DRAFT
DO NOTCITE OR QUOTE
EPA/600/R-09/099A
September 2009
CLIMATE CHANGE IMPACTS ON HUMAN HEALTH DUE
TO CHANGES IN AMBIENT OZONE CONCENTRATIONS
Draft Report
NOTICE
THIS DOCUMENT IS A PRELIMINARY DRAFT THIS INFORMATION IS
DISTRIBUTED SOLELY FOR THE PURPOSE OF PRE-DISSEMINATION PEER
REVIEW UNDER APPLICABLE INFORMA TION QUALITY GUIDELINES IT HAS NOT
BEEN FORMALLY DISSEMINATED BY THE U.S. ENVIRONMENTAL PROTECTION
AGENCY. IT DOES NOT REPRESENT AND SHOULD NOT BE CONSTRUED TO
REPRESENT ANY AGENCY DETERMINATION OR POLICY.
Global Change Research Program
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
EXECUTIVE SUMMARY
Reports from the Intergovernmental Panel on Climate Change (IPCC) and the U.S.
National Research Council (NRC) have stated that future climate change has the potential
to cause air quality degradation via climate-induced changes in meteorology and
atmospheric chemistry, posing challenges to the U.S. air quality management system and
the effectiveness of its pollution mitigation strategies.
For the past several years, the Global Change Research Program (GCRP) in EPA's Office
of Research and Development (ORD), in partnership with EPA's Office of Air and
Radiation (OAR) and the academic research community, has been evaluating the potential
consequences of global climate change for air quality in the United States. An overview
report of the initial phases of this effort, Assessment of the Impacts of Global Change on
Regional U.S. Air Quality: A Synthesis of Climate Change Impacts on Ground-Level
Ozone, describing the results from studies using linked climate change and air quality
models to simulate the possible range of changes in ozone (03) concentrations across the
United States associated with future climate change, was released in April, 2009.
A second EPA GCRP report, Land-Use Scenarios: National-Scale Housing-Density
Scenarios Consistent with Climate Change Storylines, released in June 2009, describes the
Integrated Climate and Land-Use Scenarios (ICLUS) project. Under ICLUS, a number of
high-resolution, spatially explicit population projections consistent with assumptions in
the IPCC Special Report on Emissions Scenarios (SRES) social, economic, and
demographic storylines were developed for the United States.
The work described here builds on these two reports. In this current project, we take the
next step of examining the potential indirect impacts of climate change on the health of a
future U.S. population (c. 2050) via its direct impact on 03 concentrations. This analysis
considers the health impacts associated with O3 changes induced only by future climate
change. To achieve this, modeling scenarios were designed to simulate the response of 03
to global climate change alone without changes in anthropogenic emissions of ozone
precursors (e.g., due to future air quality management efforts and/or future economic
growth).
Because of the extreme complexity of the coupled climate-air quality-health system, and
the many uncertainties present at each step of analysis, it is most useful to frame this study
as a sensitivity analysis. Therefore, we have attempted to assess the sensitivity of modeled
human health impacts to assumptions about, and modeling and methodological choices
for, the following key inputs:
• Climate-induced changes in future meteorological conditions;
• Corresponding changes in O3 concentrations in response to these meteorological
changes;
• The size, and geographic distribution across the United States, of the affected
population;
11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
• The concentration-response (C-R) relationships that link O3 levels to specific
health outcomes;
• The fraction of the year over which O3 is assumed to affect health (i.e., the "O3
season").
The Environmental Benefits Mapping and Analysis Program (BenMAP), EPA's premier
air pollution benefits analysis model, was the system used to integrate the diverse climate,
O3, and population scenarios to estimate the changes in adverse health effects resulting
from climate-induced O3 concentration changes. BenMAP contains within it a database of
C-R functions from the epidemiological literature. Each O3 C-R function is an estimate of
the relationship between ambient O3 concentrations and a population health effect (e.g.,
premature mortality or hospital admissions for respiratory illnesses). For several of the
health effects that have been associated with exposure to ambient O3, more than one C-R
function has been reported in the epidemiological literature. There is no one "correct" set
of C-R functions to use to estimate 03-related adverse health effects, and EPA has used
different sets of functions in different benefits analyses involving 03, as both methods and
available functions have evolved over time. For this analysis we looked to the benefit
analysis for the most recent 03 National Ambient Air Quality Standards (NAAQS)
Regulatory Impact Analysis (RIA), completed in 2008.
Using these C-R relationships, along with the scenarios of ambient O3 concentrations and
population around 2050, BenMAP estimates the number of cases in the population of each
03-related adverse health effect attributable to climate change in each grid cell (30 km x
30 km) of the conterminous United States. National-level impacts, as well as impacts in
three broad regions - the Northeast, the Southeast, and the West were delineated for this
analysis.
The major conclusions of this report are as follows:
• Looking across all combinations of climate change/air quality models, population
projections, 03 season definitions, and C-R functions for all-cause premature
mortality considered in our analysis, estimates of national 03-related all-cause
premature mortality around 2050 attributable to climate change range from -1,092
to 4,240 - that is, from over 1,000 cases of 03-related premature mortality avoided
because of climate change to over 4,200 cases attributable to climate change.
Despite this range, the large preponderance of the estimates are positive,
suggesting that, all else being equal, climate change would be likely to increase the
incidence of 03-related all-cause premature mortality in 2050.
• The source of the greatest uncertainty at the national level appears to be the
particular climate change/air quality models used.
• The choice of population projection also made a significant difference, although
only about half that of climate change/air quality scenario at the national level.
• It is important to take into account that the size of the population exposed to 03
will increase by a future year. Failing to do so will result in estimates that are
substantially biased downward. In one case, for example, of the almost 400-case
difference in estimates of 03-related premature mortality produced by the two
111
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
population projection extremes (holding climate change/air quality model
constant), 67 percent (or 265 deaths) was due to the difference between the result
produced by the Census_2000 population "projection" (175 deaths), which
assumes no population change from the year 2000, and the next highest result,
produced by the Woods & Poole projection (439 deaths).
• Not only is the total population exposed to O3 in a future year important, but the
age (and geographic) distribution of that population can also make a significant
difference in the estimated impact of climate change on 03-related adverse health
effects. For example, the number of 03-related deaths estimated using the
population projection with the greatest total population (424.8 million) was less
than the number estimated using a different population projection with a smaller
total of only 386.7 million. This is because about 26 percent of the latter
population is 65 or older compared to only about 21 percent of the former,
resulting in more premature deaths.
• The national results can mask important regional differences. The Northeast
showed the most consistent level of 03-related premature mortality across the
climate/air quality scenarios used in this study, while in the Southeast the
estimated premature mortality impacts varied significantly across the scenarios.
The West generally showed the smallest impacts across the scenarios, largely due
to the smaller projected populations compared to the Northeast and Southeast.
• A climate-induced extension of the O3 season later into the fall and earlier into the
spring has the potential to significantly increase the incidence of adverse health
outcomes.
