).216-217'218
210 Schweikert, A., P. Chinowsky, X. Espinet, and M. Tarbert, 2014: Climate change and infrastructure impacts: comparing the impact on roads in
ten countries through 2100. Procedia Engineering, 78, 306-316.
211 Chinowsky, P., A. Schweikert, G. Hughes, C.S. Hayles, N. Strzepek, K. Strzepek, and M. Westphal, 2015: The impact of climate change on road
and building infrastructure: a four-country study. International Journal of Disaster Resilience in the Built Environment, 6, 382-396.
212 The hydrologic movement of water across a road surface, also known as overtopping due to a flooded waterway, is not directly modeled in
this analysis.
213 Freeze-thaw related impacts affect the sub-surface components of roads while temperature-related damage is limited to the surface.
214 To maintain service, the level of maintenance applied can vary overtime, and can therefore be larger or smaller than the historic level from
the reference period.
215 The analysis assumes that for a given climate stressor, proactive adaptation prevents the need for future climate-induced maintenance.
216 Chinowsky, P. and C. Arndt, 2012: Climate Change and Roads: A Dynamic Stressor-Response Model. Review of Development Economics, 16,
448-462.
217 Espinet, X., A. Schweikert, N. van den Heever, and P. Chinowsky, 2016: Planning resilient roads for the future environment and climate
change: quantifying the vulnerability of the primary transport infrastructure system in Mexico. Transport Policy, 50, 78-86.
218 Neumann, J.E., J. Price, P. Chinowsky, L. Wright, L. Ludwig, R. Streeter, R. Jones, J.B. Smith, W. Perkins, L. Jantarasami, and J. Martinich, 2014:
Climate change risks to US infrastructure: impacts on roads, bridges, coastal development, and urban drainage. Climatic Change, 131, 97-109.
80
-------
INFRASTRUCTURE
Roads
Table 10.1. Summary of Modeled Damages and Proactive Adaptation Measures for U.S. Roads
Climate
Stressor
Road Type
Impacts
Response Measures
Temperature
Paved
Surface degradation and increased
roughness due to thermal cracking
and rutting.
Change asphalt mix to include binder
with appropriate temperature
performance.
Unpaved
Not Modeled*
N/A
Gravel
Not Modeled*
N/A
Precipitation
Paved
Erosion of base and sub-base due to
infiltration as well as increased
cracking.
Modify binder/sealant application
and increase depth of base layer.
Unpaved
Erosion of surface and development
of rutting.
Upgrade to gravel or paved road.A
Gravel
Erosion of base due to subsidence
resulting in uneven surface.
Increase thickness of gravel and sub-
base to improve strength and allow
for better drainage.
Freeze-Thaw
Paved
Degradation of base layer due to soil
heaving, and increased surface
damage due to settling and
movement.
Modify design to increase surface
density and reduce infiltration.
Unpaved
Not Modeled*
N/A
Gravel
Not Modeled*
N/A
*The effects of the temperature and freeze-thaw climate stressors on gravel and unpaved roads are likely
inconsequential and are therefore not modeled.
AWhile the accepted method for adapting unpaved roads is to upgrade to a paved surface, newer and potentially
less-costly approaches exists that are not widely established, and therefore not included in the modeling.
10.4 RESULTS
Through the end of the century, climate change is projected to result in $230 billion and $150 billion in
reactive adaptation costs to U.S. roads under RCP8.5 and RCP4.5, respectively (2015-2099, $2015,
discounted at 3%, five-GCM average). Across the five climate models, cumulative costs range from $59
to $530 billion under RCP8.5 and from $75 to $350 billion under RCP4.5. The largest impacts are
estimated under the HadGEM2-ES model, which are the hottest climate projections analyzed, while the
smallest impacts are seen under the coolest model, GISS-E2-R. As shown in Table 10.2, reactive
adaptation costs are dominated by paved roads and are higher under RCP8.5 than under RCP4.5 in all
but one of the five models (GISS-E2-R). On a per-lane-mile basis, projected costs are highest for paved
roads ($37,000 under RCP8.5 and $24,000 under RCP4.5), followed by gravel roads ($4,500 under
RCP8.5 and $3,800 under RCP4.5) and unpaved roads ($2,200 under RCP8.5 and $1,800 under RCP4.5).
81
-------
INFRASTRUCTURE
Roads
Table 10.2. Cumulative Change in Reactive Adaptation Costs
The table presents the estimated change in reactive adaptation costs for the period 2015-2099 relative
to the reference period (1950-2013) (billions $2015, discounted at 3%, five-GCM average).
GCM
Road Type
RCP8.5
RCP4.5
Paved
$160
$67
CanESM2
Gravel
$7.7
$5.0
Unpaved
$3.7
$2.4
TOTAL
$170
$75
Paved
$240
$150
CCSM4
Gravel
$2.9
$1.7
Unpaved
$1.4
$0.9
TOTAL
$250
$150
Paved
$50
$74
GISS-E2-R
Gravel
$6.4
$8.0
Unpaved
$3.1
$3.9
TOTAL
$59
$86
Paved
$510
$340
HadGEM2-ES
Gravel
$9.4
$9.1
Unpaved
$4.5
$4.4
TOTAL
$530
$350
Paved
$120
$74
MIROC5
Gravel
$3.5
$1.1
Unpaved
$1.7
$0.6
TOTAL
$130
$75
Paved
$220
$140
5-GCM Average
Gravel
$6.0
$5.0
Unpaved
$2.9
$2.4
TOTAL
$230
$150
Figure 10.1 presents the estimated annual per-lane-mile reactive adaptation costs in 2050 and 2090 at
the regional level, broken down by climate stressor and RCP. Temperature-related impacts dominate in
all regions, particularly in the Northeast, Southeast, and Midwest, and are consistently higher under
RCP8.5 compared to RCP4.5. Impacts related to precipitation are smaller, but generally increase from
2050 to 2090. Partially offsetting these impacts, the freeze-thaw stressor is projected to result in
negative costs (savings) compared to the reference period in all regions and under all scenarios. This is
82
-------
INFRASTRUCTURE
Roads
due to the projected shift in freeze zone status for a large portion of the country, from moderate-freeze
to no-freeze zones. The shift in these areas significantly reduces the maintenance costs for freeze-thaw
costs relative to the reference period. As shown in Figure 10.1, these savings are projected to be
particularly high in the Northeast. Although not shown in the figure, the largest total reactive adaptation
costs are projected to occur in the Southeast and Midwest, which is partially due to the comparatively
higher number of lane miles in these regions and also to greater climate stress.
Figure 10.1. Change in Annual Per-Lane-Mile Reactive Adaptation Costs
The graphs show changes in reactive adaptation costs in 2050 (2040-2059) and 2090 (2080-2099)
relative to the reference period (1950-2013). Results represent the five-GCM average and are presented
in thousands of $2015, undiscounted.
NORTHERN PLAINS
MIDWEST
NORTHWEST
NORTHEAST
SOUTHEAST
SOUTHWEST
SOUTHERN PLAINS
Temperature
Precipitation
Freeze-Thaw
83
-------
INFRASTRUCTURE
Roads
Potential for Adaptation to Reduce Impacts
Table 10.3 presents the cumulative change in costs for 2015-2099 relative to reference period (1950-
2013) with reactive and proactive adaptation. Across all stressors and road types, proactive adaptation
is projected to decrease costs by 98% under RCP8.5 and 83% under RCP4.5 relative to the scenario with
reactive adaptation. For paved roads, proactive adaptation reduces temperature-related costs by 68%
and 59% under RCP8.5 and RCP4.5, respectively, and reduces precipitation-related costs by 58% and
47%, respectively. For gravel and unpaved roads, precipitation-related costs are higher with proactive
adaptation than with reactive adaptation. This is because the options for proactively adapting unpaved
roads to increased precipitation risks are limited to upgrading the roads to paved or gravel, which are
both very expensive. Proactive adaptation for gravel roads is also very expensive, as it essentially
involves reconstructing the road with enhanced structural capacity. Costs associated with the freeze-
thaw stressor do not change significantly between the reactive and proactive adaptation scenarios. In
the proactive adaptation scenario, total, cumulative, discounted costs are higher under RCP4.5 than
under RCP8.5 because the freeze-thaw related savings are greater under RCP8.5.
84
-------
INFRASTRUCTURE
Roads
Table 10.3. Cumulative Change in Costs with Reactive and Proactive Adaptation
The table presents cumulative change in costs with reactive and proactive adaptation for the 2015-2099
period relative to the reference period (1950-2013) in billions $2015, discounted at 3%, for the five-GCM
average.
RCP8.5
RCP4.5
Reactive
Adaptation
Proactive
Adaptation
Reactive
Adaptation
Proactive
Adaptation
Temperature
Paved
$300
$95
$190
$78
Gravel
N/A*
N/A*
N/A*
N/A*
Unpaved
N/A*
N/A*
N/A*
N/A*
Subtotal
$300
$95
$190
$78
Freeze-Thaw
Paved
-$120
-$120
-$77
-$80
Gravel
N/A*
N/A*
N/A*
N/A*
Unpaved
N/A*
N/A*
N/A*
N/A*
Subtotal
-$120
-$120
-$77
-$80
Precipitation
Paved
$37
$15
$30
$16
Gravel
$6
$7
$5
$6
Unpaved
$3
$6
$2
$6
Subtotal
$46
$28
$37
$28
Total
Paved
$220
-$8
$140
$14
Gravel
$6
$7
$5
$6
Unpaved
$3
$6
$2
$6
TOTAL
$230
$5
$150
$26
*The effects of the temperature and freeze-thaw climate stressors on gravel
and unpaved roads are likely inconsequential and are therefore not modeled.
Figure 10.2 shows the change in total projected costs in 2050 and 2090 relative to the reference period
with reactive and proactive adaptation, distributed across climate stressors. Temperature- and
precipitation-related costs are significantly reduced in the proactive adaptation scenario relative to the
reactive adaptation scenario, while freeze-thaw related savings do not change significantly between the
two scenarios.
85
-------
INFRASTRUCTURE
Roads
Figure 10.2. Change in Annual Costs for U.S. Roads with Reactive and Proactive Adaptation
The graphs present the change in annual costs for reactive and proactive adaptation in 2050 (2040-2059)
and 2090 (2080-2099) relative to the historic reference period (1950-2013) in billions $2015,
undiscounted, for the five-GCM averages.
$30
$25
$20
$15
$10
$5
Reactive Adaptation
¦
1
¦
$30
$25
$20
$15
$10
$5
$0
$5
-$10
Proactive Adaptation
RCP8.5
RCP4.5
RCP8.5
RCP4.5
2050
2090
I Precipitation 1 Temperature ¦ Freeze-Thaw
10.5 DISCUSSION
The analysis estimates that climate change will result in increased costs of maintaining, repairing, and
replacing roads, which is consistent with the findings of the assessment literature.219 In particular, the
analysis projects high costs associated with temperature- and precipitation-related impacts to paved
roads. Total annual costs in 2090 are estimated at $20 billion under RCP8.5 and $8.1 billion under
RCP4.5 ($2015, undiscounted, five-GCM average). With well-timed proactive adaptation, the analysis
projects savings of $7.3 billion under RCP8.5 and $3 billion under RCP4.5 compared to the reference
period. A previous study using a similar approach and different climate scenarios found that the
estimated costs through 2100 were $10 billion under a high emissions scenario and $2.6 billion under a
219 Schwartz, H. G., M. Meyer, C. J. Burbank, M. Kuby, C. Oster, J. Posey, E. J. Russo, and A. Rypinski, 2014: Ch. 5: Transportation. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 130-149. doi:10.7930/J06QlV53.
86
-------
INFRASTRUCTURE
Roads
global GHG mitigation scenario (discounted at 3%).220 The difference between the current results and
these previous estimates reflect two key differences between the two analyses. First, the savings
reflected in the current results are due to the changes in the freeze-thaw stressor, as described above,
which were modeled differently in the previous analysis. Second, the climate models used in the
previous analysis project significantly wetter conditions across the U.S. compared to the models used in
the current analysis, resulting in larger precipitation-related costs for unpaved roads.
The large reductions in costs due to proactive adaptation in this study are estimated under a scenario
assuming well-timed and effective adaptation. As described in the Approach section, examples of
proactive adaptation strategies include changing asphalt mixes to use binders with better temperature
performance, or using gravel on unpaved roads that are subject to increasingly heavy precipitation. This
proactive scenario is useful for evaluating how costs related to climate change impacts could be
reduced. It is worthwhile to note, however, that the timing of road maintenance is important, and
delays or deferred maintenance can decrease the potential effectiveness of adaptation, yielding smaller
reductions in total costs than those reported under the proactive adaptation scenario which assumes
well-timed investments to maintain levels of service.
Implementation of well-timed adaptation measures to maintain service levels is a potentially overly
optimistic assumption given that infrastructure investments are oftentimes delayed and underfunded.
Significant cases of delayed maintenance can result in road closure, which would lead to large public
costs (e.g., increased travel time) not reported here. In addition, for unpaved roads, the effects of
changes in precipitation are likely dependent on the amount of traffic on the road, which is not explicitly
captured in the analysis. However, advancements in technology and changes in driving behavior are not
directly modeled in the analysis, and could have long-term implications on the vulnerability of the road
network to climate change. Lastly, among the three climate stressors examined in the analysis, freeze-
thaw is the most complex and the most uncertain. The analysis assumes that areas fall neatly into
climate zones with specific freeze-thaw risks (i.e., no-freeze or moderate-freeze) and that road
maintenance decisions are made accordingly. In reality, areas that are close to the border between no-
freeze and moderate-freeze zones will need to manage for some freeze events, which would lead to
larger costs than those reported here. Specifically, there is a 61-70% increase in maintenance costs from
no-freeze zones to moderate-freeze zones, so the cost for no-freeze zones can increase quickly if freeze
events do in fact occur.
220 EPA, 2015: Climate Change in the United States: Benefits of Global Action. U.S. Environmental Protection Agency, Office of Atmospheric
Programs, EPA 430-R-15-001.
87
-------
INFRASTRUCTURE
Bridges
11. BRIDGES
11.1 KEY FINDINGS
• By 2050, an estimated 4,600 inland bridges across the contiguous U.S. are projected to become
vulnerable each year under RCP8.5. Under RCP4.5, this estimate is reduced to 2,500. By 2090, 6,000
bridges are projected to become vulnerable each year under RCP8.5, while 5,000 would be
vulnerable under RCP4.5.
• National average annual proactive maintenance or rehabilitation costs under RCP8.5 are estimated
at $1.7 billion by 2050 and $1.0 billion by 2090. Costs are reduced under RCP4.5 to $1.5 billion each
year in 2050 and $510 million each year in 2090.
11.2 INTRODUCTION
Road bridges are a central component of the U.S. transportation system. With the average U.S. bridge
now over 40 years old, however, vehicles cross structurally deficient bridges over 2 million times a
day.221 Similar to other transportation infrastructure, bridges are vulnerable to a range of threats from
climate change.222 Currently, most bridge failures caused by extreme events are due to scour, where
swiftly moving water removes sediment from around bridge structural supports, weakening or
destroying their foundations.223 Increased flooding and long-term river flow changes caused by climate
change are expected to increase the frequency of bridge scour, further stressing the aging U.S.
transportation system.
11.3 APPROACH
The analysis estimates impacts on inland bridges that span bodies of water in the contiguous U.S.
resulting from projected changes in peak flows from 100-year, 24-hour precipitation events in two
future eras: 2050 (2035-2064) and 2090 (2070-2099),224 as modeled by five GCMs under RCP8.5 and
RCP4.5. Using data from the National Bridge Inventory, this method quantifies the costs associated with
two levels of perfect-foresight responses for bridges determined to be vulnerable as a result of climate
change: (1) the application of riprap to stabilize bridges, and (2) the strengthening of bridge piers and
abutments with additional concrete. The analysis assumes that riprap is required when projected peak
flows from a 100-year, 24-hour storm increase by 20%. Concrete strengthening is required when peak
flows increase by 60% for bridges on non-sandy soils and by 100% for bridges on sandy soils. This study
requires an estimate of peak flows from rainfall events and simulation of nonlinear watershed
processes, accounting for watershed land use, soil type, and topography. The method adopted is the
221 DOT, cited 2017: National Bridge Inventory. United States Department of Transportation, Federal Highway Administration. Available online
at https://www.fhwa.dot.gov/bridge/nbi.cfm
222 Schwartz, H. G., M. Meyer, C. J. Burbank, M. Kuby, C. Oster, J. Posey, E. J. Russo, and A. Rypinski, 2014: Ch. 5: Transportation. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 130-149. doi:10.7930/J06QlV53.
223 Briaud J.L., Hunt B.E. (2006) Bridge scour and the structural engineer. Structure December:58-61.
224 The era referred to as 2090 is not centered on 2090 because the climate data was only available through 2099 and therefore the 30-year
period required forthe analysis had to begin in 2070.
88
-------
INFRASTRUCTURE
Bridges
U.S. Department of Agriculture's Natural Resources Conservation Service TR-20 model, used to convert
24-hour rainfall "design-storm" depths to peak flows, consistent with Wright et al. (20 1 2).225,226
Based on the projections of bridge vulnerability, the analysis evaluates a response scenario in which
bridges are proactively rehabilitated to avoid service disruption caused by climate-induced changes in
extreme river flow.227 Projected costs in this scenario include the costs of riprap installation and
concrete strengthening based on engineering data from the reference period. Importantly, this analysis
assumes perfect foresight, in that bridges are only rehabilitated if they are known to be threatened by a
near-term river flow level that crosses one of the thresholds described above. This scenario may
underestimate potential bridge damages, as the costs of proactive, well-timed rehabilitation are likely
far lower than the costs associated with repairing or reconstructing bridge failures, and because this
analysis does not estimate the damages associated with delays or disruption from loss of use. Also, this
analysis focuses on the incremental effects due to climate change, and does not estimate the additional
costs associated with retrofitting bridges that were structurally vulnerable in the reference period (i.e.,
there may be deficient bridges that are not projected to be rehabilitated because the climate
projections do not suggest that they will be subjected to damaging high river flows).
For more information on the CIRA approach and results for the bridges sector, please refer to Neumann
et al. (2014)228 and Wright et al. (2012).229
11.4 RESULTS
Figure 11.1 shows the estimated percentage of bridges identified as vulnerable to climate change in
each four-digit HUC of the contiguous U.S. In 2050 (2035-2064), the majority of HUCs across the U.S. are
projected to contain 20% or fewer vulnerable bridges under both RCPs. By 2090 (2070-2099), there are a
greater number of HUCs with 40% or more vulnerable bridges, particularly under RCP8.5. Table 11.2
summarizes the annual numbers of vulnerable bridges by region. By 2050, approximately 4,600 bridges
are projected to be vulnerable each year under RCP8.5.230 Under RCP4.5, this number is reduced by 46%
to 2,500. Under both RCPs, the Southeast is projected to experience the highest number of vulnerable
bridges in 2050, followed by the Midwest. By 2090, 6,000 bridges are projected to be vulnerable each
year under RCP8.5, and this number is reduced to 5,000 per year under RCP4.5.231 The Midwest is
projected to experience the highest number of vulnerable bridges in 2090 under both RCPs.
225 Wright, L., P. Chinowsky, K. Strzepek, R. Jones, R. Streeter, J. Smith, J. Mayotte, A. Powell, L. Jantarasami, and W. Perkins, 2012: Estimated
effects of climate change on flood vulnerability of U.S. bridges. Mitigation and Adaptation Strategies for Global Change, 17, 939-955, doi:
10.1007/sl 1027-011-9354-2.
226 For this analysis, the ratio of peak precipitation that is used as an input to TR-20 is slightly different than past applications; it reflects
identification of a 100-yr, 24-hour storm over a longer period (1980-2009; 30 years rather than 20 years) and also by fitting an extreme value
Type 1 (Gumbel) distribution to the 30 year set of annual maximum precipitation values. The use of an extreme value Type 1 distribution differs
from past applications, such as Wright et al. (2012), which have used the Log Pearson Type III distribution. The update in method for identifying
the 100-year 24-hr precipitation event in each HUC reflects a desire to better match, and to not statistically overfit, the statistical characteristics
of the precipitation distributions.
227 Bridge overtopping, whereby extreme riverflows rise higherthan bridge decks, are an important effect not directly modeled in this analysis.
228 Neumann, J., J. Price, P. Chinowsky, L. Wright, L. Ludwig, R. Streeter, R. Jones, J.B. Smith, W. Perkins, L. Jantarasami, and J. Martinich, 2014:
Climate change risks to U.S. infrastructure: impacts on roads, bridges, coastal development, and urban drainage. Climatic Change, 131, 97-109,
doi: 10.1007/S10584-013-1037-4.
229 Wright, L., P. Chinowsky, K. Strzepek, R. Jones, R. Streeter, J. Smith, J. Mayotte, A. Powell, L. Jantarasami, and W. Perkins, 2012: Estimated
effects of climate change on flood vulnerability of U.S. bridges. Mitigation and Adaptation Strategies for Global Change, 17, 939-955, doi:
10.1007/sl 1027-011-9354-2.
230 Across the contiguous U.S., the analysis models impacts on a total of 440,000 bridges.
231 The same bridge may be considered vulnerable in both 2050 and 2090; for example, a bridge may be subject to peak flows that surpass the
threshold for riprap strengthening in 2050, and then in 2090 it may become subject to peak flows surpassing the threshold for concrete
strengthening.
89
-------
INFRASTRUCTURE
Bridges
Figure 11.1. Percentage of Bridges identified as Vulnerable to Climate Change
Estimated percentage of bridges in each four-digit HUC of the contiguous U.S. identified as vulnerable
under each RCP in 2050 (2035-2064) and 2090 (2070-2099).
2050
2090
Percent of Bridges Identified as Vulnerable
0% - 10%
\ 1
11%-20%
~
21% - 40%
~
41%-60%
61%-80%
¦¦
81% - 100%
RCP8.5
RCP4.5
90
-------
INFRASTRUCTURE
Bridges
Table 11.1. Projected Number of Vulnerable Bridges per Year
Estimated number of bridges in each region identified as vulnerable each year by 2050 (2035-2064) and
2090 (2070-2099) under each RCP. Values represent averages of the five GCMs. Totals may not sum due
to rounding.
2050 2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
510
350
570
390
Southeast
1,400
750
1,600
1,200
Midwest
1,300
600
1,700
1,500
Northern Plains
260
160
410
430
Southern Plains
810
420
1,100
1,000
Southwest
160
120
360
260
Northwest
120
83
200
160
National Total
4,600
2,500
6,000
5,000
Table 11.2 presents the average proactive maintenance costs in 2050 and 2090. For the five-GCM
average, annual costs under RCP8.5 are estimated at $1.7 billion by 2050 and $1.0 billion by 2090.
Projected annual costs are reduced under RCP4.5 to $1.5 billion in 2050 and $510 million in 2090. Costs
are smaller in 2090 than in 2050 under both RCPs because many bridges require repairs due to climate
changes by 2050, and once repaired, are less susceptible to extreme river flow impacts in 2090. Of the
five GCMs, GISS-E2-R, HadGEM2-ES, and MIROC5 project the highest impacts and CCSM4 projects the
lowest impacts.
Table 11.2. Projected Proactive Maintenance Costs to U.S. Bridges Across Climate Models
Average annual costs (millions) in the contiguous U.S. in 2050 (2035-2064) and 2090 (2070-2099)
(undiscounted, $2015).
2050
2090
GCM
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
$1,700
$1,500
$1,100
$560
CCSM4
$950
$1,100
$670
$310
GISS-E2-R
$1,500
$1,500
$1,300
$390
HadGEM2-ES
$2,000
$1,700
$1,100
$740
MIROC5
$2,200
$1,600
$800
$530
5-GCM Average
$1,700
$1,500
$1,000
$510
91
-------
INFRASTRUCTURE
Bridges
Table 11.3 presents the estimated proactive maintenance costs at national and regional levels. At a
national scale, projected proactive maintenance costs under RCP8.5 are estimated at $1.4 billion per
year by 2050 and $1.1 billion by 2090, while under RCP4.5 costs are reduced to $1.2 billion per year by
2050 and $590 million by 2090. The Midwest and the Southeast incur the highest adaptation costs to
maintain bridge service in both eras under both RCPs. Proactive maintenance costs are projected to be
the smallest in the Northern Plains and Northwest, mostly due to the smaller number of bridges in those
regions. Across the majority of regions, impacts are reduced under RCP4.5 relative to RCP8.5 (Table
11.3).
Table 11.3. Regional Proactive Maintenance Costs for Vulnerable Bridges
Average annual costs (millions) by region in 2050 (2035-2064) and 2090 (2070-2099) for the five-GCM
average (undiscounted, $2015).
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
$220
$180
$120
$77
Southeast
$430
$340
$300
$150
Midwest
$430
$390
$270
$110
Northern Plains
$89
$91
$42
$25
Southern Plains
$300
$300
$180
m
00
-oo-
Southwest
$120
$95
$54
$37
Northwest
m
00
-oo-
$71
1
m
-oo-
$22
National Total
$1,700
$1,500
$1,000
$510
11.5 DISCUSSION
The findings regarding near-term bridge vulnerability and proactive maintenance costs due to
unmitigated climate change are consistent with the findings of the assessment literature,232 but this
work provides quantification of those risks in a consistent manner for a full lower 48 state domain. It is
important to consider several limitations of the analysis. The analysis considers the effects of climate
change on inland bridges, not coastal bridges, and also focuses only on high streamflow risks, and not
other climatic stresses (e.g., extreme temperature) or synergistic effects of climate with other stresses,
and therefore is likely an underestimate of future impacts of climate change on the nation's total bridge
inventory. In addition, although there will likely be significant changes to the nation's bridges over the
course of the century—some bridges will be strengthened for reasons separate from climate change
risks, some will deteriorate, some will be removed, and new bridges will be added—this analysis
estimates costs based on the existing bridge inventory in its current state. Further, this analysis assumes
that proactive, well-timed adaptation will be taken to maintain the current level of bridge service. In
reality, some bridges will likely fail in the future due to a combination of delayed maintenance and
inadequate design to address future climate risks, resulting in loss of use and the associated public costs,
232 Schwartz, H. G., M. Meyer, C. J. Burbank, M. Kuby, C. Oster, J. Posey, E. J. Russo, and A. Rypinski, 2014: Ch. 5: Transportation. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 130-149. doi:10.7930/J06QlV53.
92
-------
INFRASTRUCTURE
Bridges
such as increased traffic and delays. Finally, the adaptation option evaluated here only consider a class
of actions that could reduce physical impacts at the bridge facility. Other adaptation options to reduce
the consequences of those physical impacts - such as re-routing of road traffic or building in other forms
of network flexibility - could also be considered, and might in some cases be more cost-effective than
bridge strengthening.
93
-------
INFRASTRUCTURE
Rail
12. RAIL
12.1 KEY FINDINGS
• Increasing temperatures are projected to result in significant damages to the U.S. rail system. In
response to increased risks of rail cracking, rail operators will be forced to reduce speeds, causing
economic damages associated with delays to freight and passenger rail. Average cumulative
discounted damages through 2100 are estimated at $50 billion under RCP8.5 and $40 billion under
RCP4.5.
• Well-timed proactive adaptation is projected to reduce average cumulative discounted costs
through 2100 to $12 billion under RCP8.5 and $4.5 billion under RCP4.5.
12.2 BACKGROUND
The U.S. rail network is a critical component of the nation's infrastructure system, connecting U.S.
consumers with agricultural, economic, logistics, and manufacturing centers across the nation and the
world.233 Climate change affects the rail network principally through projected temperature increases
across the U.S. Passenger and freight tracks are susceptible to damage during periods of extreme heat,
which are expected to increase in frequency as a result of climate change. Specifically, when exposed to
temperatures outside of the range of normal operating conditions, steel rail expands and can undergo a
displacement or buckling called a "sun kink," increasing the risk of derailments and leading to costly
maintenance expenditures and train delays.
12.3 APPROACH
The purpose of the analysis it to determine the potential risk of climate change to the Class I rail
network in the U.S., which comprises 140,000 rail miles operated by seven railroad companies and
carrying both freight and passenger trains.234 To model the existing rail network, the analysis relies on
geospatial data from the National Transportation Atlas Database (NTAD) for active main line and sub
main line track.235 Average daily train traffic volume is estimated based on highway-rail crossing data
from the Federal Railroad Administration's (FRA's) Office of Safety Analysis.236,237
The analysis uses the Infrastructure Planning Support System (IPSS) tool, which incorporates engineering
knowledge, stressor-response algorithms, and climate projections, to quantify potential vulnerabilities
to the rail system resulting from climate change.238 The tool quantifies the costs of reactive adaptation
and proactive adaptation under RCP8.5 and RCP4.5 and for each of the five GCMs, and represent
impacts above and beyond what is spent on periodic maintenance. The reactive adaptation costs are
233 DOT, cited 2016: Freight Rail Overview. United States Department of Transportation, Federal Railroad Administration. Available online at
https://www.fra.dot.gov/Page/P0528
234 DOT, cited 2016: Freight Rail Today. United States Department of Transportation, Federal Railroad Administration. Available online at
https://www.fra.dot.gov/Page/P0362
235 DOT, cited 2016: Bureau of Transportation Statistics: National Transportation Atlas Databases 2015. [Available online at:
http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national transportation atlas database/2015/index.htmll
236 FRA's Office of Safety Analysis provides data on daily highway-rail crossings for over 150,000 unique highway-rail crossings. Based on these
data, the study estimated the average daily volume of train traffic per grid cell.
237 DOT, cited 2016: Highway-Rail Crossings. United States Department of Transportation, Federal Railroad Administration, Office of Safety
Analysis. Available online at http://safetvdata.fra.dot.gov/OfficeofSafetv/publicsite/Querv/gxrtallvl.aspx
238 Chinowsky, P., and C. Arndt, 2012: Climate change and roads: a dynamic stressor-response model. Review of Development Economics, 16,
448-462, doi: 10.1111/j.l467-9361.2012.00673.x
94
-------
INFRASTRUCTURE
Rail
associated with delays resulting from increased temperatures under climate change, as current rail
safety guidelines require reduced speed and traffic in areas where extreme temperatures are occurring
or predicted. Delays are first quantified in minutes and then converted to dollars using a methodology
that estimates the cost of delays for freight trains to the railroad company and the public.239 The costs of
delays include costs to the railroad companies (including the costs of crew, cars, locomotives, lading,
and fuel), and costs to the public include costs of locomotive emissions attributed to additional
operational time and car traffic delay at railroad crossings.240
The study also quantifies the costs of proactive adaptation measures that reduce the risk of rail line
damage and the associated temperature-based delays.241 The proactive adaptation measure modeled is
the FRA-proposed installation and use of temperature sensors to identify the times and locations when
speed and traffic reductions are required due to local conditions.242 This is in contrast to the current
practice of widespread restrictions over a predetermined number of hours, which corresponds to a
broader set of delays. For more information on the approach to estimating impacts on rail
infrastructure, please see Chinowsky et al. (2017).243
12.4 RESULTS
The projected cumulative reactive adaptation costs to the U.S. rail network are substantial, estimated at
$50 billion under RCP8.5 and $40 billion under RCP4.5 for the five-GCM average (2016-2099, $2015,
discounted at 3%). Table 12.1 shows the projected annual reactive adaptation costs for 2050 and 2090
for the five GCMs and the five-GCM average. As shown, costs are consistently higher in 2090 than in
2050 under both RCPs and across all five models. For the five-GCM average, annual costs in 2090 are
$5.5 billion and $3.5 billion (undiscounted $2015) under RCP8.5 and RCP4.5, respectively. Projected
costs are largest under the HadGEM2-ES model and smallest under the GISS-E2-R model, which,
respectively, represent the hottest and coolest GCMs of the five analyzed.
239 Lovett, A.H., C.T. Dick, and C.P. Barkan, 2015: Determining freight train delay costs on railroad lines in North America. In: Proceedings of the
International Association of Railway Operations Research (IAROR) 6th International Conference on Railway Operations Modelling and Analysis,
Tokyo, Japan. Available online at http://railtec.illinois.edu/articles/Files/Conference%20Proceedings/2015/Lovett-et-al-2015-IARQR.pdf
240 The analysis quantifies the costs of conventional pollutants excluding CO2.
241 In this scenario with proactive adaptation, impacts include both the costs of the adaptation measure as well as any damages resulting from
climate change that are not prevented by proactive adaptation.
242 Kish, A. and G. Samavedam, 2013: Track Buckling Prevention: Theory, Safety Concepts, and Applications. United States Department of
Transportation, Federal Railroad Administration. Technical Report No. D0T/FRA/0RD-13/16. Available online at
https://www.fra.dot.gov/eLib/details/L04421
243 Chinowsky, P., J. Helman, S. Gulati, J. Neumann, and J. Martinich, 2017: Impacts of Climate Change on Operation of the US Rail Network.
Transport Policy, doi: 10.1016/j.tranpol.2017.05.007.
95
-------
INFRASTRUCTURE
Rail
Table 12.1. Projected Annual Reactive Adaptation Costs to the U.S. Rail System
The table presents the change in reactive adaptation costs in 2050 (2040-2059) and 2090 (2080-2099)
relative to the reference period (1950-2013) (billions $2015, undiscounted).
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
$1.9
$1.6
$6.1
$3.8
CCSM4
$1.7
$1.3
$5.1
$3.2
GISS-E2-R
$1.3
$1.1
$4.0
$2.4
HadGEM2-ES
$2.2
$1.8
$6.6
$4.4
MIROC5
$1.6
$1.6
$5.8
$3.7
5-GCM Average
$1.8
$1.5
$5.5
$3.5
Figure 12.1 displays the average annual reactive adaptation costs in 2050 and 2090 under both RCPs at
the half-degree grid cell level (approximately 34 square miles). The white areas in the maps represent
areas where no Class I rail is present in addition to where the costs of climate change are estimated to
be near zero. The highest projected costs are mainly concentrated in the Northeast, Midwest, and
Southwest, particularly under RCP8.5. These impacts are due to the relatively higher rail network density
and/or the projected increases in temperature relative to the temperature at which the rails were
designed to operate.
96
-------
INFRASTRUCTURE
Rail
Figure 12.1. Average Annual Reactive Adaptation Costs to the U.S. Rail Network
The maps display the change in reactive adaptation costs relative to the reference period (1950-2013) for
the five-GCM average ($2015, undiscounted) in 2050 (2040-2059) and 2090 (2080-2099).
2050
2090
$25,001-$50,000
$50,001 - $75,000
$75,001-$100,000
$100,001-$250,000
$250,001 - $500,000
$500,001 - $1,000,000
$1,000,001 - $5,000,000
$5,000,001 - $10,000,000
$10,000,001 - $52,000,000
Potential for Proactive Adaptation to Reduce Impacts
As described in the Approach section, the analysis also quantifies the impacts of climate change on the
rail system in a scenario where proactive adaptation measures are implemented to reduce the
temperature-delay effect on the rail system. Table 12.2 shows the estimated cumulative costs of climate
change by region with reactive and proactive adaptation.244 As shown, impacts are reduced significantly
at the national level when proactive adaptation measures are taken. For the five-GCM average,
estimated cumulative costs are reduced from $50 billion to $12 billion (77%) under RCP8.5 and from $40
billion to $4.5 billion (89%) under RCP4.5, for savings of $39 billion and $35 billion, respectively. At the
regional level, reactive adaptation costs are highest in the Southeast and Southern Plains under both
244 As described in the Approach section, impacts in the scenario with proactive adaptation include both the costs of making proactive
adaptation measures and the climate-change related damages that are not prevented bythe modeled adaptation.
97
-------
INFRASTRUCTURE
Rail
RCPs. Proactive adaptation reduces these costs by 73% and 79%, respectively, under RCP8.5 and by 84%
and 91%, respectively, under RCP4.5.
Table 12.2. Projected Cumulative Costs to U.S. Rail with Reactive and Proactive Adaptation
The table presents the cumulative reactive and proactive adaptation costs to the U.S. rail system by
region for the period 2016-2099 relative to the reference period (five-GCM average, billions $2015,
discounted at 3%).
Total Costs (Billions $2015)
Costs Per Rail Mile (Thousands $2015)
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Reactive Adaptation
Northeast
$8.7
$7.1
$410
$330
Southeast
$10
$7.7
$260
$200
Midwest
$4.6
$3.6
$100
$78
Northern Plains
$1.4
$1.0
$85
$62
Southern Plains
$14
$11
$620
$500
Southwest
$6.5
$5.2
$170
$130
Northwest
$5.2
$4.1
$600
$470
National Total
$50
$40
$290
$230
Proactive Adaptation
Northeast
$1.6
$0.55
$77
$26
Southeast
$2.8
$1.2
$72
$31
Midwest
$0.63
$0.24
$14
$5
Northern Plains
$0.53
$0.23
$33
$14
Southern Plains
$2.9
$1.0
$130
$44
Southwest
$1.4
$0.60
$72
$30
Northwest
$1.6
$0.72
$180
$82
National Total
$12
$4.5
$67
$26
98
-------
INFRASTRUCTURE
Rail
12.5 DISCUSSION
This analysis projects significant costs for the U.S. rail system associated with both reactive adaptation
to increasing temperatures under climate change, which is consistent with the assessment literature.245
Depending on the climate scenario selected and climate model used, the increase in cumulative reactive
adaptation costs relative to the reference period range from $27 to $62 billion by 2099 (discounted at
3%) (see Appendix A.9). The study suggests that the use of sensor technology combined with changes in
operating policy could reduce delays by limiting temperature-based speed restrictions for specific
locations. These proactive adaptations could reduce costs to $1.1 to $26 billion by 2099 (discounted at
3%), depending on the climate scenario and model used.
Although national-scale analysis of climate change impacts on rail has not been done in the U.S., a
recent study suggests that costs of climate-change related delays are projected to increase significantly
across Europe under RCP8.5.246 The study projects that Southern Europe will experience the highest
increased risk for rail track buckling.
The proactive adaptation evaluated in this study is not the only approach to reduce train delays caused
by climate change. Continuing innovations in track management and potential changes in track
materials may provide additional opportunities. In addition, since rail lines must be replaced every 50 to
60 years, there may be scheduled opportunities to use more resilient infrastructure. Rail lines that
anticipate implementing new rail technologies, such as high-speed rail, or that focus on specific types of
freight, may implement new technologies optimized for those options.
Although the focus of this study was on temperature effects, additional climate change considerations
can affect the vulnerability of the rail system. Precipitation changes could result in flooding that affect
bridge or railbed stability, and thus require additional investment to stabilize the infrastructure.
Similarly, increased threats from wildfires and hurricanes could exacerbate potential vulnerabilities.
245 Schwartz, H.G., M. Meyer, C.J. Burbank, M. Kuby, C. Oster, J. Posey, E. J. Russo, and A. Rypinski, 2014: Ch. 5: Transportation. Climate Change
Impacts in the United States: The Third National Climate Assessment, J.M. Melillo, Terese (T.C.) Richmond, and G.W. Yohe, Eds., U.S. Global
Change Research Program, 130-149. doi:10.7930/J06QlV53.
246 Nemry, F. and H. Demirel, 2012: Impacts of Climate Change on Transport: A focus on road and rail transport infrastructures. JRC Scientific
and Policy Reports. European Commission. Available online at http://ftp.irc.es/EURdoc/JRC72217.pdf
99
-------
INFRASTRUCTURE
Alaska Infrastructure
13. ALASKA INFRASTRUCTURE
13.1 KEY FINDINGS
• Under RCP8.5, climate-driven reactive adaptation costs (repairs to maintain service) to Alaska
infrastructure are estimated at $4.5 billion through 2100 (cumulative, discounted). Under RCP4.5,
cumulative reactive adaptation costs are reduced to $3.7 billion.
• The distribution of reactive adaptation costs varies across the state, with the largest effects
projected for the interior and south central regions of Alaska.
• Road flooding associated with increased precipitation is projected to be the largest source of
reactive adaptation costs, followed by impacts to buildings associated with permafrost thaw.
Smaller costs are estimated for airports, railroads, and pipelines.
• Well-timed, proactive adaptation is projected to dramatically reduce total economic impacts relative
to the reactive adaptation scenario.
13.2 BACKGROUND
In recent decades, the rate of temperature rise across the Arctic has been twice the global average. Sea
and land ice has diminished, while coastal erosion and permafrost thaw have increased.247 Climate
change increases the vulnerability of infrastructure by creating additional strains on structures beyond
what is expected from normal conditions and use. Permafrost thaw and subsequent ground subsidence,
particularly where permafrost is ice-rich, negatively affect buildings, roads, railroads, pipelines, and oil
and gas infrastructure. Warmer temperatures can also alter the frequency of freeze-thaw cycles,
affecting foundations and underground infrastructure stability. As climate change continues, the extent
of infrastructure damage, as well as the costs to maintain, replace, and adapt the built environment, are
expected to rise.
13.3 APPROACH
This analysis estimates two types of adaptation costs associated with climate change impacts to public
infrastructure in Alaska:
• Reactive adaptation- in the form of rehabilitation and repairs to infrastructure in response to
climate-driven damages, with the goal of maintaining current levels of service and ensuring that
infrastructure is functional through its intended lifespan. The estimated reactive adaptation costs
represent the incremental change in maintenance costs due to climate change relative to those
simulated for the baseline period (1950-1999). Many parts of the Alaskan public infrastructure
network are under-maintained today, which can increase overall vulnerability to climate change.
However, this analysis focuses on the additional impacts due to climate change independent of this
underlying vulnerability.
• Proactive adaptation- in the form of investments and modifications of infrastructure to improve its
resiliency and reduce vulnerability prior to the occurrence of climate change-related damages. This
approach assumes that well-designed, well-timed adaptation measures are taken early in the
247 Christensen, J.H., K. Krishna Kumar, E. Aldrian, S.-l. An, I.F.A. Cavalcanti, M. de Castro, W. Dong, P. Goswami, A. Hall, J.K. Kanyanga, A. Kitoh,
J. Kossin, N.-C. Lau, J. Renwick, D.B. Stephenson, S.-P. Xie and T. Zhou, 2013: Climate Phenomena and their Relevance for Future Regional
Climate Change. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and
P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
100
-------
INFRASTRUCTURE
Alaska Infrastructure
century and continue to provide economic benefits into later eras. Proactive adaptation cost
estimates represent both the costs of the adaptive actions and repair costs associated with impacts
that are not avoided by proactive adaptation.
The reactive and proactive adaptation costs are estimated using the Infrastructure Planning Support
System (IPSS) software tool, which integrates stressor-response algorithms, engineering data on the
Alaska public infrastructure network, and climate projections.248,249 For this analysis, IPSS analyzes the
following specific infrastructure assets: roads, buildings, airports, railroads, and pipelines.250 The
numbers, locations, and levels of use of these infrastructure types were compiled from numerous
sources to create the inventory of public infrastructure used for this analysis. The IPSS model accounts
for climate change impacts unique to northern latitudes, including near-surface permafrost thaw,
extreme freeze-thaw dynamics, and the effects of precipitation and precipitation-induced flooding. Due
to unique climate conditions and infrastructure present in Alaska, these infrastructure types are
separately modeled and reported from those for the contiguous U.S. (see Roads and Rail sections).
The analysis uses the SNAP Alaska climate projections251,252 described in the Modeling Framework
section of this Technical Report. Reactive and proactive adaptation costs are estimated for 2030 (2020-
2039), 2050 (2040-2059), 2070 (2060-2079), and 2090 (2080-2099), and represent the incremental
change in expenditures associated with projected climate change for each relevant environmental
stressor and infrastructure type analyzed.253 As such, the effect of climate change can be isolated from
maintenance costs in the reference period (1950-1999). IPSS simulations were aggregated to the scale of
Alaska boroughs, which are shown in Figure 13.1. For more information on the approach used in this
analysis, please refer to Melvin etal (2016).254
248 Schweikert, A., P. Chinowsky, X. Espinet, and M. Tarbert, 2014: Climate change and infrastructure impacts: comparing the impact on roads in
ten countries through 2100. Procedia Engineering, 78, 306-316.
249 Chinowsky, P., A. Schweikert, G. Hughes, C.S. Hayles, N. Strzepek, K. Strzepek, and M. Westphal, 2015: The impact of climate change on road
and building infrastructure: a four-country study. International Journal of Disaster Resilience in the Built Environment, 6, 382-396.
250 The approach described in Melvin et al. (2016) also developed a novel approach to generate first-order estimates of projected coastal
erosion rates and evaluated how alternative climate scenarios may influence erosion in twelve coastal communities where immediate actions
to manage erosion or relocate have been recommended. Because this coastal erosion component did not feed into the economic damage
calculations, these results are not included in this Technical Report.
251 Climate variables used include projected minimum and maximum annual temperature, precipitation, and change in mean annual ground
temperature, which were then used to project active layer thickness, compared to reference period permafrost and ground ice content.
2 52 As noted in the Modeling Framework section, the SNAP downscaled database contains projections for five climate models, two of which
(CCSM4 and GISS-E2-R) overlap with the five LOCA GCMs applied throughout the other sectors of this Technical Report. The results presented in
this section represent the mean results for CCSM4 and GISS-E2-R only, and therefore differ modestly from the results of the five SNAP models
reported in Melvin et al. (2016).
253 For some infrastructure-type/climate-stressor combinations, no adaptation measures are modeled.
2 54 Melvin, A.M., P. Larsen, B. Boehlert, J.E. Neumann, P. Chinowsky, X. Espinet, J. Martinich, M.S. Baumann, L. Rennels, A. Bothner, D.J.
Nicolsky, and S.S. Marchenko, 2016: Climate change damages to Alaska public infrastructure and the economics of proactive adaptation.
Proceedings of the National Academies of Sciences, doi:10.1073/pnas.1611056113.
101
-------
INFRASTRUCTURE
Alaska Infrastructure
Figure 13.1. Distribution of Current Permafrost across Alaska's Boroughs
North Slope
Northwest
Arctic
Yukon-Koyukuk
Nome®"
Fairbanks
inuska-Sugitna
Bethel
Valdez-
:ordova
Dillingham
Yakytet-Angoon
ninsula
luneau
Haine
^d^Kodiak
Island
Absent
Continuous
Discontinuous
Isolated
Sporadic
Aleutians Wes
Lake and
Aleutians East,;
Wrangell-
¦etersburg
Permafrost Distribution
Sitka
Prince of Wal
Outer Ketchikan
Ketchikan
Gateway
13.4 RESULTS
The reactive adaptation (repair) cost estimates presented in this section represent the incremental
change in expenditures due to climate change that are required to maintain current levels of service and
allow infrastructure to be functional through its intended lifespan. Total cumulative reactive adaptation
costs resulting from projected climate change this century are estimated at approximately $4.5 billion
under RCP8.5 and $3.7 billion under RCP4.5 (Table 13.1). Under RCP8.5 and RCP4.5, flooding (associated
with changes in precipitation) accounts for about 44% and 46%, respectively, of costs while near-surface
permafrost thaw is responsible for 38% and 35%, respectively. Repair costs from precipitation account
for about 17% and 19%, respectively, of cumulative costs. The largest total costs are projected for roads
($2.5 billion and $2.1 billion for RCP8.5 and RCP4.5, respectively) and buildings ($1.5 billion and $1.3
billion, respectively). However, the environmental stressors responsible for the reactive costs differ,
with approximately 75% of road damages caused by flooding and 85% of building damages by near-
surface permafrost thaw under both RCPs. Airports, railroads, and pipelines account for a smaller
fraction of the overall public infrastructure inventory, which contributed to considerably lower projected
reactive adaptation costs, collectively accounting for less than 12% of total costs under both RCPs.
Figure 13.2 shows the regional distribution of cumulative costs through 2100, while Figure 13.3 shows
projected annual costs by infrastructure type for four future time periods.
102
-------
INFRASTRUCTURE
Alaska Infrastructure
Table 13.1. Costs under Reactive and Proactive Adaptation
Cumulative costs are presented for the period 2015-2099 for each infrastructure type and environmental
stressor considered (millions $2015, discounted at 3%). Values may not sum due to rounding.
Reactive Adaptation (Repair) Costs
RCP
Flooding
Permafrost
Thaw
Precipitation
Freeze-Thaw
Total
Roads
RCP8.5
$1900
$100
$530
-$13
$2500
RCP4.5
$1600
-$1
$490
-$16
$2100
Buildings
RCP8.5
RCP4.5
Not Modeled
$1300
$1100
$120
$110
Not Modeled
$1500
$1300
Airports**
RCP8.5
$150
$140
$100
-$4
$380
RCP4.5
$120
$97
$92
-$4
$310
Railroads
RCP8.5
RCP4.5
Not Modeled
$130
$30
Not Modeled
Not Modeled
$130
$30
Pipelines
RCP8.5
RCP4.5
Not Modeled
$15
-$4
Not Modeled
Not Modeled
$15
-$4
Total
RCP8.5
$2000
$1700
$750
-$17
$4500
RCP4.5
$1700
$1300
$690
-$20
$3700
Proactive Adaptation Costs*
RCP
Flooding
Permafrost
Thaw
Precipitation
Freeze-Thaw
Total
Roads
RCP8.5
RCP4.5
$320
$300
Reactive*
$320
$310
Reactive*
$730
$590
Buildings
RCP8.5
RCP4.5
Not Modeled
Reactive*
$5.6
$5.1
Not Modeled
$1400
$1100
Airportsf
RCP8.5
RCP4.5
$44
$44
Reactive*
$71
$64
Reactive*
$250
$200
Railroads
RCP8.5
RCP4.5
Not Modeled
Reactive*
Not Modeled
Not Modeled
$130
$30
Pipelines
RCP8.5
RCP4.5
Not Modeled
Reactive*
Not Modeled
Not Modeled
$15
-$4
Total
RCP8.5
$360
$1700
$400
-$17
$2500
RCP4.5
$340
$1300
$380
-$20
$2000
'Includes the sum of the costs for proactively adapting those infrastructure units where impacts are projected to occur, plus any
impacts incurred to infrastructure units where adaptation was not applied. For infrastructure type-climate stressor
combinations where adaptation was not modeled, the reactive adaptation cost estimates were used in the calculations shown
in the total columns in the 'Proactive Adaptation Costs' section.
f Airports values include the sum of expenditures for airport buildings and runways.
~Proactive adaptation costs were quantified for these infrastructure types, however, adaptation was found to be more
expensive than reactive adaptation repairs, and therefore, reactive values are used when calculating total costs.