At this stage in the development of our scientific understanding of climate change and its
potential impact on air pollution-related human health, it would be unwise to rely on any
one model or any one population projection. This may be the most important "take away"
message of our analysis. The different model combinations can produce widely varying
results, particularly at the regional level, in some cases leading to fundamentally different
conclusions about the overall impact of climate change on 03-related health effects. This
has a number of implications for the development of meaningful analyses to assess the
range of benefits associated with responses to climate change. However, while there is a
very wide range of results, including some that suggest that climate change would
decrease the incidence of 03-related mortality, the large preponderance of results across
the different climate change/air quality models and population projections suggest that, all
else being equal, climate change would produce an increase in 03-related adverse health
effects in 2050.
iv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
AUTHORS, CONTRIBUTORS, AND REVIEWERS
The Global Change Research Program, within the National Center for
Environmental Assessment, Office of Research and Development, is responsible for
publishing this report. This document was prepared by Abt Associates under Contract #
EP-D-08-100, Work Assignment 0-07. Ms. Anne Grambsch served as the Work
Assignment Manager, provided overall direction and she contributed as an author. Dr.
Chris Weaver and Phil Morefield provided technical assistance that involved obtaining
climate, air quality and population data, processing the data for use in the health effects
model, and contributing as authors to the report.
AUTHORS
Ellen Post
Hardee Mahoney
Jin Huang
Ana Bel ova
Emily Connor
Abt Associates Inc.
4550 Montgomery Ave.
Bethesda, MD 20814
Anne Grambsch
Chris Weaver
Phil Morefield
Global Change Research Program
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
REVIEWERS
This report benefited from an internal EPA review. The authors thank the
following individuals for their thoughtful and constructive comments and suggestions.
Doug Grano
Neal Fann
Office of Air and Radiation
Office of Air Quality Planning and Standards
Meredith Kurpius
Region 9
Britta Bierwagen
Janet Gamble
Office of Research and Development
National Center for Environmental Assessment
Alice Gilliland
Office of Research and Development
National Exposure Research Lab
Barbara Glen
Office of Research and Development
National Center for Environmental Research
v
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
TABLE OF CONTENTS
LIST OF TABLES viii
LIST OF FIGURES ix
LIST 01 ABBREVIATIONS xi
1. INTRODUCTION 1-1
2. BACKGROUND 2-1
3. OVERVIEW OI METHODS 3-1
4. MODELING CLIMATE CHANGE AND CORRESPONDING CHANGES IN
AMBIENT OZONE CONCENTRATIONS 4-1
4.1. The Harvard University Research Effort 4-1
4.2. The Carnegie-Mellon University Research Effort 4-3
4.3. The EPA NERL Research Effort 4-3
4.4. The University of Illinois Research Effort 4-4
4.5. The Washington State University Research Effort 4-6
4.6. The GIT-NESCAUM-MIT Research Effort 4-7
4.7. A Summary of Climate Change/Air Quality Research Efforts 4-8
5. POPULATION PROJECTIONS 5-1
5.1. Extrapolation of Woods and Poole Population Projections to 2050 5-1
5.2. ICLUS Population Projections to 2050 5-2
5.2.1. ICLUS population projection A1 5-4
5.2.2. ICLUS population projection A2 5-4
6. MODELING HUMAN HEALTH IMPACTS OF PREDICTED CHANGES IN
AMBIENT OZONE CONCENTRATIONS 6-1
6.1. An Overview of BenMAP 6-1
6.2. The Structure of an Air Pollutant Benefit Analysis 6-2
6.3. Estimation of the 03-Related Human Health Impacts of Climate Change 6-4
6.3.1. Ambient O3 concentrations: Adjustment of modeled with- and without-
climate-change O3 concentrations in 2050 6-5
6.3.2. Concentration-response functions 6-8
6.3.3. Baseline incidence rates 6-10
6.3.4. Populations 6-13
6.3.5. Summary of key features of the analysis 6-13
6.3.6. Assessing and characterizing uncertainty 6-14
7. RESULTS AND DISCUSSION 7-1
7.1. National Results 7-1
7.1.1. Mortality 7-1
7.1.2. Morbidity 7-4
vi
-------
1 7.2. Uncertainty 7-6
2 7.2.1. Influence of 03 Changes from the Climate Change/Air Quality Models 7-
3 6
4 7.2.2. Influence of Projected Population Changes 7-8
5 7.3. Regional Results 7-12
6 7.4. Extension of the 03 Season 7-16
7 7.5. Conclusions 7-18
8 8. REFERENCES 7-20
9
10
vii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
LIST OF TABLES
Table
2-1
Table
4-1
Table
4-2
Table
5-1
Table
6-1
Table
7-1
Summary of Selected Studies Projecting Impacts of Climate Change-Related
Ozone Health Impacts 2-5
Summary of Regional Climate and O3 Modeling Systems 4-8
Summary of Global Climate and 03 Modeling Systems Used in This Analysis
4-10
ICLUS Population Projection: Demographic Characteristics 5-3
Summary of Concentration-Response Functions Used to Estimate Climate
Change-Related Impacts of 03 on Human Health 6-9
Estimated National 03-Related Incidence of Morbidity Attributable to Climate
Change in 2050 During the 03 Season, Taken to be June, July, and August,
Based on Different Combinations of Climate Change/Air Quality Model and
Population Projection 7-5
viii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
LIST OF FIGURES
Figure 3-1. The Structure of the Analysis of 03-Related Impacts on Human Health
Attributable to Climate Change 3-2
Figure 3-2. Regions of the United States Defined for this Analysis 3-3
Figure 4-1. 2050s-Minus-Present Differences in Simulated Summer Mean MDA8 O3
Concentrations (in ppb) for the (a) NERL; (b) GNM; (c) Illinois 1; (d) Illinois
2; and (e) WSU experiments 4-9
Figure 6-1. Example of Grid Cells and Monitors 6-6
Figure 6-2. Illustration of Voronoi Cells Around the Same Grid Cell Center 6-7
Figure 7-1. Estimated National 03-Related Cases of All-Cause Mortality in 2050 O3
Season (Defined as June, July, and August) Due to Climate Change: Impact
of Climate Change/Air Quality Model, Population Projection, and C-R
Function 7-2
Figure 7-2. Indicators and Abbreviations for Climate Change/Air Quality Models and
Population Projections in Figures 7-3
Figure 7-3. Estimated National (VRelated Non-Accidental Mortality (Based on Bell et
al., 2004) in 2050 03 Season (Defined as June, July, and August) Due to
Climate Change: Impact of Climate Change/Air Quality Model, for Each
Population Projection 7-7
Figure 7-4. Legend for Climate Change/Air Quality Models in Figures 7-7
Figure 7-5. Estimated National 03-Related Non-Accidental Mortality (Based on Bell et
al., 2004) in 2050 O3 Season (Defined as June, July, and August) Due to
Climate Change: Impact of Population Projection, for Each Climate
Change/Air Quality Model 7-9
Figure 7-6. Legend for Population Projections in Figures 7-9
Figure 7-7. Age Distributions of ICLUS Al and ICLUS A2 Population Projections to
the Year 2050 7-11
Figure 7-8. Estimated National and Regional 03-Related Non-Accidental Mortality
(Based on Bell et al., 2004) in 2050 O3 Season (Defined as June, July, and
August) Due to Climate Change: Impact of Climate Change/Air Quality
Model, for Each Population Projection 7-14
Figure 7-9. Estimated National and Regional 03-Related Non-Accidental Mortality
(Based on Bell et al., 2004) in 2050 O3 Season (Defined as June, July, and
August) Due to Climate Change: Impact of Population Projection, for Each
Climate Change/Air Quality Model 7-15
Figure 7-10. Estimated National 03-Related Cases of All-Cause Mortality in 2050 O3
Season (Bell et al., 2005) Due to Climate Change: Impact of O3 Season
Definition 7-17
IX
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
LIST OF ABBREVIATIONS
AGCM
Atmospheric General Circulation Model
AOGCM
Atmosphere-Ocean General Circulation Model
AQ
air quality
BC
boundary conditions
BEIS
Biogenic Emissions Inventory System
CAA
Clean Air Act
CAM
Community Atmosphere Model
CACM
Caltech Atmospheric Chemistry Mechanism
CICE
The Los Alamos Sea Ice Model
CCM3
Community Climate Model version 3
CCSM
Community Climate System Model
CSIM
Community Sea Ice Model
CLM
Community Land Model
CMAQ
Community Multiscale Air Quality Model
C-R
concentration-response
CMIP
Coupled Model Intercomparison Project
CTM
Chemical Transport Model
EC
elemental carbon
ENSO
El Nino-Southern Oscillation
GCM
General Circulation Model
GCTM
Global Chemical Transport Model
GISS
Goddard Institute for Space Studies
GMAO
Global Modeling and Assimilation Office
HadCM3
Hadley Centre Coupled Model
IC
initial condition
IGSM
Integrated Global System Model
LANL
Los Alamos National Laboratory
LWC
liquid water content
MDA8
Maximum Daily 8-hour Average Ozone Concentration
MM
Mesoscale Model
MM5
Mesoscale Model (Version 5)
MARKAL
MARKet Allocation Model
MOSIS
Meteorology Office Surface Exchange Scheme
MPMPO
Model to Predict the Multiphase Partitioning of Organics
NAAQS
National Ambient Air Quality Standard
NCAR
National Center for Atmospheric Research
NH4+
ammonium ion
N03-
nitrate ion
OC
organic carbon
03
ozone
OGCM
Oceanic General Circulation Model
PAN
peroxyacetylnitrate
PBL
planetary boundary layer
x
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
PCM
Parallel Climate Model
PCTM
PCM/CCSM Transition Model
POP
Parallel Ocean Program
RACT
reasonably available control technology
RCM
Regional Climate Model
RCMS
Regional Climate Modeling System
RCTM
Regional Chemical Transport Model
RH
relative humidity
RRF
relative reduction factor
PM2.5
particulate matter with aerodynamic diameter below 2.5 [j,m
SIP State
Implementation Plan
SAPRC
statewide air pollution research center
SMOKE
Sparse Matrix Operator Kernel Emissions
SOA
secondary organic aerosols
S02
sulfur dioxide
S04=
sulfate ion
SRES
special report on emissions scenarios
SST
sea surface temperature
THC
thermohaline circulation
TKE
turbulent kinetic energy
UKMO
United Kingdom Meteorology Office
voc
volatile organic compound
XI
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
CLIMATE CHANGE IMPACTS ON HUMAN HEALTH VIA
CHANGES IN AMBIENT OZONE CONCENTRATIONS
1. INTRODUCTION
There is now a substantial and growing literature on the potential impacts of climate
change that may occur in the absence of efforts to mitigate the atmospheric accumulation
of greenhouse gases due to global emissions and other factors (e.g., deforestation). The
recent Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report
(AR4) found that "warming of the climate system is unequivocal, as is now evident from
observations of increases in global average air and ocean temperatures, widespread
melting of snow and ice, and rising global average sea level" (IPCC, 2007). The IPCC
also found that "most of the observed increase in globally averaged temperatures since
the mid-20th century is very likely due to the observed increase in anthropogenic
greenhouse gas concentrations." Furthermore, of particular importance for the U.S.
Environmental Protection Agency's (EPA's) mission to protect human health and the
environment was the IPCC finding that "future climate change may cause significant air
quality degradation by changing the dispersion rate of pollutants, the chemical
environment for ozone and aerosol generation and the strength of emissions from the
biosphere, fires and dust. The sign and magnitude of these effects are highly uncertain
and will vary regionally."
Discussion of the potential sensitivity of air quality to climate change has increased in
recent years. In 2001, the National Research Council (NRC) posed the question "To what
extent will the United States be in control of its own air quality in the coming decades?"
noting that".. .changing climatic conditions could significantly affect the air quality in
some regions of the United States ..." and calling for the expansion of air quality studies
to include investigation of how U.S. air quality is affected by long-term climatic changes
(NRC, 2001). A subsequent NRC report emphasized that the U.S. air quality management
system must be "flexible and vigilant" to ensure the effectiveness of pollution mitigation
strategies in the face of climate change (NRC, 2004).
The Global Change Research Program (GCRP) in EPA's Office of Research and
Development (ORD) has been evaluating the potential consequences of global climate
change and climate variability for air and water quality, aquatic ecosystems, and human
health in the United States. In an initial report, Assessment of the Impacts of Global
Change on Regional U.S. Air Quality: A Synthesis of Climate Change Impacts on
Ground-Level Ozone (U.S. EPA, 2009a), the GCRP provides air quality managers and
scientists with timely and useful information about the potential effects of climate change
on air quality in the United States. This report, written in partnership with EPA's Office
of Air and Radiation (OAR), describes the multidisciplinary research efforts using linked
climate change and air quality models to simulate the possible range of changes in ozone
1-1
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
(O3) concentrations across the United States as a result of the future meteorological
changes associated with modeled climate change scenarios.
The GCRP assessment was designed to be carried out in two phases. In the first phase,
modeling systems were used to consider the sensitivity of air quality responses to global
climate change alone; this includes direct meteorological impacts on atmospheric
chemistry and transport, and the effect of these meteorological changes on climate-
sensitive natural emissions of pollutant precursors, such as volatile organic compounds
(VOCs) and nitrogen oxides (NOx), but not changes in anthropogenic emissions of these
pollutants (e.g., due to future air quality management efforts and/or future economic
growth). The second phase, now ongoing, is tackling the additional complexities of
integrating the effects of changes in anthropogenic emissions, in the U.S. and worldwide,
with the climate-only impacts investigated in the first phase.
In a second report, Land-Use Scenarios: National-Scale Housing-Density Scenarios
Consistent with Climate Change Storylines (U.S. EPA, 2009b), the GCRP considers the
interactions between climate change and changes in land use, observing that land use
could exacerbate or alleviate climate-change effects. Noting that it is important to use
land-use scenarios that are consistent with the assumptions underlying recognized
international climate-change scenarios, this report describes its Integrated Climate and
Land-Use Scenarios (ICLUS) project. The ICLUS project developed several population
projections based on the IPCC Special Report on Emissions Scenarios (SRES) social,
economic, and demographic storylines. The GCRP adapted these storylines to the United
States and modified U.S. Census Bureau population and migration projections to be
consistent with these storylines.