103
-------
INFRASTRUCTURE
Alaska Infrastructure
Figure 13.2. Cumulative Reactive Adaptation Costs to Infrastructure by Borough
Cumulative costs (2015-2099, discounted at 3%) to infrastructure and per capita costs for each borough
across Alaska under RCP8.5 and RCP4.5,
RCP8.5 RCP4.5
Total
Costs
Millions of $2015
I <0 10-50 100-300 >500
0-10 50-100 300-500
Per-Capita
Costs
Thousands of$2015
~ 50
104
-------
INFRASTRUCTURE
Alaska Infrastructure
Figure 13.3. Projected Annual Reactive Adaptation Costs by Infrastructure Type
Annual costs to each infrastructure type (roads, A; buildings, B; airports, C; railroads, D; and pipelines, E)
for the four study eras and two RCPs. Values represent the mean annual undiscounted costs for the
twenty years included in each time period. Note the difference in scale among panels.
o
CM
W
J<
C
o
o
u
~ro
z>
c
c
<
o
rN
-------
INFRASTRUCTURE
Alaska Infrastructure
Figure 13.4. Effect of Adaptation on Vulnerability
Annual reactive adaptation costs to roads and runways (A) and buildings (B) from flooding (blue) and
precipitation (purple) under the two RCPs. Percentages represent the percent savings in expenditures
from proactive adaptation compared to mean estimated reactive costs. Percentages greater than 100
indicate instances where estimated proactive adaptation costs fell below the reference period
maintenance costs. Note the difference in scale between panels.
rsi
-t/v
c
o
o
u
75
13
C
C
<
Roads and Runways
All Buildings
150
125
5-
100
4-
3-
2-
97%
97%
100%
100%
101% 102%
100% n
I"!—I |103
78%
81%
Flooding RCP8.5
Flooding RCP4.5
Precipitation RCP8.5
Precipitation RCP4.5
38%
34%
98%
97%
43%
,38%
98%
100%
43%
47%
101%
101%
43%
i in
2030 2050 2070 2090
2030 2050 2070 2090
13.5 DISCUSSION
Damages to Alaska public infrastructure from climate change are projected to be large and widespread.
This analysis did not estimate impacts to ports, electricity transmission structures, telecommunications,
and other infrastructure types, whose inclusion would provide a more comprehensive evaluation of
potential vulnerabilities and associated damages. Quantification of loss of use impacts would also
inform estimates of potential damages, and could be particularly meaningful in Alaska where there is a
lack of infrastructure redundancy across most of the state.
Many previous assessments and studies have recognized the risks of permafrost thaw to infrastructure
in Alaska;255,256 however, few studies have sought to quantify the physical and economic risks to
multiple infrastructure types in response to a broad list of climate stressors. The results presented for
this sector are similar in both direction and magnitude to other studies that have projected climate-
related increases in costs in Alaskan infrastructure. Those studies project an estimated $50 million
255 Chapin, F. S., Ill, S. F. Trainor, P. Cochran, H. Huntington, C. Markon, M. McCammon, A. D. McGuire, and M. Serreze, 2014: Ch. 22: Alaska.
Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe,
Eds., U.S. Global Change Research Program, 514-536, doi:10.7930/J00Z7150.
256 U.S. Arctic Research Commission, 2003: Climate change, permafrost, and impacts on civil infrastructure. U.S. Arctic Research Commission,
Permafrost Task force Report, Special Report 01-03. Available online at https://storage.googleapis.com/arcticgov-
static/publications/other/permafrost.pdf
106
-------
INFRASTRUCTURE
Alaska Infrastructure
annually in costs by 2080 from a subset of climate stressors for the roads and electricity sectors,257 and
approximately $7.3-15 billion through 2080 (note cumulative estimate) above 'normal' operations and
maintenance due to permafrost thaw, flooding, and coastal erosion impacts on a variety of
infrastructure types.258
257 Cole, H., V. Colonell, and D. Esch, 1999: The Economic Impact and Consequences of Global Climate Change on Alaska's Infrastructure. In
Assessing the Consequences of Climate Change for Alaska and the Bering Sea Region, summarized workshop proceedings (Center for Global
Change and Arctic System Research, University of Alaska Fairbanks), pp 43-57.
258 Larsen, P.H., S. Goldsmith, O. Smith, M. Wilson, K. Strzepek, P. Chinowsky, and B. Saylor, 2008: Estimating future costs for Alaska public
infrastructure at risk from climate change. Global Environmental Change-Human and Policy Dimensions, 18, 442-457.
107
-------
INFRASTRUCTURE
Urban Drainage
14. URBAN DRAINAGE
14.1 KEY FINDINGS
• Climate change is projected to result in costs for adapting urban drainage infrastructure to increased
runoff associated with more intense rainfall events.
• Under RCP8.5, annual average adaptation costs of 10-, 25-, and 50-year storms in 100 major U.S.
cities in 2090 are estimated at $2.5, $3.9, and $5.6 billion, respectively. Projected costs in 2090
under RCP4.5 are lower ($1.6, $2.7, and $4.1, respectively). Inclusion of all U.S. cities would likely
increase costs.
• Under both RCP8.5 and RCP4.5, projected weighted average costs for a 50-year storm event are
highest in the Southern Plains at $460,000 and $230,000 per square mile, respectively. High
adaptation costs are also projected for 50-year storm events in the Southeast under RCP8.5
($380,000 per square mile).
14.2 BACKGROUND
Urban drainage systems capture and treat stormwater runoff and prevent urban flooding. During storm
events, the volume of runoff flowing into drainage systems and the ability of these systems to manage
runoff depend on a variety of site-specific factors, such as the imperviousness of the land area in the
drainage basin. Changes in storm intensity associated with climate change have the potential to
overburden drainage systems, which may lead to flood damage, disruptions to local transportation
systems, discharges of untreated sewage to waterways, and increased human health risks from
waterborne illness and fish kills.259 In areas where precipitation intensity increases significantly,
adaptation investments may be necessary to prevent runoff volumes from exceeding system capacity.
14.3 APPROACH
The analysis estimates the costs of proactive adaptation for urban drainage systems in 100 major coastal
and non-coastal cities of the contiguous U.S. to meet future demands of increased runoff associated
with more intense rainfall under climate change. Adaptive actions focus on the use of best management
practices to limit the quantity of runoff entering stormwater systems and maintain current level of
service (i.e., proactive adaptation to avoid damages), instead of expanding formal drainage networks of
basins and conveyance systems. These best management practices generally include temporary storage
above or below ground (e.g., bioswails, retention ponds), or infiltration (e.g., permeable pavement), and
are based on EPA guidelines and construction cost estimates.260
While many site-specific factors influence the effect of climate change on a given drainage system, this
analysis uses a streamlined approach that allows for the assessment of potential impacts in multiple U.S.
cities under future climate scenarios RCP4.5 and RCP8.5. Specifically, the analysis uses a reduced-form
approach for projecting changes in flood depth and the associated costs of flood prevention, based on
259 Trtanj, J., L. Jantarasami, J. Brunkard, T. Collier, J. Jacobs, E. Lipp, S. McLellan, S. Moore, H. Paerl, J. Ravenscroft, M. Sengco, and J. Thurston,
2016: Ch. 6: Climate Impacts on Water-Related Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific
Assessment. U.S. Global Change Research Program, Washington, DC, 157-188, doi:10.7930/J03F4MH.
260 See the following for more information: Price, J., L. Wright, C. Fant, and K. Strzepek, 2014: Calibrated Methodology for Assessing Climate
Change Adaptation Costs for Urban Drainage Systems. Urban Water Journal, 13, doi:10.1080/1573062X.2014.991740.
108
-------
INFRASTRUCTURE
Urban Drainage
an approach derived from EPA's Storm Water Management Model (SWMM). The simplified approach261
yields impact estimates in units of average adaptation costs per square mile for a total of 100 cities
across the contiguous U.S. (see Figure 14.1) for three categories of 24-hour storm events (those with
precipitation intensities occurring every 10, 25, and 50 years—metrics commonly used in infrastructure
planning) and two future time periods: 2050 (2040-2059) and 2090 (2080-2099).262 Inclusion of all U.S.
cities with stormwater conveyance systems would provide a more comprehensive characterization of
future impacts. The analysis also assumes that the systems are able to manage runoff associated with
historical climate conditions, and estimates the costs of implementing the adaptation measures
necessary to manage increased runoff due to climate change.263 For more information on the CIRA
approach and results for the urban drainage sector, please refer to Neumann et al. (2014)264 and Price et
al. (2014).265,266
14.4 RESULTS
Table 14.1 presents the projected proactive adaptation costs for urban drainage infrastructure in the
100 modeled cities across the contiguous U.S. In 2050, the projected costs are slightly higher under
RCP4.5 than RCP8.5 in a few cases, particularly with the CCSM4 model. For example, for 25- and 50-year
storms, costs under RCP4.5 are higher than RCP8.5 for CCSM4, GISS_E2_R (50-year storm only), MIROC
5, and the five-GCM average.267 However, by 2090, the climate signal is clearer, and the projected
adaptation costs are higher under RCP8.5 for all three storm types and across all five models with the
one exception of GISS_E2_R for the 50-year storm. For the five-GCM average, annual costs in 2090 for
RCP8.5 for 10-, 25-, and 50-year storms are estimated at $2.5, $3.9, and $5.6 billion ($2015),
respectively. Projected costs under RCP4.5 are lower ($1.6, $2.7, and $4.1, respectively).
261 Although more detailed models, such as SWMM, are often used by municipalities for local stormwater management planning, applying these
models across the 100 cities examined in this analysis was not practicable.
262 The analysis assumes that adaptation investments are made at the beginning of each time period analyzed.
263 Because this analysis does not model damages due to urban flooding or combined sewer overflow events, a reactive adaptation scenario
was not modeled. For estimates of damages due to Inland Flooding, please see that section.
264 Neumann, J., J. Price, P. Chinowsky, L. Wright, L. Ludwig, R. Streeter, R. Jones, J. Smith, W. Perkins, L. Jantarasami, and J. Martinich, 2014:
Climate change risks to U.S. infrastructure: impacts on roads, bridges, coastal development, and urban drainage. Climatic Change, 131, 97-109,
doi:10.1007/sl0584-013-1037-4.
265 Price, J., L. Wright, C. Fant, and K. Strzepek, 2014: Calibrated Methodology for Assessing Climate Change Adaptation Costs for Urban
Drainage Systems. Urban Water Journal, 13, doi:10.1080/1573062X.2014.991740.
266 The results presented here apply the methods described in Price et al. (2014) and Neumann et al. (2014), with an expansion of the number
of cities modeled.
267 This is likely due to the fact that the analysis relies on climate projections for extreme events, and in some cases the changes in extremes
under RCP4.5 are higherthan under RCP8.5.
109
-------
INFRASTRUCTURE
Urban Drainage
Table 14.1. Projected Annual Proactive Adaptation Costs for Urban Drainage Infrastructure in 100
Cities in the U.S.
Total costs for 10-, 25-, and 50-year storms under RCP8.5 and RCP4.5 in 2050 (2040-2059) and 2090
(2080-2099) (billions $2015, undiscounted).
2050
2090
10-year
25-year
50-year
10-year
25-year
50-year
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
$1.5
$1.4
$2.4
$2.4
$3.5
$3.5
$2.5
$1.3
$3.7
$2.0
$5.2
$2.9
CCSM4
$1.0
$1.3
$1.5
$2.6
$2.1
$4.2
$1.5
$1.2
$2.3
$2.0
$3.3
$3.3
GISS_E2_R
$1.7
$1.5
$2.9
$2.8
$4.3
$4.5
$2.5
$1.6
$3.9
$3.4
$5.8
$5.9
HadGEM2
$2.0
$1.8
$3.2
$3.0
$4.6
$4.6
$3.3
$2.3
$5.0
$3.6
$7.0
$5.1
MIROC5
$2.0
$1.7
$3.0
$3.2
$4.2
$4.9
$2.7
$1.6
$4.5
$2.4
$6.6
$3.4
5-GCM
Average
$1.7
$1.6
$2.6
$2.8
$3.7
$4.3
$2.5
$1.6
$3.9
$2.7
$5.6
$4.1
Figure 14.1 presents the projected costs for each of the 100 modeled cities, aggregated to seven regions
used in the NCA4. The costs presented in Figure 14.1 are weighted average costs per square mile.268 As
shown, for a 10-year storm, costs are projected to be highest in the Northwest under both RCP8.5 and
RCP4.5 ($300,000 and $240,000 per square mile, respectively). In all other regions except for the
Southeast, the projected costs for a 10-year storm are less than half of the costs in the Northwest. For a
25-year storm, costs are projected to be highest in the Southeast and Southern Plains under RCP8.5, at
approximately $290,000 and $280,000, respectively, per square mile. Under RCP4.5, costs for a 25-year
storm are projected to be highest in the Northwest (approximately $220,000 per square mile).
As expected, the highest projected costs for urban drainage infrastructure are associated with a 50-year
storm. Under both RCP8.5 and RCP4.5, projected costs for a 50-year storm are highest in the Southern
Plains at $460,000 and $230,000, respectively, per square mile. The second highest costs for a 50-year
storm under RCP8.5 are projected to occur in the Southeast ($380,000 per square mile) while the
second highest costs under RCP4.5 are projected to occur in the Northern Plains ($200,000 per square
mile).
268 The adaptation costs per square mile, calculated by city, storm, scenario, and year, were aggregated to the regions used in the Fourth
National Climate Assessment and weighted by area. For example, for a region with 2 cities, each with an area of 100 square miles, each city's
area is divided by the sum of the areas, resulting in a proportion value of 0.5 for each city. This proportion value is then multiplied by each
calculation of per-square-mile adaptation costs (calculated by storm, scenario, and year) to produce a weighted average adaptation cost per
square mile.
110
-------
INFRASTRUCTURE
Urban Drainage
Figure 14.1. Projected Regional Proactive Adaptation Costs for Urban Drainage Infrastructure
Weighted average per-square-mile adaptation costs in 2090 (2080-2099) for 10-, 25-, and 50-year
storms under RCP8.5 and RCP4.5 (five-GCM average, $2015, undiscounted). Costs for each of the 100
modeled cities (shown) are aggregated to the NCA4 regions.
NORTHERN PLAINS
MIDWEST
NORTHWEST
RCP8.5
5400,000
S 300.000
5400,000
5300,000
RCP4.5
5400,000
5200,000
5200,000
5300.000
5100.000
5100,000
5200,000
5100.000
10-year 25-year 50-year
10-year 25-year 50-year
10-year 25-year 50-year
NORTHEAST
5400,000
5300,000
5200,000
5100,000
10-year 25-year 50-year
SOUTHEAST
5400,000
5300,000
5200,000
SOUTHERN PLA NS
5100.000
5400,000
10-year 25-year 50-year
5300,000
5200,000
5100.000
10-year 25-year 50-year
SOUTHWEST
5400,000
5300,000
$200,000
5100,000
. 1 k
50
m m m
10-year 25-year 50-year
111
-------
INFRASTRUCTURE
Urban Drainage
14.5 DISCUSSION
The results presented above suggest that climate change will stress the nation's aging water
infrastructure to varying degrees by location and over time. These results are consistent with the
findings of the assessment literature, which describes much of the country's current drainage
infrastructure as already overwhelmed during heavy precipitation and high runoff events - an impact
that is projected to be exacerbated as a result of climate change, land-use change, and other factors.
There are several important considerations worth noting. First, the use of best management practices in
this analysis to address all future increases in runoff may be overly optimistic, and therefore might
underestimate total risk. In some instances, these practices may not be sufficient to handle all increases
in stormwater volume, and therefore construction and expansion of existing conveyance networks may
be necessary under a future climate. In addition, this analysis only estimated adaptation costs for 100
cities; inclusion of all U.S. cities with stormwater conveyance systems would provide a more
comprehensive characterization of future impacts and would likely increase costs substantially.
112
-------
INFRASTRUCTURE
Coastal Property
15. COASTAL PROPERTY
15.1 KEY FINDINGS
• A large area of U.S. coastal land and property is at risk of inundation from sea level rise, and an even
larger area is at risk of damage from storm surge, which will intensify as sea levels continue to rise.
• Without adaptation, cumulative discounted damages to coastal property in the contiguous U.S. are
estimated at $3.6 trillion through 2100 under both RCPs. Damages under RCP4.5 are reduced by $92
billion compared to RCP8.5.
• Well-timed adaptation measures significantly reduce cumulative discounted costs to an estimated
$820 billion under RCP8.5 and $800 billion under RCP4.5. In comparison, reductions in damages
under RCP4.5 are modest, with the majority of benefits projected to occur late in the century.
• Projected sea level rise and storm surge have environmental justice implications. In the example of
Tampa Bay, nearly all of the area inhabited by the most socially vulnerable is projected to be
abandoned as opposed to approximately half of the area inhabited by the least socially vulnerable.
15.2 BACKGROUND
Coastal areas in the U.S. are some of the most densely populated, developed areas in the nation, and
contain a wealth of natural and economic resources. Sea level rise threatens to inundate many low-lying
coastal areas and increase flooding, erosion, wetland habitat loss, and saltwater intrusion into estuaries
and freshwater aquifers. Climate change will increase exposure risk to coastal flooding due to increases
in extreme precipitation and in hurricane intensity and rainfall rates, as well as sea level rise and the
resulting increases in storm surge.269 Rising temperatures are causing ice sheets and glaciers to melt and
ocean waters to expand, contributing to global sea level rise at increasing rates. The combined effects of
sea level rise and other climate change factors, such as increased intensity of storms, may cause rapid
and irreversible change across the contiguous U.S. coastline. Nationally important assets, such as ports,
tourism and fishing sites, water supply and energy infrastructure, and evacuation routes in already-
vulnerable coastal locations, are increasingly exposed to sea level rise, storm surges, inland flooding, and
erosion.270 In the Gulf Coast, home to more than 50% of the nation's petroleum refining capacity,
impacts to liquid fuels infrastructure can disrupt distribution and availability of products, with effects on
consumer prices.
15.3 APPROACH
This analysis identifies at-risk coastal property and land-based energy infrastructure across the
contiguous U.S. and projects costs incurred due to sea level rise and storm surge, with and without
adaptation.271 Coastal property considered in the analysis includes residential, commercial, industrial,
institutional, and government properties, as well as energy infrastructure and facilities. Importantly,
impacts to other coastal assets (e.g., transportation and telecommunication infrastructure) and
ecological resources are not estimated.
269 Bell, J.E., S.C. Herring, L. Jantarasami, C. Adrianopoli, K. Benedict, K. Conlon, V. Escobar, J. Hess, J. Luvall, C.P. Garcia-Pando, D. Quattrochi, J.
Runkle, and C.J. Schreck, III, 2016: Ch. 4: Impacts of Extreme Events on Human Health. The Impacts of Climate Change on Human Health in the
United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 99-128, doi: 10.7930/J0BZ63ZV.
270 Moser, S. C., M. A. Davidson, P. Kirshen, P. Mulvaney, J. F. Murley, J. E. Neumann, L. Petes, and D. Reed, 2014: Ch. 25: Coastal Zone
Development and Ecosystems. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.)
Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 579-618, doi:10.7930/J0MS3QNW.
271 The additional risks to coastal flooding events caused by precipitation are not included in this analysis.
113
-------
INFRASTRUCTURE
Coastal Property
The EPA's National Coastal Property Model is used to estimate how areas along the coast may respond
to sea level rise and storm surge and calculates the economic impacts of adaptation decisions (i.e.,
damages due to climate change associated with costs of protection strategies or lost value of inundated
property). The approach uses four primary responses to the threat of climate change: beach
nourishment, property elevation, shoreline armoring, or property abandonment. The model projects an
adaptation response for areas at risk based on sea level rise, storm surge height, property value, and
costs of protective measures. The model is also run assuming no adaptation to compare the risks of
inaction with the net costs of adaptation.
This analysis uses regional sea level rise scenarios based on projections from NOAA (2017), described in
the Modeling Framework section of this Technical Report. The National Coastal Property Model then
uses a tropical cyclone simulator272 and a storm surge model273 to estimate the joint effects of sea level
rise and storm surge for East and Gulf Coast sites, and an analysis of historic tide gauge data to project
future flood levels for West Coast sites.274
The adaptation responses projected by the National Coastal Property Model are developed using a cost-
benefit framework comparing the costs of protection relative to the property value. Developed using a
simple metric to estimate potential adaptation responses in a consistent manner for the entire
coastline, the estimates presented here should not be construed as recommending any specific policy or
adaptive action. Further, additional adaptation options not included in this analysis, such as marsh
restoration, may be appropriate, and potentially more cost-effective, for some locales. Importantly, this
analysis assumes that development in the coastal flood plan remains fixed at current locations, with
growth in economic value at those locations consistent with past trends in national property
appreciation. Increased development at the extensive margin in coastal communities, which follows
observed patterns over the past several decades, or faster rates of property appreciation in coastal
versus inland sites, could compound the economic impacts of sea level rise and storm surge. For more
information on the National Coastal Property Model, please refer to Neumann et al. (2014)275 and
Neumann et al. (2014).276
15.4 RESULTS
Sea level rise and storm surge pose increasingly large risks to coastal property, including costs associated
with property abandonment, residual storm damages,277 and protective adaptation measures, such as
property elevation, beach nourishment, and shoreline armoring. As shown in Figure 15.1, cumulative
damages to coastal property across the contiguous U.S. are significantly reduced if protective adaptation
measures are implemented compared to a scenario where no adaptation occurs. Without adaptation,
272 Emanuel, K., R. Sundararajan, and J. Williams, 2008: Hurricanes and Global Warming: Results from Downscaling IPCC AR4 Simulations.
Bulletin of the American Meteorological Society, doi:10.1175/BAMS-89-3-347.
273 Jelesnianski, C.P., J. Chen, and W.A. Shaffer, 1992: SLOSH: Sea, lake, and overland surges from hurricanes. NOAA Technical Report NWS 48,
National Oceanic and Atmospheric Administration, U. S. Department of Commerce, Washington, DC.
274 Tebaldi, C., B. Strauss, and C. Zervas, 2012: Modeling sea-level rise impacts on storm surges along U.S. coasts. Environmental Research
Letters, 7, 014032, doi:10.1088/1748-9326/7/l/014032.
275 Neumann, J., K. Emanuel, S. Ravela, L. Ludwig, P. Kirshen, K. Bosma, and J. Martinich, 2014: Joint Effects of Storm Surge and Sea-level Rise on
U.S. Coasts. Climatic Change, 129, 337-349, doi: 10.1007/sl0584-014-1304-z.
276 Neumann, J., J. Price, P. Chinowsky, L. Wright, L. Ludwig, R. Streeter, R. Jones, J.B. Smith, W. Perkins, L. Jantarasami, and J. Martinich, 2014:
Climate change risks to U.S. infrastructure: impacts on roads, bridges, coastal development, and urban drainage. Climatic Change, 131, 97-109,
doi:10.1007/sl0584-013-1037-4.
277 Residual damages in this analysis are those that cause property damages smaller than the value of the property and any potential protective
measures. Therefore the model estimates that the economically efficient response is to incur these smaller damages.
114
-------
INFRASTRUCTURE
Coastal Property
cumulative damages under RCP8.5 are estimated at $3.6 trillion through 2100 (discounted at 3%),
compared to $820 billion in the scenario where cost-effective adaptation measures are implemented.
Under RCP4.5, costs without adaptation are estimated at $3.6 trillion through 2100 (discounted 3%) (a
reduction of $92 billion relative to RCP8.5), compared to $800 billion with adaptation.278
Figure 15.1. Cumulative Costs of Sea Level Rise and Storm Surge to Coastal Property
Costs are presented with and without adaptation under RCP8.5 and RCP4.5279 in trillions of $2015,
discounted at 3%.
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
RCP8.5 - Costs with Adaptation RCP8.5 - Costs without Adaptation
RCP4.5 - Costs with Adaptation RCP4.5 - Costs without Adaptation
Figure 15.2 shows the total estimated costs (including the value of abandoned property, costs of
protective adaptation measures, and residual costs of storm surge damage) for 17 key sites under both
RCPs. Costs vary across sites primarily due to the value of property at risk and the severity of the storm
surge threats. For example, adaptation costs are comparatively higher in sites such as Tampa and Miami,
where there are many high-value properties in low-lying areas and high levels of storm surge are
projected in the future.
278 Global sea level rise is similar under the RCPs scenarios through mid-century. It is not until the second half of the century when the benefits
of reduced sea level rise under RCP4.5 become apparent, which are more heavily affected by discounting. In addition, some of the effects on
coastal property are due to land subsidence which is assumed to occur at an equal rate under the sea level rise projections of the two RCPs.
279 The step-wise nature of the graph is due to the fact that the analysis evaluates storm surge risks every ten years, beginning in 2005.
115
-------
INFRASTRUCTURE
Coastal Property
Figure 15.2. Projected Costs to Coastal Property of Sea Level Rise and Storm Surge
Costs are shown for 17 multi-county coastal areas (see map below) that were modeled for sea level rise
and storm surge impacts and potential adaptation responses through 2100 (billions $2015, discounted at
3%).
$100
¦ RCP8.5
^ RCP4.5
Washington
Oregon
SE Massachusetts
• New York City Area
Coastal NJ and DE
Northern California
Virginia Beach. VA
/ Wilmington. NC Area
Charleston. SC Area
Southern California
Galveston. TX Area
Jacksonville. FLArea
New Orleans, LA Area
""—-Ta mpa. FL Are a
Miami. FLArea
Mobile, AL Area \
Pensacda. FLArea
116
-------
INFRASTRUCTURE
Coastal Property
Table 15.1 presents the breakdown of projected costs for the 17 sites under RCP8.5.280 In general, the
largest costs are associated with shoreline armoring and incurred residual storm surge damages.
Table 15.1. Projected Costs of Sea Level Rise and Storm Surge Damages
Projected costs are for the period 2000-2100 under RCP8.5 at 17 sites, and are presented in millions
$2015, discounted at 3%. See Figure 15.2 for the geographic extent of each coastal area.
Coastal Area
Value of
Abandoned
Property
Cost of
Armoring
Cost of
Nourishment
Cost of
Elevation
No
Adaptation
(Storm
Surge
Damage)
Total
Costs
SE Massacusetts
$4,000
$8,900
$530
$140
$790
$14,000
Charleston, SC area
$23,000
$4,100
$640
$8
$1,400
$30,000
Galveston, TX area
$30,000
$2,200
$330
$28
$2,300
$34,000
Jacksonville, FL area
$19,000
$10,000
$1,800
$420
$6,200
$37,000
Miami, FL area
$57,000
$13,000
$1,600
$2,000
$23,000
$96,000
Mobile, AL area
$4,700
$900
$420
00
T—1
¦uy
$1,600
$7,600
New Orleans, LA area
$29,000
$2,600
$440
$0
$3,100
$35,000
New York City are
$15,000
$3,100
$150
$110
$2,000
$20,000
Northern California
$1,000
$450
$50
$7
$110
$1,600
Coastal NJ and DE
$19,000
$9,600
$0
$480
$2,300
$31,000
Oregon (parts of)
$450
$430
$28
$2
$54
$960
Pensacola, FL area
$3,100
$4,400
$600
$150
$1,300
$9,600
Southern California
$480
$1,800
$570
$35
$410
$3,300
Tampa, FL area
$63,000
$7,600
$580
$150
$7,600
$79,000
Virginia Beach VAarea
$930
$1,200
$320
$5
$870
$3,300
Washington (state)
$690
$730
$0
$0
$190
$1,600
Wilmington, NC area
$1,400
$2,300
$630
$160
$1,100
$5,600
Environmental Justice Case Study: Tampa Bay
The analysis also explores the potential impact of sea level rise and storm surge on socially
disadvantaged populations in the Tampa Bay area. The approach, based on the methodology described
in Martinich et al. (2013),281 quantifies how sea level rise and storm surge risks are distributed across
280 The projected costs under RCP4.5 are not significantly different than the costs under RCP8.5 (approximately 1% lower across the 17 sites).
281 Martinich, J., J. Neumann, L. Ludwig, and L. Jantarasami, 2013: Risks of sea level rise to disadvantaged communities in the United States.
Mitigation and Adaptation Strategies for Global Change, 18, 169-185, doi: 10.1007/sll027-011-9356-0.
117
-------
INFRASTRUCTURE
Coastal Property
different socioeconomic populations; how these populations are likely to respond; and what adaptation
costs (i.e., property damage and protection investments) will potentially be incurred. The analysis uses
the Social Vulnerability Index (SoVI)282 to identify socially vulnerable coastal communities. It calculates
census tract-level SoVI values based on data on gender, age, race, employment, and wealth from the
2010 Census and 2014 American Community Survey.
Figure 15.3 presents the at-risk areas in each SoVI category under RCP8.5. As shown, 96% of the area
inhabited by the high-vulnerability population is likely to be abandoned as opposed to 54% of the area
inhabited by the low-vulnerability population. Results under RCP4.5 show similar patterns.
Figure 15.3. Tampa Bay Areas at Risk from Sea Level Rise and Storm Surge
The chart displays the areas (in square miles) within each SoVI category at risk from sea level rise and
storm surge through 2100 under RCP8.5 in the Hillsborough and Pinellas Counties of Florida. The chart
identifies the adaptation measures for each SoVI category.
120
100
<-1.5 -1.4 to -0.5 -0.4 to 0.5 0.6 to 1.5 >1.5
(Low Vulnerability) (High Vulnerability)
¦ No Adaptation (Storm Surge Damage) ¦ Property Abandonment
¦ Shoreline Armoring ¦ Property Elevation
¦ Beach Nourishment
15.5 DISCUSSION
Projections of increasing risks to coastal property of sea level rise and storm surge, and of the potential
for adaptation to reduce overall costs, are consistent with the findings of the assessment literature,283
and other recent reports. Assuming no protective measure are taken, The American Climate
282 Cutter, S., B.J. Boruff, and W.L. Shirley, 2003: Social Vulnerability to Environmental Hazards. Social Science Quarterly, 84, 242-261.
283 Moser, S. C., M. A. Davidson, P. Kirshen, P. Mulvaney, J. F. Murley, J. E. Neumann, L. Petes, and D. Reed, 2014: Ch. 25: Coastal Zone
Development and Ecosystems. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.)
Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 579-618. doi:10.7930/J0MS3QNW.
118
-------
INFRASTRUCTURE
Coastal Property
Prospectus284 found that $66-106 billion worth of current coastal property will likely be below mean sea
level by 2050 under RCP8.5 ($62-85 billion under RCP4.5), which could grow to $238-507 billion by 2100
($175-339 billion under RCP4.5). Values from the American Climate Prospectus presented are
undiscounted, at 2011 property prices, using mean sea level measures, while the estimates presented in
this section are based on mean high water levels (as adopted by other analyses, such as Strauss et al.
2015285), and using projected property prices that grow with projected GDP per capita. A recent
Congressional Budget Office report286 found that annual hurricane damages to coastal development,
considering both flooding and wind damages, currently amount to approximately $28 billion, but that by
2075, the figure could reach approximately $39 billion. Results are undiscounted, and incorporate both
adaptation to these risks and a projection of future property prices. This CBO analysis estimated the
joint effects of hurricanes and increased coastal development on future damages, concluding that
roughly 45% of their projected increase is attributable to climate change, and 55% to coastal
development. Both of these efforts rely on coastal damage functions from a model of insured losses
developed by Risk Management Solutions, so not all aspects of the approach used in those studies,
some of which reflect proprietary data and methods, can be compared to those adopted in this analysis.
It should also be noted that none of these analyses, including the results of this section, incorporate
findings from recent Antarctic research287 suggesting that 6 feet or more of global sea level rise may be
possible this century under RCP8.5. Consideration of this possibility would roughly double most of the
estimates presented in this section, and may have an unknown (nonlinear) effect on property damages.
The effect of global GHG mitigation in reducing damages and adaptation costs is not significant in this
study, and is likely underestimated for several reasons. In terms of projected damages, global sea level
rise is similar under the RCPs scenarios through mid-century. It is not until the second half of the century
when the benefits of reduced sea level rise under RCP4.5 become apparent. In terms of adaptation
costs, when considering the total present value estimates under the RCPs, avoided adaptation costs
accrued in later years are more heavily affected by discounting.288 In addition, costs under both RCPs are
projected with the assumption that coastal areas will implement cost-efficient and well-timed
adaptation measures in response to risks. Since many parts of the coastline are not sufficiently
protected today, and because adaptation measures that are taken are oftentimes not well-timed,
estimates for this sector likely underestimate damages.289 Also, this analysis holds constant the level of
development in the coastal floodplain, which likely leads to underestimates of reported risk.290 Finally,
the inclusions of impacts of sea level rise and storm surge on other coastal assets (e.g., transportation
and telecommunication infrastructure) and ecological resources would increase total potential damages.
284 Hsiang, S., R. Kopp, D. Rasmussen, M. Mastrandrea, A. Jina, J. Rising, R. Muir-Wood, P. Wilson, M. Delgado, S. Mohan, K. Larsen, T. Houser,
2014. American Climate Prospectus: Economic Risks in the United States. Goldman School of Public Policy Working Paper. Available online at
https://gspp.berkelev.edu/research/working-paper-series/american-climate-prospectus-economic-risks-in-the-united-states
285 Strauss, B.H., S. Kulpa, and A. Levermann, 2015: Carbon choices determine US cities committed to futures below sea level. Proceedings of
the National Academy of Sciences, 112, 13508-13513, doi: 10.1073/pnas.1511186112.
286 Dinan, T., 2016: CBO's Approach to Estimating Expected Hurricane Damage: Working Paper 2016-02. Congressional Budget Office. Available
online at https://www.cbo.gov/publication/51610
287 DeConto, R. M., and D. Pollard, 2016: Contribution of Antarctica to past and future sea-level rise. Nature, 531, 591-597.
288 Without discounting, the cumulative effect of mitigation is larger, reducing impacts by about 8% (from $1.2 trillion to $1.1 trillion), and the
annual benefits rise from approximately $450 million in 2050 to nearly $1.6 billion in 2100.
289 However, climate change amplification of flood risk may also trigger proactive adaptation, such as people choosing not to move into flood-
prone areas, or where zoning or insurance markets present clear barriers to moving there. These dynamics are not captured in this analysis.
290 Dinan, T., 2016: CBO's Approach to Estimating Expected Hurricane Damage: Working Paper 2016-02. Congressional Budget Office. Available
online at https://www.cbo.gov/publication/51610
119
-------
ELECTRICITY
16. ELECTRICITY DEMAND AND SUPPLY
16.1 KEY FINDINGS
• Higher temperatures due to climate change are projected to increase demand for electricity to meet
cooling needs across the contiguous U.S.
• Electricity demand rises under both RCPs. Nationally, electricity demand increases 2.4-2.9% in 2050
under RCP8.5 and 1.7-2.0% under RCP4.5, though increases in regional demand vary.
• Electric power sector capacity and generation will increase to meet these higher demands, resulting
in national cumulative costs of tens to hundreds of billions of dollars through mid-century. Warming
temperatures are projected to increase national power system costs by 2.4-4.0% under RCP8.5 and
1.8-3.1% under RCP4.5 through 2050.
• The Southeast is projected to experience the highest costs associated with meeting increased
electricity demands, with high costs also projected in the Northeast, Midwest, and Southern Plains.
16.2 BACKGROUND
Electricity is an essential element of modern life. It lights and cools our homes, powers our appliances
and electronics, supports the production of goods and services, and enables critical infrastructure
services such as water treatment and telecommunications. The generation of electricity in the U.S., most
of which comes from fossil fuels, also contributes to climate change, accounting for approximately 30
percent of U.S. greenhouse gas emissions.291
As air temperatures rise due to climate change, electricity demands for cooling are expected to increase
in every U.S. region.292 Higher summer temperatures, particularly during heat waves, will likely increase
peak electricity demand, placing more stress on the electricity grid and increasing electricity costs.
Although the majority of U.S. residential and commercial cooling demand is met with electricity, less
291 U.S. Department of Energy, 2014: Electric Power Monthly: Table 1.1. Net Generation by Energy Source: Total (All Sectors), 2004-February
2014.
292 Dell, J., S. Tierney, G. Franco, R. G. Newell, R. Richels, J. Weyant, and T. J. Wilbanks, 2014: Ch. 4: Energy Supply and Use. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 113-129. doi:10.7930/J0BG2KWD
120
-------
ELECTRICITY
Electricity Demand and Supply
than nine percent of heating demand is met with electricity.293,294 Therefore, although higher average
temperatures are expected to reduce electricity demands for heating, net electricity use is projected to
increase under climate change. Meeting these higher demands for electricity has cost and operating
implications for the electric sector as additional capacity and generation are required. At the same time,
higher temperatures reduce the capacity of both thermal power plants and transmission lines.295
16.3 APPROACH
The analysis projects how rising temperatures under climate change will affect electricity demand in the
contiguous U.S., and how system costs in the electric power sector will change in response to these
shifts in demand. To estimate changes in demand, the approach applies downscaled temperature
projections from the five GCMs under RCP8.5 and RCP4.5 to two models of the U.S. electric power
sector:
• Regional Electricity Deployment System Model (ReEDS): a technology-rich model of the deployment
of electric power generation technologies and transmission infrastructure for the contiguous
U.S.296-297
• Global Change Assessment Model (GCAM-USA, referred to as "GCAM" hereafter): a detailed,
service-based building energy model for the 50 U.S. states298-299-300-301
The models project changes in electricity demand as functions of changes in heating and cooling degree-
days (HDDs/CDDs).302 To assess the effect of rising temperatures, changes in heating and cooling degree
days and electricity demand are compared to a control scenario that assumes temperatures do not
change over time but does incorporate future population growth.303
In addition to estimating the change in electricity demand, this analysis assesses impacts on the U.S.
electricity sector's supply side using the same two models. The models project changes in the
293 U.S. Department of Energy, U.S. Energy Information Administration, 2009: 2009 Residential Energy Consumption Survey (RECS): Table CE4.1.
Available online at http://www.eia.gov/consumption/residential/data/2009/index.cfm
294 U.S. Department of Energy, U.S. Energy Information Administration, 2003: 2003 Commercial Buildings Energy Consumption Survey (CBECS):
Tables E1A and E3A. Available online at http://www.eia.gov/consumption/commercial/data/2003/index.cfm
295 Dell, J., S. Tierney, G. Franco, R. G. Newell, R. Richels, J. Weyant, and T. J. Wilbanks, 2014: Ch. 4: Energy Supply and Use. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 113-129. doi:10.7930/J0BG2KWD
296 Eurek, K., W. Cole, D. Bielen, N. Blair, S. Cohen, B. Frew, J. Ho, V. Krishnan, T. Mai, B. Sigrin and D. Steinberg, 2016: Regional Energy
Deployment System (ReEDS) Model Documentation. Version 2016. National Renewable Energy Laboratory Technical Report NREL/TP-6A20-
67067. [Available online at www.nrel.gov/docs/fyl7osti/ 67067.pdf]
297Sullivan, P., J. Colman, and E. Kalendra, 2015: Predicting the Response of Electricity Load to Climate Change. National Renewable Energy
LaboratoryTechnical Report NREL/TP-6A20-64297. Available online at www.nrel.gov/docs/fvl5osti/64297.pdf
298 Iyer, G. et al., 2017: U.S. electric power sector transitions required to achieve deep decarbonization targets: Results based on a detailed
state-level model of the U.S. energy system. PNNL Report (forthcoming).
299 Kyle, P., L. Clarke, F. Rong, and S.J. Smith, 2010: Climate policy and the long-term evolution of the U.S. buildings sector. The Energy Journal,
31, 145-172.
300 Zhou, Y., J. Eom, and L. Clarke, 2013: The effect of global climate change, population distribution, and climate mitigation on building energy
use in the U.S. and China. Climatic Change, doi:10.1007/sl0584-013-0772-x.
301 Zhou Y., L. Clarke, J. Eom, P. Kyle, P. Patel, S. Kim, J.A. Dirks, E.A. Jensen, Y. Liu, J.S. Rice, L.C. Schmidt, T.E. Seiple, 2014: Modeling the effect
of climate change on U.S. state-level buildings energy demands in an integrated assessment framework. Applied Energy, 113, 1077-1088.
302 HDDs and CDDs are one way to measure the influence of temperature change on energy demand. They measure the difference between
outdoor temperatures and a temperature that people generally find comfortable indoors. These measurements suggest how much energy
people might need to use to heat and cool their homes and workplaces. The approach used a fixed balance point of 65°F.
303 The HDD/CDD values were smoothed using 4th degree polynomial curves to capture long-term climate effects as opposed to interannual
variability.
121
-------
ELECTRICITY
Electricity Demand and Supply
generation and generation mix needed to meet increasing demand due to future warming. The two
models also estimate the corresponding system costs comprised of capital (e.g., expenditures related to
bring new capacity online), operations and maintenance, and fuel costs over time. Note that ReEDS
includes the effects of transmission capacity losses due to high air temperatures. In both the demand
and supply analyses of this section, the ReEDS model is simulated through 2050, while the GCAM model
runs through 2100. While ReEDS' electricity demand path is fixed, the GCAM demand path responds to
changes in power prices. For more information on the methodology for estimating impacts to electricity
demand and supply, please refer to McFarland et al. (20 15).304
In an extension to the approach taken in McFarland (2015), the ReEDS analysis incorporates the effects
of changes in precipitation on hydropower generation. Changes in hydropower generation are
estimated using the US Basins modeling framework, a linked water systems model designed to evaluate
the impacts of climate change on water resources.305,306 Precipitation and temperature from each
RCP/GCM combination are inputs into: (a) a rainfall-runoff model (CLIRUN-II), which is used to simulate
monthly runoff; and (b) a water demand module, which projects the water requirements of the
municipal and industrial (M&l) and agriculture sectors across the GCMs. With these runoff and demand
projections, US Basins produces a time series of reservoir storage, release, and allocation to the various
demands in the system, which include M&l, agriculture, transboundary flows, and hydropower. See
Figure A10-1 of the Appendix to this Technical Report for projected changes in flow aggregated to the
scale of four-digit hydrologic unit codes. The changes in hydropower generation estimated by the US
Basins framework are then applied to ReEDS. For more information on the hydropower methodology,
please refer to Boehlert et al. (20 16).307,308
16.4 RESULTS
The projected changes in regional CDD and HDD over time and across the GCMs are shown in Figures
16.1 and 16.2, respectively. The increase in CDDs are more pronounced in the northern regions, rising by
a factor of 3 to 5 by 2100 under RCP8.5 versus a doubling in the southern regions over the same
timeframe. The smaller change in the southern regions is primarily due to the higher number of CDDs in
the reference year. The change in CDD under RCP4.5 is more modest, rising by factor of less than two in
304 ReEDS and GCAM employ different techniques and underlying datasetsfor estimating the sensitivity of electricity demand to changes in
HDD/CDD. See: McFarland, J., Y. Zhou, L. Clarke, P. Sullivan, J. Colman, W. Jaglom, M. Colley, P. Patel, J. Eom, S. Kim, G. Kyle, P. Schultz, B
Venkatesh, J. Haydel, C. Mack, and J. Creason, 2015: Impacts of rising air temperatures and emissions mitigation on electricity demand and
supply in the United States: a multi-model comparison. Climatic Change, doi: 10.1007/sl0584-015-1380-8.
305 Boehlert, B., K.M. Strzepek, Y. Gebretsadik, R. Swanson, A. McCluskey, J. Neumann, J. McFarland, and J. Martinich, 2016: Climate change
impacts and greenhouse gas mitigation effects on U.S. hydropower generation. Applied Energy, 183, 1511-1519.
306 Boehlert, B., K.M. Strzepek, S.C. Chapra, C. Fant, Y. Gebretsadik, M. Lickley, R. Swanson, A. McCluskey, J. Neumann, J. Martinich, 2015:
Climate change impacts and greenhouse gas mitigation effects on US water quality. Journal of Advances in Modeling Earth Systems, 7, 1326-
1338.
307 Boehlert, B., K.M. Strzepek, Y. Gebretsadik, R. Swanson, A. McCluskey, J. Neumann, J. McFarland, and J. Martinich, 2016: Climate change
impacts and greenhouse gas mitigation effects on U.S. hydropower generation. Applied Energy, 183, 1511-1519.
308 While not presented in detail in this section, additional sensitivity cases were run in ReEDS to assess the effects of climate-induced changes
in thermo-electric cooling water availability on system costs. The climate-induced changes to thermal cooling water availability in ReEDS are
determined through iteration with the US Basins model (See Water Quality section of this Technical Report for more information). Whereas
ReEDS constrains only quantities of water withdrawn in each season and balancing area, US Basins represents detailed hydrology and plant-
specific cooling water constraints such as thermal discharge limits. ReEDS scenarios are first run with cooling water availability based solely on
the evolution of the hydrological system as projected by US Basins. The resulting electric power sector generation and capacity expansion in
ReEDS are then used as input to a second iteration of the US Basins model, which evaluates the feasibility of ReEDS results given its additional
thermal cooling water constraints. Any constraint violations are fed back to ReEDS by multiplying the initial water available by the fraction of
ReEDS proposed water demand that is met in US Basins to form updated cooling water availability data. ReEDS is then run with this second
iteration of cooling water availability data to produce the final results.
122
-------
ELECTRICITY
Electricity Demand and Supply
the northern regions and under 1.5 in the southern regions by 2100. Higher temperatures reduce HDDs
by approximately 30% in the northern regions and 50% in the southern regions by 2100 under RCP8.5.
Because electricity comprises less than 10 percent of heating demand, changes in HDD have only a small
effect on electricity demand.309
Figure 16.1. Change in Cooling Degree-Days
Results shown by region through 2100 as compared to 2005 (2005=1), with a value of 2 representing a
100% increase in CDDs.
NORTHWEST
NORTHERN PLAINS
NORTHEAST
MIDWEST
2050
2050
2100
2050
2100
SOUTHWEST
2050
2100
SOUTHERN PLAIN!
SOUTHEAST
2050
2100
RCP8.5
2050
RCP4.5
3i® HDD reductions will have a larger effect on reducing natural gas and fuel oil consumption during the winter months, however those effects
are not modeled here.
123
-------
ELECTRICITY
Electricity Demand and Supply
Figure 16.2. Change in Heating Degree-Days
Results shown by region through 2100 as compared to 2005 (2005=1), with a value of 0.5 representing a
50% decrease in HDDs.
NORTHWEST
NORTHERN PLAINS
0.9
0.9
MIDWEST
0.7
NORTHEAST
0.6
0.9
0.7
0.9
0.5
0.6
0.7
0.5
2050
0.6
0.4
0.6
2050
2100
0.5
0.5
0.4
SOUTHWEST
0.4r
2050
2050
2100
0.9
SOUTHERN PLAINS
SOUTHEAST
0.6
0.9
0.5
0.4
0.7
2050
2100
0.6
0.5
0.4
RCP8.5
RCP4.5
2050"
2100 04
2050
At the national level, both the ReEDS and GCAM models estimate that the HDD/CDD changes result in
average increases in electricity demand of 2.9% and 2.4%, respectively in 2050 under RCP8.5 (Figure
16.3). Under RCP4.5, average change in demand in 2050 increases across both models by 2.0% and
1.7%, respectively. National average electricity demand is projected by GCAM to increase 5.3% under
RCP8.5 in 2090, and 2.4% under RCP4.5. Figure A.10.2 of the Appendix to this Technical Report provides
projected changes in regional demand. As shown there, the largest regional increases in demand by
2050 occur in the Northeast, Southeast, Midwest, and Southern Plains in the ReEDS model; the largest
increases in demand in the GCAM model occur in the Southeast, Southern Plains, and Northwest,
especially by 2090.
124
-------
ELECTRICITY
Electricity Demand and Supply
Figure 16.3. Percent Change in National Electricity Demand
Values across RCPs and the five GCMs are shown relative to a control scenario without climate change
for the year 2050 (ReEDS and GCAM) and 2090 (GCAM only).
7%
6%
5%
4%
3%
2%
1%
0%
RCP8.5 RCP4.5 RCP8.5 RCP4.5 RCP8.5 RCP4.5
2050 2050 2090
ReEDS GCAM
GCM
RCP8.5
CanESM2
CCSM4
GISS-E2-R
HadGEM2-ES
MIROC5
5-Model Average
To meet the increase in demand, ReEDS and GCAM primarily expand generation of natural gas, non-
hydro renewables, and, to a lesser extent, nuclear in GCAM (Figure 16.4). However, the CanESM2
climate model, which projects large increases in precipitation in the Northwest and Southwest, leads to
an expansion of hydropower in 2050, by as much as 12%, under both climate scenarios in the ReEDS
results.
Figure 16.4. Change in National Electricity Generation by Technology Type
Values across RCPs and the five GCMs are shown for 2050 (ReEDS and GCAM) and 2090 (GCAM only) and
are relative to a control scenario without climate change.
0.6
S 0-5
>-
cl 0.4
-C
i 0.3
0.2
0.1
-0.1
00 (N
5 2
LO 1/1
RCP 8.5
RCP4.5
2050
ReEDS
I Hydro ¦ Non-Hydro Renewables
llllllliilll
OC
-------
ELECTRICITY
Electricity Demand and Supply
Discounted cumulative system costs by region and RCP from both power system models are shown in
Table 16.1. In the ReEDS model, projected national system costs increase by 4.0% under RCP8.5 and
3.1% under RCP4.5 through 2050,310 with most regions showing increases in costs. Notably, system costs
in ReEDS are estimated to decrease in the Northwest by 6.9% and 7.0% through 2050 under RCP8.5 and
RCP4.5, respectively, due to increased hydropower availability. Under the GCAM model, national system
costs are projected to increase by 2.4% and 1.8% under RCP8.5 and RCP4.5, respectively, through
2050—a smaller increase due in part to GCAM's less-sensitive demand response to temperature change.
System costs in GCAM are estimated to increase by 3.4% and 2.0% through 2100 under RCP8.5 and
RCP4.5, respectively. In both models and under all scenarios, the Southeast is projected to experience
some of the highest costs associated with meeting changing electricity demand. For example, increased
cumulative costs of $57 billion and $15 billion through 2050 are projected under RCP8.5 in the ReEDS
and GCAM models, respectively. High costs are also projected in the Northeast, Midwest, and Southern
Plains. The Northern Plains is projected to have the lowest increases in cumulative costs.
Table 16.1. Projected Change in Cumulative System Costs
Values relative to a control scenario without climate change are shown by region and the national
(contiguous U.S.) in billions of discounted (3%) $2015 and percent change. Values represent average of
the five GCMs over the 2015-2050 period (ReEDS and GCAM) and 2015-2100 (GCAM only). Totals may
not sum due to rounding.