The work described in this report builds on the work described in these two preceding
GCRP reports. In this current project we take the next step, examining the potential
indirect impacts of climate change on the health of a future U.S. population via its direct
impact on O3 concentrations that adversely affect human health. Our analysis is based on
only the first phase of the GCRP climate and air quality assessment (U.S. EPA, 2009a) -
i.e., it considers the health impacts associated with O3 responses to global climate change
alone, not including changes in anthropogenic emissions of ozone precursor pollutants
(e.g., due to future air quality management efforts and/or future economic growth).
Following the climate change/air quality modeling efforts on which the current effort
builds, we focus on future years around 2050.
To achieve its mission to protect human health and the environment, EPA implements a
variety of programs under the Clean Air Act that reduce ambient concentrations of air
pollutants. Secondary pollutants such as O3 are not emitted directly into the atmosphere:
instead they are created by chemical reactions between NOx and VOCs in the presence of
heat and sunlight. These pollutants are emitted from a variety of sources, including motor
vehicles, chemical and power plants, refineries, factories, and consumer and commercial
products, as well as natural sources such as vegetation, lightning, and biological
processes in the soil. EPA's efforts have been successful: between 1980 and 2007,
emissions of VOCs and NOx decreased by 50 and 39 percent, respectively, even though
1-2
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
gross domestic product increased 124 percent, vehicle miles traveled increased 103
percent, and energy consumption increased 30 percent (U.S. EPA, 2008). Air pollution,
however, including O3 pollution, continues to be a widespread public health and
environmental problem in the United States, with peak level 03 concentrations in
numerous counties still exceeding the National Ambient Air Quality Standards (NAAQS)
for O3,1 and with health effects ranging from increased premature mortality to chronic
impacts on respiratory and cardiovascular health (e.g., see Jerrett et al., 2009).
Significant regional variability already exists in ground-level O3 under current climate
conditions. A large body of observational and modeling studies have shown that 03
concentrations tend to be especially high where the emissions of VOCs and NOx are also
large, and that 03 concentrations increase even more when meteorological conditions
most strongly favor net photochemical production - persistent high pressure, stagnant air,
lack of convection, clear skies, and warm temperatures (e.g., U.S. EPA, 1989; NRC,
1991; Cox and Chu, 1993; Bloomfield et al., 1995; Morris et al., 1995; Sillman and
Samson, 1995; EPA, 1999; Thompson et al., 2001; Camalier et al., 2007; among many
others). Consequently, the O3 NAAQS are most often exceeded during summertime hot
spells in places with large natural and/or anthropogenic NOx and VOC emissions (e.g.,
cities and suburban areas).
Since climate change may alter weather patterns, and, hence, potentially increase the
frequency, duration, and intensity of 03 episodes in some regions, this has the potential to
create additional challenges for air quality managers. However, the links between long-
term global climate change and 03 changes is not necessarily straightforward, reflecting a
balance among multiple interacting factors (U.S. EPA, 2009a). For example, the current
relationship between temperature and 03 does not necessarily provide a basis for
predicting O3 concentrations in a warmer future climate, since higher temperatures are
often correlated with other important drivers of O3 such as sunlight and stagnation (NRC,
1991; U.S. EPA, 2009a).
As noted above, this analysis focuses on the time period around 2050, following the
climate change/air quality modeling efforts on which the current effort builds. Section 2
of this report gives background information on the literature relevant to the analysis
discussed here. Section 3 provides an overview of the methods used in our analysis. The
methods used in each component of the analysis are described in more detail in the
subsequent sections of the report. In particular, Section 4 describes the climate
change/air quality models that were used to simulate changes in O3 concentrations across
the United States resulting from various modeled climate change scenarios. Section 5
describes the five population projections used in the analysis. Section 6 describes how
we modeled the human health impacts resulting from the climate-induced changes in O3
concentrations. Finally, Section 7 presents and discusses the results of our analysis.
1 Currently set at 75 parts per billion (ppb) for the 8-hour NAAQS.
1-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
2. BACKGROUND
There are many papers in the literature that describe the potential relationships between
climate change, air quality, and human health in general terms. Several of these studies
have outlined steps for estimating the health impacts resulting from climate change.
Other studies have used climate change and air quality models of varying formulations
and complexities, along with health impact functions, to estimate air pollution-related
health impacts expected to result from climate change.
Patz, et al. (2000) summarized research resulting from the National Assessment of the
Potential Consequences of Climate Variability and Change (NAPCCVC) under the U.S.
Global Climate Change Research Program. They identified five categories of health
outcomes that are likely to be affected by climate change: temperature-related morbidity
and mortality; health effects of extreme weather events (such as storms, tornadoes,
hurricanes, and precipitation extremes); air pollution-related health effects; water- and
food-borne diseases; and vector- and rodent-borne diseases. The National Assessment's
categorization of potential health effects generally agrees with the categorization
described in the IPCC's Fourth Assessment Report (IPCC, 2007).
The analysis described in this report focuses only on air pollution-related - in particular,
03-related - health effects. Based on the NAPCCVC document, Bernard et al. (2001)
outlined the pathways through which climate change may affect exposure to air pollutants
that have been associated with adverse health effects. Climate change may affect
exposures to air pollution by affecting:
• weather and pollutant transport and transformations;
• anthropogenic emissions (mitigative and/or adaptive actions);
• natural sources of air pollutant emissions (such as biogenic VOCs); and
• the distribution and types of airborne allergens (Bernard, et al., 2001).
Increased temperatures and sunlight due to climate change might impact the
development, transport, and dispersion of O3. Higher temperatures can accelerate
photochemical reactions that form O3 in the troposphere (Bernard, et al., 2001). Forests,
shrubs, grasslands and other natural sources of VOCs may emit greater quantities at
higher temperatures. Higher temperatures may also increase soil microbial activity which
may lead to an increase in NOx (Bernard, et al., 2001). However, as O3 is formed by
complex secondary reactions dependent upon the amount of sunlight and relative NOx
and VOCs levels, O3 levels do not always increase with increasing temperature. In
addition, changing temperatures can have an impact on the mixing height or wind speed
and direction. Therefore, the impact of increased temperatures and other meteorological
changes on O3 must be evaluated using atmospheric models that simulate the
photochemistry and physical advection and diffusion processes that influence ambient O3
levels across regional scales.
2-1
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
The climate change and air quality models used for the analysis of Ch-related health
effects due to climate change, described in this report, incorporate weather and pollutant
transport and transformation mechanisms as well as mechanisms for potential increases
in biogenic VOC emissions, based on the simulated future meteorology. However, the
present analysis does not incorporate the other two pathways noted above -
anthropogenic emissions and airborne allergens - through which climate change can
influence air-pollution-related health outcomes. For completeness, we discuss these
briefly below.