2050
2100
ReEDS
GCAM
GCAM
Region
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
$35
5.2%
$30
4.4%
$2.6
1.1%
$1.9
0.8%
$7.4
1.7%
$4.1
0.9%
Southeast
$57
4.3%
$43
3.2%
$15
3.6%
$11
2.7%
$36
4.6%
$21
2.7%
Midwest
$37
4.5%
$29
3.6%
$5.6
1.9%
$4.1
1.4%
$13
2.4%
$7.9
1.4%
Northern Plains
$2.0
1.4%
$2.0
1.3%
$0.2
0.5%
$0.2
0.5%
$0.5
1.4%
$0.3
0.9%
Southern Plains
$27
4.4%
$21
3.4%
$5.7
2.7%
$4.4
2.1%
$16
4.1%
o
T—1
-c/>
2.6%
Southwest
00
T—1
-c/>
4.0%
$15
3.2%
$4.7
2.1%
$3.8
1.7%
$14
3.3%
$8.8
2.0%
Northwest
-$6.1
-6.9%
-$6.2
-7.0%
$1.2
2.2%
$0.9
1.8%
$4.3
5.3%
$2.4
2.9%
National Total
$170
4.0%
$130
3.1%
$35
2.4%
$26
1.8%
$92
3.4%
$55
2.0%
16.5 DISCUSSION
Consistent with findings of the assessment literature,311 the analysis presented in this section shows that
rising temperatures due to climate change are projected to increase demand for electricity and affect
the operation and planning of the power system.312 The results are also consistent with another recent
national-scale study,313 which found that average electricity demand in the residential and commercial
sectors increase by 2.3-4.9% by 2050 under RCP8.5 and 1.2-4.1% under RCP4.5.
It is important to note several limitations of this analysis when interpreting the above results. First, the
two electric power system models only resolve the system to annual or seasonal levels. This coarse
310 When thermo-cooling constraints are added to ReEDS simulations, the cumulative, discounted system costs from 2015 through 2050
increase by up to $10-13 billion nationally.
311 Dell, J., S. Tierney, G. Franco, R. G. Newell, R. Richels, J. Weyant, and T. J. Wilbanks, 2014: Ch. 4: Energy Supply and Use. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 113-129. doi:10.7930/J0BG2KWD
312 Importantly, the analyses described in this Technical Report do not examine implementation of any specific policy.
313 Rhodium Group, 2014: American Climate Prospectus: Economic Risks in the United States. Input to the Risky Business Project. Available
online at http://climateprospectus.org/publications/
126
-------
ELECTRICITY
Electricity Demand and Supply
temporal resolution does not fully capture the potential effects of frequent temperature peaks in
electricity demand that hourly or sub-hourly dispatch models are designed to address. As such, the
present analysis underestimates some impacts of climate change on the power system. Second, the
approach only estimates electricity demand, and does not consider impacts on demand for other fuel
sources used in residential cooling or heating, such as oil, natural gas, or wood. Third, consistent with
the focus of the Technical Report to estimate the impacts of climate change on U.S. sectors, this analysis
does not estimate costs associated with reducing greenhouse gas emissions from the electric power
sector.314 Fourth, costs on power systems and the distribution of electricity (i.e., power interruptions)
due to changes in extreme weather events (e.g., high winds, tropical storms) are not estimated in this
analysis. Finally, although the estimates described in this sector do not include impacts to the electric
power system of changes in cooling water availability for thermo-electric generation, a side-case
modeled only in ReEDS indicates that discounted national system costs from 2015 through 2050 would
increase by 6% to 10% to adapt to water constraints.
314 See McFarland, J., Y. Zhou, L. Clarke, P. Sullivan, J. Colman, W. Jaglom, M. Colley, P. Patel, J. Eom, S. Kim, G. Kyle, P. Schultz, B Venkatesh, J.
Haydel, C. Mack, and J. Creason, 2015: Impacts of rising airtemperatures and emissions mitigation on electricity demand and supply in the
United States: a multi-model comparison. Climatic Change, doi: 10.1007/sl0584-015-1380-8.
127
-------
WATER RESOURCES
17. INLAND FLOODING
17.1 KEY FINDINGS
• The frequency of historical "100-year" floods across the contiguous U.S. is projected to increase over
the 21st century, with approximately twice as many flood events projected under RCP8.5 compared
to RCP4.5 by the end of the century.
• Under RCP8.5, projected annual damages from "100-year" floods approximately double over the
21st century, whereas estimated flood damages under RCP4.5 change only modestly. By 2100, the
difference between projected damages under RC8.5 and RCP4.5 is estimated to be approximately $4
billion per year.
• Changes in flood frequency and increases in associated flood damages are not evenly distributed
geographically through the U.S. The most significant difference between projected inland flood
damages under RCP8.5 and RCP4.5 is in the Southeast, where the difference between the two
scenarios approaches $2 billion per year by the end of the century.
17.2 INTRODUCTION
Extreme precipitation events have intensified in recent decades across most of the U.S., and this trend is
projected to continue.315 Heavier downpours can result in more extreme flooding and increase the risk
of costly damages.316 Flooding affects human health and safety, property, infrastructure, and natural
resources. In the U.S., inland flooding caused over 4,500 deaths between 1959 and 2005 and flood-
related property and crop damages averaged nearly $8.5 billion per year317 from 1981 to 2011.318 The
potential for increased damages is large, given that climate change is projected to continue to increase
the frequency of extreme precipitation events and amplify risks from non-climate factors such as
315 Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates, 2014: Ch. 3: Water
Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J.M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program, 69-112. doi:10.7930/J0G44N6T.
316 Walsh, J., D. Wuebbles, K. Hayhoe, J. Kossin, K. Kunkel, G. Stephens, P. Thorne, R. Vose, M. Wehner, J. Willis, D. Anderson, S. Doney, R. Feely,
P. Hennon, V. Kharin, T. Knutson, F. Landerer, T. Lenton, J. Kennedy, and R. Somerville, 2014: Ch. 2: Our Changing Climate. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 19-67. doi:10.7930/J0KW5CXT.
317 Based on the National Weather Service database, the median value is approximately $4.5 billion per year, while the annual values over this
period range from $500 million to $55 billion.
318 Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates, 2014: Ch. 3: Water
Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J.M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program, 69-112. doi:10.7930/J0G44N6T.
128
-------
WATER RESOURCES
Inland Flooding
expanded development in floodplains, urbanization, and land-use changes. People in flood-prone
regions are expected to be at greater risk of exposure to flood hazards due to climate change.319
17.3 APPROACH
This analysis evaluates how climate change could affect precipitation-driven inland flooding damages in
the contiguous U.S. Catchment hydrology is simulated using the variable infiltration capacity (VIC)
hydrologic model, driven by downscaled precipitation fields from five GCMs under RCP8.5 and RCP4.5.
The VIC model simulates the range of hydrologic processes relevant to generating runoff, including
interception on the forest canopy, evapotranspiration, water storage and melt from snowpack,
infiltration, and direct runoff. For this analysis, VIC-projected runoff is routed through a national-scale
river network using a tool that incorporates both hillslope and river channel processes. See Figure A.ll.l
of the Appendix to this Technical Report for a map showing the average annual maximum flows across
the contiguous U.S. in the reference period (2001-2020). For each of the 10 GCM/RCP combinations in
the hydrologic model output, the time-series of annual maximum flow is then extracted at each of the
approximately 57,000 stream segments in the contiguous U.S. To estimate future flood frequency and
damages through 2100, the analysis compares the full transient of future annual streamflow maxima for
each GCM/RCP combination to the modeled 1% annual exceedance probability (AEP) flood event in the
reference period (i.e., the "100-year flood"). At each stream segment, an ensemble average probability
of exceeding the 1% AEP event in each year is calculated by tabulating the fraction of models
experiencing a flood and smoothing these probabilities over a 20-year moving window. The analysis
then links these time and ensemble-averaged flood probabilities to the assets exposed within each
floodplain to calculate projected annual damages.
Asset damages resulting from 1% AEP floods are calculated using data from a tool under development
for the U.S. Army Corps of Engineers (USACE). The tool compiles all of the 100-year (1% AEP) floodplains
mapped by the U.S. Federal Emergency Management Agency (FEMA) and estimates the distribution of
flood depths within each floodplain using digital topographic data from the U.S. Geologic Survey's
National Elevation Dataset (NED). Floodplain polygons are intersected with residential and commercial
building inventories from FEMA's HAZUS-MH software to estimate the number of buildings exposed to
inundation within each floodplain. Damages to buildings exposed to the 1% AEP flood event are
estimated using published depth-damage functions from the USACE and FEMA, and assuming that
assets are evenly distributed within all urbanized portions of the floodplain/census block intersection
areas. See Figure All-2 of the Appendix to this Technical Report for a map of total expected damages
from a 1% AEP flood event in each 12-digit HUC of the contiguous U.S. The analysis uses a Monte Carlo
approach to estimate both a mean and a variance in regional and national-scale flood damages
throughout the 21st century. One thousand 100-year time-series of flood damages in the CONUS are
simulated using the ensemble average probability of exceeding the 1% AEP event at each node in each
year. For each flood event, expected damages are tabulated within the 12-digit HUC corresponding to
319 Bell, J.E., S.C. Herring, L. Jantarasami, C. Adrianopoli, K. Benedict, K. Conlon, V. Escobar, J. Hess, J. Luvall, C.P. Garcia-Pando, D. Quattrochi, J.
Runkle, and C.J. Schreck, III, 2016: Ch. 4: Impacts of Extreme Events on Human Health. The Impacts of Climate Change on Human Health in the
United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 99-128.
129
-------
WATER RESOURCES
Inland Flooding
that node.320 For more information on the approach and results for flooding damages, please refer to
Wobusetal. (2017).321
17.4 RESULTS
Climate change is projected to increase the frequency of inland flooding in most watersheds of the U.S.
As shown in Figure 17.1, the annual number of 100-year floods across the contiguous U.S. across all five
GCMs averages approximately 500 events from 2000-2020. This average number of floods increases
substantially to approximately 1,300 events per year by 2100 under RCP8.5, with a smaller increase to
approximately 600 events per year under RCP4.5. Climate-driven changes in flood risk are not evenly
distributed across the contiguous U.S. See Figure A.11.3 of the Appendix to this Technical Report for
maps of projected changes in the frequency of historical 100-year flood events based on the full CMIP-5
ensemble results.
320 Data on built assets were taken from FEMA's HAZUS-MH General Building Stock inventory, which provides estimates of the number and
aggregate dollar value of multiple types of residential, commercial, and industrial buildings for each Census block. For the developed portion of
each Census block/floodzone intersection, damage estimates were created using depth-damage functions from USACE and FEMA. A separate
depth-damage function was used for each of 28 different categories of buildings (e.g., residential one-story homes without a basement). Each
depth-damage function describes the percent loss as a function of depth. The depth-damage functions were applied to the aggregate value for
each building category within each NFHL-Census block intersection, using the depth exposure results described above. See Wobus et al. (2017)
for additional detail and citations to these inventories and functions. These estimates represent physical damages to residential, commercial,
and industrial buildings, but do not include the costs of reconstruction or rehabilitation. Further, this analysis is not linked with the Urban
Drainage analysis described in this Technical Report, which quantifies the effect of proactive adaptation measures (best management practices)
in reducing urban runoff entering stormwater systems.
321 Wobus, C., E. Gutmann, R. Jones, M. Rissing, N. Mizukami, M. Lorie, H. Mahoney, and J. Martinich, 2017: Modeled changes in 100 yearflood
risk and asset damages within mapped floodplains of the contiguous United States. Natural Hazards and Earth System Sciences, doi:
10.5194/nhess-2017-152.
130
-------
WATER RESOURCES
Inland Flooding
Figure 17.1. Number of 100-Year Floods
In each plot, black dots are the median value across the five GCMs throughout the contiguous U.S. in
each year of the 21st century, thick blue bars are the middle 50% of models, whiskers extend to the 95th
percentile of values, and dots represent outliers. Thick black lines are five-year moving averages across
all models.
4000
RCP8.5
3500
3000
2500
2000
1500
1000
500
S- 4000
RCP4.5
z 3500
3000
2500
2000
1500
1000
500
2080
2100
2060
2040
2020
2000
Year
Figure 17.2 below shows the full time-series of projected changes in flood damages across the
contiguous U.S. through 2100, generated by combining changes in frequency of flooding at each stream
segment with the asset exposure and damage associated with a 100-year event in each floodplain. As
shown, changes in flood damages broadly mimic changes in flood frequency, with Figure 17.2 also
highlighting the difference in trajectories under the two RCPs. While the RCP8.5 and RCP4.5 pathways
are generally similar through mid-century, the trajectories begin to diverge in the latter half of the 21st
century. Under RCP8.5, estimated annual flood damages increase from approximately $3.0 billion in the
early 21st century to over $8.1 billion by 2100. Projected damages under RCP4.5 increase modestly to
approximately $4.3 billion per year by 2050 and remain at this level through the end of the century.
131
-------
WATER RESOURCES
Inland Flooding
Figure 17.2. Projected National Flood Damages within 100-Year Flood Zones
Thin grey lines represent results across 1,000 simulations of damages under RCP8.5 (light grey) and
RCP4.5 (dark grey). Red and blue lines are means of simulations for the two RCPs. All results represent
the average of the five GCMs.
15
c
o
CO
in
° 10
CD
CT>
ft!
£
rt3
O
TJ
O
O
CU
TJ
5
c
o
2000
2020
2040 2060
Year
2080
2100
RCP8.5
RCP4.5
Similar to changes in flood frequency, the increasing flood damages under RCPS.5 relative to RCP4.5 are
not evenly distributed through the U.S. Figure 17.3 shows the time-series of average annual damages in
each NCA4 region, calculated from the 1,000 Monte Carlo simulations described above. Regional
changes shown in Figure 17.3 are primarily driven by the projected changes in precipitation-based flood
risk. As shown in Figure 17.3 and Table 17.1, the most significant difference between projected flood
damages under the RCPs is in the Southeast, where the difference between the two trajectories
approaches $2 billion per year by the end of the century, though substantial inter-annual variability is
also observed.
132
-------
WATER RESOURCES
Inland Flooding
1/1
c
o
CD
lo
0
CM
V>
iyi
01
CD
ru
Q
Figure 17.3. Average Annual Damages by Region
Results represent the average of the five GCMs under RCP8.5 and RCP4.5.
4
o
4
2
0
2000
2020 2040 2060 2080 2100
Year
¦ Midwest
¦ Northeast
Northern Plains
¦ Northwest
¦ Southeast
Southern Plains
¦ Southwest
1 1 1 1 1 1 1 1
RCP4.5
Table 17.1. Annual Damages from Flooding
Results for 2050 (2040-2059) and 2090 (2080-2099) represent the average annual damages from
flooding of the five GCMs in billions of $2015.
2050
2090
Region
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
$1.6
$0.62
$1.5
$0.95
Southeast
$1.7
$1.8
$3.1
$1.5
Midwest
$0.67
$0.54
$0.69
$0.83
Northern Plains
$0.06
$0.07
$0.06
$0.08
Southern Plains
$0.51
$0.81
$0.90
$0.38
Southwest
$0.45
$0.36
$1.5
$0.41
Northwest
$0.10
$0.13
$0.28
$0.17
National Total
$5.1
$4.3
$8.1
$4.3
133
-------
WATER RESOURCES
Inland Flooding
17.5 DISCUSSION
Increasing risks of flooding associated with climate change are consistent with the findings of the
assessment literature322 and other national scale analyses projecting flooding damages using different
methodologies.323,324 Projecting changes in local flood risk at a national scale can be challenging, and as
such, several important caveats are important to note. First, the downscaling method used in
developing the climate projections for this analysis was designed in part to improve GCM resolution of
historical precipitation, but this method also introduces artifacts associated with the transition from the
historical to future period (see Wobus et al. 2017 for more detail). An improved representation of
precipitation extremes would improve confidence in the results. Second, the method is limited by
available data on assets exposed to inland flooding. Because the 1% AEP floodplains are the only flood
risk zones consistently mapped at a national scale, the approach only tabulates damages within these
mapped floodplains, therefore missing damages in other areas, including pluvial (direct rainfall-driven)
flooding. Third, this approach does not estimate flood damages from more frequent floods (e.g., 10- and
50-year events), nor does it account for the fact that larger flood events above the historical "100-year"
event threshold will cause more significant damages than smaller ones. These factors combined signify
that the results presented in this section are likely underestimates to total potential damages. Fourth,
the estimated damages do not include impacts on human health or economic disruption, which are
likely to increase the total economic impacts from flooding. Finally, the approach does not account for
future adaptations to protect against changing flood risk; nor does it account for population growth or
increasing development within flood risk zones. Future demographic and infrastructure changes could
either increase or decrease damages from flooding in the future: flood protection could decrease
damages,325 while increases in development in the floodplain could increase them. Without reasonable
means of predicting future flood development and protection changes, the analysis assumes that
floodplain development and protection will on average remain static.
322 Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates, 2014: Ch. 3: Water
Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J.M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program, 69-112. doi:10.7930/J0G44N6T.
323 Das, T., Maurer, E. P., Pierce, D. W., Dettinger, M. D., and Cayan, D. R., 2013: Increases in flood magnitudes in California under warming
climates, Journal of Hydrology, 501, 101-110.
324 Wobus, C., M. Lawson, R. Jones, J. Smith, and J. Martinich, 2013: Estimating monetary damages from flooding in the United States under a
changing climate. Journal of Flood Risk Management, 7, 217-229, doi: 10. Ill 1/jf r3.12043.
325 Climate change amplification of flood risk may also trigger proactive adaptation, such as people choosing not to move into flood-prone
areas, or where zoning or insurance markets present clear barriers to moving there. These dynamics are not captured in this analysis.
134
-------
WATER RESOURCES
Water Quality
18. WATER QUALITY
18.1 KEY FINDINGS
• Climate change is projected to have negative impacts on water quality in the U.S. as measured by
water temperature, dissolved oxygen, phosphorous, and nitrogen levels. Water quality is projected
to decrease across a large majority of the country under all models, scenarios, and time periods, but
particularly under RCP8.5 in the Northeast, Southeast, and Midwest.
• Important regional differences are observed across the two water quality models used in the
analysis, owing primarily to differences in model structure. Despite these regional differences, there
is good agreement in changes of overall water quality, both in terms of the direction and magnitude
of climate impacts.
• By 2090, national water quality damages are estimated at $4.3-4.8 billion per year under RCP8.5,
and $2.6-3.3 billion annually under RCP4.5.
18.2 INTRODUCTION
Climate change is likely to have far-reaching effects on water quality in the U.S. due to increases in river
and lake temperatures and changes in the magnitude and seasonality of river flows, both of which will
affect the concentration of water pollutants. Rising water temperatures, reduced lake mixing, and
increased biotic consumption of dissolved oxygen each reduce water quality. These physical impacts on
water quality will have potentially substantial economic impacts, since water quality is valued for
drinking water and recreational and commercial activities such as boating, swimming, and fishing.326,327
The analysis presented in this section estimates the economic value associated with changes in the
quality of recreation opportunities (e.g., swimming, boating, fishing), but does not quantify health
effects.
18.3 APPROACH
This analysis estimates climate change effects on water quality at the eight-digit HUC scale of the
contiguous U.S. using two biophysical models: Hydrologic and Water Quality System (HAWQS) and US
Basins. HAWQS advances the functionality of the widely used and accepted Soil and Water Assessment
Tool (SWAT), providing a platform for water quality modeling, primarily by minimizing the necessary
initialization time.328 This improves the ease of application to national scale analyses while still
simulating a large array of watershed processes for a defined period of record. Originally developed by
the U.S. Department of Agriculture (USDA), SWAT has been the core simulation tool for numerous U.S.
national and international assessments of soil and water resources. The model follows a broad modeling
326 Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates, 2014: Ch. 3: Water
Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J.M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program, 69-112. doi:10.7930/J0G44N6T.
327 Luber, G., K. Knowlton, J. Balbus, H. Frumkin, M. Hayden, J. Hess, M. McGeehin, N. Sheats, L. Backer, C. B. Beard, K.L. Ebi, E. Maibach, R. S.
Ostfeld, C. Wiedinmyer, E. Zielinski-Gutierrez, and L. Ziska, 2014: Ch. 9: Human Health. Climate Change Impacts in the United States: The Third
National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G.W. Yohe, Eds., U.S. Global Change Research Program, 220-256.
doi:10.7930/J0PN93H5.
328 Yen, H., P. Daggupati, M.J. White, R. Srinivasan, A. Gossel, D. Wells, and J.G. Arnold, 2016: Application of large-scale, multi-resolution
watershed modeling framework using the hydrologic and water quality system (HAWQS). Water, 8, 164, doi:10.3390/w8040164.
135
-------
WATER RESOURCES
Water Quality
sequence: (1) the landscape phase, where the primary processes are climate, soil water balance,
nutrient and sediment transport and fate, land cover, plant growth, farm management, and (2) the main
channel phase, where the main processes are river routing, sediment and nutrient transport through the
rivers and reservoirs.
US Basins is a linked water systems and water quality model designed to evaluate the impacts of climate
change on water quantity and quality outcomes.329 Precipitation and temperature from each climate
scenario are inputs into: (a) a rainfall-runoff model (CLIRUN-II), which is used to simulate monthly
runoff; and (b) a water demand model, which projects the water requirements of the municipal and
industrial (M&l) and agriculture sectors. With these runoff and demand projections, a water resources
systems model produces a time series of reservoir storage, release, and allocation to the various
demands in the system, which include M&l, agriculture, transboundary flows, and hydropower. The
water quality model (QUALIDAD330) uses managed flows and reservoir states from the water resources
systems model to simulate a number of water quality constituents in rivers and reservoirs. Since US
Basins does not include a representation of loading transport across the landscape, nonpoint loadings
from the HAWQS landscape (phosphorus, nitrogen, and biological oxygen demand) are used directly in
US Basins, equally distributed across each segment within the eight-digit HUC.
Each water quality model projects changes in water quality parameters, along with simulated changes in
river flow, which are projected for five climate models under RCP8.5 and RCP4.5, with future municipal
wastewater treatment plant loadings (point source) scaled to account for population growth.331 Changes
in overall water quality are estimated using changes in a Climate-oriented Water Quality Index (CWQI), a
metric that combines multiple pollutant and water quality measures. Four water quality parameters
(water temperature, dissolved oxygen (DO), total nitrogen, and total phosphorus) are aggregated from
the eight-digit HUC level to the Level-Ill Ecoregions, weighted by area.332 Finally, a relationship between
changes in the CWQI and changes in the willingness to pay (WTP) for improving water quality is used to
estimate the economic implications of projected water quality changes. For more information on the
approach and results for the water quality sector, please refer to Fant et al. (20 1 7),333 Boehlert et al.
(2015), and Yen etal. (2016).
18.4 RESULTS
Climate change will lead to the warming of rivers, lakes, and reservoirs across the country. Since water
temperature is primarily driven by changes in air temperature, this water quality parameter shows the
most similarity between the two water quality models (Figure 18.1).334 The differences in spatial
patterns that can be observed between the models are due to how future temperatures are simulated
329 Boehlert, B., K.M. Strzepek, S.C. Chapra, C. Fant, Y. Gebretsadik, M. Lickley, R. Swanson, A. McCluskey, J. Neumann, and J. Martinich, 2015:
Climate change impacts and greenhouse gas mitigation effects on US water quality. Journal of Advances in Modeling Earth Systems, 7, 1326-
1338. This paper includes a detailed flow diagram showing the suite of nested models involved in this analysis.
330 Chapra, S.C., 2014: QUALIDAD: A parsimonious modeling framework for simulating river basin water quality, Version 1.1, Documentation and
user's manual. Civil and Environmental Engineering Dept., Tufts University, Medford, MA.
331 Key differences between the HAWQS and US Basins modeling frameworks are described in Fant et al. (2017).
332 Designed to serve as a spatial framework for environmental resource management, ecoregions denote areas within which ecosystems (and
the type, quality, and quantity of environmental resources) are generally similar. Ecoregions were originally created to support the
development of regional biological criteria and water quality standards, and to set management goals for nonpoint source pollution.
333 Fant, C., R. Srinivasan, B. Boehlert, L. Rennels, S.C. Chapra, K.M. Strzepek J. Corona, A. Allen, and J. Martinich, 2Q17\ Climate change impacts
on US water quality using two models: HAWQS and US Basins. Water, 9, 118, doi: 10.3390/w9020118
334 Fant et al. (2017) describe the main differences in the reference period water quality parameters from both models, which can affect
comparisons of future projections, and identifies the main causes for these differences.
136
-------
WATER RESOURCES
Water Quality
in the two approaches and the influence of different projected changes in runoff and flow on water
temperature.
Since DO is largely influenced by temperature through levels of DO saturation (i.e., higher temperatures
reduce DO saturation levels, thereby reducing DO aeration),335 decreases in DO are generally projected
in the future. In both water quality models, there are consistent decreases in DO in the East (Figure
18.1). HAWQS shows large decreases in the Northeast, Southeast, and coastal regions of the Pacific
Northwest, with areas of increases in the Midwest. Conversely, US Basins shows the largest decrease in
DO in the Midwest. In both water quality models, changes in DO are largest for 2090 compared to 2050
and larger under RCP8.5 than RCP4.5.
Figure 18.1 shows percent change in total nitrogen concentrations. Both water quality models show
increases in total nitrogen in the south-central U.S. and the area around the Great Lakes. HAWQS shows
large increases in total nitrogen in the Southwest, while US Basins shows decreases in this region. As for
changes in total phosphorus concentrations, HAWQS projects increases in phosphorus levels along the
Southwest coast and Texas. In contrast, US Basins projects larger increases in the central U.S., especially
in 2090, as well as in the East. Differences in nitrogen and phosphorus concentration changes can be
explained primarily by the differences in flow changes for these two water quality models.
Reflecting changes in the underlying water quality parameters discussed above, Figure 18.2 shows
changes in CWQI across the contiguous U.S. in 2050 and 2090 under the two RCPs. The CWQI serves as a
measure of water quality; the higher the CWQI, the higher the water quality. Importantly, water quality
is projected to decrease across a large majority of the country under all models, RCPs, and time periods.
For both water quality models, changes in CWQI are more pronounced in the Northeast, Southeast, and
Midwest regions of the country, primarily due to loadings and temperature being higher in this area. For
HAWQS, changes in CWQI are largest along the East coast and in parts of the Southeast, although this
pattern can also be seen in US Basins under RCP8.5. Compared to HAWQS, US Basins tends to show
larger decreases in CWQI in the central U.S. and around the Great Lakes.
335 DO is also influenced by changes in nitrogen, phosphorus, and BOD loadings, as well as changes in flow.
137
-------
WATER RESOURCES
Water Quality
Figure 18.1. Change in Water Quality Parameters
The maps show the change in water quality parameters under RCP8.5 in 2050 (2040-2059) and 2090
(2080-2099) relative to the reference period (1986-2005). Results for each eight-digit HUC represent the
average values across the five GCMs, and are aggregated to the Level-Ill Ecoregions.
Temperature
Dissolved
Oxygen
Nitrogen
Phosphorus
HAWQS US Basins
2050 2090 2050 2090
% change
% change
% change
<123456789> change in °F
138
-------
WATER RESOURCES
Water Quality
Figure 1S.2. Changes in Mean Water Quality Index
Projected values for 2050 (2040-2059j and 2090 (2080-2099) represent changes relative to the reference
period (1986-2005) across the contiguous U.S. The results are averages across the five GCMs, and are
aggregated to the Level-Ill Ecoregions. For reference, the water quality index is based on a 100-point
scale.
HAWQS
RCP8.5 RCP4.5
2050
2090
US Basins
RCP8.5 RCP4.5
< -24 -18 -12 -6 0 6 12 18 24 >
change in water quality index
Total WTP is shown in in Table 18.1 for ail five GCMs, RCPs, time periods, and the two water quality
models. In all GCMs and water quality models, WTP decreases most for RCPS.5 compared to RCP4.5.
Under RCP8.5, water quality damages for the five-GCM average are estimated at $4.3-4.8 billion
annually by 2090, and $2.6-3.3 billion under RCP4.5 (with values solely representing recreational
impacts). The largest changes in WTP are shown in the HadGEM2-ES GCM (on average, the hottest GCM
modeled), with the least projected under GISS-E2-R (the coolest, comparatively). Table 18.2 presents
WTP results across the NCA4 regions. Under results from both water quality models, projected damages
are largest in the Northeast, Southeast, and Midwest, though substantial damages are also estimated
for the Southwest under HAWQS.
139
-------
WATER RESOURCES
Water Quality
Table 18.1. National Water Quality Damages to Recreation
The table presents estimated damages in 2050 (2040-2049) and 2090 (2080-2099). Results represent the
national willingness to pay for improving water quality each (in billions $2015) for all five GCMs under
RCP8.5 and RCP4.5.
HAWQS
US Basins
2050
2090
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
$2.0
$1.6
$5.3
$3.1
$1.9
$1.6
$4.4
$2.6
CCSM4
$1.6
$1.2
$4.8
$3.1
$1.5
$1.2
$3.9
$2.3
GISS-E2-R
$1.4
$1.1
$3.6
$2.6
$1.3
$0.8
$3.2
$1.7
HadGEM2-ES
$2.8
$2.2
$5.7
$4.2
$2.5
$2.0
$5.2
$3.5
MIROC5
$2.0
$1.7
$4.8
$3.6
$2.2
$1.9
$4.8
$3.1
5-GCM Average
$2.0
$1.6
$4.8
$3.3
$1.9
$1.5
$4.3
$2.6
Table 18.2. Regional Water Quality Damages to Recreation
The table presents estimated damages in 2050 (2040-2049) and 2090 (2080-2099). Results are shown in
billions of $2015 per year, and represent the average of the five GCMs for each RCP. Values may not sum
due to rounding.
HAWQS
US Basins
2050
2090
2050
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
$0.47
$0.40
$1.1
$0.75
$0.39
$0.33
$0.88
$0.55
Southeast
$0.45
$0.35
$1.5
$1.0
$0.53
$0.39
$1.3
$0.85
Midwest
$0.32
$0.25
$0.61
$0.33
$0.45
$0.37
$0.89
$0.51
Northern Plains
$0.05
$0.04
$0.10
$0.05
$0.06
$0.05
$0.12
$0.08
Southern Plains
$0.20
$0.14
$0.55
$0.35
$0.24
$0.19
$0.52
$0.35
Southwest
$0.39
$0.33
$0.86
$0.67
$0.18
$0.15
$0.43
$0.24
Northwest
$0.07
$0.06
$0.17
$0.12
$0.05
$0.03
$0.11
$0.06
National Total
$2.0
$1.6
$4.8
$3.3
$1.9
$1.5
$4.3
$2.6
18.5 DISCUSSION
Projections that climate change will decrease river and lake water quality are consistent with the
findings of the assessment literature,336 and with a previous analysis focused on specific U.S. watersheds
to characterize the sensitivity of streamflow, nutrient, and sediment loading to a range of climate and
urban development scenarios.337 The analysis presented in this section of the Technical Report
336 Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates, 2014: Ch. 3: Water
Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G.
W. Yohe, Eds., U.S. Global Change Research Program. doi:10.7930/J0G44N6T.
337 U.S. EPA. Watershed Modeling to Assess the Sensitivity of Streamflow, Nutrient, and Sediment Loads to Potential Climate Change and Urban
Development in 20 U.S. Watersheds (Final Report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-12/058F, 2013.
140
-------
WATER RESOURCES
Water Quality
contributes to the literature on water quality impacts by using two impact models to project long-term
changes. Differences in these water quality projections between the two models point to dissimilarities
in their structure and inherent bias of each water quantity and quality modeling component. As these
are complex systems modeled over large geographic areas, inconsistencies in the outcomes of the two
models are expected. Despite regional differences in projected water quality parameters between the
two models, good agreement in water quality index was observed, both in terms of direction and
magnitude of climate impacts.
Decreases in water quality due to climate change will likely have adverse effects on human health and
the environment that are not represented in the results of this section. For example, climate change
impacts to water quality may affect ecological dynamics of freshwater systems, with cascading effects
on ecosystem services and recreational opportunities.338 Also, this analysis only considers four water
quality parameters, and omits other constituents, such as sediment and heavy metals, that may be
affected by changes in the climate system. In addition, the methods underlying the analysis do not
consider the effects of climate change-induced extreme events on water quality, such as increased
siltation and runoff following wildfire events. Finally, the analysis considers only a subset of all use/non-
use values linked to water quality changes, therefore the damages reported here are likely
underestimates.
Simulation results illustrate a high degree of variability in the response of different streamflow and water quality attributes to climate change
throughout the nation. Results also illustrate sensitivity to methodological choices such as different approaches for downscaling global climate
change simulations and use of different watershed models.
338 Please seethe Harmful Algal Blooms section of this report, which projects changes in bloom activity and resulting effects on recreation.
141
-------
WATER RESOURCES
Municipal and Industrial Water Supply
19. MUNICIPAL AND INDUSTRIAL WATER SUPPLY
19.1 KEY FINDINGS
• Climate change is projected to alter precipitation patterns across the contiguous U.S., when
combined with changes in temperature, will likely cause water shortages in some areas while
increasing water availability in others.
• Estimated annual welfare losses at a national level are projected to be $320 million per year by 2090
under RCP8.5 and $210 million under RCP4.5, compared to the no-climate change control scenario.
• The effects of climate change on municipal and industrial water supply and demand vary over space.
Cumulative discounted regional welfare impacts through 2100 range from a loss of $1.2 billion in the
Southern Plains to a gain of $540 million in the Southeast under RCP8.5.
19.2 INTRODUCTION
Climate change is projected to affect municipal and industrial (M&l) water supply and demand due to
changing temperatures and precipitation patterns.339 Water supplied for industrial uses can include a
broad range of activities, such as process water for manufacturing processes, industrial cleaning
operations, cooling, and food processing. In this study, the industrial category excludes uses for cooling
of thermoelectric power plants, which are discussed in the Electricity Demand and Supply section.
Municipal (sometimes called "domestic") uses of water include two broad categories: outdoor uses
(primarily for landscaping) and indoor uses (for bathing, cleaning, cooking, drinking, and sanitation).
Indoor uses tend to require a lower volume on an annual basis, but are of higher economic value to
consumers compared to outdoor uses. For example, when municipal water supplies run low, outdoor
municipal uses are often the first to be curtailed. In 2010, public water supply withdrawals for municipal
uses accounted for about 8 percent of all water withdrawals in the U.S.340 Industrial uses (excluding
thermoelectric cooling and mining) account for about 4 percent of all water withdrawals. Because of its
high economic value, water supplied to the M&l sector is often the last sector to experience water
shortages (as opposed to commercial agricultural irrigation, for example). The high value of M&l water
also means that when unmet demands do occur, the economic impacts of these shortages tend to be
large. The impacts of climate change on water supply and demands for sectors other than M&l, namely
hydropower and irrigation are captured in the Electricity Demand and Supply and Agriculture sections of
this Technical Report.
19.3 APPROACH
This analysis estimates the welfare loss to consumers of water associated with unmet M&l water
demands (i.e., shortages) from climate change in the contiguous U.S. To develop these estimates, the
analysis defines demand functions for three categories of M&l water demand: municipal indoor use,
municipal outdoor use, and industrial use. Each demand is defined by a constant elasticity of demand
339 Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates, 2014: Ch. 3: Water
Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J.M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program, 69-112. doi:10.7930/J0G44N6T..
340 Maupin, M.A., J.F. Kenny, S.S. Hutson, J.K. Lovelace, N.L. Barber, and K.S. Linsey, 2014: Estimated use of water in the United States in 2010.
U.S. Geological Survey, Circular 1405. Available online at https://dx.doi.org/10.3133/cirl405
142
-------
WATER RESOURCES
Municipal and Industrial Water Supply
where elasticity varies by category (outdoor elasticity = -0.6; indoor elasticity = -0.2; industrial elasticity=
-0.4).341,342
Total demand data values are derived from the U.S. Geological Survey (USGS) National Water-Use
Science Project.343 These data are used for baseline water demand, as well as proportions used for
allocating demand and unmet demand among use categories. Return flow fractions from the Water
Supply Stress Index (WaSSI) Ecosystem Services Model from the USDA Climate Change Resource Center
are used to assign unmet demand based on consumptive use.344 All projected demands are scaled using
population projections from EPA's Integrated Climate and Land-Use Scenarios (ICLUS) Version 2
model.345 Unmet demands at the eight-digit HUC level are distributed among the three categories of
demand based on a prioritization scheme where shortages are first expected to be absorbed by outdoor
use. If unmet demand is greater than landscape demand, the remaining shortage is absorbed by
industry. Finally, any remaining unmet demand after both outdoor and industry uses have been
exhausted is taken from indoor municipal demand.346
Surface water supply estimates, and the water balance of supply and demand for all sectors, are
generated using the US Basins model. US Basins characterizes the effects of climate change and water
infrastructure on flow volumes, as well as water balance across demand sectors, at the eight-digit HUC
level. The US Basins system employs GCM projections of precipitation and temperature as inputs to a
rainfall-runoff model, which is used to simulate monthly runoff in each of the eight-digit HUCs, and a
water demand model which projects the water requirements of the M&l and agriculture sectors. With
these runoff and demand projections, the US Basins water resources systems model produces a time
series of reservoir storage, release, and allocation to the various demands in the system, which include
341 The demand function is ln(Q) =oc —s x ln(P). For examples of previous uses of this demand function for water see: Henderson, J., C.
Rodgers, R. Jones, J. Smith, K. Strzepek, and J. Martinich, 2015: Economic impacts of climate change on water resources in the coterminous
United States. Mitigation and Adaptation Strategies for Global Change, 20, 135-157; and Cai, X., C. Ringler, and M.W. Rosegrant, 2006:
Modeling water resources management at the basin level. Methodology and application to the Maipo River Basin. Research Report 149.
International Food Policy Research Institute, Washington, DC.
342 Elasticities derived from ranges provided in: 1) Renzetti, S., 1992: Estimating the structure of industrial water demands: the case of Canadian
manufacturing. Land Economics, 396-404; 2) Espey, M., J. Espey, and W.D. Shaw, 1997: Price elasticity of residential demand for water: a meta-
analysis. Water Resources Research, 33, 1369-1374; and 3) Dalhuisen, J.M., R.J.G.M. Florax, H.L.F. de Groot, and P. Nijkamp, 2003: Price and
income elasticities of residential water demand: a meta-analysis. Land Economics, 79, 292-308.
343 U.S. Geological Survey, cited 2017: The National Water-Use Science Project. Available online at http://water.usgs.gov/watuse/wunwup.html
344 U.S. Department of Agriculture, cited 2017: WaSSI Ecosystem Services Model. Available online at https://www.wassiweb.sgcp.ncsu.edU/s
345 EPA, 2017: Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (Iclus) (Version
2). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-16/366F. Available online at
https://cfpub.epa.gov/ncea/iclus/recordisplav.cfm?deid=322479
346 Before calculating economic welfare loss, several eight-digit HUCs are removed from consideration in the analysis because the underlying
model calculating unmet demand is unable to effectively simulate complexities of the water system supplies in those locations, either due to
water connections (i.e., interbasin transfers) that are not modeled or to storage and groundwater access contingency plans for water supply in
densely populated areas. This includes all HUCs with unmet demand in the reference period of over 20 percent, all HUCs adjacent or one-
removed from a Great Lake, and any HUC with the majority of its area in a metropolitan statistical area (MSA) with a population of over 1
million people (where back up plans for unmet demand would be in place). A sensitivity analysis was run where the last group (HUCs with
majority of area in an MSA) was included in the results. The five GCM average cumulative welfare losses were $55 billion and $37 billion in
RCP8.5 and RCP4.5, respectively, about a fifteen-fold increase over the results presented in this assessment. While these areas are likely to
incur some costs due to water shortages, it is unlikely that they would seethe levels of unmet demand originally modeled that yield these large
damage results due to existing emergency water source planning typical in highly populated areas.
143
-------
WATER RESOURCES
Municipal and Industrial Water Supply
M&l, agriculture, environmental flows, transboundary flows, hydropower, and others. US Basins is
described more fully in Boehlert et al. (20 1 5).347,348
Welfare losses are calculated at the eight-digit HUC level for each of the five GCMs and two RCPs, and a
control scenario, which includes changes in demand due to population and per capita water use but no
changes in the climate. Welfare change is estimated using M&l water supply prices from over 300
utilities nationwide,349 with a willingness-to-pay cap set at five times the price. Final welfare loss
estimates due to climate change for each scenario are calculated as the difference in consumer surplus
loss between the control and climate scenario summed across the three demand categories.350
19.4 RESULTS
Future unmet M&l demands are a function of both climate change impacts on water supply and
socioeconomic changes affecting demand (e.g. population growth, population shifts, and per capita
water demand). The results of this analysis measure the incremental impacts of climate change only.
National average unmet demand projections do not change dramatically due to climate change (a 0.16%
change in unmet demand is projected annually by 2090 under RCP8.5, 5-model average). However, even
small amounts of unmet demand, often concentrated in certain areas and time periods, can produce
large welfare losses because of the high value of M&l water. As seen in Table 19.1, by 2090 the national
annual average welfare loss estimates due to unmet M&l water demands for the five-GCM average is
$320 million, with estimates for each model ranging from $190 million to $640 million under RCP8.5.
The corresponding 2090 five-GCM average is $210 million for RCP4.5, reflecting individual model results
that range from a gain of $9.0 million to a loss of $410 million. As would be expected, the GCMs
projecting generally wetter futures (CanESM2 and GISS-E2-R) show lower damages than those projecting
generally drier futures (MIROC5 and HadGEM2-ES).
Table 19.1. National Average Annual Welfare Loss
Results for 2030 (2020-2039), 2050 (2040-2059), 2070 (2060-2079), and 2090 (2080-2099) are shown in
millions of $2015. Totals may not sum due to use of two significant figures and rounding.
2030
2050
2070
2090
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
$120
$94
$68
$96
$170
$89
$220
$140
CCSM4
$100
$80
$100
$71
$140
$180
$190
$220
GISS-E2-R
-$22
$58
$26
$27
$160
$38
$190
-$9
HadGEM2-ES
$95
$120
$240
$150
$390
$180
$340
$310
MIROC5
$81
$61
$150
$240
$540
$310
$640
$410
5-GCM Average
$76
$83
$120
$120
$280
$160
$320
$210
347 Boehlert, B., K. M. Strzepek, S. C. Chapra, C. Fant, Y. Gebretsadik, M. Lickley, R. Swanson, A. McCluskey, J. E. Neumann, and J. Martinich,
2015: Climate change impacts and greenhouse gas mitigation effects on U.S. water quality. J. Adv. Model. Earth Syst., 7, 1326-1338,
doi:10.1002/2014MS000400.
348 In most of the 8-digit HUCs (or basins), the demands and sources are assumed to be within the basin. However, the water systems model
does include all major basin transfers (totaling over 100 transfers), the most significant of which are from the Colorado River to southern
California. M&l supply is driven by the water demand in each HUC, which is taken from the U.S. Geological Survey's "Estimated Use of Water in
the United States in 2005" report (Kenny et al. 2009). Municipal and industrial demands were aggregated from among the numerous categories
of demands provided by Kenny et al. Changes in M&l overtime are projected based on population changes and estimates of changes in
consumption rates. See Boehlert et al. (2015) for more detail on these data and methods. Reference: Kenny, J. F., N. L. Barber, S. S. Hutson, K. S.
Linsey, J. K. Lovelace, and M. A. Maupin (2009), Estimated use of water in the United States in 2005, U.S. Geological Survey Circular 1344, 52 p.
349 Water prices from: American Water Works Association and Raftelis Financial Consultants. 2015. 2014 Water and Wastewater Rate Survey.
350 Aflat supply curve is assumed; therefore, no producer surplus is lost due to supply shortages. The consumer surplus loss is the value (price
times quantity) of the quantity demanded less the quantity supplied.
144
-------
WATER RESOURCES
Municipal and Industrial Water Supply
Welfare impacts also vary by region, as changes in both future demands and climates vary over space.
Figure 19.1 presents the average annual welfare loss by 2050 for the 5-model average under RCP8.5 by
four-digit HUC. Lost welfare is concentrated in the Southwest and Southern Plains, while the Northern
Plains and the Northwest see welfare gains. The magnitude of the losses, however, outweighs the gains
at the national scale.
Figure 19.1. Average Annual Welfare Losses
Results for 2050 (2040-2059) and 2090 (2080-2099) represent the five-GCM average and are aggregated
to the four-digit HUC scale.
RCP 8.5
RCP 4.5
Welfare Loss ($2015)
-$100,000 to -$100
$100 to $500
| $20,000 to $500,000
| greater than $1,000,000
| less than - $100.000
I I -$100 to $100
n $500 to $20,000
| $500,000 to $1,000,000
I NCA Regions
Through the end of the century, climate change is projected to result in $3.7 billion and $3.0 billion in
national welfare loss under RCP8.5 and RCP4.5, respectively (2015-2099, $2015, five GCM average,
discounted at 3%). The impacts vary by region; as shown in Table 19.2, the Southeast and Northwest
regions are projected to see cumulative welfare gains under both RCPs. Welfare losses are largest in the
Midwest, Southern Plains, and Southwest regions. As noted earlier, these results are consistent the
precipitation projections for each of the five GCMs.
145
-------
WATER RESOURCES
Municipal and Industrial Water Supply
Table 19.2. Regional Cumulative Welfare Losses
Results are shown in millions of $2015 for the period 2015-2099 (discounted at 3%). Totals may not sum
due to rounding.
Region
RCP8.5
RCP4.5
Northeast
$730
$510
Southeast
-$540
-$580
Midwest
$1,200
$980
Northern Plains
$37
00
LO
-u~y
Southern Plains
$1,200
$860
Southwest
$1,200
$1,200
Northwest
-$12
-$17
National Total
$3,700
$3,000
19.5 DISCUSSION
The incremental effect of climate change on M&l water supply and demand averaged across the country
may appear small in absolute terms, however communities with insufficient water to meet household
demands will see large negative welfare impacts. It is important to note that climate change is one of
several factors that affect water shortages, particularly in the municipal sector. Factors such as
increased population also affect supplies. The estimates presented in this section reflect total unmet
demands for the M&l sector of between 0.87 and 2.0 percent of total demand by 2090 depending on
the GCM and scenario. The analysis projects that without climate change, there would be an unmet
demand of 0.75 percent due to population increases. Climate change therefore acts as a threat
multiplier, exacerbating existing trends toward increased water shortages.
This analysis also does not consider changes in M&l demand that may be associated with climate
change. It is possible that in drier future climates, shortages would be even greater than estimated due
to an increase in water demand, particularly for outdoor municipal uses due to reduced soil moisture
levels associated with more arid climates, but also for some industrial uses where uncovered pond
storage may be less efficient owing to higher evaporation rates.351
The national annual welfare losses projected in this study are consistent with estimated impacts in 2050
from another study estimating welfare impacts of water supply and demand.352,353 By excluding areas
with water systems not captured in the water supply model, this assessment is a lower-bound estimate
of the total impacts of climate change on M&l water supply and demand. While city residents may not
ultimately be faced with unmet demands and associated welfare losses because of back up water supply
plans, there will be a cost to water utilities of implementing these designs.
In general, the analysis does not consider the effectiveness of adaptation. One form of adaptation would
be on the supply side, in the form of infrastructure construction to provide more resiliency (e.g.,
enhanced options for interbasin transfers, enhanced emergency storage, more extensive use of
groundwater sources, conservation initiatives, or even desalination plants). While these adaptive
responses could reduce the damages reported in this section, implementing some of these options
351 However, accelerated evaporation rates from larger reservoir infrastructure are reflected in the US Basins supply calculations.
352 Henderson, J., C. Rodgers, R. Jones, J. Smith, K. Strzepek, and J. Martinich, 2015: Economic impacts of climate change on water resources in
the coterminous United States. Mitigation and Adaptation Strategies for Global Change, 20, 135-157.
353 The estimates for other years in Henderson et al. 2015 diverge from the estimates found in this analysis, likely due to the water allocation
optimization methods used in the earlier study.
146
-------
WATER RESOURCES
Municipal and Industrial Water Supply
would likely require large infrastructure investments and could interact with water right compacts in
some areas of the country. A second form of adaptation would be on the demand side, in the form of
behavioral changes to reduce demand, or changes in industrial processes to less water-intensive forms
of industry. Some of these patterns of demand change represent existing trends in water demand that
are reflected in the demand projections, but these trends could accelerate (or slow) over the long time
periods considered here. Both supply- and demand-side adaptation could in turn reduce the
vulnerability of the M&l sectors to future water shortages.
147
-------
WATER RESOURCES
Winter Recreation
20. WINTER RECREATION
20.1 KEY FINDINGS
• Climate change is projected to shorten the season lengths for downhill skiing, cross-country skiing,
and snowmobiling across the country, especially under RCP8.5. Under RCP8.5 projections for 2090,
the median closing date of downhill ski resorts nationally is more than a month earlier compared to
the reference period, moving from early April to the end of February. However, a large amount of
regional variability is projected, with the most significant reductions in season length occurring in
the upper Midwest and the Northeast, and the smallest reductions occurring in the central Rockies
and Sierras.
• Shorter season lengths will lead to reduced winter recreational opportunities at a national level.
Under RCP8.5, national downhill ski visits are projected to decrease considerably from
approximately 56 million in 2013 to 31 million by 2090. Ski visits under RCP4.5 are projected to
decrease slightly to 53 million by 2090.
• These reductions in downhill skiing visits result in an estimated $2.0 billion in annual damages (lost
revenue) by 2090 under RCP8.5. Reductions in annual cross-country skiing and snowmobiling visits
by 2090 result in $10 million and $5.5 million under RCP8.5, respectively.
• By 2090 under RCP4.5, the national monetized impact on downhill skiing is positive, totaling $130
million in benefits even though visits are modestly reduced compared to the reference period -
reflecting both adverse impacts from climate change and increased visits due to population growth.
20.2 INTRODUCTION
Warmer temperatures and changes in precipitation are expected to decrease the duration and extent of
natural snow cover across the northern hemisphere. If there are fewer days in the winter season with
sufficient snow, key components of the winter recreation industry will face challenges. While
improvements in snowmaking technology can offset declines in natural snowfall at many ski resorts,
continued warming temperatures are likely to delay opening dates with resulting effects on recreation.
In addition, adaptive responses, such as snowmaking, are less effective for cross-country skiing and
snowmobiling than for downhill skiing. Climate change-related threats to winter recreation could affect
tens of millions of visits annually, and is expected to have important implications in areas where these
industries are a substantial part of the economic activity of the region.
20.3 APPROACH
This analysis estimates the impacts of climate change on three types of winter recreation in the
contiguous U.S.: downhill skiing and snowboarding, cross-country skiing, and snowmobiling. The analysis
uses the Utah Energy Balance (UEB) model, a water and energy balance model which tracks snow-water-
equivalent (SWE), internal energy of the snowpack, and snow surface age in its simulation of snow
accumulation and melt. The model simulates natural snow accumulation and snowmelt at 247 winter
recreation sites across the contiguous U.S. using site-specific climatic and topographic characteristics of
each site. At each of the 247 modeled locations, results include a simulation of natural snowpack for the
20-year reference period (1986-2005), and future simulations for the 20-year periods centered on the
reporting years of 2050 and 2090 for each of five GCMs under RCP8.5 and RCP4.5. Snowpack was
modeled at the bottom and top elevation of each location.
148
-------
WATER RESOURCES
Winter Recreation
For downhill skiing and snowboarding, results from the natural snow model are combined with
projections of the resorts' abilities to make artificial snow in the future. These calculations are based on
a tabulation of cumulative hours with temperatures cold enough to make snow (approximately 28°F wet
bulb temperature). Because ski resorts typically need between 400-500 hours of snowmaking conditions
to open for the season, the date at which this number of hours is reached is calculated under reference
and future climate conditions (30-year periods). Opening date for downhill skiing and snowboarding was
calculated as the first date with 10 cm SWE from the UEB model, or the date to reach 450 hours of
snowmaking conditions, whichever comes earlier. The closing date was estimated as the last day with 10
cm SWE. Changes in season length for each of the recreation types are used to estimate changes in
visits, which are then monetized. For the reference period, recreational activity levels for downhill
skiing, cross-country skiing, and snowmobiling, as well as average ticket prices at ski facilities and entry
fees at national forests, are derived from a combination of National Ski Areas Association (NSAA) and
U.S. Forest Service National Visitor Use Monitoring (NVUM) data. Future recreational activity levels are
scaled into the future to account for population growth using state-level ICLUS population projections
described in Modeling Framework section of this Technical Report. For more information regarding the
approach used in this analysis to estimate climate change impacts on winter recreation, please refer to
Wobusetal. (2017).354
20.4 RESULTS
Projected changes in the length of the winter season for downhill skiing reflect the combined influence
of early season temperature changes, which modify resorts' ability to make snow, and the changes in
precipitation and temperature throughout the ski season, which control the water and energy balance
that drives the natural snowpack. Figure 20.1 shows that, in general, climate change is projected to
substantially shorten average downhill skiing season lengths at the majority of sites across in the U.S.
However, the results range from increases in a very limited number of areas (under RCP4.5, 10 sites in
2050 and 4 sites in 2090; under RCP8.5, 6 sites in 2050 and none in 2090, with a maximum value of 83%
increase relative to the reference period), to declines of more than 80% under RCP8.5 in many locations.
The projected changes in season length presented in Figure 20.1, which represent the average of the
five GCMs, are most dramatic in the Northeast and upper Midwest, and less dramatic in higher-
elevation areas of the Rockies and Sierras. The few locations with increases in season length in 2050 are
located in the upper Midwest. These increases in season length are driven primarily by projected
increases in precipitation, which offset projected increases in temperature by mid-century. However,
these effects are not as significant by the end of the century.
For most downhill skiing locations, being open for the Christmas and New Year's holidays is critical to
remaining profitable and staying in business given the disproportionate share of revenue and profit
realized in this period (note that other critical periods are the Presidents Day to Martin Luther King Day
period). Earlier closing dates could also impact skier visits during the Spring Break period, which typically
falls in March. Figure 20.2 shows that climate change is projected to have a larger impact on closing
dates than opening dates across the combinations of RCPs and the two future time-periods considered.
In the most extreme reductions, observed for the RCP8.5 projections for 2090, the median closing date
is more than a month earlier compared to the reference period, moving from early April to the end of
February. In contrast, the largest shift in the start date from the reference period, which also occurs
under RCP8.5 in 2090, involves a shift of several weeks from the initial beginning of December to the
354 Wobus, C., E.E. Small, H. Hosterman, D. Mills, J. Stein, M. Rissing, R. Jones, M. Duckworth, R. Hall, M. Kolian, J. Creason, and J. Martinich.
2017. Projected climate change impacts on skiing and snowmobiling: A case study of the United States. Global Environmental Change, doi:
10.1016/j.gloenvcha.2017.04.006.
149
-------
WATER RESOURCES
Winter Recreation
end of the month. Figure 20.3 provides the detailed summary of the season length projections for Aspen
Mountain in Colorado. As shown, projected season lengths shorten over the century and under RCP8.5,
though considerable variability is observed across the different GCMs, especially under RCP8.5 in 2090.
Figure 20.1. Average Percent Change in Downhill Ski Season Length
Estimates are based on a combination of water balance modeling and snowmaking results. Projected
values for 2050 (2040-2059) and 2090 (2080-2099) represent changes relative to the reference period
(1986-2005) across the contiguous U.S.
RCP8.5
RCP4.5
2050
2090
J m i H
I . * •
• * \
A * • ' *.
}
v
}.* M
« .
• * \
A •• •*.
• • *' t
M ' •
A# M
• «*•
•
• •
...r
* V
£i/
• A'
; *
I • w
: • # # /•
»• •
•
T #.
w
•
\
/
(
v\
i
/•' *•
• « • •
• • •
• # •»
• * • • «
•
• •
* V
A
• • \
X • • •>,
• • • • •»
M • •
•
, r »
T .
• |
>
/
J
\ \
% \
J
* ...&
- *
• • • v
>
r
V» v
* Pt!
>0%
150
-------
WATER RESOURCES
Winter Recreation
Figure 20.2. National Average Season Start and End Dates for the Downhill Ski Season
Results represent the five-GCM average for a given RCPfor 2050 (2040-2059) and 2090 (2080-2099)
relative to the reference period (1986-2005). The box and whiskers represent the distribution of results
for the downhill ski season. Mean start dates are represented by red lines at bottom of the boxes and
mean close dates are represented by the red lines at top of the box.
01-Jul
01-Jun
01-May
01-Apr
01-Mar
01-Feb
01-Jan
01-Dec
01-Nov
01-Oct
Across all modeled locations, average annual changes in cross-country skiing and snowmobiling season
lengths across the GCMs range from small increases at some locations, to declines of more than 80% at
others (Figure 20.4). In general, the most significant reductions in season length occur in the upper
Midwest and the Northeast, and the smallest reductions occur at locations in the central Rockies and
Sierras. Under RCP8.5, a substantially large fraction of the modeled locations is projected to experience
average annual reductions from their reference period season length of > 80% compared to the RCP4.5
estimates.
Reference RCP4.5 2050 RCP8.5 2050 RCP4.5 2090 RCP8.5. 2090
151
-------
WATER RESOURCES
Winter Recreation
Figure 20.3. Change in Season Length for Downhill Skiing at Aspen Mountain, CO
Results represent each GCM results for a given RCP for 2050 (2040-2059) and 2090 (2080-2099) relative
to the reference period (1986-2005) for the Aspen Mountain, CO location. The box and whiskers
represent the distribution of results for the downhill ski season length in number of days, with mean
season lengths represented by the red line.
200
<~>
ro 150
O
cn
g 100
_L
_L
c
o
I/•>
RCP 8.5 2050
JL -L
_L
JL
RCP 4.5 2050
-i i i i
_L
c
o
CO
fU
a>
6
i
_L
o
_L
RCP 8.5 2090
_L
l | |
S T -
JL
RCP 4.5 2090
5}
l/l
O
73
o
n
73
fT>
CD
3
O
n
G}
u\
00
O
73
o
r>
152
-------
WATER RESOURCES
Winter Recreation
Figure 20.4. Average Percent Change in Annual Cross-Country Skiing and Snowmobiling Season
Lengths
Results represent the five-GCM average for a given RCP for 2050 (2040-2059) and 2090 (2080-2099)
relative to the reference period (1986-2005) across the contiguous U.S.
RCP8.5 RCP4.5
yt' V. A ¦ 't' *1 »T-
/ • * ?*. "i «l*Jl * J *' *'| I •*"'i .
4 *;* * - - . . Jul* / * *.¦* v# [— . r*o<3;
• • • 1fw\ ' *i»' • • '
^ 2.**. " * * /**• *.*». * * . . ¦*.« >
2050 ,
d V I
:fl T\i> (H 14,0 TztV>5t
j
M
^ vi vr \>
V ^ —7—T-^ A
ql »J/ l '• 1
«» .
A • ¦ ' /
•
* f
i' *
h
•* * f n Tr ,¦ */
4 lis wrTv^
•• r ¦ -l /- zlj
4
TTy sVAX
\ A \ s f
\f \) i
>0%
In this approach, future population growth increases the anticipated number of winter recreation
participants and visits, which helps counteract adverse impacts of climate change. For all three
recreation types, Table 20.1 shows the regional change in visits and dollars under RCP8.5 and RCP4.5,
accounting for population growth and climate change. Under RCP8.5, national downhill ski visits
decrease considerably after adjusting for changes in climate and population from approximately 56
million in 2013 to 31 million by 2090 (monetized impact equivalent to $2.0 billion in damages). Ski visits
under RCP4.5 are projected to decrease slightly to 53 million by 2090, with an equivalent monetary
impact of $130 million.355
For cross-country skiing and snowmobiling activity, projected recreational visits under RCP8.5 decrease
in 2050 and 2090, even with increased participation due to population growth. Under RCP4.5, cross-
355 By 2090 under RCP4.5, the national monetized impact on downhill skiing is positive, even though the change in visits is modestly negative.
Although visits decrease at a national level under RCP4.5 in 2090, visits increased in the Southwest and Northern Plains regions where lift
tickets are high, driving the monetized impacts slightly positive. Those regions include major ski areas states (for example, California, Colorado,
and Utah are in the Southwest region and Montana and Wyoming are in the Northern Plains region) with high lift ticket prices of approximately
$107-124 per visit, as opposed to approximately $60 to $80 per visit for the other regions.
153
-------
WATER RESOURCES
Winter Recreation
country skiing visits increase slightly in 2050 and 2090, while snowmobiling visits decrease slightly in
2050 and increase slightly in 2090.
Table 20.1. Changes in Winter Recreational Visits and Economic Damages under Climate Change
Results represent the five-GCM average. Visits reported as millions of visits and monetized impacts are
reported as millions of $2015, undiscounted. Totals may not sum due to rounding.
Winter
Recreational
Activities
Reference/2015
Visits Dollars
(millions) ($millions)
2050
RCP8.5 RCP4.5
Change Monetized Change Monetized
in visits impact in visits impact
2090
RCP8.5 RCP4.5
Change Monetized Change Monetized
in visits impact in visits impact
Downhill
skiing
56
$5,400
-11
-$780
-6.8
-$350
-25
-$2,000
-3.2
$130
Cross-country
skiing
3.6
$32
-0.09
-$1.1
0.22
$1.6
-1.1
O
1
-oo-
1
0.63
$5.2
Snowmobiling
2.8
$13
-0.32
-$1.4
-0.11
-$0.53
-1.2
-$5.5
0.08
$0.35
Total
62
$5,400
-12
-$780
-6.7
-$340
-28
-$2,000
-2.5
$130
20.5 DISCUSSION
Physical models accounting for changes in the amount and duration of natural snow and ski resorts'
ability to make snow demonstrate that the available time for winter recreation activities will decline at
nearly all sites in the contiguous U.S. under climate change. Sites at higher elevations, such as the Rocky
Mountains and Sierras, tend to be less affected by projected changes in temperature and precipitation,
whereas sites at lower elevations, generally in the upper Midwest and New England, are more sensitive
to climate change. These findings are consistent with a number of studies that have examined how
climate change could influence seasonal snowpack in the western U.S.356,357 These results also build on
prior work related to winter recreation by combining the geographic breadth of previous studies,358 with
the detail that has historically been applied only to site- or regionally-specific studies.359,360 However, the
physical model was necessarily simplified in order to model nearly 250 specific sites across the U.S.
More site-specific analyses could improve the model depiction of snowpack at any individual resort, but
the computational demands of such an analysis would have been prohibitive at a national scale.
Several important caveats regarding the winter recreation analysis are summarized in this section. First,
pressure on downhill ski resort operators from a sequence of short seasons or seasons with poor-quality
conditions could possibly result in permanent closure. Since this analysis did not project potential ski
area closures, the estimates of downhill skiing visits are conservative. Second, this analysis does not fully
account for the different types of substitution that could arise with climate change (e.g., a switch from
skiing to mountain biking), therefore conclusions about the net effects to recreational activity cannot be
drawn. Finally, the industries supporting winter recreation are already experienced in addressing
356 Mote, P.W., A.F. Hamlet, M.P. Clark, and D.P. Lettenmaier, 2005: Declining mountain snowpack in western North America. Bulletin of the
American Meteorological Society, 86, 39.
357 Pierce, D.W. and D.R. Cayan, 2013: The uneven response of different snow measures to human-induced climate warming. Journal of Climate,
26, 4148-4167.
358 Burakowski, E. and M. Magnusson, 2012: Climate Impacts on the WinterTourism Economy in the United States. Available online at
https://www.nrdc.org/sites/default/files/climate-impacts-winter-tourism-report.pdf
359 Lazar, B., and M. Williams, 2008: Climate change in western ski areas: Potential changes in thetiming of wet avalanches and snow quality for
the Aspen ski area in the years 2030 and 2100. Cold Regions Science and Technology, 51, 219-228.
360 Dawson, J. and D. Scott, 2013: Managing for climate change in the alpine ski sector. Tourism Management, 35, 244-254.
154
-------
WATER RESOURCES
Winter Recreation
variable winter conditions and could further implement innovative adaptive approaches that to provide
greater resiliency than what is represented here.
155
-------
AGRICULTURE
21. DOMESTIC YIELD AND WELFARE EFFECTS
21.1 KEY FINDINGS
• Climate change is projected to result in decreases in national yields for most major agricultural crops
under both RCP8.5 and RCP4.5 through 2100. In general, larger declines are estimated under RCP8.5
than under RCP4.5. However, the direction and magnitude of crop yield response to climate change
also varies by geographic region.
• As a result of declining yields, crop prices generally rise in the future under both RCPs compared to a
no-climate change control scenario, with larger projected price increases under RCP8.5.
• Total cropland area increases to help meet demand in the face of falling yields, but national
production and consumption of agricultural commodities are still projected to decline.
• Higher crop prices and lower production and consumption lead to generally negative effects on total
economic welfare under both RCPs, with moderately larger losses under RCP8.5.
21.2 INTRODUCTION
The U.S. has a robust agriculture sector that produces nearly $330 billion per year in agricultural
commodities.361 The sector ensures a reliable food supply and supports job growth and economic
development. In addition, as the U.S. is currently the world's leading exporter of agricultural products,
the sector plays a critical role in the global economy.362 Agricultural production is highly sensitive to
climate conditions. Climate change will alter the spatial and temporal distribution of temperature and
precipitation as well as the frequency and severity of extreme events, such as flooding and drought.
These changes are likely to affect future agriculture productivity, and may lead to increased variability in
yield. Changes in potential yields will shift land allocation, crop mix, and production practices
throughout the U.S. These changes will affect commodity and production prices and the level of
production and consumption of these goods.
361 Hatfield, J., G. Takle, R. Grotjahn, P. Holden, R. C. Izaurralde, T. Mader, E. Marshall, and D. Liverman, 2014: Ch. 6: Agriculture. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 150-174. doi:10.7930/J02Z13FR.
362 U.S. Congress, 2013: The Economic Contribution of America's Farmers and the Importance of Agricultural Exports. United States Congress,
Joint Economic Committee, Vice Chair Amy Klobuchar. Available online at http://www.iec.senate.gov/public/ cache/files/266a0bf3-5142-4545-
b806-ef9fd78b9c2f/iec-agricu lture-report.pdf
156
-------
AGRICULTURE
Domestric Yield and Welfare Effects
21.3 APPROACH
This analysis estimates the effects of climate change on crop productivity in the contiguous U.S., and
then simulates changes in market and economic welfare outcomes in the U.S. agriculture sector. To
simulate the effects of climate change on crop productivity, the Environmental Policy Integrated Climate
(EPIC) model363,364 is used to simulate changes for eight crops: corn, soybean, wheat, alfalfa hay,
sorghum, cotton, rice, and barley. Yield potential is simulated for each crop for both rainfed and
irrigated production.365 The EPIC model captures heterogeneous yield response to climate (including
temperature, precipitation, relative humidity, wind speed, and solar radiation), which can vary
depending on regional climate, soil type, irrigation status, and C02 levels. Because production regions
may change over time in response to climate change, EPIC simulates potential cultivation and
production in areas within 100 kilometers (62 miles) of historical production regions. EPIC is driven by
changes in future climate projected by five GCMs under both RCP8.5 and RCP4.5, with comparisons to
yields in the reference period of 1986-2005. The EPIC results presented in this section include the effect
of C02 fertilization on crop yields; a sensitivity analysis of the effect of C02 fertilization on the crop yield
results from EPIC are provided in the Appendix A.12. However, the yield projections presented in this
Technical Report are based solely on one crop model. In addition, EPIC does not simulate the adverse
effects from pests, disease, and ozone, and damage due to changes in the occurrence of storms, such as
flooding, tornadoes, and hurricanes. Inclusion of these impacts on crop yields would likely result in
larger adverse effects from climate change. For more information on the methods of EPIC modeling
described in this section, please refer to Beach et al. (2015).366
The Forest and Agricultural Sector Optimization Model with Greenhouse Gases (FASOM-GHG)367,368 was
used to estimate changes in market and economic welfare outcomes in the U.S. agriculture sector.
FASOM-GHG is driven by changes in potential yield from EPIC for each of the five GCMs under the two
RCPs, and calculates changes in other crops based on the most relevant proxies. FASOM-GHG simulates
future potential landowner decisions regarding crop mix and production practices, and projects the
allocation of land over time to competing activities in the agricultural sector and the associated impacts
on commodity markets.369 Given the changes in potential yields projected by EPIC, FASOM-GHG uses an
optimization approach (of the variables listed in the previous sentence) to maximize consumer and
363 Williams, J.R., 1995: The EPIC Model. In Computer Models in Watershed Hydrology, V.P. Singh (ed.), pp. 909-1000. Highlands Ranch, CO:
Water Resources Publication.
364 Thomson, A.M., R.A. Brown, N.J. Rosenberg, R.C. Izaurralde, and V. Benson, 2005: Climate Change Impactsforthe Conterminous USA: An
Integrated Assessment. Part 3: Dryland production of grain and forage crops. Springer Netherlands, doi:10.1007/1-4020-3876-3.
365 The EPIC simulations assume that crops can be irrigated to a level that eliminates water stress. A particular concern for climate change is
that in areas where the need for irrigation is greatest due to reduction in precipitation, the supply of water for irrigation will also be reduced. To
fully consider this risk requires integration of crop modeling with hydrologic modeling for projections of future water supply, which was not
modeled in this biophysical crop yield analysis.
366 Beach, R., Y. Cai, A. Thomson, X. Zhang, R. Jones, B. McCarl, A. Crimmins, J. Martinich, J. Cole, and B. Boehlert, 2015: Climate change impacts
on US agriculture and forestry: benefits of global climate stabilization. Environmental Research Letters, 10, doi: 10.1088/1748-
9326/10/9/095004.
367 Beach, R., D. Adams, R. Alig, J. Baker, G. Latta, B. McCarl, B. Murray, S. Rose, and E. White, 2010: Model documentation forthe Forest and
Agricultural Sector Optimization Model with Greenhouse Gases (FASOMGHG): Draft Report. Prepared for U.S. Environmental Protection
Agency. Available online at: http://agecon2.tamu.edu/people/facultv/mccarl-
bruce/papers/1959FASQMGHG%20Model%20Documentation PR Aug2010.doc
368 Importantly, the agriculture-only version of FASOM-GHG was used in this analysis. As such, the interacting effects between agriculture and
forestry, including the movement of land from forest to agriculture and vice versa, are not captured here. See Beach et al. (2015) for an
example of a FASOM-GHG analysis looking at both the agriculture and forestry sectors.
369 FASOM-GHG is an intertemporal optimization model, which means decisions today are made with expectations of future potential
conditions, including net returns and climate change effects, and thus can optimize near term land owner behavior. The model also simulates
changes in agricultural commodities beyond those modeled in EPIC (eight crops).
157
-------
AGRICULTURE
Domestric Yield and Welfare Effects
producer surplus over time, which are the measures of economic welfare reported in the Results
section.370,371 The model is constrained such that total production is equal to total consumption, total
U.S. land use remains constant, and non-climate drivers in the agriculture sector are consistent between
the scenarios to isolate the effect of climate change.372 Although the EPIC simulations assume that crops
can be irrigated to a level that eliminates water stress, the FASOM-GHG simulations include shifts in
water availability for irrigation based on data obtained from the water balance framework described in
the Water Supply and Demand section of this report. Finally, this analysis does not reflect climate
change impacts on international agriculture, which would also affect domestic relative returns to
different uses of land and trade patterns and therefore affect land use decisions. These potential
impacts are separately discussed in the following section of this report. For more information on the
approach to using FASOM-GHG for estimating market and economic welfare impacts, please refer to
Beach et al. (2015).373
21.4 RESULTS
Climate change is projected to have an overall adverse impact on the productivity of U.S. agriculture
sector. Figure 21.1 presents the projected percent change in national crop yields through 2100
compared to reference yields under RCP8.5 and RCP4.5 (average results of the five climate models). For
all major crops, with the exception of wheat,374 unmitigated climate change under RCP8.5 is projected to
result in lower yields by the end of the century compared to reference yield rates (though cotton yields
are higher than the reference until just before the end of the century). Importantly, the projected
magnitude of this effect increases with time, suggesting that higher levels of climate change increase the
adverse effects to crop yields. Yields under RCP4.5 decline relative to the reference period for most
crops, except for cotton, wheat, and sorghum. With the exception of hay and wheat, projected yields
under RCP4.5 show smaller declines compared to those estimated for the higher forcing scenario.
Compared to RCP8.5, RCP4.5 is projected to have a significantly smaller negative effect on the future
yields of barley, corn, cotton, and rice. See Appendix A.12 for crop yield results comparing simulations
with and without C02 fertilization.
370 FASOM-GHG is optimized to maximize consumer and producer surplus in the base, but re-adjusts production and consumption patterns to
re-optimize in response to changes in potential yields.
371 Consumer and producer surplus are used to estimate impacts on total economic welfare. Consumer surplus is the monetary gain obtained by
consumers because they are able to purchase a product for a price that is less than the highest price that they would be willing to pay. Producer
surplus or producers' surplus is the amount that producers benefit by selling at a market price that is higher than the least that they would be
willing to sell for.
372 In addition, the analysis assumes no price incentives for avoiding GHG emissions or maintaining or increasing carbon sequestration in the
agriculture sector (i.e., the sector does not participate in the global GHG mitigation assumed under RCP4.5).
373 Beach, R., Y. Cai, A. Thomson, X. Zhang, R. Jones, B. McCarl, A. Crimmins, J. Martinich, J. Cole, and B. Boehlert, 2015: Climate change impacts
on US agriculture and forestry: benefits of global climate stabilization. Environmental Research Letters, 10, doi: 10.1088/1748-
9326/10/9/095004.
374 This finding is consistent with other recent reports (Marshall et al. 2015) which project an overall positive response of climate change on
wheat yields in the contiguous U.S. However, estimated yields vary by region, based on the timing and magnitude of projected changes in
temperature and precipitation. Importantly, crop model inter-comparisons have shown that projected changes in yield can vary considerably,
and grow more variable over space and time. For example, Asseng et al. (2014) show a general trend towards decreasing wheat yields in the
U.S. based on an ensemble mean of crop models. References: (1) Asseng, S., R. Ewert, P. Martre, R.P. Rotter, and D.B. Lobell, 2014: Rising
temperatures reduce global wheat production. Nature Climate Change, 5,143-147, doi: 10.1038/NCLIMATE2470, and (2) Marshall, E., M.
Aillery, S. Malcolm, and R. Williams, 2015: Climate Change, Water Scarcity, and Adaptation in the U.S. Fieldcrop Sector. United States
Department of Agriculture, Economic Research Service. Economic Research Report No. ERR-201. Available online at
https://www.ers.usda.gov/publications/pu b-details/?pubid=45496
158
-------
AGRICULTURE
Domestric Yield and Welfare Effects
Figure 21.1. Projected Percent Change in National Crop Yields
Results shown represent the average of the five GCMs under RCP8.5 and RCP4.5 compared to the
reference period (1986-2005). Results are weighted averages of the individual irrigated and rain fed
values from the EPIC model.
BARLEY
-30 -
-40
-50 -
20
10
0
-10
-20
-30
-40
-50
20
10
0
-10
-20
-30 -
-40 -
-50
CORN
RICE
7T00
¦MOO
20
10
0
-10
-20
-30
-40
-50
20
10
0
-10
-20
-30
-40
-50
20
10
0
-10
-20
-30
-40
-50
HAY
2050
SOYBEAN
2050
WHEAT
2050
2100
2100
20
10
0
-10
-20
-30
-40
-50
20
10
0
-10
-20
-30
-40
-50
SORGHUM
COTTON
2050
2100
RCP8.5
RCP4.5
Figure 21.2 shows the projected change in national yield under RCP8.5 for the three largest U.S. crops
(by area and production volume, not including hay) under the five different climate models, along with
the ensemble average. In general, there is agreement in the direction of yield effects across the GCMs,
although the magnitude of change varies by climate model and crop. In addition, the magnitude of
change, whether positive or negative, increases over time in almost all cases. The largest change from
reference yields is projected under the HadGEM2-ES model, which is the hottest model used in this
analysis, with the exception of wheat where yield changes under this GCM are the most positive.
159
-------
AGRICULTURE
Domestric Yield and Welfare Effects
Figure 21.2. Projected Percent Change in Yield
Results shown represent projections by climate model and the average of the five GCMs under RCP8.5
compared to the reference period (1986-2005).
30
20 -
2
£ 10
~ 0
bO
| -io
i! -20
c
91
2 -30
01
-40
-50
30
20
Barley
2030 2050 ¦ 2070 ¦ 2090
II-OI
Sorghum
2030 2050 ¦ 2070 ¦ 2090
> J-U
c
'Z 0
(20
m -10
" -20
c
V
a -30
-40
-50
i ¦
H
30
20
Soybean
2030 2050 ¦ 2070 ¦ 2090
c
0
-10
~ -20
-30
-40
-50
1
I"
30
20
1 0
BO
£ -10
~ -20
c
SI
a -30
-40
-50
Rice
2030 ¦ 2050 ¦ 2070 ¦ 2090
V
# ^ J? ,f
30
20
2
£ io
•i 0
60
| -io
~ -20
c
S
a -30
a
-40
-50
30
20
~ o
bp
-10
" -20
c
2090
t, fa I Lfc
Wheat
2030 2050 ¦ 2070 ¦ 2090
i j
ff
&
.# J?
160
-------
AGRICULTURE
Domestric Yield and Welfare Effects
Spatial differences in projected climate change across the regions of the contiguous U.S. drive variability
of crop yield response. Using the example of corn yields, Figure 21.3 shows the percent change by
region under the two forcing scenarios, with uncertainty bands showing the range in response across
climate models. As shown, yields under RCP8.5 are projected to decline considerably in the Midwest,
Southeast, and Southern Plains. More modest losses are estimated for the Northeast, Northern Plains,
and Southwest, with the Northwest showing moderate increases by mid-century and declines
thereafter.
Figure 21.3. Percent Change in Projected Corn Yields by NCA Region
The graphs show the percent change in projected corn yields under RCP8.5 and RCP4.5 from 2006-2100
relative to the reference period (1986-2005). The bolded lines represent the five model average and the
shaded cones reflect the minimum and maximum annual values from among the projected values across
the five GCMs.
NORTHWEST
NORTHEAST
20
NORTHERN PLAINS
20
10
MIDWEST
10
10
r
20
2050
2100
-10,
2050
21<
2050
LOO
10
-10
-10
-20
-20
-10
-30
,-20
SOUTHWEST
-40
20
10
20
SOUTH
2050
2100 20
SOUTHERN PLAINS
Ttr
10
-2«
-30
-10
-10
-40
-20
-30
-40
RCP8.5 —
RCP4.5
Figure 21.4 shows the projected change in total national acreage by crop under differing climate change
scenarios.375 In response to falling yields due to climate change, results show a general increase in total
crop acreage relative to the no-climate change control scenario, as producers expand agricultural land to
376 As an intertemporal model, FASOM-GHG makes land use decisions today are made with expectations of future potential conditions,
including future climate change effects, and it therefore can optimize near term land owner behavior. The model also simulates changes in
agricultural commodities beyond the eight crops modeled in EPIC and shown in Figure 21.4.
161
-------
AGRICULTURE
Domestric Yield and Welfare Effects
meet demands for these crop commodities.376 Across crop varieties, acreage devoted to growing barley,
hay, corn, and soybeans increase the most under RCP8.5, with small decreases in acreage for sorghum,
wheat, and, through 2050, cotton. Similar patterns of change are observed under RCP4.5, though the
magnitude of changes in acreage is generally smaller in the later parts of the century.
Figure 21.4. Average Percent Change in Total Acreage across the Eight Crop Types
Results, shown in millions of acres, are relative to a no-climate change control scenario.
60
50
„ 40
C?
J= 30
u
5 20
u
k.
0)
10
(10)
60
50
40
30
0)
GO
C
fZ
-C
U
a* 20
u
i-
-------
AGRICULTURE
Domestric Yield and Welfare Effects
century (particularly 2020-2035) is the very large and rapid reduction in hay yields during this period.
Not only does hay experience substantially more negative yield impacts than the other crops simulated
using EPIC, but the majority of the yield reductions happen much earlier. As a result, land is being
diverted from other crops to hay production, thus limiting the supply and driving up the price of wheat,
potatoes, rice, barley, and other crops. As the yield reductions for hay begin to flatten out, the price
impacts start to fall.377 Prices then tend to increase again later in the century as the yield impacts for
other major crops become more negative. Finally, these changes in land allocation, crop mix, and
production practices in turn affect GHG emissions from agriculture. Figure A.12.2 in Appendix A.12
shows the estimated changes in cumulative GHG emissions from the agriculture sector.
Figure 21.5. Percent Change in Crop Price
Values represent national changes in all crops aggregated into a single index, relative to the no-climate
change control scenario.
13
11
9
7
5
3
1
2025
2050
2100
1
3
5
GCM
RCP8.5
RCP4.5
CanESM2
CCSM4
GISS-E2-R
HadGEM2-ES
MIROC5
5-Model Average
The changes in crop prices and the level of production and consumption of agricultural products have
important implications for the economic welfare of consumers and commodity producers. The FASOM-
GHG model estimates these effects through changes in consumer and producer surplus, as summarized
in Table 21.1. As shown in Table 21.1, climate change under both RCPs is projected to result in
substantial decreases in total economic welfare (well-being) in the agriculture sector through 2100, with
greater losses under RCP8.5. The cumulative, discounted (3%) decrease in total welfare is estimated at
$190 ($160-260) billion through 2100 under RCP8.5, and $180 ($160-230) billion under RCP4.5.
However, decreases in crop yield and the resulting price increases result in cumulative gains in producer
surplus through the end of the century.378 But as indicated by the total welfare estimates, the declines in
377 FASOM-GHG incorporates assumptions about yield improvements taking place overtime. All climate change yield impacts are implemented
as relative changes in yields compared with these projections. Because yields are improving in the control scenario, there is a general trend
towards reduced pressure on land resources overtime. There is also more flexibility to reallocate land and change production practices farther
into the future. Other things being equal, climate change impacts of a given magnitude that occur earlier in the century will be more difficult for
the agricultural sectorto adjust to and will result in larger price increases.
378 Falling crop yields and resulting rising crop prices also drive up livestock prices and reduce production and consumption of livestock
products, which is part of the overall decline in economic welfare from the agriculture sector as modeled in FASOM-GHG.
163
-------
AGRICULTURE
Domestric Yield and Welfare Effects
projected consumer surplus are larger than the increases in producer surplus, resulting in negative
welfare effects overall.
Table 21.1. Cumulative Change in Welfare
Values represent the cumulative change in welfare across the 2010-2100 period relative to the reference
period (1986-2005), and are discounted at 3% using billions of $2015. Totals may not sum due to
rounding.
Consumer Surplus
Producer Surplus
Total
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
-$290
-$260
$120
to
00
-$180
-$170
CCSM4
-$280
-$260
$92
$99
-$180
-$170
GISS_E2_R
-$230
-$240
00
to
$89
-$160
-$160
HadGEM2_ES
-$440
-$350
$190
$120
-$260
-$230
MIROC5
-$270
-$280
$97
$93
-$180
-$190
5-GCM Average
-$300
-$280
$110
$97
-$190
-$180
21.5 DISCUSSION
The results of this sectoral analysis are consistent with published studies focused on agricultural
impacts. Projections of adverse yields resulting from unmitigated climate change, large regional
differences in crop response, increasing vulnerability of water supplies for irrigation, and the ability of
adaptation (via crop mix and land use management changes) to reduce adverse effects are consistent
with the findings of the major climate science assessments.379 In addition, the yield projections are
generally consistent with findings of other recent analyses.380,381
Several important limitations of this analysis should be noted. First, the methodology does not reflect
climate change impacts on international agriculture and related price and trade effects, which would
also affect relative estimated returns to different uses of land and trade patterns and therefore affect
land use decisions.382 For example, incorporating negative impacts from climate change on yields in the
rest of the world would tend to drive up global prices and make U.S. exports more competitive.
Implications of international effects on U.S. agriculture are explored in the following section of this
Technical Report. Second, the use of just one crop process model and one market model, both of which
contain their own structural uncertainties, should be noted given the importance of these general
uncertainties raised in recent model inter-comparisons.383 Third, as mentioned above, this analysis does
not consider interactions with and resulting market or GHG emissions effects from the forestry sector,
which could impact net results. Fourth, several uncertainties in the FASOM-GHG analysis remain
regarding issues such as future potential changes in crop technology, energy and land use policies, and
other interactions that could affect market outcomes. Finally, this analysis omits important aspects of
379 Hatfield, J., G. Takle, R. Grotjahn, P. Holden, R. C. Izaurralde, T. Mader, E. Marshall, and D. Liverman, 2014: Ch. 6: Agriculture. Climate Change
Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global
Change Research Program, 150-174. doi:10.7930/J02Z13FR.
380 Marshall, E., M. Aillery, S. Malcolm, and R. Williams, 2015: Climate Change, Water Scarcity, and Adaptation in the U.S. Fieldcrop Sector. U.S.
Department of Agriculture, Economic Research Service, ERR-201.
381 EPA, 2015: Climate Change in the United States: Benefits of Global Action. United States Environmental
Protection Agency, Office of Atmospheric Programs, EPA 430-R-15-001.
382 Leclere, D., P. Havlik, S. Fuss, E. Schmid, A. Mosnier, B. Walsh, H. Valin, M. Herrero, N. Khabarov, and M. Obersteiner, 2014: Climate change
induced transformations of agricultural systems: insights from a global model. Environmental Research Letters, 9, 124018, doi: 10.1088/1748-
9326/9/12/124018.
383 Rosenzweig, C., J. Elliot, D. Deryng, A.C. Ruane, C. Muller, A. Arneth, K.J. Boote, C. Folberth, M. Glotter, N. Khabarov, K. Neumann, F. Piontek,
T.A. Pugh, E. Schmid, E. Stehfest, H. Yang, and J.W. Jones, 2013: Assessing agricultural risks of climate change in the 21st century in a global
gridded crop model intercomparison. Proceedings of the National Academy of Sciences, 111, 3268-73, doi: 10.1073/pnas.1222463110.
164
-------
AGRICULTURE
Domestric Yield and Welfare Effects
climate change impacts to agriculture, including damages from extreme weather events, wildfire, and
changes in weeds, pests, disease, and ozone damage. Collectively, these effects would likely result in
larger yield losses than those estimated in this section.
165
-------
AGRICULTURE
U.S. and Global Agriculture Interactions
22. U.S. AND GLOBAL AGRICULTURE
INTERACTIONS
22.1 KEY FINDINGS
• Climate change impacts are projected to result in increased prices and decreased production and
consumption for corn, soy, and wheat under all scenarios when compared with the control. Under
RCP8.5, projected U.S. yields are lower, production declines are steeper, and prices rise more
sharply than under RCP4.5 by 2050.
• Climate-driven changes to global agriculture will affect the U.S. agriculture market. The net
difference in impacts between the U.S.-only and global scenarios is relatively small for corn and
soybeans, but U.S. wheat output under the global scenarios results in a significantly smaller net
change in crop area and total production, and higher net price impacts due to the relatively smaller
global market share the U.S. commands for wheat than for corn or soybeans.
22.2 INTRODUCTION
This assessment offers a secondary climate impacts analysis for the U.S. agricultural sector, building
upon the primary analysis described in the previous section using the FASOM-GHG modeling framework.
While FASOM-GHG provides substantial detail on U.S. production systems and has endogenous trade
flows, it does not have detailed global components and instead, holds agricultural supply functions fixed
in the rest of the world. By assuming fixed global supply, a domestic market model like FASOM-GHG can
respond to future domestic productivity changes by adjusting imports and exports of key commodities.
This model component can act as a buffer to reduce the net welfare impacts of the exogenous domestic
productivity changes. However, this approach does not recognize the impacts of climate change on
agricultural productivity in the rest of the world, nor the resulting effects on competitiveness,
international trade, and global markets. Thus, a more comprehensive summary of climate change
impacts on the U.S. agriculture sector necessitates an understanding of global interactions.
22.3 APPROACH
This analysis evaluates climate impacts on U.S. agriculture with and without accounting for climate
change impacts to agriculture in the rest of the world. This evaluation is accomplished through a
scenario design that first isolates several climate change scenarios and crop yield impacts to the U.S.
only, followed by scenarios that extend the climate impacts to the rest of the world using the Global
Biosphere Management Model (GLOBIOM), a detailed partial equilibrium model of the global
agriculture, forestry, and bioenergy sectors. GLOBIOM partitions the world into 30 economic regions, in
which a representative regional consumer optimizes consumption depending on income, preferences,
and product prices. On the production side, producers maximize their margins and the model solves for
a market equilibrium corresponding to overall welfare maximization.
GLOBIOM is combined with a global version of the Environmental Policy Integrated Climate (EPIC) crop
model to calculate the impact of climate change on the agricultural sector. EPIC is used to simulate
yields for each global location, management practice, and climatic scenario. For this study, EPIC was
applied to estimate the biophysical and environmental parameters of 18 crops for three different types
of management systems (low input rainfed, high input rainfed, and irrigated systems). In this analysis,
EPIC results are driven by global climate projections under RCP8.5 and RCP4.5 in the HadGEM2-ES
166
-------
AGRICULTURE
U.S. and Global Agriculture Interactions
GCM,384 and with full C02 fertilization effects.385 The EPIC simulations include some adjustments in crop
management intensity in response to climate, such as marginal changes to fertilizer and irrigation water
use, as well as shifts in annual planting and harvesting dates. Other larger scale adjustments (e.g.,
management intensification from high input rainfed to high input, irrigated) are determined by
GLOBIOM.
For additional information on the model structure, parameters, and the approach to analyzing
interactions between U.S. and global agriculture markets in response to climate change, please see
Havlfk et al. (2011);386 Havlfk et al. (2014);387 and Leclere et al. (2014).388
22.4 RESULTS
This analysis focuses on three primary crop groups that account for more than 50% of the current U.S.
cropland base: corn, soybeans, and wheat. Figure 22.1 shows changes in the output (yield, crop area,
production, consumption, and prices) of these three crops in 2050 relative to a no climate change
control scenario. In general, projected yields are lower, production declines are steeper, and prices rise
more sharply under RCP8.5 compared to RCP4.5.
As shown in Figure 22.1, projected corn yields under RCP8.5 decline more than 20% from the control,
even when accounting for C02 fertilization and allowing land management responses to buffer against
the climate-induced yield change. As yields decrease, more land area shifts to corn production, resulting
in corn area increases of 11%-12% under RCP8.5, but total production still falls by approximately 12%
and corn prices increase more than 70%. Under RCP4.5, net impacts are smaller, with total production
and corn area declining less than 5% relative to the control, and a corresponding net price increase
ranging approximately 15%-17%. While the impact on net yield for corn is close to zero under RCP4.5,
corn area declines relative to the control, a result partially driven by improved productivity and
utilization of grasslands for livestock feeding as a substitute to traditional feed grains (note that the
proportionate change in consumption of corn for feed is higher than consumption for food).
Under RCP8.5, projected impacts on U.S. soybean systems are similar in direction, though slightly larger
in total magnitude relative to corn. Under this scenario, projected soybean yields decline more than 25%
and prices rise more than 75%. Impacts on U.S. soybeans under RCP4.5 are more notable than for corn,
with yields declining approximately 13%-15% relative to the control. Similar to corn, crop area, total
384 Globally-downscaled datasets using CMIP5 GCMs with the full suite of climate variables needed to run EPIC are limited. Of the available data
from the Inter-Sectoral Impact Model Intercomparison Project (ISI-MIP: www.isimip.org). the HadGEM2-ES model matches one of the five
GCMs used throughout the sectors of this Technical Report, and was thus selected for this GLOBIOM analysis. As shown in the domestic
FASOM-GHG results, the HadGEM2-ES projections for the contiguous U.S. resulted in the largest changes in yields for major crops, the largest
increase in crop price index, and the most negative welfare impacts. At a global level, the HadGEM2-ES model shows one of the warmest
responses to changes in radiative forcing among the CMIP5 models. Thus, by choosing to focus on this GCM, this analysis ensures a relatively
strong global climate signal to help identify the relative impacts of accounting for impacts in the rest of the world on U.S. agriculture.
385 The effect of net CO2 fertilization on crop yields is still debated in the literature (Tubiello et al., 2007), so EPIC simulations were run to
produce crop yield impact estimates assuming both "no" and "full" CO2 fertilization. Scenarios applied to this study assume full fertilization.
Source: Tubiello, F.N., J.S. Amthor, K.J. Boote, M. Donatelli, W. Easterling, G. Fischer, R.M. Gifford, M. Howden, J. Reilly, and C. Rosenzweig,
2007: Crop response to elevated CO2 and world food supply - A comment on "Food for Thought..." by Long et al., Science 312:1918-1921, 2006.
European Journal of Agronomy, 26, 215-223.
386 Havlik, P., U.A. Schneider, E. Schmid, H. Bottcher, S. Fritz, R. Skalsky, K. Aoki, S. De Cara, G. Kindermann, F. Kraxner, S. Leduc, I. McCallum, A.
Mosnier, T. Sauer, and M. Obersteiner, 2011: Global land-use implications of first and second generation biofuel targets. Energy Policy, 39,
5690-5702.
387 Havlik, P., H. Valin, M. Herrero, M. Obersteiner, E. Schmid, M.C. Rufino, A. Mosnier, P.K. Thornton, H. Bottcher, R.T. Conant, S. Frank, S. Fritz,
S. Fuss, F. Kraxner, and A. Notenbaert, 2014: Climate change mitigation through livestock system transitions. Proceedings of the National
Academy of Sciences, 111, 3709-3714.
388 Leclere, D., P. Havlik, S. Fuss, E. Schmid, A. Mosnier, B. Walsh, H. Valin, M. Herrero, N. Khabarov, and M. Obersteiner, 2014: Climate change
induced transformations of agricultural systems: insights from a global model. Environmental Research Letters, 9, 1748-9326.
167
-------
AGRICULTURE
U.S. and Global Agriculture Interactions
production, and consumption all decline for soybeans under RCP4.5, with prices increasing
approximately 16%-17%.
U.S. wheat production systems see the largest projected yield impact in the scenarios that included
impacts of global agriculture. Wheat yields decline up to 32% under RCP8.5 and approximately 18%-24%
under RCP4.5 relative to the control. Production and total area decline substantially (more than 35%
under RCP4.5 and more than 58% for RCP8.5 for production; more than 15% under RCP4.5 and more
than 39% under RCP8.5 for total area), especially when compared to corn and soybeans. However,
consumption and price impacts are relatively less severe than price impacts seen for corn and soybeans.
Prices rise less than 41% for both RCP8.5 and RCP4.5. This is due in part to supply-side adjustments and
shifting trade patterns in the rest of the world that buffer against productivity losses from the U.S.
wheat system. As the U.S. share of global corn and soybean production and exports is much larger than
the share for wheat, corn and soybean prices are more significantly impacted by the climate change
scenarios as the rest of the world is less able to respond to changes in U.S. output.
Figure 22.1. Percent Change in U.S. Crop Output
GLOBIOM results are presented (from left to right) as climate change impacts in 2050 on: yield, crop
area, production, consumption, consumption for food, consumption for feed, and prices. All values are
relative to a no climate change control. Lighter colors represent impacts for the U.S.-only climate change
scenarios; darker bars represent global impacts.
Soybean
Wheat
RCP8.5
RCP4.5
Global
The net difference in impacts is relatively small for corn and soybeans when comparing the U.S.-only and
global scenarios. Projected domestic price impacts for corn and soybeans do increase when global
agriculture effects are accounted for, but this shift results in a net increase of less than 10% for both
crops relative to the U.S.-only scenarios. For wheat, however, this relative difference between global
168
-------
AGRICULTURE
U.S. and Global Agriculture Interactions
and domestic estimated impacts is much greater. Unlike corn and soybeans, the net change in crop area
and total production of wheat is significantly smaller for the global scenarios than when only U.S.
impacts are considered, as U.S. wheat production reacts to climate-induced productivity changes in the
rest of the world more than it does for corn and soy. When global climate change effects on agriculture
are accounted for, net price impacts for U.S. wheat is approximately 35% higher than the projected price
impact under the U.S.-only scenario for RCP4.5, and 70% higher for RCP8.5.
22.5 DISCUSSION
Results of this analysis provide global perspective as a complement to the domestic FASOM-GHG
modeling presented in the previous section by directly evaluating the importance of including global
agricultural impacts of climate change on U.S. agricultural outputs relative to a scenario in which U.S.
impacts are evaluated in isolation (i.e. holding climate in the rest of the world to reference levels). While
the estimated net impact of moving from national to global impact scenarios is fairly small for some crop
groups like corn and soybean, the changes could be substantial for crops like wheat.
The primary reason from this difference among crop types is that U.S. wheat production commands a
smaller global market share than corn or soybeans. U.S-produced corn and soybeans account for
roughly 37% and 33% of global production and 46% and 44% of total global exports from all regions,
respectively.389 Conversely, U.S. wheat production only accounts for approximately 10% of global
production and 17% of global exports. Any given crop with a higher total share of production and
exports would see greater net effects on prices given local climate change impacts, regardless of
production shifts in the rest of the world. Thus, any significant change to U.S. corn and soybean
production systems would have important implications for global market effects, leading to higher
overall price impacts and less difference between the U.S.-only and global agriculture scenarios.
The global wheat market is less reliant on U.S. wheat production overall, so even a large shift in U.S.
production under the U.S.-only scenarios does not result in similarly scaled price impacts. Commanding
a smaller market share globally offers greater flexibility for U.S. wheat producers and consumers to
adapt to lower yields induced by climate change by shifting production and export patterns.
These results demonstrate that future climate impacts analyses on U.S. agriculture could benefit from
additional consideration of global impacts and trade adjustments, as global and domestic-only results
may differ according to different crop types and relative country production and export volumes. This
specific type of comparison between U.S. and global agriculture impacts is unique, as it has not been
addressed in the literature using a detailed global model like GLOBIOM. Nonetheless, these estimated
impacts are generally in line with previously published estimates that considered global scenarios (but
not U.S. impacts only).390
There are a few important caveats to note regarding this analysis. First, exogenous crop yield impacts
were produced using a single biophysical model (global EPIC), an approach that does not capture the
potential structural uncertainties that have been highlighted in the literature.391 Second, the scenario
389 Food and Agriculture Organization of the United Nations, cited 2016: FAOSTAT statistics database. Available online at
http://www.fao.org/faostat/en/
390 Brown, M.E., J.M. Antle, P. Backlund, E.R. Carr, W.E. Easterling, M.K. Walsh, C. Ammann, W. Attavanich, C.B. Barrett, M.F. Bellemare, V.
Dancheck, C. Funk, K. Grace, J.S.I. Ingram, H. Jiang, H. Maletta, T. Mata, A. Murray, M. Ngugi, D. Ojima, B. O'Neill, and C. Tebaldi. 2015. Climate
Change, Global Food Security, and the U.S. Food System. 146 pages. Available online at
http://www.usda.gov/oce/climate change/FoodSecuritv2015Assessment/FuIIAssessment.pdf
391 Rosenzweig, C., J. W. Jones, J. L. Hatfield, Alex Ruane, K. J. Thornburn, J. M. Antle, G. C. Nelson, C. Porter, S. Janssen, B. Basso, F. Ewert, D.
Wallach, G. Baigorria, and J. M. Winter, 2013: The Agricultural Model Intercomparison and Improvement Project (AgMIP): Protocols and pilot
studies. Papers in Natural Resources. University of Nebraska - Lincoln, School of Natural Resources. Available online at
http://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1448&context=natres papers
169
-------
AGRICULTURE
U.S. and Global Agriculture Interactions
design was developed using underlying climate data from one GCM (HadGEM2-ES), and thus does not
capture the uncertainty in regional climate projection across climate models. It is also important to
caveat that there are key structural differences between GLOBIOM and FASOM-GHG that can influence
the magnitude of net agricultural sector impacts, specifically with each model's treatment of time
dynamics. FASOM-GHG is an intertemporal optimization model, and land management changes in a
given period can be made in anticipation of future climate and productivity changes. GLOBIOM is a
recursive dynamic framework, so management changes reflect a contemporaneous response to the
exogenous yield shock and do not reflect expectations of the future. Finally, scenarios included in this
analysis assume international trade consistent with current patterns.
170
-------
ECOSYSTEMS
23. CORAL REEFS
23.1 KEY FINDINGS
• Coral reefs are already disappearing due to climate change and other non-climate stressors.
Temperature increases, including the increasing frequency and intensity of high-temperature
bleaching events, and ocean acidification are projected to further reduce coral cover in the future.
• Extensive loss of shallow corals is projected by 2050 for major U.S. reef locations. Modest delays in
Hawaiian coral reef loss are projected under RCP4.5 compared to RCP8.5, but the lower emissions
scenario provides little benefit to coral cover in South Florida and Puerto Rico, as these reefs have
already passed critical thresholds of ecosystem change.
• Coral reef recreation is projected to decline considerably under both scenarios, though slightly less
under RCP4.5.