Anthropogenic emissions of air pollutants may occur as a result of actions to either
mitigate or adapt to climate change. Mitigative actions, such as implementation of
strategies to reduce C02 emissions, may have the additional benefit of also reducing
criteria air pollutant levels, resulting in short- and longer-term human health benefits
(Bell, et al., 2008; Cifuentes, et al., 2001). These benefits (also known as co-benefits or
ancillary benefits) may be substantial. Including them in the analysis, however, would
require additional assumptions about greenhouse gas control policies. Adaptive actions
in response to climate change may include, for example, increased fossil fuel burning to
satisfy demand for electricity needed for air conditioning purposes.
Climate change may also affect air pollution-related health outcomes by altering the
distribution and types of airborne allergens. Higher temperatures and potentially higher
C02 levels may themselves result in earlier onset of the pollen season and greater pollen
production (Kinney, 2008), thereby increasing the prevalence and severity of asthma and
related allergic diseases (Shea, 2008). Given that air pollution may facilitate penetration
of allergens into the lungs, increases in levels of airborne allergens may magnify the
health effects of air pollutants.
Past literature has noted substantial uncertainty surrounding the likely response of future
air pollutant concentrations to climate change, because ambient concentrations of these
air pollutants are the result of meteorological conditions, natural systems, and human
activities (Bernard et al., 2001). In an update to the NAPCCVC, Ebi et al. (2006)
continued to report on studies finding both increased and decreased O3 concentrations,
depending upon the locations and scenarios considered. The divergent results reflect
differences in model assumptions as well as a number of factors influencing O3 levels
(Ebi et al., 2006).
Exposure to O3 may result in several minor and severe adverse health outcomes,
including decreased lung function, increased airway reactivity, lung inflammation,
emergency room visits and hospitalizations for respiratory illnesses, and premature
mortality (see, e.g., Bernard, et al., 2001; Bell et al., 2004; Bell et al., 2005; Levy et al.,
2005; Burnett et al., 2001; Moolgavkar et al., 1997; Jerrett et al., 2009). Several studies
have examined the potential effects of climate change on 03-related morbidity and
mortality by linking climate change and air quality models.
Knowlton et al. (2004) used an integrated modeling framework to assess 03-related
health impacts in future decades. They linked a global climate model (GCM, created by
2-2
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
the Goddard Institute for Space Studies) with a regional climate model (the Mesoscale
Model 5; MM5). The modeling domain was comprised of 36-km grid cells across the 31
counties in the New York metropolitan region, covering an area of 33,600 km2 and a
population of 21 million. The linked GCM/MM5 models provided inputs to an air
quality model (the Community Multi-scale Air Quality model; CMAQ), which was used
to estimate daily 1-hour maximum 03 concentrations for five summers (June - August) in
the 1990s and 2050s. Changes in greenhouse gas emissions for the 2050s were adopted
from the A2 Standard Reference Emission Scenario described by the IPCC, which
predicted a 1.6-3.2°C temperature increase in the 2050s compared with the 1990s.
Population and age structure were held constant at year 2000 level and distribution.
Considering climate change alone, there was a median 4.5 percent increase in (Vrelated
premature mortality across the New York metropolitan area. The authors found that
incorporating O3 precursor emissions did not have a significant impact on the results, but
incorporating population growth did.
Bell (2007) followed the framework of Knowlton et al. (2004) but expanded the
geographic scope of the analysis to 50 U.S. cities. This study again focused specifically
on the impact of altered climate on 03 and human health. Potential changes in
anthropogenic emissions (other than greenhouse gas emissions) were not considered.
The scenarios employed projected overall increases in 03 concentration levels in the
targeted future year, with larger increases in cities that currently experience high
pollution. Across the 50 U.S. cities considered, the simulated summertime daily 1-hour
maximum O3 concentration increased by 4.8 ppb, on average. The average number of
days per summer season on which the 8-hour regulatory ozone standard was exceeded
increased by a factor of 1.7. Total daily (Vrelated premature mortality was estimated to
increase by 0.11- 0.27 percent, depending upon the concentration-response function
employed.
Ebi and McGregor (2008) describe several other studies, all using different assumptions
regarding climate scenarios, models, time intervals, baseline conditions, and population
projections. Although it is difficult to compare results across studies with varying
assumptions, we briefly summarize three studies described in Ebi and McGregor (2008).
Hwang et al. (2004) found an average increase in O3 peaks of 2.0 - 3.2 ppb in the 2050s
(2.1-2.7°C temperature increase), with a corresponding increase in daily mortality
ranging from 0.08 to 0.46 percent, depending on the concentration-response function
used. Increases were similarly found for hospital admissions. Anderson et al. (2001)
found increases in O3 to result in 20 percent more premature deaths in 2050 in the United
Kingdom (corresponding to a 0.89-2.44°C temperature increase). West et al. (2007)
examined 10 world regions and found large increases in O3, with a population-weighted
average of 9.4 ppb, in the year 2030 under the IPCC A2 Standard Reference Emission
Scenario. Of these three studies, West et al. (2007) was the only study to consider
population growth. All results were sensitive to specification of O3 thresholds in the
concentration-response relationships used to estimate O3 health effects.
2-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Tagaris et al. (2009) used regional CMAQ air pollutant modeling for the years 2001 and
2050, as well as the Goddard Institute for Space Studies global and the MM5 regional
climate models for the year 2050 to estimate PM2.5 and O3 concentrations. Health effects
were assessed using The Environmental Benefits Mapping and Analysis Program
(BenMAP), the air pollution health impact model also used in our analysis. The authors
did not estimate population projections for the year 2050; instead, they used 2001
population levels. Their simulations showed that two-thirds of the U.S. were adversely
affected by climate change-driven air quality-related health effects. Like other analyses,
the authors noted that both positive and negative impacts varied geographically.
Nationally, they estimated approximately 4,000 additional premature deaths due to PM2.5
and 300 due to O3. They also found that in almost a dozen states the increased premature
mortality due to increased 03 levels was offset by reduced premature mortality due to
decreased PM2.5 They also noted large uncertainties, however, including those
associated with emissions projections to simulate future climate, meteorological
forecasting, downscaling from large scale to small scale models, PM2.5 speciation,
pollutant-pollutant interactions, temperature-pollutant interactions, and concentration-
response functions.
The model-based scenarios of potential impacts of climate change on 03-related health
effects from several studies conducted in the U.S. are summarized below in Table 2-1.
There is a substantial literature focusing on temperature-related health impacts.
Moreover, these impacts may include not only the direct effects of temperature on human
health, but also an indirect effect via the influence of temperature on air pollutant-related
health effects - i.e., there may be synergistic effects of temperature and air pollution on
human health. A few studies have reported such synergistic relationships between
temperature and 03-related health outcomes. In their analysis of 60 large eastern U.S.
communities during April to October, 1987-2000, Ren et al. (2008a) found that
temperature modified 03-premature mortality associations and that such modification
varied across geographic regions. In particular, they found that in the northeast region a
10-ppb increment in ozone was associated with an increase of 2.22 percent, 3.06 percent,
and 6.22 percent in mortality at low, moderate, and high temperature levels, respectively.
However, such a pattern was not apparent in the southeast region. Ren et al. (2008b)
found similar patterns for cardiovascular mortality, O3, and temperature during the
summer in 95 large U.S. communities.
Although the analysis described in this report does not address the climate change-related
health impacts of temperature, such a focus, including possible synergies between
temperature and O3 in their impacts on human health, would be a natural sequel to the
current analysis.
2-4
DRAFT: Do Not Quote or Cite
-------
Table 2-1. Summary of Selected Studies Pro jecting Impacts of Climate Change-Related Ozone Health Impacts
US
Geographic
Area
Health effect(s)
C-R Function / Air
Quality Model
Climate scenario
Temperature
increase and
baseline
Population projections and other
assumptions
Main results
Reference
New York
metropolitan
region
Ozone-related
deaths by county
C-R function from
published epidemiologic
literature.
Gridded ozone
concentrations from
CMAQ.
GISS driven by
SRES A2,
downscaled using
MM5. 2050s.
1.6-3.2°C in
2050s
compared with
1990s.
Population and age structure held
constant at year 2000. Assumes no
change from US EPA 1996 national
emissions inventory and A2-
consistent increases in NOx and VOCs
by 2050s.
A2 climate only: 4.5 percent increase in
ozone-related deaths
A2 climate and precursors: 4.4 percent
increase in ozone-related deaths
Knowlton et
al. 2004
50 cities,
eastern states
Ozone-related
hospitalizations
and deaths
C-R function from
published epidemiologic
literature.
Gridded ozone
concentrations from
CMAQ.
GISS driven by
SRES A2,
downscaled using
MM5. 2050s.
1.6-3.2°C in
2050s
compared with
1990s.
Population and age structure held
constant at year 2000. Assumes no
change from US EPA 1996 national
emissions inventory and A2-
consistent increases in NOx and VOCs
by 2050s.
68 percent increase in average number
of days/summer exceeding the 8-hr
regulatory standard, resulting in 0.11-
0.27 percent increase in nonaccidental
mortality and an average 0.31 percent
increase in cardiovascular disease
mortality
Bell et al.
2007
Los Angeles
and San
Diego
regions,
California
Ozone-related
hospitalizations
and deaths
C-R function from
published epidemiologic
literature.
Gridded ozone
concentrations.
HadCM3 driven
by SRES A2,
downscaled using
MM5, then a
photochemical
model in the
2050s and 2090s.
2.1-2.7°C in
2050s, and
4.6-5.5°C in
2090s.
Population and age structure held
constant. Assumes no change from
US EPA 1997 national emissions
inventory and A2- consistent
increases in NOx and VOCs by 2050s
and 2090s.
Average increase in ozone peaks of 2.0-
3.2 ppb in the 2050s, and 3.1-4.8 ppb in
the 2090s. Increases in maximum peak
concentrations are 2- to 3-fold higher.
Percent increase in daily mortality in
the 2050s range from 0.08 to 0.46
percent depending on the exposure-
response relationship. Increases in the
2090s are 0.12-0.69 percent. Projected
increases in hospital admissions are
higher
Hwang et al.
2004
Nationwide
Ozone and PM-
related health
effects and
deaths
BenMAP for health
effects.
Gridded ozone
concentrations from
CMAQ for 2001 and 2050
GISS driven by
SRESA1B,
downscaled using
MM5. 2050s.
1.6°C in 2050s
compared with
2001.
Population and age structure held
constant. Assumes no change from
US EPA 2001 national emissions
Climate change to adversely affect air
quality in 2/3 of the US. Additional
300 ozone-mortality deaths in 2050 due
to climate change-induced ozone
increases. Impacts vary spatially.
Tagaris et al.
2009
Abbreviations: C-R, Concentration-Response; CMAQ, Community Multiscale Air Quality; GISS, Goddard Institute for Space Studies; HadCM3, a climate model from the Hadley Centre; MM5, Fifth generation
NCAR/Penn State Mesoscale Model; NOx, nitrogen oxides; SRES, Special Report on Emissions Scenarios (IPCC); VOC, volatile organic compound.
Table modified from Ebi and McGregor (2008).
2-5 DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
3. OVERVIEW OF METHODS
The basic structure of the analysis described in this report is illustrated in Figure 3-1.
Each research group exploring the potential impacts of climate change on 03
concentrations in the United States used two linked models: First, a climate change
model was used to develop scenarios of meteorological conditions within the United
States for the present day and for future change around the year 2050.
These modeled with- and without-climate change meteorological scenarios were then
input to an air quality model to simulate the ambient 03 concentrations that would result
under each scenario. Therefore, each climate change/air quality model produces a pair of
(with- and without-climate change) 03 characterizations in each cell of an air quality grid
over the United States. Although different models used different grids, for consistency
the air quality grids for all of the models were remapped to a 30 km x 30 km grid for this
analysis.
The U.S. population that will be affected by these climate change-induced changes in
ambient 03 concentrations will itself change over time, so the third component of the
analysis is a projection of the population - its size, geographic distribution, and
composition - to the target year of the analysis, 2050. Population projections were made
at the county level and interpolated to the grid cell level, using the same grid that was
used by the climate change/air quality models.2
Finally, we used BenMAP (Abt Associates Inc., 2008), EPA's premier air pollution
benefits analysis model, to estimate the changes in adverse health effects predicted to
result from the changes in ambient 03 concentrations simulated by the climate-air quality
modeling systems. BenMAP takes as input two 03 scenarios: a with-climate-change
scenario (produced by a pair of linked climate change and air quality models) and a
without-climate-change scenario (produced by the same pair of linked models, but with
the climate change model simulating present-day climate). BenMAP contains within it a
database of concentration-response (C-R) functions from the epidemiological literature.
Each 03 C-R function is an estimate of the relationship between ambient 03
concentrations and a population health effect (e.g., premature mortality or hospital
admissions for respiratory illnesses). Using this database along with the grid cell-specific
with- and without-climate-change scenarios of ambient 03 concentrations in 2050, and
grid cell-specific population projections to 2050,
2 There is something of an internal inconsistency in this. As noted above in Section 1, our analysis is based
on only the first phase of the GCRP report (U.S. EPA, 2009a) - i.e., it considers the health impacts
associated with 03 responses to global climate change alone, including direct meteorological impacts on
atmospheric chemistry and transport, and the effect of these meteorological changes on climate-sensitive
natural emissions of pollutant precursors (such as VOCs and NOx), but not changes in anthropogenic
emissions of these pollutants (e.g., due to future air quality management efforts and/or future economic
growth). Thus, we allow for an increase by the year 2050 in the population exposed to 03 in which adverse
health effects can occur, but we do not allow for a corresponding increase in anthropogenic emissions of 03
precursor emissions. That will be part of the second phase of the EPA analysis.