23.2 BACKGROUND
Coral reefs, including those found in Flawai'i and the Caribbean, are unique ecosystems that are home to
large numbers of marine plant and animal species. They also provide vital fish spawning habitat, protect
shorelines, and are valuable for recreation and tourism. Flowever, shallow-water coral reefs are highly
vulnerable to climate change.392 High water temperatures can cause coral to expel the symbiotic algae
that provide nourishment and vibrant color for their hosts. This coral bleaching can cause the coral to
die, especially when bleaching events occur consecutively. In addition, ocean acidification (ocean
chemistry changes due to elevated emissions of C02) can reduce the availability of calcium carbonate in
seawater that is needed to build and maintain coral skeletons.
23.3 APPROACH
This analysis examines the physical and economic impacts of climate change (temperature effects) and
ocean acidification on shallow-water coral reefs in Hawai'i, South Florida, and Puerto Rico. Using the
COMBO (Coral Mortality and Bleaching Output) model,393-394 the analysis first estimates declines in coral
reef cover (a measure of coral reef health and density) using projections of future sea surface
392 Melillo, Jerry M., Terese (T.C.) Richmond, and Gary W. Yohe, Eds., 2014: Climate Change Impacts in the United States: The Third National
Climate Assessment. U.S. Global Change Research Program, 841 pp. doi:10.7930/J0Z31WJ2.
393 Buddemeier, R.W., P.L. Jokiel, K.M. Zimmerman, D.R. Lane, J.M. Carey, G.C. Bohling, and J.A. Martinich, 2008: A modeling tool to evaluate
regional coral reef responses to changes in climate and ocean chemistry. Limn Oceanogr Methods, 6, 395-411, doi: 10.4319/lom.2008.6.395.
394 Buddemeier, R.W., D.R. Lane, and J.A. Martinich, 2011: Modeling regional coral reef responses to global warming and changes in ocean
chemistry: Caribbean case study. Climatic Change, 109, 375-397, doi:10.1007/sl0584-011-0022-z.
171
-------
ECOSYSTEMS
Coral Reefs
temperature (from the five GCMs) and ocean acidification (i.e., carbonite saturation state with respect
to aragonite) under RCP8.5 and RCP4.5. The effects of future bleaching events, driven by projected
changes from the climate models, are also estimated. Next, the analysis quantifies the economic
impacts associated with coral reef cover loss based on declines in reef-based recreation. The analysis
estimates these impacts using a benefit-transfer approach; that is, it draws on reef-related recreation
benefits measured in previously published studies conducted at a range of coral reef sites to estimate
the value of reef-related recreation benefits in the areas considered in this study.395 As shown in
Appendix A. 13, COMBO was also simulated using explicit modeling of three types of coral that have
different responses to bleaching: feeders, switchers, and optimizers. Each coral type has its own
bleaching threshold, mortality, and growth parameters, which together help account for variability in
biological response to future bleaching events. For more information on the approach for the coral reef
sector, please refer to Lane et al. (20 13)396 and Lane et al. (2014).397
23.4 RESULTS
For major U.S. reefs, projections under RCP8.5 (average of five GCMs) show extensive bleaching and
dramatic loss of shallow coral cover by 2050, and near complete loss by 2100. In Hawai'i, coral cover is
projected to decline from 38% in 2010 to approximately 11% by 2050, with further declines thereafter
(Figure 23.1). In South Florida and Puerto Rico, where present-day sea surface temperatures are already
close to bleaching thresholds and where these reefs have historically been affected by non-climate
stressors, coral is projected to disappear even faster.
Some of the projected biological and economic impacts of climate change on coral reefs in the U.S. are
delayed, but not avoided, under RCP4.5. Figure 23.1 shows projected change in percent coral reef cover
from 2010 to 2100 in Hawai'i, South Florida, and Puerto Rico under RCP8.5 and RCP4.5. These results
represent the average results across the five GCMs, and are displayed using a consistent y-axis scale to
show differences in initial coral cover. In Hawai'i, the decline in reef cover slows modestly under RCP4.5
compared to RCP8.5, but both scenarios suggest substantial reductions. In South Florida and Puerto
Rico, the RCP4.5 scenario is likely insufficient to avoid multiple bleaching and mortality events by 2020,
and coral cover declines thereafter nearly as fast as under RCP8.5.
395 Lane, D.R., R.C. Ready, R.W. Buddemeier, J.A. Martinich, K.C. Shouse, and C.W. Wobus, 2013: Quantifying and valuing potential climate
change impacts on coral reefs in the United States: Comparison of two scenarios. PLoS ONE, 8, e82579, doi:10.1371/journal.pone.0082579.
396 Ibid.
397 Lane, D., R. Jones, D. Mills, C. Wobus, R.C. Ready, R.W. Buddemeier, E. English, J. Martinich, K. Shouse, and H. Hosterman, 2014: Climate
change impacts on freshwater fish, coral reefs, and related ecosystem services in the United States. Climatic Change, 131, 143-157, doi:
10.1007/sl0584-014-l 107-2.
172
-------
ECOSYSTEMS
Coral Reefs
Figure 23.1. Average Change in Percent Coral Reef Cover
Results show change in percent coral cover under RCP8.5 and RCP4.5 for the five-model average
South Florida
40
35
® -»r,
> 30
o
- 25
8 20
1«
u
fc 10
o.
2010 2030 2050 2070 2090
Hawai'i
40
35
130
if 25
(0
o 20
15
c
30
o
H 25
o 20
v
S 10
Q.
2010 2030 2050 2070 2090
Figure 23.2. Change in Percent Coral Reef Cover by GCM
Results show change in percent coral cover under RCP8.5 and RCP4.5 for each GCM
Puerto Rico
25
20
o
^ 15
re
10
o_
2010
2030
2050
2070
2090
12
South Florida
10
8
6
4
2
0
2010 2030 2050 2070 2090
Year
40
Hawai'i
o
re
u
2010 2030 2050 2070 2090
GCM
RCP8.5
RCP4.5
CanESM2
CCSM4
GISS-E2-R
HadGEM2-ES
Ml ROCS
Figure 23.2 shows projected changes in coral cover from 2010 to 2100 under each GCM/RCP
combination for the three regions (note differences in scaling of y-axis). Four of the five GCMs show
similar results over time. The GISS-E2-R model, which projects comparatively smaller increases in sea
surface temperatures over time under both RCPs compared to the other GCMs, estimates a delay in the
timing of coral loss. This effect is most notable in projected changes in Hawaiian coral cover.
The projected loss of coral reef cover will have substantial effects on reef-based recreation (Table 23.1).
Under most RCP/GCM combinations, more than 90% of the value of recreation in the reference period is
lost by the end of the century. Across all three regions, an estimated $140 billion (discounted 3%) in
reef-based recreation is projected to be lost through 2100 under RCP8.5, and $130 billion under RCP4.5.
More than half of these losses are projected for South Florida, which has larger levels of tourism for
173
-------
ECOSYSTEMS
Coral Reefs
reef-based recreation. Table 23.2 provides cumulative estimates for the differences between the RCPs.
Including the economic value of other services provided by coral reefs, such as shoreline protection and
fish-rearing habitat, would provide a more comprehensive understanding of the total economic value of
these declines in habitat.
Table 23.1. Cumulative Value of Lost Reef-Based Recreation
Results are presented through 2100 in 2015$, discounted at 3%.398 Estimates for Puerto Rico only
represent recreational effects for permanent residents, and therefore are both smaller (listed in millions)
and not directly comparable to the results for the other locations which include visits from nonresident
tourists. Totals may not sum due to rounding.
HAWAI'I
SOUTH FLORIDA
PUERTO RICO
TOTAL
(Billions $)
(Billions $)
(Millions $)
(Billions $)
GCM
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
$57
$55
$96
$97
$9.1
$9.6
$150
$150
CCSM4
$41
$43
$95
$95
$8.5
$8.4
$140
$140
GISS-E2-R
$14
$6.0
$86
$70
$8.3
$7.1
$99
$76
HadGEM2-ES
$51
$60
$100
$99
$9.3
$9.2
$150
$160
MIROC5
$47
$38
$93
$94
$9.2
$9.6
$140
$130
5-GCM Average
$42
$40
$95
$91
$8.9
$8.8
$140
$130
Table 23.2. Cumulative Difference in Value of Reef-Based Recreation
Results are presented as the difference between RCP8.5 and RCP4.5 at major U.S. coral reefs through
2100 in millions of 2015$, discounted at 3%. Due to rounding, values may not equate to differences
between RCP-specific results shown in Table 23.1.
GCM
HAWAI'I
SOUTH FLORIDA
PUERTO RICO
TOTAL
CanESM2
$1,400
$1,200
-$0.47
$2,600
CCSM4
-$2,200
-$290
$0.11
-$2,500
GISS-E2-R
$8,000
$15,000
$1.2
$23,000
HadGEM2-ES
-$8,700
$4,800
$0,051
-$3,900
MIROC5
$9,200
-$1,700
-$0.36
$7,500
5-GCM Average
$1,500
$3,900
$0.11
$5,400
23.5 DISCUSSION
The findings described above suggest very substantial impacts to U.S. coral reefs within the coming
decades. Drastic decline in coral reef cover, indicating the exceedance of an ecosystem threshold, is
likely to have significant ecological and economic consequences at regional levels. The projections of
398 Values calculated by comparing recreation provided by available coral cover in each year to a no-climate change control scenario which
assumes constant coral cover (based on reference period values) but includes changes in population and the effects of economic discounting.
174
-------
ECOSYSTEMS
Coral Reefs
shallow coral loss for major U.S. reefs are consistent with the findings of the assessment literature,399-400
with other studies/01,402 and, most importantly, with what has been observed in reefs across the U.S.
over the past 15 years. Unlike other sectors of this Technical Report where the climate change signal
emerges from natural variability over the course of the next 25 years, the most severe impacts to coral
reefs are occurring now.403
Importantly, the impacts modeled in COMBO do not include non-climate stressors, such as overfishing
or water pollution, which have significantly affected U.S. reef ecosystems over the past 40 years. In
addition, the COMBO analysis does not account for the ability of large-scale reefs to contain important
refugia for resilient corals that could potentially be used in coral restoration efforts. Together, these
factors can adjust the estimates presented in this report downwards or upwards.404 See Lane et al.
(2013)405 for additional information regarding caveats to the modeling of coral reefs using the COMBO
model.
399 Doney, S., A. A. Rosenberg, M. Alexander, F. Chavez, C. D. Harvell, G. Hofmann, M. Orbach, and M. Ruckelshaus, 2014: Ch. 24: Oceans and
Marine Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 557-578, doi:10.7930/J0RF5RZW.
400 Romero-Lankao, P., J.B. Smith, D.J. Davidson, N.S. Diffenbaugh, P.L. Kinney, P. Kirshen, P. Kovacs, and L. Villers Ruiz, 2014: North America. In:
Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change [Barros, V.R., C.B. Field, D.J. Dokken, M.D. Mastrandrea, K.J. Mach, T.E.
Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L.White (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1439-1498.
401 Donner, S.D., 2009: Coping with commitment: Projected thermal stress on coral reefs under different future scenarios. PLOS ONE,
doi:10.1371/journal.pone.0005712.
402 Veron, J.E.N., O. Hoegh-Guldberg, T.M. Lenton, J.M. Lough, D.O. Obura, P. Pearce-Kelly, C.R. Sheppard, M. Spalding, M.G. Stafford-Smith, and
A.D. Rogers, 2009: The coral reef crisis: The critical importance of <350 ppm CO2. Marine Pollution Bulletin, 58, doi:
10.1016/j.marpolbul.2009.09.009.
403 National Oceanographic and Atmospheric Administration, cited 2016: Coral Health and Monitoring Program. Available online at:
http://www.coral.noaa.gov/
404 Recent research suggests that reefs experiencing smaller levels of non-climate stressors (e.g., water pollution and overfishing) are equally
vulnerable to bleaching events. See Hughes, T.P., et al., 2017: Global warming and recurrent mass bleaching of corals. Nature, 543,
doi:10.1038/nature21707.
405 Lane, D.R., R.C. Ready, R.W. Buddemeier, J.A. Martinich, K.C. Shouse, and C.W. Wobus, 2013: Quantifying and valuing potential climate
change impacts on coral reefs in the United States: Comparison of two scenarios. PLoS ONE, 8, e82579, doi:10.1371/journal.pone.0082579.
175
-------
ECOSYSTEMS
Shellfish
24. SHELLFISH
24.1 KEY FINDINGS
• Harvests of five types of U.S. shellfish are projected to decline substantially by the end of the
century due to ocean acidification. Declines in supply are substantially higher under RCP8.5 as
compared to RCP4.5.
• Decreases in supply are projected to result in price increases for six types of shellfish by the end of
the century, ranging from 24% to 230% depending on the species and region from which they are
harvested. Under RCP8.5, percent change in the price of oysters is projected to increase by 33% in
the Northeast, 140% in the Southeast, and 150% in the Northwest by 2090.
• Cumulative discounted losses in consumer welfare are $230 million under RCP8.5 and $140 million
under RCP4.5.
24.2 INTRODUCTION
The ocean absorbs about one quarter of the C02 released into the atmosphere by human activities.
Although the ocean's ability to absorb C02 prevents atmospheric levels from climbing even higher,
measurements made over the last few decades have demonstrated that marine C02 levels have risen,
leading to an increase in acidity.406 Ocean acidification is projected to adversely affect a number of
valuable marine ecosystem services by making it more difficult for many organisms to form shells and
skeletons.407 Some shellfish are highly vulnerable to ocean acidification408 and impacts to these species
are expected to negatively affect consumers, the fishing industry, and the broader economy. Certain
species have high commercial value; for example, oysters, clams, and scallops supplied approximately
170 million pounds of U.S. seafood each year between 1990-2010 valued at $400 million.409
24.3 APPROACH
This analysis models biophysical and economic impacts of ocean acidification on several shellfish species
of the contiguous U.S. under RCP8.5 and RCP4.5. The biophysical impacts are estimated using C02 and
sea surface temperature projections from five GCMs to simulate seawater chemistry conditions through
the 21st century. These conditions are then used to estimate how the growth rates of oysters, scallops,
geoducks, quahogs, clams, and mussels410 will change over time along the contiguous U.S. coastline. The
economic analysis uses the projected growth rates of these species to estimate changes to the U.S.
supply of shellfish in three U.S. regions: the Northeast, the Southeast (which includes the Gulf of
406 EPA, 2016: Climate change indicators in the United States, 2016. Fourth Edition. United States Environmental Protection Agency, EPA430-R-
16-004. Available online at www.epa.gov/climate-indicators
407 Doney, S., A. A. Rosenberg, M. Alexander, F. Chavez, C. D. Harvell, G. Hofmann, M. Orbach, and M. Ruckelshaus, 2014: Ch. 24: Oceans and
Marine Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 557-578. doi:10.7930/J0RF5RZW.
408 In early life stages, some species will have higher mortality rates and more developmental abnormalities under acidification conditions
expected over the next several decades. In addition, adult shellfish tend to grow more slowly and have thinner, more fragile shells under these
conditions.
409 NOAA, cited 2017: Annual Commercial Landing Statistics (1990-2010). National Oceanic and Atmospheric Administration, Office of Science
and Technology. Available online at http://www.st.nmfs.noaa.gov/commercial-fisheries/commercial-landings/annual-landings/index
410 The consumer demand model described in Moore and Griffiths (2017) includes mussels, however, the supply of mussels is held constant in
this analysis. This is because mussels did not exhibit a statistically significant reaction to increasing CO2 concentrations (and falling aragonite
saturation state) in the biophysical experiments underlying the methodology.
176
-------
ECOSYSTEMS
Shellfish
Mexico), and the Northwest. Not all species are present in each region. A two-stage consumer demand
model of the shellfish market, described in Moore and Griffiths (2017)411 but expanded to capture
regional differences (e.g. Richards et al. 1997),412 projects changes in prices and consumer behavior
under the ten RCP/GCM combinations, and estimates changes in consumer welfare. By considering
impacts to five of these species (excluding mussels), this approach estimates just a fraction of the
potential economic damages from ocean acidification. For more information on the approach to
estimating the economic impacts of ocean acidification in the shellfish market, see Moore and Griffiths
(2017) and Moore (2015).413
24.4 RESULTS
Continued ocean acidification is estimated to reduce the supply of oysters, scallops, geoducks, quahogs,
and clams (Figure 24.1). Under RCP8.5, supplies decrease by the following amounts by the end of the
century (average of 2080-2099): 50% of oysters, 55% of scallops, 54% of geoducks, 45% of quahogs, and
35% of clams. Under RCP4.5, national supplies decrease by the following amounts by the end of the
century: 24% of oysters, 27% of scallops, 11% of geoducks, 9.0% of quahogs, and 6.1% of clams.
These decreases in supply are projected to result in price increases in all species and in all regions (Table
24.1). The largest increases in price by the end of the century compared to the reference period (2010)
are projected to occur under RCP8.5 in quahogs in the Northeast (a 230% increase) and Southeast (a
180% increase), geoducks in the Northwest (a 220% increase), oysters in the Northwest and Southeast
(150% and 140% increase, respectively), and scallops in the Northeast and Southeast (increases of
130%). Figure 24.2 shows the regional percent change in the price of oysters, where there is projected
catches for all three regions.
411 Moore, C. and C. Griffiths, 2017: Welfare Analysis in a Two-Stage Inverse Demand Model: An Application to Harvest Changes in the
Chesapeake Bay. Empirical Economics. 181:1-26. DOI 10.1007/s00181-017-1309-3.
412 Richards, T.J., Van Ispelen, P. and Kagan, A., 1997: A two-stage analysis of the effectiveness of promotion programs for US apples. American
Journal of Agricultural Economics, 79(3), pp.825-837. The Richards et al. 1997 study uses two-stage budgeting to examine consumer demand
for apples from three different countries. This analysis takes the same approach to examine the demand for shellfish harvested in different
regions within the contiguous U.S.
413 Moore, C., 2015: Welfare estimates of avoided ocean acidification in the US mollusk market. Journal of Agricultural and Resource Economics,
40, 50-62.
177
-------
ECOSYSTEMS
Shellfish
Table 24.1. Percent Increase in U.S. Shellfish Price
The table presents the estimated percent change in shellfish price from 2010 to 2090 (2080-2099). Note
that not all species are commercially harvested in each region.
RCP8.5
RCP4.5
Northeast
Clam
89%
39%
Mussel
74%
30%
Oyster
33%
24%
Quahog
230%
59%
Scallop
130%
51%
Southeast
Oyster
140%
48%
Quahog
180%
51%
Scallop
130%
52%
Northwest
Geoduck
220%
48%
Oyster
150%
54%
178
-------
ECOSYSTEMS
Shellfish
Figure 24.1. Percent Change in U.S. Shellfish Supplies
The graphs present the estimated percent change in national shellfish supply from 2011 to 2099 under
RCP8.5 and RCP4.5 for the five-GCM average.
Clams
Quahog
0.10
0.00
-0.10
-0.20
-0.30 -
-0.40 -
-0.50 -
-0.60
-0.70
Geoduck
Oyster
-0.20
-0.30
-0.40
-0.70
-0.10
-0.20
-0.30
-0.40
-0.50
-0.60
-0.70
Scallop
-0.10
-0.20
-0.30
-0.40
-0.50
-0.60
-0.70
-0.10
-0.20
-0.30 -
-0.40
-0.50 -
-0.60 -
RCP8.5
RCP4.5
179
-------
ECOSYSTEMS
Shellfish
Figure 24.2. Percent Change in the Price of Oysters
The graph presents the estimated percent change in the price of oysters by region under RCP8.5 and
RCP4.5 from 2011 to 2099
250%
200%
150%
100%
50%
0%
2050
2090
-50%
RCP8.5 Southeast/Gulf
RCP8.5 Northeast
RCP8.5 Northwest
RCP4.5 Southeast/Gulf
RCP4.5 Northeast
RCP4.5 Northwest
Changes in consumer welfare are reported as compensating surplus lost (Table 24.2). Cumulative losses
in consumer welfare in response to decreasing supplies of these five shellfish are estimated at $140
million under RCP4.5 (ranging from $46 to $220 million) and $230 million under RCP8.5 (ranging from
$95 to $360 million).
Table 24.2. Cumulative Losses in Consumer Welfare
The table presents the cumulative losses from 2010 to 2099 on the five species analyzed (millions 2015$,
discounted at 3% to 2015) under RCP8.5 and RCP4.5for each GCM and the five-GCM average.
Model
RCP8.5
RCP4.5
CanESM2
$95
$46
CCSM4
$97
$66
GISS-E2-R
$360
$200
HadGEM2-ES
$300
$180
MIROC5
$310
$220
5-GCM Average
$230
$140
180
-------
ECOSYSTEMS
Shellfish
24.5 DISCUSSION
Warming waters and more acidic conditions due to climate change are projected to decrease shellfish
supply in the U.S., increasing prices and decreasing consumer welfare, particularly under RCP8.5. These
projections are consistent with the findings of the assessment literature, which describe reduced growth
and survival of U.S. shellfish stocks due to continued ocean acidification.414 Demand for shellfish is
projected to increase through the end of the century with a growing population and rising incomes,
exacerbating the economic impacts in this sector. As prices rise in response to the contracting supply,
consumers will substitute away from the affected species toward other less-preferred commodities.
Climate change will also influence losses in shellfish catch through mechanisms beyond just loss in
supply. For instance, as ocean temperatures have increased, the average center of biomass for 105
marine fish and invertebrate species has already shifted northward by about 10 miles and moved an
average of 20 feet deeper between 1982 and 2015.415 This approach isolates impacts due to
acidification, but does not evaluate the impacts of other stressors on shellfish supply over time,
including overfishing pressures, losses due to nutrient and eutrophication issues (including coastal
acidification), disease or contamination. Furthermore, the monetized shellfish losses reported here do
not include impacts on the fishing industry, nor on local economies that rely on these activities.
414 Doney, S., A. A. Rosenberg, M. Alexander, F. Chavez, C. D. Harvell, G. Hofmann, M. Orbach, and M. Ruckelshaus, 2014: Ch. 24: Oceans and
Marine Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 557-578. doi:10.7930/J0RF5RZW.
415 EPA, 2016: Climate change indicators in the United States, 2016. Fourth Edition. United States Environmental Protection Agency, EPA430-R-
16-004. [Available online at www.epa.gov/climate-indicatorsl
181
-------
ECOSYSTEMS
Freshwater Fish
25. FRESHWATER FISH
25.1 KEY FINDINGS
• Under RCP8.5, coldwater fisheries are projected to be lost in many areas of the U.S. over the course
of the 21st century, especially in the mountain regions of the Northwest, Southwest, and the
Northeast through Appalachia. By 2090, coldwater recreational fishing days are estimated to decline
nationally by more than 90 million days per year under RCP8.5 and almost 67 million days per year
under RCP4.5.
• Large losses of suitable stream habitat are projected for warmwater species, such as small and large
mouth bass, across the Southern Plains, Northern Plains, and Midwest by the end of the century.
Habitats suitable for rough water species, such as catfish and carp, are projected to increase in these
regions.
• Lost recreational fishing values (for all fishing guilds) in 2090 compared to the reference period are
approximately $3.1 billion annually under RCP8.5 and $1.7 billion annually under RCP4.5.
Cumulative discounted losses are $45 billion under RCP8.5 and $38 billion under RCP4.5.
25.2 INTRODUCTION
Freshwater fishing is an important recreational activity that contributes significantly to local economies
in many parts of the country. In 2011 alone, more than 27 million people in the U.S. spent a total of $25
billion on over 365 million freshwater recreational fishing trips.416 Most fish species thrive only in certain
ranges of water temperature and stream flow conditions. For example, trout and salmon can only
tolerate coldwater streams, while shad and largemouth bass thrive in warmwater habitats. Climate
change threatens to disrupt these habitats and affect certain fish populations through higher stream
temperatures and changes in river flow.417
25.3 APPROACH
This analysis projects the impacts of climate change on the distribution of habitat suitable for freshwater
fish across the U.S. and estimates the economic implications of these changes. Water temperature and
streamflow changes are simulated for the two RCPs using five GCMs to estimate changes in suitable
habitat for three types of freshwater fishery guilds: coldwater, warmwater, and rough (species tolerant
to the warmest stream temperatures). Each fishery type represents a categorization of individual
species based on their tolerance for different river and stream water temperatures, which directly
affects dissolved oxygen concentrations and other parameters that affect suitability. The coldwater fish
guild contains species that are the least tolerant to increasing stream temperatures, and are therefore
the most vulnerable to climate change.
416 U.S. Department of the Interior, U.S. Fish and Wildlife Service, and U.S. Department of Commerce,
U.S. Census Bureau, 2014: 2011 National Survey of Fishing, Hunting, and Wildlife-Associated Recreation. Available online at
http://www.census.gov/prod/2012pu bs/fhwll-nat.pdf
417 Groffman, P. M., P. Kareiva, S. Carter, N. B. Grimm, J. Lawler, M. Mack, V. Matzek, and H. Tallis, 2014: Ch. 8: Ecosystems, Biodiversity, and
Ecosystem Services. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 195-219. doi:10.7930/J0TD9V7H.
182
-------
ECOSYSTEMS
Freshwater Fish
COLOWATER FISHERY EXAMPLES WARMWATER FISHERY EXAMPLES I ROUGH FISHERY EXAMPLES
Largemouth Bass
Smallmouth Bass
Trout
Bluegill
Salmon
Catfish
Results from habitat modeling considering projected changes in both water temperature and
streamflow serve as input to an economic model to analyze the impacts of habitat change on number of
fishing days and the value of recreational fishing.418 Accounting for population growth, the model
estimates fishing behavior as the likelihood that an adult in a particular state is an angler and the
likelihood that an angler fishes for species in each fishery type. The fishing value for each fishery type is
derived by multiplying the number of fishing days by the value of a fishing trip.419 Reported values of the
change in the number of annual fishing days represent the change in "supply of fishing days/' or the
impact on availability of freshwater fish to support fishing trips, not any change in "demand for fishing
days," or impact on recreational behavior. For more information on the approach for the freshwater fish
sector, please refer to Lane et al. (2014)420 and Jones et al. (2012).421
25.4 RESULTS
Increasing stream temperatures and changes in stream flow are likely to transform many habitats that
are currently suitable for coldwater fish into areas that are only suitable for warm or rough water
species that are less recreationally valuable. Figure 25.1 shows the projected changes in potential
freshwater fish habitat from the reference period to 2090. Under RCP8.5, coldwater fisheries are
estimated to be limited almost exclusively to the Mountainous West by 2090, nearly disappearing from
the Northeast through Appalachia in all five of the GCMs. In addition, substantial portions of Florida and
central states, including Texas, Oklahoma, and Kansas, shift from warmwaterto rough habitat. The
losses of coldwater and warmwater habit are largest under the HadGEM2~ES climate model, while the
smallest shifts are projected under the GISS-E2-R GCM. Importantly, projected shifts from coldwater and
warmwater habitat are reduced under RCP4.5 compared to RCP8.5.
41®The approach used in this Technical Report slightly modifies the approach described in Jones et al. (2012) by using a ratio of change in
suitable habitat at the 8-digit hydrologic unit code level to drive the recreational fishing behavior model.
419 Jones, R., C. Travers, C. Rodgers, B. Lazar, E. English, J. Upton, J. Vogel, K. Strzepek, and J. Martinich, 2012: Climate Change Impacts on
Freshwater Recreational Fishing in the United States. Mitigation and Adaptation Strategies for Global Change, 18, 731, doi: 10.1007/sll027-
012-9385-3.
420 Lane, D., R. Jones, D. Mills, C, Wobus, R.C. Ready, R.W. Buddemeier, E. English, J. Martinich, K. Shouse, and H. Hostel man, 2014: Climate
change impacts on freshwater fish, coral reefs, and related ecosystem services in the United States. Climatic Change, 131, 143-15 7, doi:
10.1007/sl0584-014-l 107-2.
421 Jones, R., C. Travers, C. Rodgers, B. Lazar, E. English, J. Lipton, J. Vogel, K. Strzepek, and J. Martinich, 2012: Climate Change Impacts on
Freshwater Recreational Fishing in the United States. Mitigation and Adaptation Strategies for Global Change, 18, 731, doi: 10.1007/sll027-
012-9385-3.
183
-------
ECOSYSTEMS
Freshwater Fish
Climate change is projected to have a significant impact on freshwater recreational fishing in the
contiguous U.S. Between 2011 and 2050, coldwater fishing days are estimated to decline nationally by
67 million days per year (five-model average, with a range of 54-80 million days per year across the
GCMs) under RCP8.5 and 62 million days per year (47-75 million days per year) under RCP4.5. Lost
fishing days under RCP8.5 becomes significantly larger by 2090, when coldwater fishing days are
estimated to decline nationally by 90 million days per year (73-104 million days per year) under RCP8.5.
Coldwater fishing days decline by almost 67 million days per year under RCP4.5 in 2090. A high number
of lost coldwater fishing days occur in mountainous regions by 2090 under both RCPs, including the
Northeast, the Northwest, and the Southwest.
184
-------
ECOSYSTEMS
Freshwater Fish
Figure 25.1. Projected Impact of Climate Change on Potential Freshwater Fish Habitat
The maps show the projected changes in habitat from the reference period centered on 2011 (labeled
"tCurrent") to 2090 (labeled "Projected") under RCP8.5 and RCP4.5.
(a) Projected changes for CanESM2, CCSM4, and GISS-E2-R
RCP8.5 RCP4.5
Current Cold, Projected Cold
Current Cold. Projected Rough
Current Cold, Projected Warm
Current Warm, Projected Warm
Current Warm, Projected Rough
CanESM2
CCSM4
185
-------
ECOSYSTEMS
Freshwater Fish
(b) Projected changes for HadGEM2-ES and M1R0C5
RCP8.5 RCP4.5
Current Cold, Projected Cold
Current Cold, Projected Rough
| Current Cold, Projected Warm
| Current Warm, Projected Rough
Current Warm, Projected Warm
Lost recreational fishing values (for all fishing guilds) in 2090 compared to 2011 are $3.1 billion annually
under RCP8.5 and $1.7 billion annually under RCP4.5 (undiscounted). Cumulative losses through 2100 to
national recreational value are estimated at $45 billion under RCP8.5 and $38 billion under RCP4.5,
respectively (discounted at 3%). Cumulative losses for coldwater fishing only are estimated at $100
billion under RCP8.5 and $93 billion under RCP4.5 (discounted at 3%). These results (Table 25.1) reflect
tradeoffs in economic losses from coldwater fishing, spatially-varied gains and losses in warmwater
fishing, and gains in rough fishing.
HadGEM2-ES
MIROC5
186
-------
ECOSYSTEMS
Freshwater Fish
Table 25.1. Change in National Value of Recreational Fishing
Results are presented in millions of $2015, discounted at 3% for 2011-2100. Values may not sum due to
rounding.
All Fishing
Coldwater Fishing
RCP8.5
RCP4.5
RCP8.5
RCP4.5
CanESM2
-$42,000
-$28,000
-$110,000
-$93,000
CCSM4
-$49,000
-$35,000
-$110,000
-$95,000
GISS-E2-R
$680
-$9,400
-$85,000
-$75,000
HadGEM2-ES
-$85,000
-$64,000
-$120,000
-$110,000
MIROC5
-$50,000
-$51,000
-$96,000
-$88,000
5-GCM Average
-$45,000
-$38,000
-$100,000
-$93,000
25.5 DISCUSSION
Warming waters and changes in stream flows due to climate change are projected to alter the
distribution of freshwater fisheries across the country. The projected loss of coldwater fish habitat and
expansion of rough fisheries are consistent with the conclusions of the assessment literature, which find
that as temperatures rise and precipitation patterns change, many fish species (such as salmon, trout,
and char) will be lost from lower-elevation streams.422 Modeling altered stream flows as well as
increased temperature, Wenger et al. (2011)423 projects an overall loss of 47% of habitat for four trout
species in the western U.S. by 2080 under a moderate GHG emissions scenario (SRES A1B). A recent
review of 31 peer-reviewed studies found that observed changes in climate are already altering the
abundance, growth, recruitment, and habitat ranges of some North American inland fish populations,
particularly coldwater species.424 Varying degrees of physiological impacts of climate change on fish,
such as decreased cardiorespiratory performance, compromised immune function, and altered
reproductive behaviors, have also been observed.425
As the implications of changes to the distribution of freshwater fisheries extend beyond recreational use
by humans, including effects on food chains and ecosystem services, this analysis underestimates the
avoided economic impacts projected under RCP4.5. Climate change will also influence losses in
freshwater fishing through mechanisms beyond just loss of habitat. For instance, changing temperatures
can affect the timing of when stream and lake waters freeze or thaw, which can affect the length of
open water fishing season, potentially extending fishing efforts in northern regions.426 This analysis does
422 Groffman, P. M., P. Kareiva, S. Carter, N. B. Grimm, J. Lawler, M. Mack, V. Matzek, and H. Tallis, 2014: Ch. 8: Ecosystems, Biodiversity, and
Ecosystem Services. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 195-219. doi:10.7930/J0TD9V7H.
423 Wenger, S.J., D.J. Isaak, C.H. Luce, H.M. Neville, K.D. Fausch, J.B. Dunham, D.C. Dauwalter, M.K. Young, M.M. Eisner, B.E. Rieman, A.F.
Hamlet, and J.E. Williams, 2011: Flow regime, temperature, and biotic interactions drive differential declines of trout species under climate
change. PNAS, 108,14175-14180, doi: 10.1073/pnas.ll03097108.
424 Lynch, A.J., B.J.E. Myers, C. Chu, L.A. Eby, J.A. Falke, R.P. Kovach, T.J. Krabbenhoft, T.J. Kwak, J.L. Lyons, C.P. Paukert, and J.E. Whitney, 2016:
Climate Change Effects on North American Inland Fish Populations and Assesmblages. Fisheries, 41, 346-361, doi:
10.1080/03632415.2016.1186016.
425 Whitney, J.E., R. Al-Chokhachy, D.B. Bunnell, C.A. Caldwell, S.J. Cooke, E.J. Eliason, M. Rogers, A.J. Lunch, and C.P. Paukert, 2016:
Physiological Basis of Climate Change Impacts on North American Inland Fishes. Fisheries, 41, 332-345, doi: 10.1080/03632415.2016.1186656.
426 Paukert, C.P., A.J. Lynch, and J.E. Whitney, 2016: Effects of Climate Change on North American Inland Fishes: Introduction to the Special
Issue. Fisheries, 41, doi: 10.1080/03632415.2016.1187011.
187
-------
ECOSYSTEMS
Freshwater Fish
not evaluate impacts to fisheries in lakes and reservoirs, which have different vulnerabilities to climate
change, can be thermally stratified or provide other climate refugia for fish, and are also oftentimes
heavily managed (i.e., water level and fish stocking). Changes in land use, water demands for irrigation
or commercial uses, water quality, and diseases, parasites or invasive species will also interact with
climate impacts to affect aquatic ecosystems and freshwater fishing.
188
-------
ECOSYSTEMS
Wildfire
26. WILDFIRE
26.1 KEY FINDINGS
• Under RCP8.5, wildfire acres burned in the contiguous U.S. are projected to remain consistent with
rates observed over the past several decades, but moderately decrease under RCP4.5, with changes
under both scenarios driven by shifts in vegetation overtime. In Alaska, burned acreage is projected
to increase under both RCPs, especially under RCP8.5.
• Under both RCPs, the largest levels of future wildfire activity are projected to occur in the
southwestern parts of both the contiguous U.S. and Alaska.
• Through 2100, the cumulative, discounted wildfire response costs in the contiguous U.S. and Alaska
under RCP8.5 are estimated at $24 billion. Other impacts, such as property damage or health effects
from decreased air quality, are not estimated, but would significantly increase economic damages.
• Compared to the more severe climate change scenario, RCP4.5 is projected to reduce the
cumulative area burned by wildfires in the contiguous U.S. and Alaska over the course of the 21st
century by approximately 60 million acres. The corresponding avoided response costs are estimated
at $75 million (cumulative, discounted).
26.2 INTRODUCTION
Terrestrial ecosystems in the U.S. provide a wealth of goods and services such as timber, wildlife habitat,
erosion management, water filtration, recreation, and aesthetic value. Climate change threatens these
ecosystems as heat, drought, and other disturbances have already led to an increased frequency of large
wildfires, as well as longer durations of individual wildfires and longer wildfire seasons in the western
U.S.427 Wildfires can damage property, disrupt ecosystem services, destroy timber stocks, impair air
quality, and result in loss of life.428 In the last decade (2006-2015), approximately 7 million acres of
forest have burned each year due to wildfires in the contiguous U.S. and Alaska, and the federal
government has spent approximately $1.5 billion per year on wildfire suppression 429 Additionally,
wildfires release carbon stored in terrestrial ecosystems, potentially further accelerating climate
change.430-431
427 Fann, N., T. Brennan, P. Dolwick, J.L. Gamble, V. Ilacqua, L. Kolb, C.G. Nolte, T.L. Spero, and L. Ziska, 2016: Ch. 3: Air Quality Impacts. The
Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington,
DC, 69-98, doi: 10.7930/J0GQ6VP6.
428 Groffman, P. M., P. Kareiva, S. Carter, N. B. Grimm, J. Lawler, M. Mack, V. Matzek, and H. Tallis, 2014: Ch. 8: Ecosystems, Biodiversity, and
Ecosystem Services. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 195-219. doi:10.7930/J0TD9V7H.
429 National Interagency Fire Center, 2016: Federal Firefighting Costs (Suppression Only). Available online at
https://www.nifc.gov/firelnfo/firelnfo documents/SuppCosts.pdf
430 Groffman, P. M., P. Kareiva, S. Carter, N. B. Grimm, J. Lawler, M. Mack, V. Matzek, and H. Tallis, 2014: Ch. 8: Ecosystems, Biodiversity, and
Ecosystem Services. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond,
and G. W. Yohe, Eds., U.S. Global Change Research Program, 195-219. doi:10.7930/J0TD9V7H.
431 Joyce, L. A., S. W. Running, D. D. Breshears, V. H. Dale, R. W. Malmsheimer, R. N. Sampson, B. Sohngen, and C. W. Woodall, 2014: Ch. 7:
Forests. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program, 175-194. doi:10.7930/J0Z60KZC.
189
-------
ECOSYSTEMS
Wildfire
26.3 APPROACH
This analysis projects wildfire activity and future response costs in the contiguous U.S. and Alaska using
models developed and calibrated for each geographic area. To simulate the effects of climate change on
areas burned by wildfires in the contiguous U.S., the analysis uses the MC2 dynamic global vegetation
model (DGVM) developed and run by the U.S. Forest Service's (USFS) Pacific Northwest Research
Station. The model simulates changes in future terrestrial ecosystem vegetative cover, including shifts in
vegetation types over time, and burned area across the contiguous U.S. in the 21st century, excluding
consideration of the proportion of a cell assumed to be in developed or in agricultural land use types.432
The MC2 model is driven by changes in future climate (e.g., temperature, precipitation, humidity) based
on the climate projections of five GCMs under two RCP scenarios.
The projected impacts of wildfires are summarized by scenario and geographic area,433 and then
monetized using average wildfire response costs for each region based on data from the National
Wildfire Coordinating Group.434 These costs include expenditures associated with labor (e.g., fire crews)
and equipment (e.g., helicopters, bulldozers) that are used for fire-fighting. Importantly, the analysis
adjusts projected changes in fire regime over time to account for fire suppression tactics. However, the
economic costs associated with the endogenous fire suppression tactics within MC2 are not accounted
for in the valuation results presented in this section (i.e., only wildfires that occur in spite of the
endogenous suppression are quantified and valued, therefore resulting in an underestimate of total
suppression costs). For more information on the MC2 model, including the wildfire module, and
calibration used in this analysis, please refer to Drapek et al. (2015)435 and Conklin et al. (2016).436 For
information on the approach to estimating wildfire response costs in the contiguous U.S., please refer to
Mills et al. (2014).437
The Alaska Frame-Based Ecosystem Code (ALFRESCO) model438 projects changes in wildfire activity in
Alaska. ALFRESCO is a spatially explicit model that simulates spatial processes of fire and recruitment
across the circumpolar arctic/boreal zone. The model combines disturbance events, seed dispersal, and
succession on a landscape at a spatio-temporal scale appropriate for investigating effects of climatic
change. In Alaska, wildfire suppression activities are prioritized based on spatially-delineated zones
called fire management option (FMO) regions. Table 26.1 provides a brief summary of characteristics of
432 A static layer of current agricultural lands based on the National Land Cover Dataset is removed from the vegetative mapping, along with a
dynamic layer of developed (urban and suburban) lands based on the ICLUSv2 projections described in the Modeling Framework section of this
Technical Report. For more information on the National Land Cover Dataset, see: Homer, C., J. Dewitz, L. Yang, S. Jin, P. Danielson, G. Xian, J.
Coulston, N. Herold, J. Wickham, and K. Megown, 2015: Completion of the 2011 National Land Cover Database for the conterminous United
States - Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, 81, 345-354.
433 Over the century-long modeled period, the same area can potentially burn in the future if the vegetation has recovered sufficiently to build
combustible fuel, and if appropriate conditions are met to cause ignition.
434 National Wildfire Coordinating Group, 2011: Historical incident ICS-209 reports. Available online at http://fam.nwcg.gov/fam-
web/hist 209/report list 209
435 Drapek, R.J., J.B. Kim, and R.P Neilson, 2015: Continent-wide Simulations of a Dynamic Global Vegetation Model overthe United States and
Canada under Nine AR4 Future Scenarios. Global Vegetation Dynamics: Concepts and Applications in the MCI Model, 73-90.
436 Conklin, D.R., J.M. Lenihan, D. Bachelet, R.P. Neilson, and J.B. Kim, 2016: MCFire model technical description. Gen. Tech. Rep. PNW-GTR-926.
Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. See also: Kim, J.B., Monier, E., Sohngen, B.,
Pitts, G., et al., 2017: Assessing climate change impacts, benefits of mitigation, and uncertainties on major global forest regions under multiple
socioeconomic and emissions scenarios. Environmental Research Letters, doi: 10.1088/1748-9326/aa63fc.
437 Mills, D., R. Jones, K. Carney, A. St Juliana, R. Ready, A. Crimmins, J. Martinich, K. Shouse, B. DeAngelo, and E. Monier, 2014: Quantifying and
Monetizing Potential Climate Change Policy Impacts on Terrestrial Ecosystem Carbon Storage and Wildfires in the United States. Climatic
Change. doi:10.1007/sl0584-014-1118-z.
438 Rupp, T.S., A.M. Starfield, and F.S. Chapin III, 2000: A frame-based spatially explicit model of subarctic vegetation response to climatic
change: comparison with a point model. Landscape Ecology, 15, 383-400, doi: 10.1023/A:1008168418778.
190
-------
ECOSYSTEMS
Wildfire
the four FMOs. The analysis relies on a configuration of ALFRESCO that considers the spatial distribution
of FMOs and adjusts the probability of fire spread between cells (for example, fire is less likely to spread
in FMOs with higher suppression priority).439 Projected annual area burned under RCP8.5 and RCP4.5 in
two GCMs (CCSM4 and GISS-E2-R)440 is aggregated by FMO to calculate response costs using historical
(2002-2013) data from the Alaska Interagency Coordination Center and the National Fire and Aviation
Management web application. For more information on the approach and results to estimating wildfire
response costs in Alaska, please refer to Melvin et al. (2017).441
Table 26.1. Characteristics of the Four Fire Management Option Regions in Alaska
FMO
Characteristics
Area
(millions of
acres)
Critical
Highest priority areas where the risks to human life, residences, and
community-dependent infrastructure are greatest. Immediate action is
taken to suppress all wildfires.
3.2
Full
High-valued areas, but where risks to human life or inhabited property are
low. Fires are usually suppressed.
53
Modified
Areas where risks to humans and infrastructure are relatively low and
management decisions are designed to balance area burned with
suppression costs. Suppression occurs if there is high fire danger.
42
Limited
Remote areas with a low density of valuable property, allowing for natural
fire dynamics and associated ecological processes. In general, no actions
are taken to suppress these fires.
264
26.4 RESULTS
Contiguous U.S.
Figure 26.1 shows the annual acres burned by wildfires in the contiguous U.S. through 2100 driven by
climate projections from five GCMs under two RCPs, as well as for an average across the climate models.
The large inter-annual variability reflects simulated periods of fuel accumulation followed by seasons of
large wildfire activity - a trend similar to the variability observed over the past several decades.442 In
general, projected annual acres burned across the contiguous U.S. under RCP8.5 remain consistent with
439 Two hundred replicate simulations for the 1901-2100 time period are simulated for each of the ten GCM x RCP combinations. Because
ALFRESCO is a stochastic model, each replicate simulation produced a different spatial and temporal pattern of wildfire occurrence. The median
values across the two hundred simulations are presented in this Technical Report, which allows for greater statistical certainty in projected
outcomes for each FMO.
440 As described in the Modeling Framework section of this Technical Report, the SNAP downscaled climate projections only contained two
GCMs (two of five available) which overlapped with the climate models being used forthe contiguous U.S. Therefore, this section presents
wildfire modeling results using these two GCMs. See Melvin et al. (2017) for results using all five of the SNAP GCMs.
441 Melvin, A.M., J. Murray, B. Boehlert, J.A. Martinich, L. Rennels, and T.S. Rupp, 2017: Estimating wildfire response costs in Alaska's changing
climate. Climatic Change Letters, doi: 10.1007/sl0584-017-1923-2. Available online at http://link.springer.eom/article/10.1007%2Fsl0584-017-
1923-2
442 National Interagency Fire Center, 2016: Federal Firefighting Costs (Suppression Only). Available online at
https://www.nifc.gov/firelnfo/firelnfo documents/SuppCosts.pdf
191
-------
ECOSYSTEMS
Wildfire
levels observed over the past several decades, but decrease slightly under RCP4.5. MC2 output based ori
the MIROC5 climate projections show the largest level of wildfire activity (a projected 320 million acres
through 2100), while activity is smallest under the HadGEM2-ES GCM (250 million acres through 2100),
Figure 26.1. Projected Wildfire Activity
Estimated annual acres burned (millions) by wildfire in the contiguous U.S. over the course of the 21st
century under RCP8.5 and RCP4.5 in the five GCMs, with the five-GCM average shown in black,443
CCSM4
GISS-E2-R
HadGEM2-ES
MIROC5
5-Model Average
CanESM2
2090
0
2006
RCP8.5
2030 2050 2070
20
18
16
14
12
10
8
6
4
2_
0.
2006
2030
RCP4.5
2050 2070 2090
The projections of future wildfire activity vary considerably by region. Figure 26.2 shows the projected
change in average annual acres burned in 2050 and 2090 compared to the reference period (1986-
2005). As shown, parts of the Southwest (e.g., Colorado and Nevada) and New England are projected to
experience the highest levels of wildfire activity by the end of the century, though the spatial patterns of
change vary by GCM (see Figure A14.1 of the Appendix for GCM-specific maps). In some regions that
experience large levels of wildfire activity today, such as California, the MC2 model projects a shift
toward fewer acres burned, due to vegetation converting to types that burn less frequently. These
modeling results may indicate the exceedance of an ecosystem threshold, whereby historically-
dominant plant species are replaced with those having less-frequent fire return intervals. These shifts
result in lower levels of future burning than would be anticipated in an analysis that does not include
these dynamic changes.
,43 To enable the reader to more clearly see differences amongst the lines for each GCM, the following two values were excluded from the
graphic: the CCSM4 estimate for 2024 under RCP4.5 (27 million acres) and the MIROC5 estimate for 2034 under RCP4.5 (21 million acres).
These values are included in all other cumulative impact and economic results in this section.
192
-------
ECOSYSTEMS
Wildfire
Figure 26.2. Projected Change in Wildfire Activity
Projected change in average annual acres burned across the contiguous U.S. under RCP8.5 and RCP4.5 by
mid-century (2040-2059) and end of century (2080-2099) compared to the reference period (1986-2005).
Results shown represent the average of the five GCMs. Acres burned include all vegetation types and are
calculated at a cell resolution of l/16th of a degree, and aggregated to 1A degree for mapping purposes.
Agricultural and developed lands are removed.
2090
Change in Acres Burned
-28.000 to-10,000
-9,999 to -5,000
-4,999 to -2,500
-2,499 to 0
1 to 2,500
2,501 to 5,000
5,001 to 11,000
| | Agricultural and Developed Lands
RCP8.5
2050
RCP4.5
Table 26.2 shows the cumulative acres burned and discounted wildfire response costs by region and for
the contiguous U.S. Cumulatively through the end of the century, approximately 330 million acres are
projected to burn under RCP8.5 and 290 million under RCP4.5 (average of the five GCMs). The difference
between the RCP8.5 and RCP4.5 estimates represents a 12% reduction in cumulative wildfire acreage in
the contiguous U.S. due to reduced levels of climate change. In terms of economic effects, the
cumulative wildfire response costs are estimated at $23 biiiion through 2099 under RCP8.5 and $23
billion under RCP4.5 (2015$, discounted at 3%), with the difference between the RCPs equaling $55
million. At a regional level, the Southwest is projected to experience the largest amount of wildfire
activity and incur the highest response costs through 2099. Conversely, the Southeast is projected to
have the smallest level of wildfire activity and subsequent response costs.
193
-------
ECOSYSTEMS
Wildfire
Table 26.2. Projected Acres Burned and Response Costs
Cumulative values represent the average of the five GCMs over the 2006-2099 period by region and the
national (contiguous U.S.) total. Totals may not sum due to rounding.
Cumulative Acres Burned
(millions of acres)
Cumulative Response Costs
(millions of $2015, discounted at 3%)
Region
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Northeast
36
28
$640
$600
Southeast
15
13
$230
$300
Midwest
26
21
$480
$440
Northern Plains
36
33
$3,300
$3,700
Southern Plains
14
11
$380
$640
Southwest
150
150
$13,000
$13,000
Northwest
47
40
$5,400
$4,800
National Total
330
290
$23,000
$23,000
Alaska
Climate change is one of several factors that will affect the distribution, extent, and cost of wildfires in
Alaska throughout this century. Figure 26.3 shows the projected change in average annual acres burned
in Alaska under RCP8.5 and RCP4.5 based on outputs of the ALFRESCO model. The relative change in
mean annual acres burned between the reference period (1970-2005) and 2050 and 2090 indicates an
increase in area burned under RCP8.5 across much of the state. The largest increase in area burned
under RCP8.5 is projected in the Southwest and western parts of the Far North, while results under
RCP4.5 project a smaller increase in the Southwest and a decline in acres burned across much of rest of
the State. Under all scenarios and timeframes, the eastern parts of the State show a decrease in wildfire
activity, and the North Slope is generally projected to experience a slight increase in burned area. The
projected shifts in burning over time are influenced by changes in vegetation composition, where forests
historically dominated by black spruce are replaced with deciduous forest. This transition can influence
the flammability of forested areas, including the modeled fire return intervals within ALFRESCO.444
444 See Melvin et al. (2017) for more detail.
194
-------
ECOSYSTEMS
Wildfire
Figure 26.3. Projected Impact of Climate Change on Alaska Wildfire Activity
Change in average annual acres burned under RCP8.5 and RCP4.5 (mean of two GCMs) by mid-century
(2040-2059) and end of century (2080-2099) compared to the reference period (1970-2005). Results are
presented at a 1A degree cell resolution.