3-1
DRAFT: Do Not Quote or Cite
-------
Figure 3-1. The Structure of the Analysis of 03-Related Impacts on Human Health Attributable to
Climate Change
Climate Change Model
Meteorology in Meteorology in
2050 without 2050 with climate
climate change change
I I
Air Quality Model
03 concentrations in 03 concentrations in
"without climate change" "with climate change"
2050 scenario 2050 scenario
Population
Projection for 2050
Baseline Health
Effects Incidence
Rates
Suite of
Concentration-
Response
Functions for O3
BenMAP: Model of
Human Health
Impacts of
Changes in Air
Pollution
Changes in Incidence of
Adverse Health Effects in
2050 due to Climate Change-
Related Changes in 03
3-2
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
BenMAP estimates the adverse health impacts - i.e., the number of cases in the
population in 2050 of each 03-related adverse health effect - attributable to climate
change in each grid cell of the grid covering the United States. National impacts are
calculated by summing up the grid cell-specific impacts; regional impacts are similarly
calculated by summing up the grid cell-specific impacts for cells within specified regions
of the country. Three broad regions were delineated for this analysis, as shown in Figure
3-2 below: the Northeast (defined as east of 100 degrees west and north of 36.5 degrees
north latitude (the Missouri compromise)); the Southeast (defined as east of 100 degrees
west and south of 36.5 degrees north); and the West (defined as everything west of 100
degrees west longitude).
Figure 3-2. Regions of the United States Defined for this Analysis
For several of the health effects that have been associated with exposure to ambient 03,
more than one C-R function has been reported in the epidemiological literature. There is
no one "correct" set of C-R functions to use to estimate 03-related adverse health effects,
and EPA has used different sets of functions in different benefits analyses involving 03,
as both methods and available functions have evolved over time. For this analysis we
looked to the benefit analysis for the most recent O3 NAAQS Regulatory Impact Analysis
(RIA), completed in 2008 (U.S. EPA, 2008b).
03 is typically measured by air quality monitors on an hourly basis. There are various
ways in which O3 concentrations can be characterized. Among these are the 24-hour
average, the daily 1-hour maximum, and the daily 8-hour maximum, which is the basis of
the current O3 NAAQS. All of the air quality models included in this analysis used the
daily 8-hour maximum, and the metric input to BenMAP is the daily 8-hour maximum
averaged over all days in the O3 season. This results in a single value for each grid cell
for each of the two air quality scenarios (with-climate change and without-climate
change) per model.
There is substantial uncertainty surrounding each of the inputs to our analysis,
particularly because it focuses so far in the future -
3-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
• the meteorological conditions that will result from the accumulation of
greenhouse gases in the atmosphere,
• the corresponding changes in O3 concentrations,
• the size, as well as the age and geographic distribution of the population that will
be affected, and
• the relationships between adverse health effects in that population and (future) O3
concentrations.
Even the definition of "the O3 season" in 2050 is uncertain,3 and this definition will affect
the results of the analysis, since the longer the season, the greater the period of time over
which 03-related adverse health effects can occur.
There has recently been much emphasis on the uncertainty surrounding several of the
inputs to the typical air pollution benefits analysis (see, for example, National Research
Council, 2001). The uncertainty surrounding climate change is even greater.4 The
uncertainty in our analysis is thus substantial and comes from multiple sources.
Assessing and characterizing this uncertainty is therefore an important part of our
analysis. While some of this uncertainty - in particular, the statistical uncertainty
surrounding estimated coefficients in C-R functions - can easily be quantified, much of it
cannot. Each climate change model simulation is an attempt to approximate a future
complex reality, just as each air quality simulation is an attempt to approximate a future
complex reality, contingent on the future reality approximated by the linked climate
change model. Each population projection is an attempt to approximate the size,
geographic distribution, and composition of the U.S. population over forty years into the
future.
Because of the extreme complexity involved, and because we are trying to approximate a
plausible reality substantially far in the future, we cannot quantitatively assess much of
this uncertainty. Even assigning probabilities to the different models (representing our
subjective assessments about the relative accuracy with which each approximates a future
reality) is premature. Because of this, we have chosen to present our analysis as a series
of "sensitivity analyses" or "what if' scenarios designed to assess the impact of the
various assumptions and modeling approaches on the results of the analysis. The goal of
the analysis, then, is to present a range of predicted human health impact levels, and
illustrate how different (uncertain) inputs to the analysis affect the output.5
The key features of the analysis for which there is more than one plausible option are:
3 Most of the models included in the analysis identified the 03 season as June, July, and August. The
results of at least one of the models, however, suggested that a possible consequence of climate change may
be not only an increase in 03 concentrations during the 03 season, but an expansion of the 03 season as
well.
4 This is true, as time goes on, not so much about the reality of climate change, but about the specifics of it.
5 This approach - using different input models/assumptions/values and showing how the results change is
a sensitivity analysis approach, not an uncertainty analysis, because it doesn't assign probabilities to the
different input models/assumptions/values. It incorporates uncertainty into the analysis in the sense that it
illustrates the uncertainty surrounding the output value resulting from the uncertainty surrounding input
values.
3-4
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
• the climate change model (or equivalently, the description of the meteorological
conditions in the "state of the world" in 2050 with climate change as opposed to
the "state of the world" in 2050 without climate change);
• the air quality model (or equivalently, the grid cell-specific O3 concentrations in
2050 corresponding to each set of meteorological conditions);
• the size and geographic distribution of the population in 20506;
• the span of the 03 season in 2050; and
• the relationship between each adverse health effect in that future population and
O3 concentrations under each future predicted scenario (i.e., the C-R functions).7
Because the climate change and air quality models are linked, we discuss the choice of
these models in combination. The seven model combinations included in our analysis are
discussed in Section 4. The five different population projections included in the analysis
are discussed in Section 5. Because most of the models included in the analysis
identified the O3 season as June, July, and August, we used that definition as one of two
alternative choices in our analysis. Because at least one of the models suggested that
climate change may result in an expansion of the O3 season, we considered as an
alternative choice an O3 season extending from May through September.8
The C-R functions used in the analysis are discussed in Section 6.3.2. As noted above,
we used the suite of C-R functions EPA used for the benefit analysis for the recently
completed O3 NAAQS RIA. For some of the health endpoints that have been associated
with O3 there is only one C-R function available. For some, EPA pooled several C-R
functions (as described in Section 6.3.2). For others, however, EPA used two or more C-
R functions and presented results separately based on each. This illustrates not only the
uncertainty surrounding estimated coefficients in individual C-R functions resulting from
statistical error, but also the often-substantial differences between function-specific
results for the same health endpoint. In the case of premature mortality, for example,
there are four different C-R functions included in the suite of functions used to assess
impacts.
For a single (Vrelated health effect using a single C-R function, we thus produce 7
(climate change/air quality models) x 5 (population projections) x 2 (O3 season
definitions) = 70 different estimates of (Vrelated impacts due to climate change in 2050.
For a health effect such as premature mortality, for which we have four separate C-R
functions, we produce 70 x 4 = 280 different estimates. The analysis thus allows us to
examine the potentially wide range of possible estimates of (Vrelated health effects
6 Like the population size and distribution in 2050, the baseline incidence rates in 2050 for the relevant
health effects are also uncertain. However, while we projected mortality rates to 2050, we did not
incorporate this uncertainty in our analysis. This is another source of potentially important uncertainty that
could be incorporated in a subsequent extension of the analysis described here.