RCP8.5
RCP4.5
2050
2090
-f -
r"T *,4"
r- v''
(
Ik
vm*y
\
it i
' v
*-
¦
;'%v
Change in #
of Acres
4000
3000
2000
1000
0
-1000
-2000
-3000
-4000
As shown in Table 26.3, the analysis projects approximately 120 million burned acres through the end of
the century under RCP8.5, and 98 million burned acres under RCP4.5 (average of the five GCMs). The
difference in area burned across Alaska between the two RCPs averages approximately 22 million acres
through 2100. Projected cumulative, discounted wildfire suppression costs are estimated at $1.1 billion
(2015$) under both RCPS.5 and RCP4.5. The difference in projected costs between the two RCPs
averages approximately $20 million through 2100, with RCP4.5 providing modest benefits. An important
limitation of the valuation approach used in this analysis is that per-acre response costs only represent
federal costs, which are a subset of total suppression costs.445
Differences in response costs across the FMOs are influenced by the size of each FMO, the projected
area burned, and the response cost per unit area. While the Limited FMO covers the largest area and
contained over half the projected cumulative acres burned, this FMO accounted for only about 18% of
the projected incurred costs due to the low response cost per acre. In contrast, the Full FMO, whose
wildfires have larger response costs per acre, accounted for only 17% of the cumulative area burned, but
about 65% of incurred costs. Southwest Alaska contains a large area designated as Full and showed the
largest projected relative increase in burning (Figure 26.3), likely driving this finding.
445 For instance, between 2009-2015, an average of 68% of annual costs were incurred by the state. This indicates that the cost per acre values
used are likely an underestimate of what will be realized in thefuture.
195
-------
ECOSYSTEMS
Wildfire
Table 26.3. Projected Cumulative Area Burned and Wildfire Response Costs in Alaska
Results represent values for the 2006-2100 periodand are shown for each FMO, along with the State
total. Totals may not sum due to rounding.
Cumulative Acres Burned by FMO
(in millions)
GCM
RCP
Critical
Full
Modified
Limited
State Total
CCSM4
RCP8.5
0.72
23
20
89
130
110
RCP4.5
0.66
19
17
77
GISS-E2-R
RCP8.5
0.57
18
16
71
110
86
RCP4.5
0.47
15
13
58
Average
RCP8.5
0.65
21
18
80
120
98
RCP4.5
0.57
17
15
68
Cumulative Response Costs by FMO
(in millions of $2015, discounted at 3%)
CCSM4
RCP8.5
$53
$840
$180
$230
$1,300
$1,400
RCP4.5
$57
$900
$200
$250
GISS-E2-R
RCP8.5
$36
$570
$130
$160
$890
$750
RCP4.5
$32
$490
$100
$130
Average
RCP8.5
$45
$710
$160
$200
$1,100
$1,100
RCP4.5
$45
$700
$150
$190
Contiguous U.S. and Alaska Combined
Table 26.4 provides the combined cumulative acres burned and wildfire response costs in the
contiguous U.S. and Alaska through the end of the century.
Table 26.4. Projected Cumulative Area Burned and Wildfire Response Costs
Cumulative values for the contiguous U.S. and Alaska represent the average of the five GCMs over the
2006-2099 period. Totals may not sum due to rounding.
Cumulative Acres Burned (in
millions)
Cumulative Response Costs (in
millions of $2015, discounted at 3%)
RCP8.5
450
$24,000
RCP4.5
390
$24,000
Difference
60
$75
26.5 DISCUSSION
To place the estimates of this Technical Report in context with observations with recent history,
approximately 7 million acres of forest burned each year between 2006-2015 due to wildfires in the
contiguous U.S. and Alaska, and the federal government spent about $1.5 billion per year on
196
-------
ECOSYSTEMS
Wildfire
suppression.446 During this same period, the MC2 and ALFRESCO models project 3.9 million burned acres
and the economic valuation methods estimate $510 million in response costs each year.447 As such,
these modeling approaches appear to underestimate burning in the historic period by a factor of
approximately two, and response costs by a factor of three.448 Therefore, the values reported in this
section should be treated as conservative estimates. Reasons for this discrepancy include the fact that
endogenous fire suppression tactics within MC2 are not captured in the valuation methodology used for
estimating response costs in the contiguous U.S., and that the valuation approach used to estimate
response costs in Alaska is based on data of federal expenditures that does not include state costs,
which can equal 50% or more of total expenditures.
Few empirical studies have sought to quantify future wildfire response costs at these large geographic
scales, but some recent studies provide a basis for general comparison. Econometric modeling based on
observed wildfire trends449 found that federal wildfire suppression costs in the contiguous U.S. under a
high GHG emissions scenario will increase by 46% (31%-69%) by mid-century compared to reference
expenditures and 39% (15%-77%) by late-century. The vegetation modeling-based approach used in the
analysis of this Technical Report estimates a 12% decrease (36% decrease to 7% increase) in annual
average wildfire suppression costs for the contiguous U.S. under RCP8.5 in 2050 (compared to the
reference), and a 18% decrease (l%-53%) in 2090. Differences in results are likely attributed to the
methods for simulating future wildfire (i.e., statistically versus dynamically in a vegetation model) and
the use of different GCMs. Recent research has investigated the differences between statistical and
dynamic vegetation modeling approaches, and found that including changes in vegetation and the
drought-fire dynamics are important, and may challenge the assumption that warming climates will
result in increased burned area.450 However, additional research is needed to fully understand the
structural differences between modeling approaches and their subsequent effects on results.
As a comparison for the Alaska results, a recent study on Canadian wildfire response costs under climate
change estimated average annual costs to increase by 119% (compared to the reference period) under
RCP8.5 by the end of the century and 60% under RCP4.5.451 The results presented in this Technical
Report indicate that response costs in Alaska will increase by approximately 68% in 2090 under RCP8.5
(compared to the reference period) and 14% under RCP4.5. Differences in results between the two
studies are likely attributed to the different geographic regions analyzed, different underlying wildfire
methods, and the use of different GCMs.
Several caveats to the results presented in this section are important to consider. First, the wildfire
valuation analyses do not quantify human health impacts associated with worsened air quality, property
446 National Interagency Fire Center, 2016: Federal Firefighting Costs (Suppression Only). Available online at
https://www.nifc.gov/firelnfo/firelnfo documents/SuppCosts.pdf
447 These results are for RCP8.5, though values under RCP8.5 and RCP4.5 do not differ much in these early years.
448 Dynamic vegetation and wildfire models simulate vegetation and fire interactions on a spatially broad, long-term basis. The interaction
between fire and vegetation is non-linear, and dynamic overtime. Fire can have threshold effects on vegetation, and vice-versa. Wildfires are
highly complex and stochastic processes. For example, a) wildfire ignitions are not only stochastic, but in many parts of the country, the
patterns are highly correlated with anthropogenic activities; b) wildfire spread and severity is highly dynamic and dependent on fine spatial
scale patterns of fuels, topography and weather, and c) fire suppression is an active policy in many parts of the country, although particular
suppression activities are not consistent across the country. Forthese reasons, a model simulating general patterns of wildfire occurrence and
effects on a continental scale is unlikely to perfectly replicate historical observations.
449 Office of Management and Budget, 2016: Climate Change: Fiscal Risks Facing the Federal Government. Executive Office of the President.
450 McKenzie, D., J.S. Littell, 2016: Climate change and the eco- hydrology of fire: Will area burned increase in a warming western USA?
Ecological Applications: 27(l):26-36.
451 Hope, E.S., D.W. McKenney, J.H. Pedlar, B.J. Stocks, and S. Gauthier, 2016: Wildfire Suppression Costs for Canada under a Changing Climate.
PLoSONE, 11, e0157425, doi:10.1371/journal.pone.0157425
197
-------
ECOSYSTEMS
Wildfire
damage and loss, and loss of recreation, all of which could have large economic implications. Second,
continued expansion of development into the wildland-urban interface could increase the number of
human-ignited fires and the size of the area designated for higher suppressive action. The wildfire
modeling conducted does not account for these interactions. Third, the MC2 and ALFRESCO models do
not account for the effects of pest infestations (e.g., pine bark beetles) and tree disease, which can
make trees susceptible to burning and affect wildfire activity. Fourth, the analyses do not assume any
shifts in the general approach to wildland fire management, such as changes in suppression
technologies or fuels management, which could affect future burned acreage and response costs.
Finally, new evolving research indicates that climate warming may increase lightning strikes,452 which
could influence the frequency of lightning-caused fires in the future if increases in precipitation do not
serve to counteract these effects.
4 52 Romps, D.M., J.T. Seeley, D. Vollaro, and J. Molinari, 2014: Projected increase in lightning strikes in the United States due to global warming.
Science, 346, 851-854, doi: 10.1126/science.l259100.
198
-------
ECOSYSTEMS
Carbon Storage
27. CARBON STORAGE
27.1 KEY FINDINGS
• Carbon flow projections in the contiguous U.S. demonstrate high inter-annual variability, with the
magnitude and even directionality (from sinks to sources) of impacts varying over time under both
RCP8.5 and RCP4.5.
• Though carbon flows show substantial regional variation, overall national terrestrial ecosystem
carbon storage is projected to increase by 3.0 billion metric tons under RCP8.5 and 0.36 billion
metric tons under RCP4.5 through the end of the century. The Northwest is projected to experience
the largest increase in stored carbon through 2100, while the Northeast and Midwest experience
losses under both RCPs.
27.2 INTRODUCTION
Terrestrial ecosystems influence and are influenced by climate change through their important role in
the global carbon cycle. These ecosystems capture and store carbon from the atmosphere, with
different systems and plant species storing carbon over various timeframes, which can reduce
atmospheric concentrations of carbon dioxide and related climate impacts. However, they can also act
as a source, releasing carbon through decomposition, wildfires, and other forms of combustion (e.g.,
bioenergy, open burning). Terrestrial ecosystems in the U.S., which include forests, grasslands, and
shrublands, are currently a net carbon sink.453 Forest carbon storage has increased over the past several
decades due to net increases in forest area and improved forest management, as well as higher
productivity rates and longer growing seasons driven by climate change.454
While warming temperatures and rising carbon dioxide levels can increase grassland productivity and
carbon sequestration, grassland carbon is also sensitive to precipitation and may shift from sinks to
sources in some regions in response to drought.455 Climate-driven changes in the distribution of
vegetation types, wildfire, pests, and disease are affecting, and will continue to affect, U.S. terrestrial
ecosystem carbon storage.456
27.3 APPROACH
This analysis simulates climate change effects on terrestrial vegetative carbon storage in the contiguous
U.S. Changes in carbon storage and annual flows are calculated using the MC2 dynamic global
453 Land use, land-use change, and forestry (LULUCF) activities in 2014 resulted in a net increase in C stocks (i.e., net CO2 removals) of 787.0
MMT CO2 Eq. (214.6 MMT C). This represents an offset of approximately 11.5 percent of total (i.e., gross) greenhouse gas emissions in 2014.
Emissions from land use, land-use change, and forestry activities in 2014 are 24.6 MMT CO2 Eq. and represent 0.4 percent of total greenhouse
gas emissions. Total C sequestration in the LULUCF sector increased by approximately 4.5 percent between 1990 and 2014. Source: EPA, 2016:
Forest sections of the Land Use, Land Use change, and Forestry chapter, and Annex. In: U.S. Environmental Protection Agency, Inventory of U.S.
Greenhouse Gas Emissions and Sinks: 1990-2013. EPA 430-R-15-004.
454 Galloway, J. N., W. H. Schlesinger, C. M. Clark, N. B. Grimm, R. B. Jackson, B. E. Law, P. E. Thornton, A. R. Townsend, and R. Martin, 2014: Ch.
15: Biogeochemical Cycles. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.)
Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 350-368. doi:10.7930/J0X63JT0.
455 Scott, R. L., J. A. Biederman, E. P. Hamerlynck, and G. A. Barron-Gafford, 2015: The carbon balance pivot point of southwestern U.S. semiarid
ecosystems: Insights from the 21st century drought. Journal of Geophysical Research-Biogeosciences, 120, 2612-2624.
456 Joyce, L. A., S. W. Running, D. D. Breshears, V. H. Dale, R. W. Malmsheimer, R. N. Sampson, B. Sohngen, and C. W. Wood, 2014: Ch. 7:
Forests. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W.
Yohe, Eds., U.S. Global Change Research Program. doi:10.7930/J0Z60KZC.
199
-------
ECOSYSTEMS
Carbon Storage
vegetation model developed and run by the USFS' Pacific Northwest Research Station.457 The model
simulates changes in future terrestrial vegetative growth and cover (e.g., grasses, shrubs, hard and
softwood forests), including shifts in vegetation types over time, and burned area across the contiguous
U.S. from 2015 to the end of the century,458 excluding consideration of the proportion of a cell assumed
to be in developed or in agricultural land use types.459 MC2 is driven by changes in future climate (e.g.,
temperature, precipitation, humidity) based on climate projections of five GCMs under two RCP
scenarios for four future periods using a 20-year averaging window around the representative year:
2030 (2020-2039), 2050 (2040-2059), 2070 (2060-2079), and 2090 (2080-2099). Projected vegetative
cover estimates and the resulting carbon flux from year to year represent changes in climate,
biogeography, biogeochemistry, and wildfire dynamics. Projected annual changes in terrestrial carbon
flow for non-agricultural, non-developed lands across the contiguous U.S. are summarized by scenario
and geographic area in this Section of the Technical Report.
Projected changes in carbon storage have been monetized previously using the social cost of carbon
dioxide.460-461,462 This analysis did not consider the effects of future changes in ozone, pests, and disease,
which could influence the ability of U.S. terrestrial ecosystems to store carbon. For more information on
the MC2 model and calibration used in this analysis, please refer to Drapek et al. (2015)463 and Conklin
et al. (2016).464 For information on the approach to valuing changes in carbon storage, please refer to
Mills et al. (2014).465
27.4 RESULTS
Carbon flow (or flux) projections in the contiguous U.S. fluctuate and are highly variable by GCM, with
the magnitude and even directionality (from sinks to sources) of impacts varying over time under both
RCPs (Figure 27.1). The large inter-annual variability in carbon flow likely reflects similarly high inter-
annual variability in annual acres burned by wildfires under both RCPs (see Figure 27.1 in Wildfire
section), for which MC2 simulates periods of fuel accumulation (carbon storage) followed by seasons of
457 Drapek, R.J., J.B. Kim, and R.P Neilson, 2015: Continent-wide Simulations of a Dynamic Global Vegetation Model overthe United States and
Canada under Nine AR4 Future Scenarios. Global Vegetation Dynamics: Concepts and Applications in the MCI Model, 73-90.
458 Although the MC2 simulations under each RCP/GCM combination start in the year 2006, results in this section begin in 2015 due to the fact
that values are discounted back to 2015 (consistent with the approach across this Technical Report), and because the application of the
dynamic developed lands layer begins in 2010.
459 A static layer of current agricultural lands based on the National Land Cover Dataset is removed from the vegetative mapping, along with a
dynamic layer of developed (urban and suburban) lands based on the ICLUSv2 projections described in the Modeling Framework section of this
Technical Report. For more information on the National Land Cover Dataset, see: Homer, C., J. Dewitz, L. Yang, S. Jin, P. Danielson, G. Xian, J.
Coulston, N. Herold, J. Wickham, and K. Megown, 2015: Completion of the 2011 National Land Cover Database for the conterminous United
States - Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, 81, 345-354.
460 Mills, D., R. Jones, K. Carney, A. St Juliana, R. Ready, A. Crimmins, J. Martinich, K. Shouse, B. DeAngelo, and E. Monier,. 2014: Quantifying and
Monetizing Potential Climate Change Policy Impacts on Terrestrial Ecosystem Carbon Storage and Wildfires in the United States. Climatic
Change, 131, 163-178, doi:10.1007/sl0584-014-1118-z.
461 Economic estimates of carbon storage changes are not reported in this Technical Report in response to the March 28, 2017 Executive Order
on Promoting Energy Independence and Economic Growth, which withdrew the U.S. Government's social cost of carbon dioxide estimates.
462 For a review of the latest science and recommendations related to estimating the social cost of carbon dioxide, see the January 2017 report
of the National Academies of Sciences: http://sites.nationalacademies.org/dbasse/becs/valuing-climate-damages/
463 Drapek, R.J., J.B. Kim, and R.P Neilson, 2015: Continent-wide Simulations of a Dynamic Global Vegetation Model overthe United States and
Canada under Nine AR4 Future Scenarios. Global Vegetation Dynamics: Concepts and Applications in the MCI Model, 73-90.
464 Conklin, DR, JM Lenihan, D Bachelet, RP Neilson, and JB Kim, 2016: MCFire model technical description. Gen. Tech. Rep. PNW-GTR-926.
Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station.
465 Mills, D., R. Jones, K. Carney, A. St Juliana, R. Ready, A. Crimmins, J. Martinich, K. Shouse, B. DeAngelo, and E. Monier,. 2014: Quantifying and
Monetizing Potential Climate Change Policy Impacts on Terrestrial Ecosystem Carbon Storage and Wildfires in the United States. Climatic
Change, 131, 163-178, doi:10.1007/sl0584-014-1118-z.
200
-------
ECOSYSTEMS
Carbon Storage
large wildfire activity (carbon loss) - a trend similar to the variability observed over the past several
decades.466'467
Figure 27.1. Projected Annual Carbon Flow
Estimated annual carbon flow (billions of metric tons) in the contiguous U.S. over the course of the 21st
century under RCP8.5 and RCP4.5 in the five GCMs, with the five-GCM average shown in black.468
CanESM2
CCSM4
GISS-E2-R
HadGEM2-ES
MIROC5
5-Model Average
2030
2090
1.0
0.5
0
-0.5
-1.0
2015
2030
2050 2070
2090
-1.0
2015
RCP8.5
2050 2070
There is also substantial regional variation in vegetative carbon. Figure 27.2 shows changes in annual
average carbon stored in the 20-year eras around 2050 and 2090 across the contiguous U.S. compared
to the reference period (see Figure A15.1 of the Appendix for GCM-specific maps). At a national level,
carbon flows from vegetation is positive in 2050, with an increase in average carbon storage of 28
million metric tons under RCP8.5 and 9.8 million metric tons under RCP4.5. These values increase by
2090 to 88 million metric tons under RCP8.5 and 11 million metric tons under RCP4.5 in 2090. The
amount of carbon stored by vegetation in Southeast, Southwest, Northwest, and Northern Plains
increases under both RCPs, particularly later in the century, while carbon storage is projected to
generally decrease under both RCPs in the Midwest and Northeast. The amount of carbon stored by
vegetation in the Southern Plains increases under RCP8.5, but decreases for RCP4.5.
466 National Interagency Fire Center, 2016: Federal Firefighting Costs (Suppression Only). Available online at
https://www.nifc.eov/fireliifo/fireliifo documents/SuppCosts.pdf
467 EPA, 2.016: Forest sections of the Land Use, Land Use change, and Forestry chapter, and Annex. In: U.S. Environmental Protection Agency,
Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2013. EPA 430-R-15-004.
468 Carbon flow losses appearing every decade (e.g., in 2020, 2030) across all scenarios are driven by the re-estimation of developed lands and
the application of this mask layer to the vegetation modeling.
201
-------
ECOSYSTEMS
Carbon Storage
Figure 27.2. Projected Percent Change in Carbon Stored
Projected percent change in the five-GCM average annual carbon stock across the contiguous U.S. under
RCP8.5 and RCP4.5 by mid-century (2040-2059) and end of century (2080-2099) compared to the
reference period (1986-2005). Changes in carbon stock calculated at a cell resolution of 1/16th of a
degree and converted to 1A degree for mapping purposes. Agricultural and developed lands are omitted
(shown in gray).
2050
2090
Percent Change in Carbon Stock
-17 to -8
1H -7 to o
1 to 10
| 11 to 20
21 to 30
| 31 to 45
| Agricultural and Developed Lands
Table 27,1 shows the cumulative changes in carbon storage from 2015-2099 at regional and national
levels, along with the monetized value of these changes. Through the end of the century, national
terrestrial ecosystem carbon storage is projected to increase by 3.0 billion metric tons under RCP8.5 and
0.36 billion metric tons under RCP4.5. Moderate losses of carbon storage are projected for the
Northeast and Midwest under both RCPs (approximately 0.6 billion metric tons) and also in the Southern
Plains under RCP4.5 (Figure 27.2 and Table 27.1). The Northwest is projected to have the largest
cumulative increase in vegetative carbon of 1.3 billion metric tons under RCP8.5 by the end of the
century.
RCP8.5
RCP4.5
202
-------
ECOSYSTEMS
Carbon Storage
Table 27.1. Projected National and Regional Carbon Storage
Values represent the average of the five GCMs over the 2015-2099 period. Totals may not sum due to
rounding.
Change in Carbon Storage
(billions of metric tons)
Region
RCP8.5
RCP4.5
Northeast
-0.56
-0.60
Southeast
0.94
0.17
Midwest
-0.60
-0.59
Northern Plains
0.55
0.28
Southern Plains
0.32
-0.12
Southwest
1.00
0.37
Northwest
1.30
0.85
National Total
3.0
0.36
27.5 DISCUSSION
The observed rise in global atmospheric carbon dioxide concentration is lower than would be expected
given anthropogenic greenhouse gas emissions, demonstrating continued uptake of carbon by the
ocean and terrestrial vegetation over the last 50 years.469 The findings of this analysis are consistent with
recent literature470, which projects increases in forest and grassland carbon stocks across North America
in the future, a trend that also follows observations over the last few decades.471 However, uncertainty
remains regarding the magnitude of carbon storage changes across the contiguous U.S., with projections
in the literature depending upon the modeling approaches used, climate drivers, and assumptions
regarding future changes in landuse and land cover.
This analysis omits carbon storage changes on agricultural lands, urban landscapes, and in non-
terrestrial ecosystems, such as tidal marshes and seagrass beds. In addition, many factors beyond rising
carbon dioxide will have an influence on carbon storage in the U.S., including changes in land use,
management practices (e.g. changes in wildfire suppression and fuel management actions), and other
environmental factors that may limit or enhance plant growth or result in shifts in vegetation type. As
forests make up approximately 80% of the aggregate North American carbon sink for atmospheric
carbon, deforestation could have large impacts on U.S. carbon storage.472 Please refer to Mills et al.
(2014) for complete discussion about the limitations and uncertainty associated with the analysis.473
Many of the caveats described in the Wildfire section also apply to the carbon storage results presented
in this section. For instance, continued expansion of development into the wildland-urban interface
469 Ballentyne, A.P., C.B. Alden, J.B. Miller, P.P. Tans, and J.W.C. White, 2012: Increase in observed net carbon dioxide uptake by land and oceans
during the past 50 years. Nature, 488, 70-72.
470 Raczka, B.M., K.J. Davis, D. Huntzinger, R.P. Neilson, B. Poulter, A.D. Richardson, J. Xiao, I. Baker, P. Ciais, T. F. Keenan, B. Law, W.M. Post, D.
Ricciuto, K. Schaefer, H. Tian, E. Tomelleri, H. Verbeeck, and N. Viovy, 2013: Evaluation of continental carbon cycle simulations with North
American flux tower observations. Ecological Monographs, 83, 531-556.
471 EPA, 2016: Forest sections of the Land Use, Land Use change, and Forestry chapter, and Annex. In: U.S. Environmental Protection Agency,
Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2103. EPA 430-R-15-004.
472 Ibid.
473 Mills, D., R. Jones, K. Carney, A. St Juliana, R. Ready, A. Crimmins, J. Martinich, K. Shouse, B. DeAngelo, and E. Monier, 2014: Quantifying and
Monetizing Potential Climate Change Policy Impacts on Terrestrial Ecosystem Carbon Storage and Wildfires in the United States. Climatic
Change, 131, 163-178, doi:10.1007/sl0584-014-1118-z.
203
-------
ECOSYSTEMS
Carbon Storage
could increase the number of human-ignited fires and the size of the area designated for higher
suppressive action. Climate change may also impact vegetative diseases or pest infestations (e.g., pine
bark beetles). These impacts are not included in the modeling. See the Wildfire section for additional
limitations of the modeling approach.
204
-------
SYNTHESIS OF RESULTS
28. NATIONAL SUMMARY
Figure 28.1 and tables 28.1 and 28.2 provide an overview of the national-scale results presented
throughout this Technical Report. Focusing on physical effects (Table 28.1) and economic impacts
(Figure 28.1 and Table 28.2), these summaries present the estimated annual effects of climate change in
the U.S. under RCP8.5 and RCP4.5 in the years 2050 and 2090 for the impact sectors considered in this
Technical Report. Although not available for all sectors, cumulative impacts for the entire 21st century
would likely be much larger than the annual estimates presented. In addition, the individual monetized
estimates are not aggregated, as only a subset of climate change impacts is quantified in this report.
Importantly, many of the reported values do not estimate and monetize the full extent of potential
impacts from climate change on that sector, and as such, the results should be treated as conservative.
For example, the air quality analysis only estimates economic damages from mortality caused by
changes in ozone, omitting effects from changes in other air pollutants and morbidity effects, while the
extreme temperature mortality analysis only includes 49 major cities, covering about one third of the
population. In addition, interactive effects across sectors are only modeled in several instances (e.g.,
irrigation for agriculture informed by water supply/demand model), therefore omitting potentially
important compounding impacts. Please refer to the sectoral sections of this report describing the
individual modeling efforts for detailed information on the results, a summary of the methodologies
used, and references to the supporting peer-reviewed literature.
As shown in Figure 28.1 and Tables 28.1 and 28.2, annual impacts and damages are projected to
increase over time and are generally larger under RCP8.5 compared to RCP4.5. Projected impacts on
extreme temperature mortality, outdoor labor, and coastal property are the most economically
significant under both time periods and RCPs. Estimated impacts on air quality and road infrastructure
are also large. For the wildfire sector, climate change, on average, is projected to result in a decrease in
economic impacts in the future. It is important to note that while the magnitude of estimated economic
impacts for some of the sectors is relatively small, many of the physical impacts have significant societal
or iconic values.
205
-------
SYNTHESIS OF RESULTS
National Summary
28.1. Annual Damages from Climate Change
Mean estimates of annual climate change damages are shown in $millions for RCP8.5 and RCP4.5 in 2050 and 2090. The three graphs are on
different scales to capture the range of impacts. Note that graph (c) includes negative damages (benefits). The data underlying this graphic can
be found in Table 28.2.
«/) 140000
c
O
E
LO 105000
& 70000
(U
E
fU
O
"to 35000
•
|$35b
•
•
o
•
—o—
o
• o
o
•
9
J$500m
o
•
•
o
•
8
s
0
§
•
O
8
•
v %, v v \
% V X
\ \ \ V
v * % \
\» ¦%-„ V ^
XV\\XN
V\\
%
O RCP8.5 2050 o RCP4.5 2050 # RCP8.5 2090 # RCP4.5 2090
206
-------
SYNTHESIS OF RESULTS
National Summary
Table 28.1. Projected Annual Physical Impacts of Climate Change across U.S. Sectors Analyzed
Only some of the sectoral analyses produced discrete physical metric estimates, in contrast to Table 28.2 that provides economic impacts across
all sectors of this report Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages
compared to the reference period. Unless noted at the bottom of Table 28.2, upper and lower bounds are based on values across the GCMs. See
notes at the bottom of Table 28.2 for additional sector-specific information.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality: # Deaths
790
(420 to 1,200)
550
(300 to 810)
240
(NA)
1,700
(920 to 2,500)
1,200
(630 to 1,700)
500
(NA)
Extreme Temperature Mortality:
# Deaths (thousands)
3.4
(2.3 to 5.9)
2.6
(1.7 to 3.9)
0.88
(0.17 to 2.0)
9.3
(5.4 to 13)
3.9
(2.4 to 7.4)
5.4
(2.8 to 8.1)
Labor: Lost Labor Hours
(millions)
880
(500 to 1,400)
700
(380 to 1,100)
180
(-24 to 290)
1,900
(1,000 to 2,700)
970
(620 to 1,500)
910
(420 to 1,300)
Aeroallergens: ED visits
(thousands)
1.2
(0.068 to 1.8)
0.90
(0.19 to 1.6)
0.30
(-0.12 to 0.83)
2.5
(0.87 to 3.5)
1.1
(-0.081 to 1.9)
1.4
(0.95 to 1.9)
Harmful Algal Blooms: # Days
above 100k cells/mL
9.2
(5.4 to 15)
8.4
(6.8 to 13)
0.71
(-0.88 to 2.3)
15
(6.5 to 24)
9
(2.7 to 15)
5.7
(2.2 to 11)
West Nile Virus:
# Cases (thousands)
1.3
(0.92 to 1.8)
1.0
(0.72 to 1.4)
0.23
(0.19 to 0.33)
3.3
(2.0 to 4.6)
1.7
(1.2 to 2.4)
1.6
(0.81 to 2.2)
INFRASTRUCTURE
Bridges: # Vulnerable Bridges
(thousands)
4.6
(3.3 to 6.1)
2.5
(1.6 to 3.5)
2.1
(0.88 to 4.1)
6.0
(2.4 to 8.8)
5.0
(3.1 to 6.3)
0.99
(-0.67 to 3.2)
WATER RESOURCES
Winter Recreation:
Lost Visits (millions)
12
(-3.3 to 20)
-4.5
(-10 to-1.1)
16
(-2.2 to 30)
28
(4.7 to 38)
2.5
(-18 to 14)
25
(23 to 30)
AGRICULTURE
Agriculture:
% Decrease in Corn Yields
(example crop)
5.8%
(-1.6% to 17%)
3.6%
(-3.8% to 12%)
2.2%
(-6.3% to 6.6%)
17%
(6.7% to 28%)
4.5%
(-3.1% to 15%)
13%
(8.0% to 22%)
207
-------
SYNTHESIS OF RESULTS
National Summary
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Coral Reefs, HI: % Lost Cover
70%
(11% to 97%)
64%
(7.3% to 94%)
5.6%
(-11% to 17%)
96%
(88% to 98%)
79%
(26% to 97%)
16%
(-1.2% to 63%)
Coral Reefs, FL: % Lost Cover
95%
(93% to 97%)
94%
(89% to 96%)
1.6%
(-1.1% to 8.5%)
97%
(95% to 98%)
96%
(95% to 97%)
0.57%
(0.10% to 1.3%)
Coral Reefs, PR: % Lost Cover
93%
(89% to 95%)
93%
(85% to 96%)
-0.80%
(-6.7% to 9.7%)
95%
(92% to 98%)
97%
(96% to 97%)
-1.4%
(-3.9% to 0.88%)
Freshwater Fish: Lost Coldwater
Fishing Days (millions)
67
(54 to 80)
62
(47 to 75)
5.4
(3.3 to 6.8)
90
(73 to 100)
67
(55 to 80)
23
(18 to 29)
Shellfish: %Decrease in Oyster
Supply (example species)
23%
(22% to 24%)
16%
(14% to 17%)
7.1%
(5.2% to 8.1%)
48%
(46% to 50%)
22%
(20% to 24%)
25%
(24% to 26%)
Wildfire: Acres Burned (millions)
-0.55
(-1.9 to 0.50)
-1.8
(-2.6 to -0.69)
1.2
(0.093 to 2.1)
-0.36
(-1.9 to 0.38)
-2.1
(-2.8 to-1.5)
1.7
(0.49 to 2.3)
Carbon Storage: Metric Tons Lost
(millions)
-28
(-110 to 25)
-10
(-42 to 24)
-18
(-71 to 19)
-88
(-200 to -33)
-11
(-34 to 36)
-77
(-180 to -22)
Note: "NA" indicates analyses where GCM-specific results are not available.
208
-------
SYNTHESIS OF RESULTS
National Summary
Table 28.2. Projected Annual Economic Impacts of Climate Change across U.S. Sectors Analyzed
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are based on values across the GCMs. Values shown in millions of undiscounted $2015.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$9,800
($880 to $28,000)
$6,900
(-$900 to $21,000)
$2,900
(NA)
$26,000
(-$2,200 to $78,000)
$18,000
($1,600 to $51,000)
$8,000
(NA)
Extreme Temp.
Mortality
$43,000
($28,000 to $73,000)
$32,000
($21,000 to $48,000)
$11,000
($2,100 to $25,000)
$140,000
($82,000 to $200,000)
$59,000
($37,000 to $110,000)
$82,000
($42,000 to $120,000)
Labor
$44,000
($25,000 to $70,000)
$35,000
($19,000 to $56,000)
$9,000
(-$1,200 to $15,000)
$160,000
($87,000 to $220,000)
$80,000
($52,000 to $120,000)
$75,000
($35,000 to $110,000)
Aeroallergens
$0.59
($0,033 to $0.90)
$0.44
($0,092 to $0.80)
$0.14
(-$0,059 to $0.40)
$1.2
($0.43 to $1.7)
$0.52
(-$0,040 to $0.93)
$0.70
($0.47 to $0.91)
Harmful Algal
Blooms
$79
($42 to $170)
$64
($30 to $150)
$15
(-$17 to $49)
$200
($130 to $390)
$110
($54 to $230)
$89
($22 to $180)
West Nile Virus
$1,100
($780 to $1,500)
$870
($610 to $1,200)
$200
($160 to $280)
$3,300
($2,000 to $4,700)
$1,800
($1,200 to $2,500)
$1,600
($820 to $2,200)
INFRASTRUCTURE
Roads
$9,500
($2,800 to $23,000)
$6,500
($2,700 to $16,000)
$2,900
(-$680 to $7,200)
$20,000
($7,000 to $37,000)
$8,100
($3,300 to $20,000)
$12,000
($3,700 to $17,000)
Bridges
$1,700
($950 to $2,200)
$1,500
($1,100 to $1,700)
$220
(-$140 to $650)
$1,000
($670 to $1,300)
$510
($310 to $740)
$490
($270 to $910)
Rail
$1,800
($1,300 to $2,200)
$1,500
($1,100 to $1,800)
$270
(-$17 to $410)
$5,500
($4,000 to $6,600)
$3,500
($2,400 to $4,400)
$2,000
($1,600 to $2,300)
Alaska
Infrastructure
$180
($170 to $180)
$120
($110 to $130)
$60
($55 to $66)
$170
($130 to $220)
$82
($80 to $84)
$92
($49 to $140)
Urban Drainage
$3,700
($2,100 to $4,600)
$4,300
($3,500 to $4,900)
-$600
(-$2,100 to $63)
$5,600
($3,300 to $7,000)
$4,100
($2,900 to $5,900)
$1,500
(-$110 to $3,200)
Coastal Property
$75,000
(NA)
$69,000
(NA)
$6,800
(NA)
$120,000
(NA)
$92,000
(NA)
$26,000
(NA)
ELECTRICITY
Electricity
Demand and
Supply
$3,200
($2,700 to $4,200)
$2,000
($1,400 to $2,600)
$1,200
($390 to $1,600)
$9,200
($6,500 to $11,000)
$3,400
($2,300 to $5,000)
$5,800
($3,500 to $7,600)
209
-------
SYNTHESIS OF RESULTS
National Summary
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
WATER RESOURCES
Municipal and
Industrial Water
Supply
$120
($26 to $240)
$120
($27 to $240)
$1.4
(-$85 to $89)
$320
($190 to $640)
$210
(-$9.3 to $410)
$100
(-$32 to $230)
Inland Flooding
$5,100
(NA)
$4,300
(NA)
$770
(NA)
$8,100
(NA)
$4,300
(NA)
$3,800
(NA)
Water Quality
$1,900
($1,300 to $2,800)
$1,500
($1,100 to $2,200)
$390
($260 to $610)
$4,600
($3,200 to $5,700)
$3,000
($1,700 to $4,200)
$1,600
($1,100 to $2,200)
Winter
Recreation
$780
(-$440 to $1,500)
-$430
(-$890 to -$38)
$1,200
(-$400 to $2,400)
$2,000
($28 to $2,900)
-$130
(-$1,900 to $830)
$2,200
($1,900 to $2,500)
AGRICULTURE
Agriculture
$8.0
($6.7 to $11)
$7.7
($6.4 to $10)
$0.22
(-$1.7 to $1.2)
$12
($11 to $13)
$11
($9.3 to $13)
$1.3
(-$0.33 to $2.0)
ECOSYSTEMS
Coral Reefs
$3,400
($1,800 to $4,200)
$3,200
($1,600 to $4,100)
$200
(-$330 to $720)
$4,100
($3,800 to $4,200)
$3,600
($1,900 to $4,100)
$500
(-$31 to $1,900)
Freshwater Fish
$1,900
(-$430 to $4,600)
$1,800
($740 to $3,100)
$35
(-$1,200 to $1,500)
$3,100
(-$410 to $5,500)
$1,700
(-$300 to $3,700)
$1,400
(-$110 to $2,100)
Shellfish
$9.1
($3.7 to $14)
$6.1
($2.1 to $9.0)
$3.0
($1.1 to $5.1)
$23
($8.9 to $35)
$10
($3.4 to $15)
$13
($5.4 to $20)
Wildfire
-$67
(-$230 to $62)
-$150
(-$280 to -$5.6)
$82
(-$11 to $180)
-$110
(-$340 to $8.6)
-$250
(-$340 to -$170)
$140
(-$6.9 to $250)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Coastal Property: Costs with no adaptation. See Modeling Framework section for a description of SLR uncertainty.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
210
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
29. RISK REDUCTION THROUGH ADAPTATION
29.1 KEY FINDINGS
• Analysis of adaptation for the infrastructure sector suggests that adaptation is an important, cost-
effective response strategy to reduce climate change impacts and enhance resilience. Under RCP8.5,
well-timed adaptation could reduce over 75% of the cumulative impacts to coastal properties,
roads, and the rail system in the U.S., resulting in hundreds of billions of dollars in cumulative
benefits this century.
• Proactive adaptation measures implemented in anticipation of future climate change risks are
projected to be more cost-effective than reactive repairs implemented after impacts have already
occurred.
• Reduced climate change under RCP4.5 generally lowers the costs of adaptation, but the timing and
magnitude varies by sector.
• For some infrastructure sectors, a portfolio of measures - reducing climate change and taking
adaptive actions - achieves the greatest benefits in reducing climate change damages. The combined
strategies can eliminate 91%, 89% and 91% of the cumulative climate change impacts that are
otherwise projected to occur this century for coastal property, roads, and rail infrastructure,
respectively.
29.2 BACKGROUND
Adaptation, along with substantial and sustained reductions in global GHG emissions, has the potential
to limit climate change risks and reduce society's vulnerability to climate change impacts.474,475 There are
a wide range of adaptation options that can vary depending on the sector, the timing of
implementation, and other factors. For example, adaptation can be a reactive response (i.e.,
implemented in response to climate change impacts that have already occurred) or proactive (i.e.,
planned and implemented in anticipation of future climate change risks). Adaptation measures can also
be categorized as engineering and technology options (e.g., new crop varieties, coastal protection
structures, improved drainage), management options (e.g., changes of crop planting dates and
locations), ecosystem-based options (e.g., green infrastructure), economic options (e.g., financial
incentives, insurance), institutional options (e.g., land use zoning laws, building codes, disaster risk
management planning), or social options to enhance adaptive capacity (e.g., social safety nets for
vulnerable population, education).476 Adaptation options may be undertaken by different actors (e.g.,
households, private sector, governments) and at various geographic scales (local, national, and
international).
474 IPCC, 2014: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland, 151 pp.
475 Bierbaum, R., A. Lee, J. Smith, M. Blair, L. M. Carter, F. S. Chapin, III, P. Fleming, S. Ruffo, S. McNeeley, M. Stults, L. Verduzco, and E. Seyller,
2014: Ch. 28: Adaptation. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.)
Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 670-706. doi:10.7930/J07HlGGT.
476 Adapted from Noble, I.R., S. Huq, Y .A. Anokhin, J. Carmin, D. Goudou, F .P . Lansigan, B. Osman-Elasha, and A. Villamizar, 2014: Adaptation
needs and options. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of
Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B., V.R. Barros, D.J. Dokken, K.J.
Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y .0. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P .R.
Mastrandrea, and L.L. White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 833-868.
211
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
29.3 APPROACH TO ADAPTATION IN THE TECHNICAL REPORT
In this Technical Report, adaptation is explicitly analyzed for a number of impact categories to evaluate
its effects on reducing climate damages, and to characterize interactions under different climate change
scenarios. These analyses use a variety of approaches and scenarios to represent adaptation (see Table
29.1).
Table 29.1. Adaptation Approaches and Measures
Sector
Analytical Approach
Adaptation Measures Analyzed
Infrastructure 1
Coastal Property
Estimates climate change damages and
conducts a cost-benefit analysis of
adaptation options in U.S. coastal areas
using the National Coastal Property Model.
Beach nourishment, property elevation,
shoreline armoring, property
abandonment.
Roads*
Estimates climate change impacts and costs
of reactive and proactive adaptation
options for the U.S. road network to
maintain current levels of service using the
Infrastructure Planning Support System
(IPSS) tool.
Change of road materials (e.g., asphalt
mix), design modification (e.g., surface
density), road upgrade (e.g., to paved or
gravel road).
Rail*
Estimates climate change impacts (delays)
and costs of reactive and proactive
adaptation to the U.S. rail network using
the IPSS tool.
Installation and use of temperature sensors
that allow for better rail monitoring and
reduced rail traffic delays.
Bridges
Evaluates climate change vulnerability on
inland bridges across the U.S., and
estimates the costs of proactive adaptation
measures.
Proactive adaptation (repairs) using rip-rap
and concrete strengthening.
Urban Drainage
Estimates the costs of proactive adaptation
for urban drainage infrastructure in 100
modeled cities across the U.S. associated
with three storm types.
Use of best management practices (e.g.,
temporary storage, and permeable
pavement).
Health
Extreme
Temperature
Mortality
Includes a sensitivity analysis assuming the
human health response to extreme
temperature in 49 U.S. cities is equal to
that of Dallas, TX today (compared to the
no-adaptation scenario's use of city-specific
temperature thresholds).
Adaptive responses embedded in empirical
damage functions (e.g., increased use of air
conditioning, institutional/societal
responses, and physiological adaptation of
human body to warmer temperatures).
* Includes values reported in the Alaska Infrastructure section.
As shown in Table 29.1, the analysis of adaptation in this Technical Report focuses on the infrastructure
sector and includes a sensitivity analysis for extreme temperature mortality. As discussed in more details
in individual chapters, the analytical approaches used to evaluate costs and effects of adaptation on
reducing climate change impacts vary across the impact categories. For example, the Coastal Property
section quantifies damages from climate change with and without adaptation (adaptation response
costs are embedded in estimated damages), allowing for analysis of the effects of adaptation on
reducing damages to coastal properties and energy installations from sea level rise and storm surge. The
Roads and Alaska Infrastructure analyses estimate the costs of climate change impacts in the form of
reactive adaptation to maintain current levels of service, and evaluate the ability of proactive adaptation
212
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
measures to improve resiliency and reduce costs. The Rail analysis quantifies the costs of reactive
adaptation associated with delays resulting from increased temperatures under climate change, and the
costs of proactive adaptation that include both the costs of implementing adaptation measures and
residual impacts. For Bridges and Urban Drainage, the analyses assume that the modeled systems will
adapt by making adjustments or investments to maintain consistent levels of service, and estimate the
costs of implementing proactive adaptation measures.
The modeled adaptation responses for these sectors have a strong focus on structural- and technology-
based adaptation options (e.g., infrastructure maintenance and upgrades). Other adaptation strategies,
such as ecosystem-based, economic or institutional measures, are not considered in the analyses of this
Technical Report. Similarly, adaptation analysis for other impact categories is not fully explored, due to
limitations in the current modeling capability, data availability, and understanding of adaptation
pathways. For Extreme Temperature Mortality, a sensitivity analysis was conducted using city-specific
temperature thresholds to explore the potential effect of both physiological (acclimatization) and
societal adaptive responses, such as increased use of air conditioning or early-warning systems.
In some other sectoral analyses of this Technical Report, behavioral or market adjustments are assumed
to occur to minimize climate change impacts and damages without deliberate interventions. Examples
include crop switching and changes in planting dates and locations in response to changes in water
availability and temperature-induced changes in the agricultural sector; increased energy consumption
for air conditioning in response to higher temperatures; and increased snowmaking to meet winter
recreation demand. In addition, autonomous ecological adaptation is also assumed and modeled in a
number of the ecosystem sectors, such as shifting distributions of freshwater fisheries in response to
changes in suitable habitat, or vegetation changes in response to changing climate that in turn affect
wildfire activity and carbon storage. In these cases, this autonomous adaptation is embedded in the
analyses and the estimated damages reflect assumptions of behavioral, market, or ecological
adjustments that take place without deliberate interventions.
29.4 RESULTS
The analysis below focuses on the infrastructure sector, including Coastal Property, Roads, Rails, Bridges,
and Urban Drainage. This sector provides adaptation analyses that allow for comparison and the
development of broad insights on the costs and effectiveness of adaptation in reducing climate change
impacts. Table 29.2 presents the summary results of these sectoral analyses.
213
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
Table 29.2. Effects of Adaptation and Reduced Climate Change on Infrastructure Damages
Results, shown in billions of $2015, represent averages across the five GCMs by sector. Values shown in
the dark blue cells represent the combined effects of reduced climate change and adaptation. Annual
average results in the 2050 and 2090 columns are undiscounted, while the cumulative benefit estimates
(through 2099) are discounted at 3%. Values may not sum due to rounding.
2050 2090
Sector
Scenarios
RCP8.5
RCP4.5
Effect of
Reduced
Climate
Change
RCP8.5
RCP4.5
Effect of
Reduced
Climate
Change
Cumulative
Effect
through
2099**
Without
adaptation
$75
$69
$6.8
$120
$92
$26
$83
Coastal
Property
With
adaptation
$8.6
$8.2
$0.45
$7.3
$5.7
$1.6
$15
Damages
avoided by
adaptation
$67
$60
$67
$110
$87
$110
$900***
Reactive
adaptation
$9.6
$6.6
$2.9
$20
$8.2
$12
$79
Roads*
Proactive
adaptation
-$1.2
-$0.94
-$0.25
-$7.3
-$3.1
-$4.2
-$20
Damages
avoided by
proactive
$11
$7.6
$10
$27
$11
$23
$200
adaptation
Reactive
adaptation
$1.8
$1.5
$0.27
$5.5
$3.5
$2
$11
Rail*
Proactive
adaptation
$0.43
$0.23
$0.20
$1.6
$0.40
$1.2
$7
Damages
avoided by
proactive
$1.3
$1.3
$1.5
$3.9
$3.1
$5.1
$46
adaptation
Bridges
With proactive
adaptation
$1.7
$1.5
$0.22
$1.0
$0.51
$0.49
Urban
Drainage
With
adaptation
$3.7
$4.3
-$0.6
$5.6
$4.1
$1.5
* These results include estimates for Alaska and therefore do not match the results presented in the Roads and Rail sections
of this Technical Report.
** Represents the cumulative effect of reduced climate change through 2099. The time period used for deriving cumulative
benefits in this table is 2015-2099 for coastal properties and roads, and 2016-2099 for rail.
*** This value is based on cumulative effects for 2015-2099 and therefore differs from the results presented in the Coastal
Property section of this Technical Report, which present results for 2000-2099. The substantial difference in the two estimates
is attributed to the fact that a large number of properties in the U.S. are estimated to already be vulnerable or to become
vulnerable soon after the first year of the simulation (the model assumes optimal adaptation with near-term foresight,
therefore leading to adaptation responses in the near-term that may otherwise be delayed).
214
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
Climate change is projected to result in significant damages to coastal property, but there is also
substantial potential to reduce the magnitude of damages through adaptation in this sector (e.g.,
abandonment, shoreline armoring and beach nourishment). Adaptation measures are estimated to
reduce 89% and 94% of climate change impacts under RCP8.5 in 2050 and 2090, respectively.
Adaptation is also projected to be significant for the U.S. roads and rail systems when comparing costs
under the proactive adaptation scenario and those under the reactive adaptation scenario. Under
RCP8.5, adaptation responses are projected to reduce 98% of the cumulative costs of climate change
($220 billion in cumulative benefits) for roads and 77% of damages ($39 billion in cumulative benefits)
for rail networks over the course of the century.
Analyses of roads and rail also suggest that proactive adaptation measures, when undertaken in
anticipation of future climate change risks, can be far more cost-effective in reducing impacts from
climate change than reactive adaptation measures. Proactive adaptation investments, made in advance
of climate change impacts, more than compensate for the cost of reactive adaptation experienced by
the road network, and reduce 76% and 71% of the cost of reactive adaptation to the rail system under
RCP8.5 in 2050 and 2090, respectively.
Reduced climate change under RCP4.5 lowers the costs of adaptation, but the timing and magnitude of
reduction varies by sector. For rail systems, RCP4.5 reduces adaptation costs by 47% and 75% in 2050
and 2090, respectively, compared to RCP8.5. For bridges, RCP4.5 reduces adaptation costs by 13% and
49% in 2050 and 2090, respectively, compared to RCP8.5. For urban drainage, reduced climate change
under RCP4.5 (compared to RCP8.5) leads to 16% higher costs in the mid-century period, but results in
26% reduction in annual adaptation costs by 2090. For the road network, RCP4.5 is projected to increase
the cost of adaptation in the mid-century (by 22% in 2050) due to reduced cost savings from responding
to impacts of freeze-thaw on paved roads, but reduce the overall cost of adaptation later in the century
(by 58% in 2090).
For some infrastructure sectors, a portfolio of measures - both reducing climate change and taking
adaptive actions - achieves the greatest benefits in reducing damages. For coastal property, roads, and
rail infrastructure, a portfolio of strategies can eliminate 91%, 89% and 91% of the climate change
impacts that are otherwise projected to occur within this century, respectively (Table 29.3).477
477 It is noted that the magnitude of aggregate damage reduction from a combination of GHG mitigation and adaptation is less than the sum of
damage reductions from adaptation only and mitigation only, suggesting there may be some tradeoffs between mitigation and adaptation
actions. However, the overall benefit of a portfolio of GHG mitigation and adaptation is greater than a mitigation-only or adaptation-only
approach.
215
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
Table 29.3. Reduction in Damages Due to Adaptation and Reduced Climate Change
Unless noted, values represent the percent reduction in economic effects based on cumulative,
discounted (3%) damages through 2099.
Sector
Damage Reduced by
Proactive Adaptation'
Damage Reduced by
Mitigation''
Damage Reduced by
Mitigation and
Adaptation"'
Coastal Property™
90%
8%
91%
Roads
98%
35%
89%
Rail
11%
21%
91%
' Damage reductions from adaptation under RCP8.5.
" Difference in damage reductions between RCP8.5 and RCP4.5 without adaptation.
111 Difference in damage reductions between RCP8.5 without adaptation and RCP4.5 with adaptation.
lvThis value is based on cumulative benefits for 2015-2099 and therefore differs from the results presented in the
Coastal Property section of this Technical Report, which present results for 2000-2099.
29.5 DISCUSSION
Analyses of adaptation responses included in this Technical Report suggest the important role and cost-
effectiveness of adaptation in reducing climate change impacts and risks. For coastal property, roads,
and rail, well-timed adaptation measures can be very effective in reducing the negative impacts from
climate change, suggesting investment opportunities in adaptation measures would avoid costly
damages and ensure the resilience of these sectors. Proactive adaptation measures can be far more
cost-effective in reducing damages than reactive measures. Moreover, a portfolio of responses to
reduce climate change and adapt would significantly limit overall damages. These findings are consistent
with the broad literature that points to the importance of strategically timed implementation of
adaptation measures in the near- to medium-terms478, and a portfolio approach to address the long-
term climate risks.479
The analysis contributes to the literature by presenting quantitative estimates of the costs and cost-
effectiveness of adaptation responses in selected infrastructure sectors in the U.S. However, several
limitations should be noted and considered for future research. First, analysis of climate change impacts
and responses should cover a broader set of sectors and adaptation measures (e.g., ecosystem-based,
economic, and institutional measures) to improve estimates of climate change impacts and costs and
benefits of adaptation. Analysis of other adaptation options and for other impact categories was not
fully explored due to limitations in the current modeling capability, data availability, and understanding
of adaptation pathways. Second, the sectoral analyses often model optimal adaptation behavior, which
assumes well-timed, cost-minimizing implementation of responses. As such, they provide upper-bound
estimates of the potential for adaptation to reduce vulnerabilities and risks. There is a need to better
understand how uncertainty of future climate change and market, behavioral, and institutional barriers
affect the implementation of adaptation measures, and costs and benefits of adaptation.480 Third, there
478IPCC, 2014: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland, 151 pp.
479 Bosello, F., C. Carraro, and E. De Cian, 2010: Climate policy and the optimal balance between mitigation, adaptation and unavoided
damage. Climate Change Economics, 1, 71-92.
480 Sussman, F., A. Grambsch, J. Li, and C.P. Weaver, 2014: Introduction to a special issue entitled Perspectives on Implementing Benefit-Cost
Analysis in Climate Assessment. Journal of Benefit-Cost Analysis, 5, 333-346.
216
-------
SYNTHESIS OF RESULTS
Risk Reduction through Adaptation
is a need to develop analytical frameworks to understand vulnerability of socioeconomic and
environmental systems from climate change combined with other stressors (e.g., development and
population pressure, land use changes, and ecosystem modifications), and evaluate the effects of
adaptation responses across scales and sectors.
217
-------
SYNTHESIS OF RESULTS
Regional Summaries
30. REGIONAL SUMMARIES
The sectoral modeling chapters demonstrate that climate change impacts will not be uniform across the
U.S., with most sectors showing a complex pattern of regional-scale impacts. The following sections and
tables provide regional summaries of the sectoral impacts described in this Technical Report using the
NCA4 regional aggregations (Figure 30.1) under RCP8.5 and RCP4.5 in the years 2050 and 2090.
Figure 30.1. NCA4 Regional Aggregations
NCA4 Regions
Northeast
Southeast '
I^H Midwest
1 Northern Plains
HH Southern Plains
Southwest
Northwest
Alaska
Hawaii
I Puerto Rico
Esri, HERE. DeLcrme. Mapmylndis. ® OperStreetMsp confributcrs, ana tbeGIS use» community
218
-------
SYNTHESIS OF RESULTS
Regional Summaries
Several considerations are important to note regarding the regional tables in the following sections.
• The annual estimates provided here are not cumulative impacts for the entire 21st century, which
would likely be much larger (see the sectoral sections for cumulative estimates where available).
• Many of the values do not estimate and monetize the full extent of potential impacts from climate
change on that sector, and as such, the results should be treated as conservative.
• The individual monetized estimates are not aggregated, as only a subset of climate change impacts
is quantified in this report. Also, many of the physical impacts have significant societal or iconic
values that are not captured in the estimated economic impact.
• Not all sectors have results in all regions. For example, the coral reef modeling will not produce
results for the Northeast and shellfish modeling will not produce results for the Midwest.
• Only some of the sectoral analyses produced discrete physical metric estimates, therefore not all
sectors of this Technical Report are represented in the physical impact tables.
• Economic valuation of changes in agriculture welfare is only available at the national level and thus
are not represented in the following tables. As described in the Carbon Storage section, monetized
values are not estimated.
• The sectoral sections of this Technical Report provide detailed information on the methods, results,
limitations, and supporting literature.
• Benefit values may not equate to differences between the RCPs due to rounding and the use of two
significant figures.
219
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.1 NORTHEAST
Using the results presented throughout the sector sections of
this Technical Report, this section summarizes the impacts
projected to occur in the Northeast.
Key Findings
• The Northeast is projected to experience some of the
largest adverse health impacts from climate change.
Damages from lost labor hours and deaths associated
with worsened air quality and increases in extreme
temperature are on the order of billions of dollars in
damage each year from climate change. Compared to
RCP8.5, 1,400 premature deaths from extreme
temperatures are projected to be avoided each year by
2090 under RCP4.5 in the Northeast, resulting in $21
billion in annual savings.
• Coastal property damages in the Northeast remain high under both climate scenarios, with
annual costs under RCP8.5 projected to rise from $10 billion in 2050 to $12 billion in 2090.
These values would decrease significantly with well-timed adaptation measures.
• In the shellfish and freshwater fish sectors, the Northeast is projected to experience some of the
largest price increases in shellfish of all regions (particularly quahogs and scallops under RCP8.5,
in response to decreases in supply), and large losses of coldwater fishing habitat, which is
projected to nearly disappear this century throughout the Appalachians. Under RCP4.5, many of
these projected losses are avoided.
• Climate change impacts on winter recreation (from lost downhill skiing and snowboarding and
cross-country skiing visits) are projected to be more than a billion dollars a year by 2090 under
RCP8.5 ($1.1 billion), or $0.7 billion each year under RCP4.5.
Discussion
Tables 30.1 and 30.2 present the estimated annual physical and economic effects of climate change in
the Northeast under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.1, annual
physical impacts of climate change in the Northeast are projected to increase over time in all sectors and
under all climate scenarios, except corn yields and acres burned by wildfire (both under RCP4.5 only)
and carbon storage. Adverse impacts are generally greater under RCP8.5 than under RCP4.5. As shown
in Figure 30.2 and Table 30.2, annual economic damages also generally increase from 2050 to 2090 and
from RCP4.5 to RCP8.5. However, some infrastructure sectors, like bridges and urban drainage, see
higher costs in 2050 than in 2090, as many types of infrastructure are already vulnerable or will soon
become vulnerable and require repair or adaptation costs early in the century.
220
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.2. Largest Damages of Climate Change in the Northeast
Annual damages for the ten sectors with the greatest projected costs in the Northeast in 2090 under
RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The difference
between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data underlying the
sectors shown in the figure, as well as all other sectors modeled in the Northeastcan be found in Table
30.2 below.
Roads
$2.4 I -48%
Extreme Temperature
Mortality
$35 | -59%
Electricity Demand
and Supply
$0.79 | -69%
Inland Flooding
$1.5 | -38%
Freshwater Fish
$0.63 | -20%
Winter
Recreation
$1.1 | -40%
Air Quality
$10 | -53%
Coastal Property
$12 I -15%
Labor
$19 | -57%
Water Quality
$1.0 | -35%
Air Quality
Extreme Temperature Mortality
Labor
Roads
Coastal Property
Electricity Demand and Supply
Inland Flooding
Water Quality
Winter Recreation
Freshwater Fish
221
-------
SYNTHESIS OF RESULTS
Regional Summaries
The Northeast is projected to experience some of the largest impacts of climate change on health,
infrastructure, shellfish and freshwater fishing. This region experiences the highest risk of increased
emergency room visits from asthma attacks related to aeroallergens, and high projected losses in
reservoir recreation from harmful algal blooms. Water quality is projected to decrease, especially under
RCP8.5, with some of the largest projected damages occurring in the Northeast region. These health
effects will likely be compounded by increases in climate-induced migration that are projected for the
Northeast. In the shellfish and freshwater fish sectors, the Northeast is projected to experience some of
the largest price increases in shellfish (particularly quahogs and scallops, in response to decreases in
supply), and large losses of coldwater fishing habitat, which is projected to nearly disappear this century
throughout the Appalachians. Northeast infrastructure is also at risk, with high projected damages to rail
and roads. Roads in this region experience high temperature related damages, especially under RCP8.5,
but also high savings from reduced maintenance costs associated with decreases in freeze-thaw
stressors. There are also large increases in electricity costs in the Northeast to meet projected demand
increases. Finally, there are significant reductions in winter recreation season length in the Northeast.
The most economically-significant impacts in the Northeast occur from coastal property loss, lost labor
hours, and deaths associated with worsened air quality and increases in extreme temperature, all of
which are on the order of billions of dollars in damage each year from climate change. There are
significant benefits for human-health related damages under RCP4.5 compared to RCP8.5, particularly in
2090. For example, 1,400 deaths from extreme temperatures would be avoided each year under RCP4.5
compared to RCP8.5 by 2090, resulting in $21 billion in annual savings. Coastal damages remain high
under both climate scenarios in Table 30.2, where no adaptation is assumed, but would decrease
significantly with proactive adaptation measures (see Coastal Property and Risk Reduction through
Adaptation sections of this Technical Report).
For additional considerations regarding the values shown in Tables 30.1 and 30.2, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
222
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.1. Projected Annual Physical Impacts of Climate Change in the Northeast
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.2 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality:
# deaths
230
(120 to 340)
200
(110 to 300)
30
(NA)
670
(360 to 980)
310
(160 to 450)
360
(NA)
Extreme Temperature
Mortality: # deaths
640
(330 to 1,200)
650
(400 to 1,100)
-7.6
(-200 to 220)
2,300
(1,100 to 3,400)
960
(630 to 2,000)
1,400
(280 to 2,400)
Labor: Lost Labor Hours
(millions)
92
(44 to 140)
82
(63 to 120)
9.8
(-19 to 36)
230
(100 to 360)
100
(46 to 190)
130
(54 to 220)
Aeroallergens: ED visits
490
(160 to 820)
480
(250 to 670)
4
(-95 to 230)
1,100
(560 to 1,300)
470
(150 to 750)
590
(400 to 780)
Harmful Algal Blooms:
# Days above 100k
cells/mL
22
(14 to 29)
20
(14 to 28)
2.5
(-3.0 to 12)
36
(22 to 47)
22
(6.7 to 34)
14
(0.060 to 27)
West Nile Virus:
# Cases
170
(110 to 230)
130
(74 to 190)
41
(25 to 69)
490
(280 to 690)
210
(120 to 340)
280
(160 to 350)
INFRASTRUCTURE
Bridges:
# Vulnerable Bridges
510
(400 to 660)
350
(190 to 620)
160
(-180 to 300)
570
(370 to 880)
390
(200 to 660)
180
(68 to 280)
WATER RESOURCES
Winter Recreation:
Lost Visits (millions)
9.3
(5.4 to 13)
-0.80
(-5.3 to 2.0)
10
(6.0 to 19)
15
(11 to 17)
9.2
(4.5 to 13)
5.7
(3.7 to 9.4)
AGRICULTURE
Agriculture:
% Decrease in Corn
Yields (example crop)
-2.9%
(-9.5% to 9.2%)
-4.1%
(-8.2% to 5.3%)
1.2%
(-2.2% to 5.1%)
6.2%
(-1.5% to 20%)
-4.9%
(-9.7% to 9.9%)
11%
(8.1% to 20%)
223
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Freshwater Fish: Cold
Fishing Days Lost
(millions)
33
(26 to 35)
32
(24 to 34)
1.0
(0.32 to 2.0)
35
(34 to 36)
33
(29 to 35)
2.2
(1.0 to 5.3)
Shellfish:
% Decrease in Oyster
Supply (example
species)
23%
(21% to 26%)
16%
(13% to 17%)
7.8%
(4.3% to 11%)
50%
(48% to 53%)
23%
(20% to 26%)
26%
(24% to 29%)
Wildfire: Acres Burned
(thousands)
1,400
(-80 to 2,600)
1,200
(-580 to 3,300)
180
(-2,600 to 2,600)
1,800
(510 to 3,200)
-370
(-1,700 to 1,800)
2,100
(320 to 3,400)
Carbon Storage: Metric
Tons Lost (millions)
6.5
(0.47 to 17)
7.6
(1.1 to 14)
-1.1
(-5.4 to 3.3)
4.1
(-14 to 16)
4.6
(2.6 to 8.7)
-0.48
(-19 to 7.9)
Note: "NA" indicates analyses where GCM-specific results are not available.
224
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.2. Projected Annual Economic Impacts of Climate Change in the Northeast
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Regional economic impacts on agriculture yield and welfare are not available. Due to rounding, benefit values may not equate to
differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$2,900
($260 to $8,200)
$2,500
($230 to $7,200)
$400
(NA)
$10,000
($910 to $29,000)
$4,700
($420 to $13,000)
$5,300
(NA)
Extreme
Temperature
Mortality
$7,900
($4,100 to $15,000)
$8,000
($5,000 to $13,000)
-$94
(-$2,400 to $2,700)
$35,000
($16,000 to
$52,000)
$15,000
($9,500 to $31,000)
$21,000
($4,300 to $36,000)
Labor
$4,600
($2,200 to $7,200)
$4,100
($3,200 to $5,900)
$490
(-$950 to $1,800)
$19,000
($8,300 to $29,000)
$8,300
($3,800 to $15,000)
$11,000
($4,500 to $18,000)
Aeroallergens
$0.24
($0,078 to $0.40)
$0.24
($0.12 to $0.33)
$0.0018
(-$0,046 to $0.11)
$0.52
($0.27 to $0.66)
$0.23
($0,071 to $0.37)
$0.29
($0.19 to $0.38)
Harmful Algal
Blooms
$8.1
($0,060 to $19)
$5.9
($0 to $17)
$2.2
(-$3.7 to $12)
$28
($18 to $32)
$15
($0.48 to $22)
$13
($5.1 to $20)
West Nile
Virus
$140
($93 to $190)
$110
($62 to $160)
$35
($21 to $58)
$500
($280 to $710)
$210
($120 to $350)
$280
($160 to $360)
INFRASTRUCTURE
Roads
$1,200
($260 to $3,200)
$830
(-$62 to $2,100)
$420
(-$410 to $1,200)
$2,400
(-$1.1 to $6,200)
$1,300
(-$23 to $3,900)
$1,200
(-$210 to $2,400)
Bridges
$220
($170 to $280)
$180
($140 to $230)
$42
($8.4 to $61)
$120
($95 to $160)
$77
($40 to $140)
$43
(-$9.3 to $71)
Rail
$160
($120 to $200)
$130
($96 to $170)
$24
($14 to $33)
$530
($360 to $640)
$320
($220 to $430)
$210
($140 to $260)
Urban
Drainage
$280
($19 to $530)
$240
($5.6 to $400)
$36
(-$380 to $340)
$160
($39 to $250)
$220
($19 to $790)
-$68
(-$630 to $240)
Coastal
Property
$10,000
(NA)
$9,700
(NA)
$720
(NA)
$12,000
(NA)
$9,800
(NA)
$1,800
(NA)
225
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ELECTRICITY
Electricity
Demand and
$240
$130
$110
$790
$240
$540
Supply
($160 to $390)
($74 to $190)
($29 to $210)
($390 to $1,200)
($100 to $520)
($60 to $760)
WATER RESOURCES
Municipal
and Industrial
$4.9
$5.8
-$0.92
$36
$11
$24
Water Supply
($0,050 to $21)
($0,096 to $19)
(-$19 to $14)
($1.7 to $81)
($0.63 to $30)
(-$0.64 to $63)
Inland
$1,600
$620
$1,000
$1,500
$950
$580
Flooding
(NA)
(NA)
(NA)
(NA)
(NA)
(NA)
Water
$430
$360
$71
$1,000
$650
$350
Quality
($260 to $610)
($210 to $520)
($6.4 to $150)
($620 to $1,300)
($370 to $940)
($250 to $440)
Winter
$720
-$86
$800
$1,100
$710
$440
Recreation
($420 to $1,000)
(-$440 to $130)
($480 to $1,500)
($850 to $1,300)
($350 to $1,000)
($290 to $730)
ECOSYSTEMS
Freshwater
$500
$580
-$84
$630
$500
$130
Fish
($290 to $610)
($510 to $630)
(-$220 to $35)
($350 to $990)
($410 to $680)
(-$100 to $420)
Wildfire
$6.0
$5.2
$0.81
$7.9
-$1.7
$9.6
(-$0.37 to $12)
(-$2.6 to $15)
(-$12 to $12)
($2.3 to $14)
(-$7.5 to $8.1)
($1.4 to $15)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Coastal Property: Costs with no adaptation. See Modeling Framework section for a description of SLR uncertainty.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
226
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.2 SOUTHEAST
Using the results presented throughout the sector
sections of this Technical Report, this section
summarizes the impacts projected to occur in the
Southeast.
Key Findings
• Projected labor losses due to extreme heat
by the end of the century total $47 billion
each year under RCP8.5, with
approximately one third of the national
projected loss occurring in this region.
These projected losses are halved under
RCP4.5.
• The Southeast is projected to experience
some of the largest adverse health
impacts. For example, projected total cases
of West Nile neuroinvasive disease grow
from approximately 100 per year in the reference period to more than 1,100 per year by 2090.
Damages associated with harmful algal blooms under RCP8.5 rise from $48 million per year in
2050 to $96 million per year in 2090, making up approximately half of the national total loss in
reservoir recreation from climate change.
• Compared to other regions, the Southeast is projected to experience the highest costs
associated with meeting increased electricity demands; by the end of the century annual costs
are estimated at $3.3 billion each year under RCP8.5 and $1.2 billion each year under RCP4.5.
• Projected damages to infrastructure are very high in the Southeast. The region experiences
some of the largest total damages to bridges and roads of all regions. Damages to coastal
property under RCP8.5 rise from $60 billion each year in 2050 to $99 billion in 2090. Under
RCP4.5, projected coastal property damages are $56 and $79 billion each year, respectively.
Discussion
Tables 30.3 and 30.4 present the estimated annual physical and economic effects of climate change in
the Southeast under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.3, annual
physical impacts of climate change in the Southeast are projected to increase over time in all sectors and
under all climate scenarios, except in some measures of air quality (ozone and aeroallergens), carbon
storage, and wildfire. Outside those sectors, adverse impacts are generally greater under RCP8.5 than
under RCP4.5. As shown in Figure 30.3 and Table 30.4, annual economic damages also generally increase
from 2050 to 2090 and from RCP4.5 to RCP8.5, though some infrastructure sectors see higher costs in
2050, as many types of infrastructure are already vulnerable or will soon become vulnerable and require
repair costs early in the century. Benefits in the wildfire sector are estimated under RCP8.5 compared to
RCP4.5.
227
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.3. Largest Damages of Climate Change in the Southeast
Annual damages for the ten sectors with the greatest projected costs in the Southeast in 2090 under
RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The difference
between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data underlying the
sectors shown in the figure, as well as all other sectors modeled in the Southeast, can be found in Table
30.4 below.
Electricity Demand
and Supply
$3.3 | -64%
Inland Flooding
$3.1 | -53%
Urban Drainage
$2.2 | -39%
Coastal Property
$99 | -20%
Labor
$47 I -50%
Roads
$6.1 | -64%
West Nile Virus
$1.2 | -61%
Water Quality
$1.4 | -32%
Extreme
Temperature
Mortality
$25 | -59%
Coral Reefs
$2.2 | -1.3%
West Nile Virus
Water Quality
Coral Reefs
Urban Drainage
Inland Flooding
Electricity Demand and Supply
Roads
Extreme Temperature Mortality
Labor
Coastal Property
The Southeast is unique in that it experiences milder air quality impacts, or even benefits, as compared
to other regions. It is one of only two regions where climate change impacts on ozone result in projected
avoided deaths (in 2050 under RCP4.5 and 2090 under RCP8.5). This is because climate-driven
meteorological changes result in conditions slightly less conducive to ozone formation, potentially due
to an increase in precipitation and wind trajectories that transport cleaner marine air into the region.
Wildfire activity and associated response costs in the Southeast are among the smallest of the regions
and this area is also projected to increase carbon stored in vegetation.
228
-------
SYNTHESIS OF RESULTS
Regional Summaries
Projected labor losses in the Southeast are the highest of all regions, totaling in the tens of billions of
dollars annually by mid-century, and making up approximately one third of the national projected loss.
While the Southeast sees a smaller increase in oak pollen season lengths than the Northeast or
Midwest, the region still experiences significant increases in asthma related emergency room visits,
particularly among children aged 0-17. These additional visits are associated with costs in 2090 of
$360,000 per year under RCP8.5 and $100,000 per year under RCP4.5. The Southeast is projected to
experience the largest adverse effects of neuroinvasive West Nile virus, particularly under RCP8.5,
where projected total cases grow from approximately 100 per year in the reference period to more than
1,100 per year by 2090. Economic impacts from harmful algal blooms are also highest in the Southeast,
under all scenarios, time periods, and growth assumptions. These regional losses make up
approximately half of the national total loss in reservoir recreation from climate change. The Southeast
is projected to experience the highest costs associated with meeting increased electricity demands, with
increased cumulative costs of $57 billion and $15 billion through 2050 under RCP8.5 under the ReEDS
and GCAM models, respectively.
The Southeast is also projected to experience important effects of climate change on infrastructure and
water resources. The region experiences the largest total damages to bridges (in 2090 under both
climate scenarios) and roads of all the NCA regions. High damages to roads are partially due to the
comparatively high number of lane miles in this area, as well as temperature related stress. Under both
RCPs, the Southeast is projected to have the highest number of vulnerable bridges in 2050 and the
second highest in 2090 of all the regions, making up roughly one third of the national total vulnerable
bridges. Cumulative costs to rail by the end of the century are also highest in the Southeast region under
both RCPs. Adaptation costs for urban drainage are second highest (behind Southern Plains) under
RCP8.5 (based on 50-year storm estimates). Water quality is projected to decrease, particularly under
RCP8.5, and associated damages in the Southeast are among the largest of all regions. However, unlike
most regions, the municipal and industrial water supply analysis projects cumulative welfare gains in the
Southeast.
The most economically-significant impacts in the Southeast occur from coastal property loss and lost
labor hours, with very high damages also occurring from deaths associated with increases in extreme
temperature and decreased air quality, all on the order of billions of dollars in damage each year from
climate change. There are significant health benefits under RCP4.5 compared to RCP8.5, particularly in
2090. Annual labor and coastal property benefits due to global GHG mitigation are $24 billion and $20
billion, respectively, in 2090. Coastal damages remain high under both climate scenarios in Table 30.4,
where no adaptation is assumed, but would decrease significantly with well-timed adaptation measures
(see Coastal Property and Risk Reduction through Adaptation sections of this Technical Report). For
inland flooding, the most significant difference between damages under RCP8.5 and those under RCP4.5
occurs in the Southeast, where the projected difference in the two trajectories is $1.6 billion per year by
the end of the century.
For additional considerations regarding the values shown in Tables 30.3 and 30.4, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
229
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.3. Projected Annual Physical Impacts of Climate Change in the Southeast
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.4 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
Benefit
Benefit
RCP8.5
RCP4.5
RCP8.5
RCP4.5
HEALTH
Air Quality:
69
-40
110
-72
88
-160
# deaths
(37 to 100)
(-21 to -59)
(NA)
(-38 to -100)
(47 to 130)
(NA)
Extreme Temperature
600
460
140
1,600
660
960
Mortality: # deaths
(360 to 1,100)
(280 to 860)
(67 to 260)
(910 to 2,100)
(350 to 1,200)
(340 to 1,500)
Labor: Lost Labor Hours
270
220
48
570
280
290
(millions)
(140 to 460)
(120 to 370)
(-2.5 to 88)
(340 to 820)
(190 to 430)
(140 to 430)
Aeroallergens:
360
290
71.0
730
210
520
ED visits
(-150 to 620)
(-77 to 620)
(-73 to 300)
(97 to 1,200)
(-220 to 530)
(280 to 730)
Harmful Algal Blooms: #
2.5
2.3
0.24
5.2
2.9
2.3
Days above 100k cells/mL
(1.1 to 5.5)
(1.5 to 4.2)
(-0.65 to 1.3)
(3.2 to 10)
(1.6 to 5.6)
(0.73 to 4.4)
West Nile Virus:
370
270
100
1,100
440
690
# Cases
(220 to 640)
(130 to 470)
(61 to 170)
(570 to 1,800)
(230 to 790)
(340 to 960)
INFRASTRUCTURE
Bridges:
1,400
750
640
1,600
1,200
430
# Vulnerable Bridges
(1,100 to 1,800)
(560 to 920)
(240 to 1,300)
(780 to 2,300)
(810 to 1,600)
(-160 to 880)
WATER RESOURCES
Winter Recreation: Lost
0.56
-0.18
0.74
0.88
0.39
0.50
Visits (millions)
(0.26 to 0.83)
(-0.40 to 0.053)
(0.51 to 1.1)
(0.57 to 1.1)
(-0.10 to 0.85)
(0.21 to 0.86)
AGRICULTURE
Agriculture:
% Decrease in Corn Yields
7.0%
3.3%
3.7%
22%
4.5%
17%
(example crop)
(0.29% to 21%)
(-0.72% to 13%)
(-2.0% to 8.9%)
(14% to 33%)
(-0.68% to 17%)
(13% to 26%)
230
-------
SYNTHESIS OF RESULTS
Regional Summaries
ECOSYSTEMS
Coral Reefs: % Lost Cover
95%
(93% to 97%)
94%
(89% to 96%)
1.6%
(-1.1% to 8.5%)
97%
(95% to 98%)
96%
(95% to 97%)
0.57%
(0.10% to 1.3%)
Freshwater Fish:
Coldwater Fishing Days
Lost (millions)
11
(6.2 to 17)
11
(6.5 to 16)
0.41
(-0.27 to 0.91)
15
(11 to 17)
11
(7.4 to 17)
4.4
(0 to 9.3)
Shellfish:
% Decrease in Oyster
Supply (example species)
21%
(20% to 23%)
14%
(13% to 16%)
7.6%
(5.9% to 8.6%)
46%
(44% to 48%)
20%
(19% to 22%)
26%
(24% to 27%)
Wildfire: Acres Burned
(thousands)
-32
(-93 to 62)
-19
(-67 to 47)
-13
(-100 to 84)
76
(-67 to 220)
-37
(-88 to 44)
110
(15 to 180)
Carbon Storage: Metric
Tons Lost (millions)
-11
(-27 to 5.8)
0.058
(-11 to 8.6)
-11
(-26 to -0.47)
-36
(-51 to-12)
-8.1
(-33 to 11)
-28
(-50 to 5.1)
Note: "NA" indicates analyses where GCM-specific results are not available.
231
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.4. Projected Annual Economic Impacts of Climate Change in the Southeast
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Regional economic impacts on agriculture are not available. Due to rounding, benefit values may not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$850
($77 to $2,400)
-$500
(-$1,400 to -$45)
$1,400
(NA)
-$1,000
(-$3,100 to -$98)
$1,300
($120 to $3,800)
-$2,400
(NA)
Extreme
Temperature
Mortality
$7,500
($4,500 to $14,000)
$5,700
($3,500 to $11,000)
$1,700
($840 to $3,200)
$25,000
($14,000 to $33,000)
$10,000
($5,300 to $19,000)
$15,000
($5,200 to $22,000)
Labor
$14,000
($7,000 to $23,000)
$11,000
($5,900 to $19,000)
$2,400
(-$130 to $4,400)
$47,000
($28,000 to $68,000)
$23,000
($16,000 to $36,000)
$24,000
($12,000 to $36,000)
Aeroallergens
$0.18
(-$0,070 to $0.30)
$0.14
(-$0,038 to $0.30)
$0,035
(-$0,036 to $0.15)
$0.36
($0,048 to $0.57)
$0.10
(-$0.11 to $0.26)
$0.26
($0.14 to $0.36)
Harmful Algal
Blooms
$48
($23 to $91)
$38
($7.6 to $78)
$9.2
($1.5 to $24)
$96
($73 to $140)
$63
($38 to $110)
$33
($7.2 to $85)
West Nile Virus
$310
($190 to $540)
$230
($110 to $400)
$85
($51 to $140)
$1,200
($590 to $1,800)
$450
($240 to $810)
$700
($350 to $980)
INFRASTRUCTURE
Roads
$3,100
($490 to $9,600)
$2,100
($200 to $6,700)
$930
(-$320 to $2,900)
$6,100
($1,400 to $13,000)
$2,200
($29 to $6,900)
$3,900
($510 to $6,200)
Bridges
$430
($260 to $580)
$340
($270 to $430)
$87
(-$24 to $170)
$300
($220 to $380)
$150
($120 to $170)
$140
($47 to $260)
Rail
$320
($250 to $400)
$280
($200 to $340)
$40
(-$22 to $68)
$950
($750 to $1,100)
$620
($480 to $750)
$340
($270 to $410)
Urban Drainage
$1,400
($1,100 to $1,500)
$1,400
($900 to $2,400)
-$32
(-$1,300 to $530)
$2,200
($1,400 to $2,800)
$1,300
($860 to $1,900)
$840
($470 to $1,600)
Coastal Property
$60,000
(NA)
$56,000
(NA)
$3,700
(NA)
$99,000
(NA)
$79,000
(NA)
$20,000
(NA)
232
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ELECTRICITY
Electricity
Demand and
Supply
$1,200
($970 to $1,600)
$720
($400 to $960)
$490
($200 to $680)
$3,300
($2,400 to $4,200)
$1,200
($900 to $1,900)
$2,100
($1,500 to $2,700)
WATER RESOURCES
Municipal and
Industrial Water
Supply
$4.4
($0.11 to $18)
$4.0
(-$0,048 to $16)
$0.43
(-$1.8 to $2.8)
$12
(-$3.6 to $32)
$1.1
(-$3.7 to $14)
$10
(-$1.2 to $29)
Inland Flooding
$1,700
(NA)
$1,800
(NA)
-$120
(NA)
$3,100
(NA)
$1,500
(NA)
$1,600
(NA)
Water Quality
$490
($320 to $820)
$370
($170 to $590)
$120
($44 to $310)
$1,400
($930 to $1,800)
$950
($560 to $1,400)
$440
($210 to $780)
Winter
Recreation
$41
($19 to $61)
-$14
(-$29 to $4.0)
$55
($38 to $82)
$65
($42 to $82)
$29
(-$7.6 to $63)
$37
($16 to $64)
ECOSYSTEMS
Coral Reefs
$2,200
($2,100 to $2,200)
$2,100
($1,900 to $2,200)
$57
(-$39 to $290)
$2,200
($2,100 to $2,200)
$2,100
($2,100 to $2,200)
$27
($5 to $63)
Freshwater Fish
$450
(-$700 to $2,000)
$450
(-$37 to $1,300)
$0,030
(-$670 to $610)
$1,100
(-$560 to $2,000)
$280
(-$600 to $1,600)
$800
($42 to $1,600)
Wildfire
-$1.2
(-$3.6 to $2.4)
-$0.73
(-$2.6 to $1.8)
-$0.51
(-$4.0 to $3.2)
$2.9
(-$2.6 to $8.5)
-$1.4
(-$3.4 to $1.7)
$4.4
($0.58 to $6.8)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Coastal Property: Costs with no adaptation. See Modeling Framework section for a description of SLR uncertainty.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
233
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.3 MIDWEST
Using the results presented throughout the sector sections of
this Technical Report, this section summarizes the impacts
projected to occur in the Midwest.
Key Findings
• The Midwest is projected to experience a large number
of annual premature deaths from increased ground-
level ozone, under both climate scenarios and in both
2050 and 2090, making up approximately half of the
national total projected premature deaths. In 2090,
damages associated with these premature deaths are
projected to be $14 billion each year under RCP8.5 and
$8.8 billion each year under RCP4.5.
• The largest increases in premature deaths from extreme temperatures are projected to occur in
the Midwest, with significant losses in labor as well. Annual damages under RCP8.5 from these
extreme temperature impacts are similar, with estimated impacts in 2050 of $9.8 billion for
both premature mortality and lost labor. These costs rise to $31 billion and $33 billion per year
by 2090, respectively.
• The Midwest is also among the regions with the largest damages to infrastructure, especially
under RCP8.5; the region is projected to incur the second highest damages to roads and bridges.
Damages from climate impacts on rail under RCP8.5 rise from $0.5 billion each year in 2050 to
$1.4 billion each year by 2090.
• Rising temperatures will increase electricity demand across the Midwest, leading to estimated
costs on the electric power system by 2090 of $1.2 billion each year under RCP8.5 and $0.43
billion each year under RCP4.5.
• Corn yields are projected to decline considerably under RCP8.5, from a 7.1% decrease in annual
yield in 2050 to an 18% decrease in annual yield by the end of the century. Under RCP4.5,
projected declines are estimated at 4.3% and 5.5%, respectively.
Discussion
Tables 30.5 and 30.6 present the estimated annual physical and economic effects of climate change in
the Midwest under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.5, annual
physical impacts of climate change in the Midwest are projected to increase over time in all sectors and
under all climate scenarios, except for acres burned by wildfire and carbon storage loss. Impacts are
greater under RCP8.5 than under RCP4.5 for all sectors but carbon storage. As shown in Figure 30.4 and
Table 30.6, annual economic damages increase from 2050 to 2090 in all sectors but bridges and wildfire.
The RCP4.5 scenario demonstrates economic benefits when compared to RCP8.5 in all sectors but
freshwater fish in 2050 and inland flooding in 2090.
234
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.4. Largest Damages of Climate Change in the Midwest
Annual damages for the ten sectors with the greatest projected costs in the Midwest in 2090 under
RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The difference
between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data underlying the
sectors shown in the figure, as well as all other sectors modeled in the Midwest, can be found in Table
30.6 below.
Rail
$1.4 | -34%
Roads
$6.0 | -48%
Air Quality
$14 | -37%
Electricity
Demand and
Supply
$1.2 | -64%
West Nile Virus
$0.46 | -54%
Water Quality
$0.75 | -44%
Urban Drainage
$0.56 | -14%
Inland Flooding
$0.69 | 20%
Air Quality B Rai'
Extreme Temperature Mortality M Urban Drainage
Labor O Electricity Demand and Supply
West Nile Virus Inland Flooding
Roads ¦ Water Quality
235
-------
SYNTHESIS OF RESULTS
Regional Summaries
The Midwest is projected to experience the highest number of annual premature deaths from increased
ground-level ozone under both climate scenarios and in both 2050 and 2090, making up approximately
half of the national total projected premature deaths. The largest increases in deaths from extreme
temperatures are projected to occur in the Midwest, with significant losses in labor due to extreme heat
as well. Though this region has the largest increase in projected cyanobacteria concentration, it is not
among the regions with the highest recreational damages associated with harmful algal blooms.
The Midwest is also among the regions with the largest damages to infrastructure, especially under
RCP8.5; the region is projected to incur the second highest damages to roads and bridges (highest
damages to bridges in 2050). There are also high increases in electricity costs in the Midwest to meet
projected increases in demand. Large losses of suitable habitat are projected for warmwater species,
such as small and largemouth bass, in this region by the end of the century. Corn yields are projected to
decline considerably under RCP8.5, and losses in carbon storage of natural vegetation are projected.
Water quality is projected to decrease, especially under RCP8.5, with damages in the Midwest among
the highest of all the regions. Welfare losses associated with municipal and industrial water supply are
among the greatest in the Midwest. Finally, while the few locations across the country that experience
increases in winter recreation length in 2050 occur in the upper Midwest, this region as a whole
experiences the most significant reductions in winter recreation season length, as the region's elevation
is comparatively lower than the Rocky or Sierra Mountain regions and therefore particularly sensitive.
The most economically-significant impacts in the Midwest occur from lost labor hours and deaths
associated with increases in extreme temperature and ozone, all of which are on the order of billions of
dollars in damage each year from climate change. There are significant benefits in the health-related
damages under RCP4.5 compared to RCP8.5, particularly in 2090. For example, 1,200 deaths from
extreme temperatures are projected to be avoided each year under RCP4.5 compared to RCP8.5 by
2090, resulting in $18 billion in annual savings.
For additional considerations regarding the values shown in Tables 30.5 and 30.6, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
236
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.5. Projected Annual Physical Impacts of Climate Change in the Midwest
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.6 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality: # deaths
380
(200 to 550)
300
(160 to 440)
80
(NA)
910
(490 to 1,300)
580
(310 to 840)
330
(NA)
Extreme Temperature
Mortality: # deaths
790
(390 to 1,800)
690
(280 to 1,300)
97
(-160 to 520)
2,000
(1,300 to 3,300)
830
(340 to 2,100)
1,200
(620 to 2,000)
Labor: Lost Labor Hours
(millions)
200
(95 to 380)
160
(75 to 310)
33
(-53 to 80)
400
(180 to 640)
200
(110 to 390)
200
(62 to 320)
Aeroallergens:
ED visits
350
(58 to 520)
130
(-350 to 570)
220
(-82 to 870)
720
(210 to 1,100)
390
(-2.4 to 770)
330
(210 to 520)
Harmful Algal Blooms: #
Days above 100k cells/mL
2.5
(1.3 to 7.3)
1.7
(0.36 to 4.5)
0.75
(-0.22 to 2.8)
7.5
(1.7 to 17)
2.8
(1.2 to 7.6)
4.7
(0.51 to 9.4)
West Nile Virus:
# Cases
170
(120 to 250)
130
(93 to 190)
35
(26 to 54)
450
(260 to 690)
210
(140 to 310)
240
(120 to 380)
INFRASTRUCTURE
Bridges: # Vulnerable
Bridges
1,300
(770 to 1,900)
600
(270 to 1,100)
720
(160 to 1,400)
1,700
(510 to 2,900)
1,500
(660 to 2,000)
230
(-160 to 1,300)
WATER RESOURCES
Winter Recreation: Lost
Visits (millions)
3.4
(0.66 to 4.8)
-0.19
(-1.6 to 0.98)
3.6
(-0.32 to 5.5)
6.4
(4.0 to 7.5)
3.2
(0.61 to 5.0)
3.2
(2.3 to 5.0)
AGRICULTURE
Agriculture:
% Decrease in Corn Yields
(example crop)
7.1%
(-0.88% to 20%)
4.3%
(-3.3% to 14%)
2.7%
(-5.4% to 7.4%)
18%
(7.7% to 30%)
5.5%
(-3.1% to 18%)
13%
(8.0% to 22%)
237
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Freshwater Fish:
Coldwater Fishing Days
Lost (millions)
12
(11 to 12)
12
(11 to 12)
0.23
(-0.030 to 0.88)
12
(12 to 13)
12
(11 to 12)
0.38
(0.17 to 0.79)
Wildfire: Acres Burned
(thousands)
-39
(-190 to 120)
-110
(-230 to 68)
67
(-220 to 250)
-76
(-180 to 41)
-170
(-220 to -92)
90
(-32 to 190)
Carbon Storage: Metric
Tons Lost (millions)
6.3
(1.1 to 14)
8.8
(3.2 to 17)
-2.5
(-11 to 3.5)
3.4
(-14 to 10)
3.6
(1.8 to 5.7)
-0.19
(-17 to 4.4)
Note: "NA" indicates analyses where GCM-specific results are not available.
238
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.6. Projected Annual Economic Impacts of Climate Change in the Midwest
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Regional economic impacts on agriculture yield and welfare are not available. Due to rounding, benefit values may not equate to
differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$4,700
($420 to $13,000)
$3,700
($330 to $11,000)
$1,000
(NA)
$14,000
($1,200 to $39,000)
$8,800
($790 to $25,000)
$5,200
(NA)
Extreme
Temperature
Mortality
$9,800
($4,800 to $22,000)
$8,600
($3,500 to $16,000)
$1,200
(-$2,000 to $6,500)
$31,000
($19,000 to $50,000)
$13,000
($5,100 to $31,000)
$18,000
($9,500 to $30,000)
Labor
$9,800
($4,800 to $19,000)
$8,200
($3,800 to $16,000)
$1,600
(-$2,700 to $4,000)
$33,000
($15,000 to $53,000)
$17,000
($9,400 to $32,000)
$17,000
($5,200 to $26,000)
Aeroallergens
$0.17
($0,029 to $0.25)
$0,064
(-$0.17 to $0.28)
$0.11
(-$0,040 to $0.43)
$0.35
($0.10 to $0.55)
$0.19
(-$0.0012 to $0.38)
$0.16
($0.10 to $0.26)
Harmful Algal
Blooms
$4.2
($0 to $27)
$2.7
($0 to $18)
$1.5
(-$0.52 to $8.5)
$27
($0 to $84)
$4.4
($0 to $25)
$22
($0 to $59)
West Nile Virus
$140
($100 to $210)
$110
($78 to $160)
$29
($22 to $46)
$460
($260 to $700)
$210
($140 to $320)
$250
($120 to $380)
INFRASTRUCTURE
Roads
$3,300
($800 to $7,500)
$2,400
($1,000 to $5,000)
$970
(-$230 to $2,400)
$6,000
($2,600 to $10,000)
$3,100
($730 to $7,200)
$2,900
($1,800 to $3,900)
Bridges
$430
($200 to $670)
$390
($230 to $500)
$45
(-$33 to $270)
$270
($160 to $380)
$110
($52 to $200)
$160
($41 to $290)
Rail
$500
($390 to $650)
$430
($310 to $530)
$68
(-$55 to $120)
$1,400
($1,000 to $1,600)
$940
($650 to $1,200)
$470
($390 to $570)
Urban Drainage
$440
($120 to $880)
$330
($160 to $600)
$110
(-$450 to $720)
$560
($80 to $840)
$480
($41 to $1,200)
$79
(-$310 to $550)
239
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ELECTRICITY
Electricity
Demand and
Supply
$460
($390 to $620)
$280
($200 to $360)
$180
($22 to $300)
$1,200
($870 to $1,400)
$430
($220 to $680)
$770
($440 to $1,000)
WATER RESOURCES
Municipal and
Industrial Water
Supply
$29
(-$9.6 to $99)
$29
(-$15 to $85)
$0.39
(-$68 to $60)
$58
(-$39 to $150)
$57
(-$37 to $110)
$0.40
(-$32 to $47)
Inland Flooding
$670
(NA)
$540
(NA)
$120
(NA)
$690
(NA)
$830
(NA)
-$140
(NA)
Water Quality
$390
($220 to $560)
$310
($150 to $480)
$73
($36 to $120)
$750
($440 to $1,100)
$420
($200 to $690)
$330
($220 to $390)
Winter
Recreation
$190
($43 to $270)
-$26
(-$110 to $47)
$220
(-$4.0 to $330)
$360
($230 to $420)
$180
($38 to $280)
$170
($130 to $280)
ECOSYSTEMS
Freshwater Fish
$180
(-$370 to $670)
$180
(-$150 to $620)
-$4.9
(-$550 to $380)
$420
(-$710 to $900)
$270
(-$460 to $660)
$150
(-$250 to $380)
Wildfire
$0.52
(-$6.0 to $7.3)
-$2.5
(-$7.6 to $5.2)
$3.1
(-$9.5 to $11)
-$1.1
(-$5.4 to $3.6)
-$5.0
(-$7.4 to -$1.7)
$3.9
(-$1.5 to $8.2)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
240
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.4 NORTHERN PLAINS
Using the results presented throughout the sector
sections of this Technical Report, this section
summarizes the impacts projected to occur in the
Northern Plains.
Key Findings
• Many of the climate impacts projected to
occur in the Northern Plains are smaller in
relative magnitude than in other regions,
owing to comparatively smaller
populations and development. For
instance, the Northern Plains is among the
regions with the lowest projected proactive
adaptation costs for bridges, the lowest
cumulative rail impacts, and the lowest projected damages from inland flooding under both
RCPs and time periods.
• Lost labor hours from changes in extreme temperature are projected to rise under RCP8.5 from
$690 million each year in 2050 to $2.6 billion each year in 2090.
• Damages associated with projected increases in fatal and non-fatal cases of West Nile
neuroinvasive disease under RCP8.5 rise from $86 million each year in 2050 to $340 million each
year in 2090.
• As climate change leads to a loss of coldwater fishing habitat and a shift in habitat suitable for
warmwater species into areas suitable for rough water species, lost freshwater fishing days in
the Northern Plains will result in $66 million in damages per year under RCP8.5 and $25 million
damages per year under RCP4.5 by the end of the century.
• Air quality, West Nile virus, roads, rail, and winter recreation are all projected to experience
economic benefits under RCP4.5 when compared to RCP8.5.
Discussion
Tables 30.7 and 30.8 present the estimated annual physical and economic effects of climate change in
the Northern Plains under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.7,
annual physical impacts of climate change in the Northern Plains are projected to increase over time in
all sectors except winter recreation (under RCP4.5 only), acres burned by wildfire, and carbon storage
(under RCP4.5 only). Physical impacts are greater under RCP8.5 than RCP4.5 for all sectors in 2050
except corn yields and carbon storage, and for all sectors in 2090, except bridges and carbon storage. As
shown in Figure 30.5 and Table 30.8, annual economic damages also generally increase from 2050 to
2090 and from RCP4.5 to RCP8.5. However, bridges, urban drainage, municipal and industrial water
supply, and inland flooding all see higher costs under RCP4.5 compared to RCP8.5 for at least one of the
two time periods. The bridges, municipal and industrial water supply, and wildfire sectors all see higher
costs in 2050 than in 2090 for at least one of the two RCPs.
241
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.5. Largest Damages of Climate Change in the Northern Plains
Annual damages for the ten sectors with the greatest projected costs in the Northern Plains in 2090
under RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The
difference between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data
underlying the sectors shown in the figure, as well as all other sectors modeled in the Northern Plains,
can be found in Table 30.8 below.
Electricity Demand arid Supply
$0.06 | -56%
Air Quality
$0.6 | -30%
Water Quality
$0.1 | -38%
West Nile Virus
$0.3 | -56%
Rail
$0.6 I -37%
Freshwater Fish
-62%
A
i
r
Labor
L
$2.6 | -46%
Inland Flooding
$0.06 | 31%
Roads
$1.4 | -55%
Urban Drainage
$0.06 | 5.5%
Air Quality
Labor
West Nile Virus
Roads
Rail
Urban Drainage
Electricity Demand and Supply
Inland Flooding
Water Quality
Freshwater Fish
242
-------
SYNTHESIS OF RESULTS
Regional Summaries
As one of the least-populated areas of the contiguous U.S., many of the climate impacts projected to
occur in the Northern Plains are milder than those projected in other NCA regions.481 Though the
Northern Plains is among the regions with the largest increase in cyanobacteria in recreational
reservoirs, it is the only region with a projected increase in recreation days under a low algal growth
scenario compared to the control scenario (population growth but no climate change). Even under a
high growth scenario, the Northern Plains are projected to have the fewest lost reservoir recreation
days of any NCA region. The Northern Plains is also among the regions with the lowest proactive
adaptation costs for bridges under both climate scenarios and time periods, the lowest cumulative rail
reactive adaptation costs under both RCPs, and the lowest projected damages from inland flooding
under both RCPs and time periods. These results are largely influenced by the comparatively small
amount of infrastructure and development in the Northern Plains compared to other regions. However,
beyond economic impacts, damages to roads, rail systems, and bridges could be particularly meaningful
in the Northern Plains where there is a lack of infrastructure redundancy.
The Northern Plains is projected to have the lowest increases in cumulative electricity supply costs of all
the regions. Carbon storage is projected to increase in the Northern Plain under both RCPs, especially in
2090, and losses in corn yields are modest under RCP8.5, but reach 12% by 2090. Modest welfare losses
are projected in the municipal and industrial water supply sector, while the Northern Plains has the
second highest damages among the regions associated with urban drainage under RCP4.5. Both of these
effects are largely driven by projected increases in annual average precipitation for this region.
The most economically-significant impacts in the Northern Plains are projected to occur from damages
to roads (in 2090 under RCP8.5) and lost labor hours. Under RCP8.5, the estimated annual economic
damages by 2090 to roads and lost labor wages are on the order of billions of dollars. There are
significantly lower damages associated with labor loss under RCP4.5 compared to RCP8.5, particularly in
2090. Air quality, West Nile virus, roads, rail, and winter recreation are all projected to experience
economic benefits under RCP4.5 when compared to RCP8.5.
As climate change leads to a loss of coldwater fishing habitat and a shift of current warmwater fishing
habitat to rough water habitat, lost freshwater fishing days in the Northern Plains will result in $66
million in damages per year under RCP8.5 and $25 million damages per year under RCP4.5 by the end of
the century. In 2090, the Northern Plains is projected to lose 0.54 million recreational visits a year under
RCP8.5, equating to $47 million per year in damages. Under RCP4.5, winter recreational visitation would
see an increase of 1.1 million visits per year by 2090 (the net effect of losses due to climate and
increasing recreation due to population growth), resulting in $75 million in annual benefits.
For additional considerations regarding the values shown in Tables 30.7 and 30.8, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
481 Noting that none of the 49 cities modeled in the Extreme Temperature Mortality section were located in the Northern Plains.
243
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.7. Projected Annual Physical Impacts of Climate Change in the Northern Plains
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.8 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
Benefit
Benefit
RCP8.5
RCP4.5
RCP8.5
RCP4.5
HEALTH
Air Quality:
23
20
3.0
42
29
13
# deaths
(12 to 33)
(11 to 30)
(NA)
(22 to 61)
(16 to 43)
(NA)
Labor: Lost Labor Hours
14
11
2.8
31
16
15
(millions)
(5.6 to 21)
(3.7 to 16)
(-5.9 to 8.6)
(14 to 42)
(9.1 to 24)
(5.1 to 20)
Harmful Algal Blooms: #
5.7
5.2
0.43
15
8.0
7.5
Days above 100k cells/mL
(0.96 to 9.7)
(2.7 to 10)
(-2.9 to 4.5)
(3.0 to 28)
(1.4 to 17)
(1.7 to 15)
West Nile Virus:
100
79
23
330
150
190
# Cases
(51 to 150)
(40 to 120)
(11 to 31)
(150 to 480)
(64 to 220)
(86 to 260)
INFRASTRUCTURE
Bridges:
260
160
100
410
430
-18
# Vulnerable Bridges
(110 to 480)
(79 to 250)
(-54 to 310)
(190 to 630)
(260 to 580)
(-240 to 230)
WATER RESOURCES
Winter Recreation: Lost
-0.47
-0.50
0.030
0.54
-1.1
1.6
Visits (millions)
(-1.2 to -0.13)
(-0.83 to -0.32)
(-0.83 to 0.34)
(-1.2 to 1.4)
(-2.2 to -0.44)
(1.0 to 1.9)
AGRICULTURE
Agriculture:
% Decrease in Corn Yields
0.94%
1.8%
-0.90%
12%
2.0%
9.5%
(example crop)
(-5.9% to 6.8%)
(-7.3% to 13%)
(-14% to 5.0%)
(-1.3% to 17%)
(-7.7% to 11%)
(2.7% to 22%)
ECOSYSTEMS
Freshwater Fish:
Coldwater Fishing Days
0.68
0.58
0.098
1.2
0.69
0.48
Lost (millions)
(0.47 to 0.91)
(0.46 to 0.65)
(-0.030 to 0.26)
(0.75 to 2.0)
(0.57 to 0.89)
(0.16 to 1.1)
244
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
Wildfire: Acres Burned
(thousands)
-41
(-230 to 130)
-100
(-300 to 92)
62
(-4.5 to 130)
-85
(-260 to 55)
-190
(-310 to 40)
100
(-19 to 220)
Carbon Storage: Metric
Tons Lost (millions)
-6.3
(-25 to 4.7)
-5.0
(-10 to 2.8)
-1.2
(-14 to 9.1)
-7.2
(-34 to 6.7)
2.7
(-3.3 to 11)
-9.8
(-31 to 3.5)
Note: "NA" indicates analyses where GCM-specific results are not available.
245
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.8. Projected Annual Economic Impacts of Climate Change in the Northern Plains
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Regional economic impacts on agriculture yield and welfare are not available. Due to rounding, benefit values may not equate to
differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$280
($25 to $810)
$250
($23 to $720)
$30
(NA)
$630
($57 to $1,800)
$440
($40 to $1,300)
$190
(NA)
Labor
$690
($280 to $1,000)
$550
($190 to $780)
$140
(-$300 to $430)
$2,600
($1,200 to $3,400)
$1,300
($750 to $2,000)
$1,200
($420 to $1,600)
Harmful Algal
Blooms
-$0.86
(-$4.2 to $3.3)
-$0.94
(-$3.9 to $1.7)
$0,076
(-$2.1 to $2.0)
$0.27
(-$4.8 to $6.1)
-$0.58
(-$5.4 to $4.8)
$0.85
(-$0.46 to $3.8)
West Nile Virus
$86
($43 to $120)
$67
($34 to $99)
$19
($9.2 to $26)
$340
($150 to $490)
$150
($65 to $230)
$190
($87 to $270)
INFRASTRUCTURE
Roads
$580
($300 to $920)
$420
($200 to $650)
$160
($63 to $280)
$1,400
($610 to $2,000)
$590
($200 to $950)
$770
($410 to $1,100)
Bridges
$89
($55 to $120)
$91
($66 to $120)
-$1.7
(-$27 to $32)
$42
($18 to $66)
$25
($14 to $38)
$17
($3.6 to $41)
Rail
$180
($130 to $220)
$160
($100 to $190)
$23
(-$18 to $53)
$570
($370 to $690)
$360
($190 to $470)
$210
($180 to $240)
Urban Drainage
$21
($0 to $42)
$42
($0 to $87)
-$21
(-$48 to $17)
$62
($0 to $130)
$65
($0 to $110)
-$3.4
(-$81 to $88)
ELECTRICITY
Electricity
Demand and
Supply
$16
($13 to $18)
$10
($7.0 to $18)
$5.8
(-$4.9 to $11)
$61
($42 to $80)
$27
($9.1 to $42)
$34
($15 to $59)
246
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
WATER RESOURCES
Municipal and
Industrial Water
$0.24
$3.0
-$2.7
$0.43
$2.1
-$1.7
Supply
(-$1.8 to $3.6)
(-$0.25 to $9.2)
(-$9.2 to $3.8)
(-$7.2 to $12)
(-$4.4 to $12)
(-$9.6 to $4.8)
Inland Flooding
$58
$74
-$16
$61
$80
-$19
(NA)
(NA)
(NA)
(NA)
(NA)
(NA)
Water Quality
$56
$46
$9.5
$110
$66
$42
($33 to $79)
($26 to $74)
($2.8 to $20)
($69 to $160)
($22 to $110)
($34 to $47)
Winter
-$27
-$23
-$3.4
$47
-$75
$120
Recreation
(-$90 to -$4.4)
(-$47 to $4.9)
(-$95 to $25)
(-$87 to $110)
(-$160 to -$27)
($77 to $140)
ECOSYSTEMS
Freshwater Fish
00
T—1
-00-
$16
$1.9
$66
$25
$41
(-$13 to $37)
($0.16 to $38)
(-$22 to $20)
($20 to $88)
(-$1.3 to $40)
($7.3 to $56)
Wildfire
-$9.5
-$24
$15
-$22
-$44
$22
(-$52 to $32)
(-$69 to $23)
($0.99 to $35)
(-$62 to $11)
(-$72 to $4.6)
(-$5.6 to $48)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
247
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.5 SOUTHERN PLAINS
Using the results presented throughout the sector sections of this
Technical Report, this section summarizes the impacts projected
to occur in the Southern Plains.
Key Findings
• The most economically-significant impacts in the
Southern Plains occur from lost labor wages and
premature deaths associated with increases in extreme
temperature, with losses in 2090 under RCP8.5 equaling
$28 billion per year and $19 billion per year, respectively.
There are significant economic benefits in labor and
deaths from extreme temperature under RCP4.5
compared to RCP8.5. For instance, in 2090 annual
avoided damages are $9.9 billion from labor and $9.8
billion for avoided premature deaths from extreme
temperature under RCP8.5.
• The Southern Plains is projected to have some of the largest increases in cyanobacteria
concentrations and associated losses in recreation of all the regions due to harmful algal
blooms. Under RCP8.5, the number of days where recreational reservoirs surpass the
cyanobacteria concentration threshold representing a very high risk of short or long-term
adverse health effects rise from an additional 11 days per year in 2050 to an additional 15 days
per year in 2090.
• Projected climate impacts to infrastructure in the Southern Plains, such as rail and urban
drainage, are among the highest of all regions. Increases in electricity costs to meet projected
increases in demand in the Southern Plains are high, rising from $0.57 billion per year in 2050 to
$1.7 billion per year by 2090 under RCP8.5.
• Corn yields are projected to decline considerably under RCP8.5 in the Southern Plains, with yield
losses in 2090 of 23% under RCP8.5 and 7% under RCP4.5.
Discussion
Tables 30.9 and 30.10 present the estimated annual physical and economic effects of climate change in
the Southern Plains under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.9,
annual impacts in the Southern Plains are projected to increase over time in all sectors, except for
carbon storage and air quality under RCP8.5, and harmful algal blooms under RCP4.5. Adverse impacts
are projected to be greater under RCP8.5 than RCP4.5 in all sectors except carbon storage and air quality
in 2090. As shown in Figure 30.6 and Table 30.10, annual economic damages also generally increase
from 2050 to 2090, with the exception of at least one of the RCPs projecting damages in air quality,
roads, bridges, urban drainage, inland flooding, and wildfire. Some infrastructure sectors are projected
to have higher costs in 2050 than in 2090, as many types of infrastructure are already vulnerable or will
soon become vulnerable and require repair or adaptation costs early in the century.
248
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.6. Largest Damages of Climate Change in the Southern Plains
Annual damages for the ten sectors with the greatest projected costs in the Southern Plains in 2090
under RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The
difference between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data
underlying the sectors shown in the figure, as well as all other sectors modeled in the Southern Plains,
can be found in Table 30.10 below.
Roads
$1.3 | -71%
Coastal Property
$1.6 | -18%
Electricity Demand and Supply
$1.7 | -58%
Labor
$28 -35%
Inland Flooding
$0.90 | -58%
Rail
$0.69 | -35%
Water Quality
$0.53 | -34%
Freshwater Fish
$0.69 | -19%
Urban Drainage
$1.9 | -29%
i Extreme Temperature Mortality | Coastal Property
| Labor I Electricity Demand and Supply
Roads H Inland Flooding
¦ Rail ¦ Water Quality
¦ Urban Drainage ¦ Freshwater Fish
The Southern Plains is one of only two NCA regions where climate change is projected to result in fewer
premature deaths from ground-level ozone (in 2050 under RCP4.5 and 2090 under RCP8.5). This is
because climate-driven meteorological changes result in conditions slightly less conducive to ozone
formation, potentially due to an increase in precipitation. The Southern Plains also have among the
lowest increase in mortality from extreme temperatures of any region under both RCPs and time
249
-------
SYNTHESIS OF RESULTS
Regional Summaries
periods. Though carbon storage is projected to increase over time under RCP8.5, it will decrease under
RCP4.5.
With one of the largest projected increases in cyanobacteria concentrations, the Southern Plains will
have some of the largest projected losses in recreation due to harmful algal blooms. Projected costs to
the rail network in the Southern Plains are among the highest with proactive adaptation and the highest
under reactive adaptation of all the NCA regions under both RCPs. Projected average damages to urban
drainage for 25-year storms are among the highest under RCP8.5 and the highest for 50-year storms
under both RCPs. Increases in electricity costs to meet projected increases in demand in the Southern
Plains are high. Corn yields are also projected to decline considerably under RCP8.5, and cumulative
welfare loss in the municipal and industrial water supply sector is highest in the Southern Plains region
under RCP8.5.
The most economically-significant impacts in the Southern Plains occur from lost labor wages and
deaths associated with increases in extreme temperature, with losses in 2090 under RCP8.5 equaling
$28 billion a year and $19 billion a year, respectively. There are significant economic benefits in labor
and deaths from extreme temperature under RCP4.5 compared to RCP8.5, particularly in 2090. In 2090,
climate change impacts on electricity demand and supply are also large, with annual damages estimated
at $1.7 billion under RCP8.5 and $0.7 billion under RCP4.5.
For additional considerations regarding the values shown in Tables 30.9 and 30.10, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
250
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.9. Projected Annual Physical Impacts of Climate Change in the Southern Plains
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.10 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit |
HEALTH
Air Quality: # deaths
3.2
(1.7 to 4.7)
-4.2
(-6.2 to -2.3)
7.4
(NA)
-37
(-54 to -20)
89
(48 to 130)
-130
(NA)
Extreme Temperature Mortality:
550
350
200
1,300
620
650
# deaths
(280 to 840)
(170 to 480)
(34 to 380)
(960 to 1,500)
(280 to 930)
(570 to 690)
Labor: Lost Labor Hours (millions)
180
(110 to 260)
140
(56 to 200)
40
(9.4 to 55)
330
(210 to 440)
210
(170 to 270)
120
(40 to 170)
Harmful Algal Blooms: # Days above
11
10
0.69
15
7.9
7.1
100k cells/mL
(5.5 to 23)
(7.9 to 18)
(-4.9 to 4.9)
(4.3 to 32)
(3.1 to 15)
(0.91 to 17)
West Nile Virus: # Cases
220
(180 to 260)
200
(160 to 220)
25
(12 to 34)
450
(350 to 530)
330
(280 to 370)
120
(67 to 160)
INFRASTRUCTURE
Bridges: # Vulnerable Bridges
810
(300 to 1,200)
420
(27 to 760)
390
(-94 to 810)
1,100
(160 to 1,700)
1,000
(470 to 1,400)
25
(-320 to 500)
AGRICULTURE
Agriculture:
% Decrease in Corn Yields (example crop)
9.4%
(0.46% to 22%)
5.7%
(-1.5% to 12%)
3.7%
(-4.0% to 10%)
23%
(13% to 30%)
7.0%
(1.7% to 14%)
16%
(10% to 25%)
ECOSYSTEMS
Freshwater Fish: Coldwater Fishing Days
Lost (millions)
Values too small to differentiate from zero
Values too small to differentiate from zero
Wildfire: Acres Burned (thousands)
-50
(-84 to -22)
-39
(-92 to 13)
-11
(-46 to 70)
140
(-8.1 to 240)
-37
(-84 to 12)
180
(37 to 280)
Carbon Storage: Metric Tons Lost
1.1
-2.0
3.1
-17
-1.1
-16
(millions)
(-7.0 to 9.0)
(-7.3 to 4.6)
(-6.8 to 16)
(-19 to -15)
(-6.9 to 5.0)
(-23 to-9.2)
Note: "NA" indicates analyses where GCM-specific results are not available.
251
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.10. Projected Annual Economic Impacts of Climate Change in the Southern Plains
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Regional economic impacts on agriculture yield and welfare are not available. Due to rounding, benefit values may not equate to
differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$40
($3.6 to $110)
-$53
(-$150 to -$4.7)
$93
(NA)
-$560
(-$1,600 to -$50)
$1,400
($120 to $3,800)
-$2,000
(NA)
Extreme
Temperature
Mortality
$6,800
($3,400 to $10,000)
$4,400
($2,200 to $5,900)
$2,400
($420 to $4,800)
$19,000
($15,000 to $23,000)
$9,400
($4,300 to $14,000)
$9,800
($8,600 to $10,000)
Labor
$8,900
($5,300 to $13,000)
$6,900
($2,800 to $10,000)
$2,000
($470 to $2,800)
$28,000
($17,000 to $36,000)
$18,000
($14,000 to $22,000)
$9,900
($3,300 to $14,000)
Harmful Algal
Blooms
$12
(-$4.1 to $29)
$13
(-$4.7 to $38)
-$0.69
(-$37 to $16)
$38
($5.6 to $98)
$20
(-$3.3 to $56)
$18
(-$3.3 to $41)
West Nile Virus
$190
($150 to $210)
$160
($130 to $190)
$21
($10 to $29)
$460
($360 to $540)
$340
($290 to $370)
$130
($69 to $170)
INFRASTRUCTURE
Roads
$420
($96 to $890)
$370
($61 to $700)
$46
(-$310 to $350)
$1,300
($630 to $2,300)
$360
($98 to $630)
$920
($410 to $1,700)
Bridges
$300
($110 to $440)
$300
($170 to $410)
$0.25
(-$100 to $110)
$180
($44 to $320)
$83
($6.1 to $170)
$100
(-$19 to $220)
Rail
$240
($180 to $300)
$210
($130 to $260)
$36
(-$14 to $61)
$690
($500 to $790)
$450
($330 to $530)
$240
($180 to $280)
Urban Drainage
$950
($170 to $1,600)
$1,600
($670 to $2,400)
-$690
(-$2,000 to $890)
$1,900
($290 to $2,600)
$1,300
($620 to $1,900)
$560
(-$1,300 to $1,600)
Coastal Property
$840
(NA)
$800
(NA)
$49
(NA)
$1,600
(NA)
$1,300
(NA)
$290
(NA)
252
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ELECTRICITY
Electricity
Demand and
Supply
$570
($490 to $690)
$390
($230 to $540)
$180
(-$15 to $270)
$1,700
($1,400 to $1,900)
$720
($460 to $900)
$990
($670 to $1,200)
WATER RESOURCES
Municipal and
Industrial Water
Supply
$48
($8.2 to $75)
$37
($15 to $68)
$11
(-$15 to $42)
$100
($27 to $190)
$63
($32 to $110)
$41
(-$4.5 to $76)
Inland Flooding
$510
(NA)
$810
(NA)
-$300
(NA)
$900
(NA)
$380
(NA)
$520
(NA)
Water Quality
$220
($98 to $350)
$170
($45 to $270)
$52
($15 to $120)
$530
($400 to $620)
$350
($260 to $470)
$180
($130 to $240)
ECOSYSTEMS
Freshwater Fish
$630
($220 to $880)
$500
($230 to $760)
$120
(-$94 to $290)
$690
($360 to $1,100)
$570
($310 to $830)
$130
($43 to $270)
Wildfire
-$4.6
(-$7.6 to -$2.1)
-$3.8
(-$8.3 to -$0.41)
-$0.82
(-$4.6 to $6.2)
$11
(-$0.87 to $21)
-$4.0
(-$7.9 to $1.2)
$15
($3.6 to $27)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
253
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.6 SOUTHWEST
Using the results presented throughout the sector
sections of this Technical Report, this section
summarizes the impacts projected to occur in the
Southwest.
Key Findings
• The Southwest is projected to experience high
levels of premature mortality associated with
extreme temperatures. Compared to RCP8.5,
RCP4.5 is projected to substantially reduce
extreme temperature damages; 1,200 deaths
from extreme temperatures would be avoided
each year under RCP4.5 by 2090, resulting in $18 billion in annual savings.
• Extreme heat in the region leads to high labor losses; in 2090, losses of high-risk labor hours are
as much as 6.5% in Southwest counties under RCP8.5. Lost wages in 2090 are estimated at $23
billion per year under RCP8.5 and $12 billion per year under RCP4.5.
• Projected reactive adaptation costs from increased temperatures on the rail system in the
Southwest are among the highest across the nation. Under RCP8.5, reactive adaptation costs
associated with temperature impacts on rail rise from $0.32 billion per year in 2050 to $1.2
billion per year by 2090.
• The Southwest is projected to be the region with the largest level of future wildfire activity
under both RCPs, particularly in Colorado and Nevada. As a result, this region will incur the
highest cumulative wildfire response costs through the end of the century, making up more than
half the national total losses.
• The Southwest is projected to be among the regions with the largest lost welfare associated
with municipal and industrial water supply, with losses reaching $110 million per year under
RCP8.5 in 2090. Large losses of coldwater fish habitat are also projected under both RCPs, with
the number of lost fishing days under RCP8.5 rising from 8.3 million per year in 2050 to 18
million per year by 2090.
Discussion
Tables 30.11 and 30.12 present the estimated annual physical and economic effects of climate change in
the Southwest under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.11, annual
impacts in the Southwest are projected to increase overtime in all sectors, except under RCP4.5 in the
air quality, winter recreation, and wildfire sectors. While physical damages are generally greater under
RCP8.5 than RCP4.5, air quality, harmful algal blooms, and carbon storage sectors project greater
damages under RCP4.5 in 2050 and greater carbon storage damages in 2090. As shown in Figure 30.7
and Table 30.12, annual economic damages also generally increase from 2050 to 2090 and from RCP4.5
to RCP8.5. However, some sectors, including air quality, harmful algal blooms, urban drainage, municipal
and industrial water supply, and freshwater fish see higher damages in RCP4.5 than RCP8.5 in at least
one time period.
254
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.7. Largest Damages of Climate Change in the Southwest
Annual damages for the ten sectors with the greatest projected costs in the Southwest in 2090 under
RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The difference
between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data underlying the
sectors shown in the figure, as well as all other sectors modeled in the Southwest, can be found in Table
30.12 below.
Labor
$23 | -48%
Extreme Temperature Mortality
$31 | -58%
Rail Coastal Property
$1.2 | -37% $1.1 | -28%
Air Quality
$1.7 | -49%
Urban Drainage
$0.68 | -7.4%
Roads
$1.7 | -82%
Electricity Demand and Supply
$1.6 | -61%
Water Quality
Inland Flooding
$1.5 | -73%
Air Quality
Extreme Temperature Mortality
Labor
Roads
Rail
Urban Drainage
Coastal Property
Electricity Demand and Supply
Inland Flooding
Water Quality
The Southwest region is projected to incur significant damages associated with increased temperatures.
Large increases in mortality from extreme temperatures are projected, particularly under RCP8.5, where
the Southwest is the region with the highest mortality. Extreme heat also leads to high labor losses; in
2090, losses of high-risk labor hours are as much as 6.5% in counties within the Southwest under
RCP8.5. The Southwest is one of the regions with the highest projected increase in future cases of
255
-------
SYNTHESIS OF RESULTS
Regional Summaries
neuroinvasive West Nile virus, particularly in 2050. Increasing temperatures will also affect the rail
system, with projected reactive adaptation costs in the Southwest among the highest across the nation,
especially under RCP8.5.
Large losses of coldwater fish habitat are projected under both RCPs, but especially under RCP8.5 by
2090. The Southwest is projected to be the region with the largest level of future wildfire activity under
both RCPs, particularly in Colorado and Nevada. This region will also incur the highest cumulative
wildfire response costs through the end of the century, making up more than half the national total
losses (cumulative values not shown in Table 30.12; see wildfire sector). While corn yield losses are
moderate and carbon storage increases under both RCPs, especially later in the century, the Southwest
is projected to be among the regions with the largest lost welfare associated with municipal and
industrial water supply. Climate change will have slight adverse effects on winter recreation season
lengths and will therefore lead to fewer visits; however, these effects do not lead to net economic
damages because of increased visits due to population growth and relatively high lift ticket prices.
The most economically-significant impacts in the Southwest under RCP8.5 stem from extreme
temperature mortality and lost labor wages, on the order of billions to tens of billions of dollars in
damages each year from climate change. RCP4.5 is projected to substantially reduce extreme
temperature and labor related damages compared to RCP8.5, particularly in 2090. For example, 1,200
deaths from extreme temperatures would be avoided each year under RCP4.5 compared to RCP8.5 by
2090, resulting in $18 billion in annual savings. Avoided costs under RCP4.5, compared to RCP8.5, are
also very high (more than $1 billion per year) for roads, winter recreation, inland flooding, and electricity
demand and supply in 2090.
For additional considerations regarding the values shown in Tables 30.11 and 30.12, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
256
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.11. Projected Annual Physical Impacts of Climate Change in the Southwest
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.12 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality:
62
71
-9.0
110
57
53
# deaths
(33 to 91)
(38 to 100)
(NA)
(59 to 160)
(30 to 83)
(NA)
Extreme Temperature
Mortality:
# deaths
850
(460 to 1,600)
400
(180 to 790)
450
(190 to 800)
2,000
(1,200 to 3,000)
830
(380 to 1,500)
1,200
(820 to 1,700)
Labor: Lost Labor
120
78
43
280
150
130
Hours (millions)
(91 to 160)
(61 to 94)
(22 to 62)
(200 to 350)
(89 to 200)
(110 to 170)
Harmful Algal
Blooms: # Days above
100k cells/mL
12
(1.5 to 19)
12
(12 to 17)
-0.26
(-4.5 to 3.2)
15
(-0.60 to 26)
12
(0.57 to 19)
3.2
(-1.2 to 7.6)
West Nile Virus:
240
230
9.0
420
380
37
# Cases
(230 to 250)
(220 to 240)
(6.8 to 10)
(390 to 440)
(360 to 390)
(30 to 48)
INFRASTRUCTURE
Bridges:
160
120
46
360
260
94
# Vulnerable Bridges
(73 to 290)
(51 to 170)
(-32 to 130)
(200 to 530)
(120 to 370)
(-86 to 290)
WATER RESOURCES
Winter Recreation:
-2.6
-2.6
0.023
1.5
-10
12
Lost Visits (millions)
(-7.8 to 0.52)
(-3.4 to-1.4)
(-6.4 to 3.6)
(-10 to 6.5)
(-19 to -6.4)
(9.0 to 14)
AGRICULTURE
Agriculture:
% Decrease in Corn
-2.3%
-4.5%
2.2%
9.8%
-3.7%
14%
Yields (example crop)
(-6.4% to 6.9%)
(-8.4% to -0.61%)
(-1.9% to 7.5%)
(4.8% to 16%)
(-8.8% to -0.19%)
(8.9% to 19%)
257
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Freshwater Fish:
Coldwater Fishing
Days Lost (millions)
8.3
(6.0 to 13)
5.8
(2.9 to 9.9)
2.5
(0.63 to 3.3)
18
(12 to 26)
8.3
(4.1 to 13)
9.9
(6.5 to 14)
Wildfire: Acres
-850
-960
100
-840
-1,100
300
Burned (thousands)
(-1,700 to -440)
(-1,400 to -300)
(-300 to 390)
(-1,700 to -56)
(-1,400 to -730)
(-250 to 670)
Carbon Storage:
Metric Tons Lost
(millions)
-9.6
(-29 to 10)
-6.7
(-20 to 5.2)
-2.9
(-12 to 10)
-20
(-37 to -3.8)
-5.5
(-14 to 7.2)
-14
(-33 to 2.4)
Note: "NA" indicates analyses where GCM-specific results are not available.
258
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.12. Projected Annual Economic Impacts of Climate Change in the Southwest
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Due to rounding, benefit values may not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$770
($69 to $2,200)
$880
($79 to $2,500)
-$110
(NA)
$1,700
($150 to $4,800)
$860
($77 to $2,500)
$840
(NA)
Extreme Temperature
Mortality
$11,000
($5,700 to $20,000)
$5,000
($2,300 to $9,800)
$5,600
($2,400 to $9,900)
$31,000
($18,000 to
$45,000)
$13,000
($5,800 to
$22,000)
$18,000
($13,000 to
$26,000)
Labor
$6,100
($4,600 to $7,800)
$3,900
($3,000 to $4,700)
$2,100
($1,100 to $3,100)
$23,000
($17,000 to
$29,000)
$12,000
($7,300 to
$16,000)
$11,000
($8,800 to
$14,000)
Harmful Algal Blooms
$7.6
($3.2 to $13)
$4.9
($0.12 to $10)
$2.7
(-$1.3 to $11)
$6.6
(-$4.5 to $15)
$7.9
($2.8 to $12)
-$1.3
(-$11 to $7.0)
West Nile Virus
$200
($190 to $210)
$190
($180 to $200)
$7.6
($5.8 to $8.8)
$420
($390 to $450)
$390
($360 to $400)
$38
($31 to $49)
INFRASTRUCTURE
Roads
$490
(-$81 to $1,000)
$240
($30 to $360)
$250
(-$110 to $660)
$1,700
($470 to $3,300)
$280
($130 to $510)
$1,400
($340 to $2,800)
Bridges
$120
($68 to $200)
$95
($53 to $120)
$30
(-$47 to $92)
$54
($18 to $110)
$37
($12 to $65)
$17
(-$2.6 to $47)
Rail
$320
($240 to $400)
$250
($190 to $310)
$66
($46 to $88)
$1,200
($860 to $1,500)
$730
($470 to $940)
$440
($380 to $530)
Urban Drainage
$570
($170 to $1,100)
$590
($470 to $810)
-$22
(-$390 to $640)
$680
($250 to $970)
$630
($240 to $1,200)
$50
(-$460 to $430)
Coastal Property
$980
(NA)
$970
(NA)
$12
(NA)
$1,100
(NA)
$790
(NA)
$310
(NA)
ELECTRICITY
Electricity Demand and
Supply
$520
($390 to $660)
$370
($230 to $470)
$150
($92 to $240)
$1,600
($1,100 to $2,000)
$620
($420 to $770)
$980
($590 to $1,200)
259
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
WATER RESOURCES
Municipal and
Industrial Water
Supply
$31
(-$3.5 to $78)
$37
($25 to $68)
-$6.9
(-$29 to $10)
$110
($11 to $200)
$79
(-$0.24 to $180)
$29
(-$49 to $200)
Inland Flooding
$450
(NA)
$360
(NA)
$90
(NA)
$1,500
(NA)
$410
(NA)
$1,100
(NA)
Water Quality
$290
($140 to $470)
$240
($14 to $370)
$44
(-$69 to $120)
$650
($350 to $940)
$460
($130 to $780)
$190
($55 to $320)
Winter Recreation
-$250
(-$780 to $67)
-$240
(-$330 to -$58)
-$13
(-$730 to $390)
$170
(-$1,000 to $700)
-$1,100
(-$2,000 to -$660)
$1,200
($920 to $1,400)
ECOSYSTEMS
Freshwater Fish
$130
(-$130 to $450)
$150
($46 to $300)
-$19
(-$170 to $150)
$170
(-$610 to $560)
$140
(-$94 to $390)
$34
(-$520 to $290)
Wildfire
-$91
(-$210 to -$28)
-$120
(-$180 to -$37)
$25
(-$29 to $55)
-$100
(-$230 to -$35)
-$160
(-$200 to -$91)
$56
(-$30 to $130)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
260
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.7 NORTHWEST
Using the results presented throughout the sector sections of
this Technical Report, this section summarizes the impacts
projected to occur in the Northwest.
Key Findings
• The Northwest is among the regions with the highest
projected acres burned and will experience the second
highest cumulative wildfire response costs, on the
order of billions of dollars through the end of the
century. Acres burned and associated response costs
are higher in 2050 than 2090, in part due to shifts in
vegetation type.
• The Northwest is projected to have the highest damages to urban drainage for 10-year storms
under both RCPs; projected estimates in all other regions besides the Southeast are less than
half the costs in the Northwest. Damages under RCP8.5 are $84 million per year in both 2050
and 2090.
• Rising atmospheric C02 concentrations and climate change leads to losses associated with
shellfish and freshwater fishing. Large decreases in the supply of geoducks and oysters (with
subsequent price increases) are projected in the Northwest due to ocean acidification. Large
losses in coldwater fishing habitats in mountain regions lead to an annual loss of 8.1 million
fishing days under RCP8.5 and 2.1 million fishing days under RCP4.5 by 2090.
• Compared to other regions, cities located in the Northwest have low mortality rates from
extreme heat, with no significant projected changes in premature deaths in the future. Though
damages associated with air quality and labor losses are on the order of billions of dollars each
year by 2090, this region has relatively low increases in ozone related premature deaths in 2050
compared to other regions.
• Climate impacts on roads under RCP8.5 are projected to rise in the Northwest from $360 million
per year in 2050 to $950 million per year by 2090.
Discussion
Tables 30.13 and 30.14 present the estimated annual physical and economic effects of climate change in
the Northwest under RCP8.5 and RCP4.5 in the years 2050 and 2090. As shown in Table 30.13, annual
impacts in the Northwest are projected to increase over time in all sectors except wildfire under both
RCPs, and carbon storage and extreme temperature mortalities under RCP8.5. Annual impacts are also
greater under RCP8.5 than under RCP4.5 in all sectors except air quality and carbon storage, both in
2050 only, and extreme temperature mortality in 2090. As shown in Figure 30.8 and Table 30.14, annual
economic damages also generally increase from 2050 to 2090 and from RCP4.5 to RCP8.5. However,
several sectors, such as bridges and wildfire, see larger damages in 2050. Some infrastructure sectors
see higher costs in 2050 as many types of infrastructure are already vulnerable or will soon become
vulnerable and require repair costs early in the century. There are also a few sectors, including extreme
temperature mortality, harmful algal blooms, municipal and industrial water supply, and inland flooding
where damages are higher under RCP4.5 than under RCP8.5 in at least one time period.
261
-------
SYNTHESIS OF RESULTS
Regional Summaries
Figure 30.8. Largest Damages of Climate Change in the Northwest
Annual damages for the ten sectors with the greatest projected costs in the Northwest in 2090 under
RCP8.5 are shown by relative circle size and with the labeled monetary value (in $billions). The difference
between RCP8.5 and RCP4.5 in 2090 is shown as the second value (in % change). The data underlying the
sectors shown in the figure, as well as all other sectors modeled in the Northwest, can be found in Table
30.14 below.
Inland Flooding
$0.3 | -36%
Winter Recreation
$0.3 | -73%
Urban Drainage
$0.1 | -10%
Coastal Property
$0.3 | -12%
Labor
$1.9 I -63%
Electricity Demand
and Supply
$0.6 | -67%
Water Quality
$0.1 | -34%
Roads
$1.0 I -69%
Air Quality
$1.4 | -68%
Rail
$0.2 | -47%
Air Quality
Labor
Roads
Rail
Urban Drainage
Coastal Property
Electricity Demand and Supply
Inland Flooding
Water Quality
Winter Recreation
262
-------
SYNTHESIS OF RESULTS
Regional Summaries
The Northwest region is projected to experience generally moderate climate impacts in some sectors
compared to some other NCA regions. For example, the Northwest will have relatively low increases in
ozone related premature deaths due to climate change in 2050. Cities located in the Northwest have
low mortality rates from extreme heat in both the reference and future years, such that there are no
significant projected changes in premature deaths from extreme heat in this region. This region also has
the smallest projected increase in neuroinvasive West Nile virus cases and relatively low costs from
harmful algal bloom damages among the regions. Projections result in relatively low inland flooding
damages, the fewest vulnerable bridges, and the lowest bridge repair costs. It is one of just two regions
projected to see cumulative welfare gains in municipal and industrial water supply under both RCPs.
Conversely, the Northwest is among the regions with the highest cumulative costs for rail in RCP8.5,
particularly when assuming that proactive adaptation measures are taken. Damages to urban drainage
are projected to be highest in the Northwest for 10-year storms under both RCPs; projected estimates in
all other regions besides the Southeast are less than half the costs in the Northwest. The Northwest is
projected to experience large decreases in the supply of geoducks and oysters (with subsequent price
increases) due to ocean acidification, and large losses in coldwater fishing habitats in mountain regions
under both RCPs. Furthermore, this region is among those with the highest projected acres burned and
will experience the second highest cumulative wildfire response costs, on the order of billions of dollars
through the end of the century (cumulative values not shown in Table 30.14; see wildfire sector).
However, the Northwest is projected to experience increases in corn yields, and the largest increase in
stored carbon through 2100 under both RCPs, which results in large benefits, particularly under RCP8.5.
The most economically-significant impacts in the Northwest are damages associated with air quality and
labor, both of which are on the order of billions of dollars each year by 2090. Several sectors, including
air quality, labor, roads, and electricity demand and supply are projected to see significant benefits
(avoided damages) under RCP4.5 compared to RCP8.5, particularly in 2090.
For additional considerations regarding the values shown in Tables 30.13 and 30.14, see the notes at the
beginning of the Regional Summaries section and the footnotes to the National Summary table. For
more detailed results, as well as background and modeling approaches, see the individual sectoral
chapters.
263
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.13. Projected Annual Physical Impacts of Climate Change in the Northwest
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. See notes at the bottom of Table 30.14 for additional sector-specific information. Due to rounding, benefit values may
not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality:
20
5.2
15
93
29
64
# deaths
(10 to 29)
(2.8 to 7.6)
(NA)
(50 to 140)
(16 to 43)
(NA)
Extreme Temperature
Mortality:
# deaths
3.3
(-0.58 to 12)
1.1
(-1.0 to 3.7)
2.2
(-1.7 to 13)
-0.38
(-1.1 to 0.94)
3.1
(0.68 to 7.8)
-3.4
(-6.9 to-1.2)
Labor: Lost Labor Hours
6.9
4.3
2.6
23
8.8
15
(millions)
(3.3 to 16)
(1.7 to 8.8)
(0.81 to 6.8)
(12 to 40)
(3.1 to 22)
(9.1 to 18)
Harmful Algal Blooms:
# Days above 100k
cells/mL
0.19
(-0.040 to 0.54)
0.13
(0.018 to 0.43)
0.060
(-0.090 to 0.38)
1.1
(-0.010 to 3.0)
0.30
(-0.040 to 1.1)
0.78
(0.040 to 2.0)
West Nile Virus:
6.5
6.4
0.11
11
11
0.49
# Cases
(6.3 to 6.6)
(6.2 to 6.5)
(0.054 to 0.15)
(11 to 11)
(10 to 11)
(0.38 to 0.61)
INFRASTRUCTURE
Bridges:
120
83
37
200
160
42
# Vulnerable Bridges
(59 to 200)
(12 to 160)
(-47 to 93)
(76 to 290)
(96 to 210)
(-36 to 130)
WATER RESOURCES
Winter Recreation:
1.5
-0.26
1.8
3.6
1.0
2.5
Lost Visits (millions)
(-0.81 to 3.1)
(-1.1 to 0.34)
(-1.1 to 3.5)
(0.37 to 5.5)
(-2.1 to 3.4)
(2.1 to 3.1)
AGRICULTURE
Agriculture:
% Decrease in Corn
-14%
-15%
1.5%
0.42%
-14%
14%
Yields (example crop)
(-22% to -8.7%)
(-20% to -11%)
(-2.3% to 5.3%)
(-10% to 10%)
(-18% to -8.3%)
(7.8% to 21%)
264
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Freshwater Fish:
Coldwater Fishing Days
Lost (millions)
2.1
(0.67 to 3.3)
0.93
(-0.60 to 1.9)
1.2
(0.75 to 1.5)
8.1
(2.4 to 15)
2.1
(0.80 to 3.7)
6.0
(0.50 to 11)
Shellfish:
% Decrease in Oyster
Supply (example
species)
26%
(24% to 29%)
21%
(16% to 27%)
5.7%
(1.9% to 8.4%)
52%
(49% to 55%)
27%
(22% to 33%)
25%
(22% to 28%)
Wildfire: Acres Burned
(thousands)
110
(-98 to 440)
56
(-64 to 190)
51
(-46 to 320)
-15
(-220 to 110)
-80
(-180 to 95)
64
(-44 to 190)
Carbon Storage: Metric
Tons Lost (millions)
-15
(-28 to -7.5)
-12
(-18 to -6.2)
-2.2
(-10 to 1.5)
-16
(-31 to -6.2)
-7.0
(-11 to -0.92)
-8.8
(-20 to -2.1)
Note: "NA" indicates analyses where GCM-specific results are not available.
265
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.14. Projected Annual Economic Impacts of Climate Change in the Northwest
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Regional economic impacts on agriculture yield and welfare are not available. Due to rounding, benefit values may not equate to
differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
HEALTH
Air Quality
$240
($22 to $690)
$65
($5.8 to $180)
$180
(NA)
$1,400
($130 to $4,000)
$450
($40 to $1,300)
$950
(NA)
Extreme
Temperature
Mortality
$41
(-$7.2 to $150)
$14
(-$13 to $46)
$27
(-$21 to $160)
-$5.8
(-$16 to $14)
$46
($10 to $120)
-$52
(-$100 to -$19)
Labor
$350
($170 to $790)
$220
($87 to $440)
$130
($41 to $340)
$1,900
($1,000 to $3,300)
$730
($260 to $1,800)
$1,200
($750 to $1,500)
Harmful Algal
Blooms
$0.18
(-$0,050 to $0.69)
$0.22
(-$0,069 to $0.73)
-$0,033
(-$0.18 to $0,061)
$3.5
(-$0.19 to $16)
$0.15
(-$0.10 to $0.61)
$3.4
(-$0.12 to $16)
West Nile Virus
$5.5
($5.3 to $5.6)
$5.4
($5.2 to $5.5)
$0,090
($0,045 to $0.13)
$11
($11 to $12)
$11
($10 to $11)
$0.50
($0.38 to $0.62)
INFRASTRUCTURE
Roads
$360
($200 to $500)
$210
($90 to $320)
$150
($71 to $280)
$950
($580 to $1,400)
$300
($160 to $450)
$660
($310 to $960)
Bridges
$83
($48 to $130)
$71
($56 to $86)
$13
(-$7.9 to $42)
$31
($18 to $51)
$22
($2.6 to $42)
$9.1
(-$11 to $30)
Rail
$45
($33 to $63)
$36
($24 to $57)
$8.7
($5.7 to $11)
$160
($96 to $230)
$89
($42 to $130)
$75
($54 to $110)
Urban Drainage
$84
($46 to $130)
$70
($45 to $93)
$14
(-$47 to $71)
$84
($61 to $120)
$75
($65 to $83)
$8.7
(-$22 to $47)
Coastal Property
$250
(NA)
$240
(NA)
$11
(NA)
$250
(NA)
$220
(NA)
$28
(NA)
266
-------
SYNTHESIS OF RESULTS
Regional Summaries
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ELECTRICITY
Electricity
Demand and
Supply
$160
($98 to $240)
$100
($54 to $150)
$57
($25 to $94)
$550
($270 to $880)
$180
($87 to $280)
$370
($190 to $610)
WATER RESOURCES
Municipal and
Industrial Water
Supply
-$0.44
(-$0.52 to -$0.37)
-$0.40
(-$0.52 to -$0.26)
-$0,039
(-$0.16 to $0.15)
-$0.27
(-$0.61 to $0.46)
-$0.33
(-$0.63 to $0.32)
$0,069
(-$0.60 to $0.57)
Inland Flooding
$100
(NA)
$130
(NA)
-$21
(NA)
$280
(NA)
$170
(NA)
$100
(NA)
Water Quality
$58
($18 to $96)
$45
(-$5.4 to $80)
$13
($1.8 to $24)
$140
($60 to $210)
$90
($9.7 to $150)
$47
($27 to $69)
Winter
Recreation
$110
(-$57 to $220)
-$44
(-$120 to $34)
$160
(-$91 to $290)
$260
($27 to $400)
$76
(-$150 to $240)
$190
($160 to $220)
ECOSYSTEMS
Freshwater Fish
-$42
(-$84 to -$8.6)
-$59
(-$100 to -$24)
$17
(-$17 to $37)
$34
(-$34 to $120)
-$56
(-$100 to -$2.9)
$90
(-$31 to $180)
Wildfire
$22
(-$20 to $110)
$7.8
(-$16 to $35)
$15
(-$18 to $71)
-$15
(-$63 to $19)
-$29
(-$61 to $22)
$14
(-$7.4 to $44)
Notes:
"NA" indicates analyses where GCM-specific results are not available.
Air Quality: Mean and upper/lower bounds based on confidence intervals from the BenMAP-CE model.
Harmful Algal Blooms: Range and mean values based on combined high and low growth scenarios.
Urban Drainage: Values represent results under the 50-year storm.
Electricity Demand and Supply: Values represent power system supply costs. Results are from the GCAM power sector model only.
Water Quality: Range and mean values based on combined results from US Basins and HAWQS.
Freshwater Fish: Values represent impacts to all three fishing guilds (coldwater, warmwater, and rough)
Wildfire: Results represent changes in both the contiguous U.S. and Alaska.
267
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.8 ALASKA
This section presents infrastructure and wildfire
projections for Alaska.
Key Findings *,
• In Alaska, wildfire acres burned each year by
2090 are projected to be 260,000 acres under
RCP8.5, leading to response costs of $11
million each year. However, a decrease in
area burned of 370,000 acres occurs under
RCP4.5, avoiding a total of 640,000 acres from
burning.
• The largest increase in area burned under RCP8.5 is projected in the southwestern part of the
state. Under all scenarios and timeframes, the eastern parts of the state show a decrease in
wildfire activity.
• Road flooding associated with increased precipitation is projected to be the largest source of
infrastructure damages in Alaska, followed by damages to buildings associated with permafrost
thaw. Smaller damages are estimated for Alaskan airports, rail, and pipelines. Overall,
infrastructure damages from climate change by 2090 are $170 million per year under RCP8.5
and $82 million per year under RCP4.5.
Discussion
As described above, only a subset of sectors described in this Technical Report were run for the state of
Alaska. As shown in Table 30.15, wildfire acres burned decrease under RCP4.5 and increase under
RCP8.5 across much of the state, with higher impacts under 2090 than 2050. The largest increase in area
burned under RCP8.5 is projected in the southwest part of the state. Under all scenarios and
timeframes, the eastern parts of the state show a decrease in wildfire activity. The benefits of RCP4.5
compared to RCP8.5 are on the order of hundreds of thousands of acres burned per year. This results in
economic benefits in terms of avoided wildfire response costs, as seen in Table 30.16.
Also shown in Table 30.16, annual damages to Alaskan infrastructure are greater under RCP8.5 than
RCP4.5, but are also slightly less in 2090 than in 2050. This finding is consistent with some infrastructure
results across regions of the contiguous U.S., as many types of infrastructure are already vulnerable or
will soon become vulnerable and require repair costs earlier in the century. Road flooding associated
with increased precipitation is projected to be the largest source of reactive repair costs, followed by
impacts to buildings associated with permafrost thaw. Smaller reactive adaptation costs are estimated
for Alaskan airports, railroads, and pipelines. Beyond economic effects, climate-driven changes to
infrastructure could be particularly meaningful in Alaska where there is a lack of redundancy across
most of the state.
For additional considerations regarding the values shown in Tables 30.15 and 30.16, see the notes at the
beginning of the Regional Summaries section. For more detailed results, as well as background and
modeling approaches, see the individual sectoral chapters.
09f
268
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.15. Projected Annual Physical Impacts of Climate Change in Alaska
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the GCMs. Not all sectoral analyses produced discrete
physical metric estimates.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Wildfire: Acres
Burned (thousands)
220
(-160 to 600)
-700
(-760 to -650)
920
(600 to 1,200)
260
(130 to 400)
-370
(-570 to -180)
640
(310 to 970)
Table 30.16. Projected Annual Economic Impacts of Climate Change in Alaska
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Due to rounding, benefit values may not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
INFRASTRUCTURE
Alaska Infrastructure
(Reactive adaptation
costs only)
$180
($170 to $180)
$120
($110 to $130)
$60
($55 to $66)
$170
($130 to $220)
$82
($80 to $84)
$92
($49 to $140)
ECOSYSTEMS
Wildfire
$10
($1.3 to $19)
-$15
(-$16 to -$14)
$25
($18 to $33)
$11
($7.0 to $15)
-$5.2
(-$9.6 to -0.74)
$16
($7.7 to $25)
269
-------
SYNTHESIS OF RESULTS
Regional Summaries
30.9 HAWAI'I AND PUERTO RICO
This section summarizes the
coral reef projections for Hawai'i
and Puerto Rico.
Key Findings
• Extensive loss of coral
reefs is projected for
Hawai'i and Puerto Rico
under all scenarios and
time periods.
o
q
>
v
• In Hawai'i, the annual
percent of coral cover
lost under RCP8.5 rises
from 70% in 2050 to 96%
in 2090. This loss leads to damages of $1.3 billion per year in 2050 and $1.9 billion per year in
2090. In 2090, RCP4.5 would avoid 16% of coral cover loss and $470 million per year compared
to RCP8.5.
• In Puerto Rico, coral reefs pass a critical ecosystem threshold in the first several decades of the
century, such that the differences between annual percent coral cover lost between time
periods or under alternative climate scenarios is small. Under RCP8.5, coral cover loss in Puerto
Rico rises from 93% in 2050 to 95% in 2090.
As described above, results for Hawai'i and Puerto Rico only cover the coral reef sector. As shown in
Table 30.17, extensive loss of coral reefs is projected for Hawai'i and Puerto Rico under all scenarios and
time periods. Damages increase overtime and are higher in RCP8.5 than in RCP4.5 in Hawai'i. As Puerto
Rican reefs pass critical thresholds for ecosystem loss, the difference between RCP8.5 and RCP4.5 are
small. Coral reef recreation is projected to decline considerably under all scenarios, though slightly less
under RCP4.5. In most cases, more than 90% of the value of reference period coral reef recreation is lost
by the end of the century. Economic impacts, as shown in Table 30.18, are on the order of billions of
dollars each year in Hawai'i, with larger damages under RCP8.5 compared to RCP4.5 and in 2090
compared to 2050. Damages are smaller in Puerto Rico, as the values only represent recreational losses
for non-tourist residents.
For additional considerations regarding the values shown in Tables 30.15 and 30.16, see the notes at the
beginning of the Regional Summaries section. For more detailed results, as well as background and
modeling approaches, see the individual sectoral chapters.
270
-------
SYNTHESIS OF RESULTS
Regional Summaries
Table 30.17. Projected Annual Physical Impact of Climate Change on Coral Reefs in Hawai'i and Puerto Rico
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Unless noted, upper and lower bounds are derived from values across the five GCMs. Not all sectoral analyses produced discrete
physical metric estimates. Due to rounding, benefit values may not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Hawaiian Coral:
70%
64%
5.6%
96%
79%
16%
% Loss Cover
(11% to 97%)
(7.3% to 94%)
(-11% to 17%)
(88% to 98%)
(26% to 97%)
(-1.2% to 63%)
Puerto Rican
Coral:
93%
93%
-0.80%
95%
97%
-1.4%
% Loss Cover
(89% to 95%)
(85% to 96%)
(-6.7% to 9.7%)
(92% to 98%)
(96% to 97%)
(-3.9% to 0.88%)
Table 30.18. Projected Annual Economic Impact of Climate Change on Coral Reefs in Hawai'i and Puerto Rico
Positive numbers represent damages due to climate change, while negative numbers represent a reduction in damages compared to the
reference period. Values shown in millions of undiscounted $2015. Unless noted, upper and lower bounds are derived from values across the
GCMs. Due to rounding, benefit values may not equate to differences between RCPs.
2050
2090
RCP8.5
RCP4.5
Benefit
RCP8.5
RCP4.5
Benefit
ECOSYSTEMS
Hawaiian Coral
$1,300
(-$240 to $1,900)
$1,100
(-$330 to $1,900)
$140
(-$290 to $420)
$1,900
($1,700 to $2,000)
$1,400
(-$120 to $1,900)
$470
(-$36 to $1,800)
Puerto Rican
Coral
$0.24
($0.22 to $0.24)
$0.24
($0.21 to $0.25)
-$0.0026
(-$0,022 to $0,032)
$0.24
($0.23 to $0.25)
$0.25
($0.25 to $0.25)
-$0.0054
(-$0,015 to $0.0033)
271
-------
Errata:
The following minor changes have been made to the report since its finalization on May 11, 2017:
• Section 3 (Air Quality): clarified source for base non-GHG emissions data.
• Section 7 (West Nile Virus): corrected Southeast and Midwest case estimates for RCP4.5 in 2090.
Matching edits made to the Regional Summaries for the Southeast (Section 30.2) and Midwest (Section
30.3).
• Section 12 (Bridges): corrected proactive adaptation response cost estimates. Matching edits made to the
National Summary (Section 28).
• Section 23 (Coral Reefs): corrected summation error for Puerto Rico recreation damages (Table 23.1).
• Updated several references to include final journal publication information (e.g., DOI numbers,
volume/page numbers), including updates to Table 2.2.
• Corrected formatting and layout of Figures 1.14 (partially obscured) and 20.1 (missing legend)
• Corrected a limited number of discrete significant figure rounding errors in Tables 30.2, 30.4, 30.6, 30.8,
30.10, 30.12, and 30.14.
• Corrected transcription errors for air quality economic damages in Table 30.4 and Figure 30.3.
v»EPA
United States
Environmental Protection
Agency
United States
Environmental Protection Agency
1200 Pennsylvania Avenue, N.W. (6207A)
Washington, DC 20460
Official Business
Penalty for Private Use $300
EPA 430-R-17-001
May 2017
------- |