7 Changes in behavior by 2050 (including, for example, increased use of air conditioning) could affect these
C-R relationships. This additional source of uncertainty is not included in the current analysis.
8 This "expanded 03 season" is the standard definition of the 03 season in many EPA analyses and
epidemiological studies focusing on 03.
3-5
DRAFT: Do Not Quote or Cite
-------
1 attributable to climate change and how the different input values/models/approaches
2 affect those estimates.
3
3-6
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
4. MODELING CLIMATE CHANGE AND CORRESPONDING
CHANGES IN AMBIENT OZONE CONCENTRATIONS
Our analysis includes seven modeling efforts of six research groups (the Illinois group
carried out two sets of runs). We give a brief description of each of these modeling
efforts below, based on the descriptions given in the EPA/ORD report (U.S. EPA, 2009a)
on which the current work builds. More detailed descriptions are given in that report.
The modeling efforts included in our analysis can be divided into two major groups: (1)
those that have primarily used global climate and chemistry models to focus on the large-
scale changes in future U.S. air quality, and (2) those that have used nested, high-
resolution, global-to-regional modeling systems to focus on the regional details of the
potential future changes. The research teams at Harvard University and Carnegie Mellon
University fall into the first category. The second category includes the research teams at
EPA's National Exposure Research Laboratory (NERL); the University of Illinois;
Washington State University; and a joint effort of Georgia Institute of Technology (GIT),
the Northeast States for Coordinated Air Use Management (NESCAUM), and the
Massachusetts Institute of Technology (MIT).9
As noted in the EPA/ORD report, each approach - the global model simulations and the
downscaled regional simulations - has its strengths and weaknesses. The global models
simulate the whole world in an internally consistent way across both climate and
chemistry, but because of computational demand must use coarse spatial resolution,
thereby potentially missing or misrepresenting key processes. Dynamical downscaling
with a Regional Climate Model (RCM) dramatically increases the resolution and process
realism for the region of interest, but at the expense of introducing lateral boundary
conditions into the simulation. The advantages and trade-offs of these two categories of
model are discussed in more detail in the EPA/ORD report.
All of the results of these modeling efforts that are used in our analysis are from
simulations that held anthropogenic emissions of precursor pollutants constant at present-
day levels but allowed climate-sensitive natural emissions of biogenic VOCs to vary in
response to the simulated climate changes. As such, these model results provide scenarios
of the changes in 03 concentrations specifically due to climate change.
4.1. The Harvard University Research Effort
In early work for this project, the Harvard research group examined the role of potential
changes in atmospheric circulation by carrying out General Circulation Model (GCM)
simulations, using the Goddard Institute for Space Studies (GISS) GCM version II', for
the period 1950-2052, with tracers representing carbon monoxide (CO) and black carbon
9 Two additional research teams, discussed in the EPA/ORD report - at Columbia University and at the
University of California, Berkeley - are in this second category but are not included in our study because
they modeled 03 changes over only a portion of the United States rather than the whole country.
4-1
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
(BC) (Mickley et al., 2004). They based the concentrations of greenhouse gases for the
historical past on observations, while future greenhouse gases followed the Alb IPCC
SRES scenario. A key result from these simulations is a future 10% decrease in the
frequency of summertime mid-latitude surface cyclones moving across southeastern
Canada and a 20% decrease in cold surges from Canada into the Midwest. Since these
events typically clear air pollution in the Midwest and Northeast, pollution episodes in
these regions increase in duration (by 1-2 days) and intensity (by 5—10% in pollutant
concentration) in the future. These simulated future circulation changes are consistent
with findings from some other groups in the broader climate modeling community; the
Harvard model also successfully reproduces the observed 40% decrease in North
American cyclones from 1950-2000.
These results are supported, and expanded upon, by more recent work from this group —
e.g., see Leibensperger et al. (2008), who found that the frequency of mid-latitudes
cyclones tracking across eastern North America in the southern climatological storm
track was a strong predictor of the frequency of summertime pollution episodes in the
eastern United States for the period 1980-2006. In addition, they found a decreasing
trend over this period in the number of cyclones in this storm track that they attributed to
greenhouse warming, consistent with a number of other observational and modeling
studies.10
Subsequent to the initial modeling effort, the Harvard group applied the GEOS-Chem
Global Chemical Transport Model (GCTM), driven by the GISS III GCM (Wu et al.,
2007), to simulate 2050s 03 air quality over the United States (Wu et al., 2008a), as well
as global tropospheric O3 and policy-relevant background O3 over the United States (Wu
et al., 2008b). For one set of simulations with this modeling system designed to isolate
the impacts of climate change alone on air quality, anthropogenic emissions of precursor
pollutants were held constant at present-day levels, while climate changed in response to
greenhouse gas increases under the IPCC Alb scenario (Wu et al., 2008a). Climate-
sensitive natural emissions, e.g., of biogenic VOCs, were allowed to vary in response to
the change in climate. In these simulations, they found that at global scales, future O3
averaged throughout the depth of the troposphere increases, primarily due to increases in
lightning (leading to additional NOx production), but near the surface increases in water
vapor generally caused O3 decreases, except over polluted continental regions. Focusing
in more detail on the United States, they found that the response of O3 to climate change
varies by region. Their results show increases in mean summertime O3 concentrations of
2-5 ppb in the Northeast and Midwest, with little change in the Southeast. The Harvard
group also found that peak O3 pollution episodes are far more affected by climate change
than mean values, with effects exceeding 10 ppb in the Midwest and Northeast.
10 Other groups, however, do not necessarily find the same decrease in future mid-latitude cyclones when
analyzing similar GCM outputs, or even the same GCM outputs downscaled using an RCM (e.g., see
Leung and Gustafson, 2005).
4-2
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
4.2.
The Carnegie-Mellon University Research Effort
The Carnegie Mellon group performed global-scale simulations of atmospheric chemistry
under present and future (2050s) climate conditions using a "unified model," i.e., the
GISS II' model modified to incorporate tropospheric gas phase chemistry and aerosols.
Ten years of both present and future climate were simulated, following the A2 IPCC
greenhouse gas emissions scenario, with anthropogenic air pollution emissions held at
present-day levels to isolate the effects of climate change. As in the Harvard project, the
effects of changes in certain climate-sensitive natural emissions were also included as
part of the "climate" changes simulated.
The Carnegie Mellon group found that a majority of the atmosphere near the Earth's
surface experiences a decrease in average O3 concentrations under future climate with air
pollution emissions held constant, mainly due to the increase in humidity, which lowers
O3 lifetimes (Racherla and Adams, 2006). Further analysis of these results on a seasonal
and regional basis found that, while global near-surface 03 decreases, a more complex
response occurs in polluted regions. Specifically, summertime O3 increases over Europe
and North America, with larger increases for the latter. A second key finding is that the
frequency of extreme O3 events increases in the simulated future climate: over the eastern
half of the United States, where the largest simulated future 03 changes occurred, the
greatest increases were at the high end of the O3 distribution, and there was increased
episode frequency that was statistically significant with respect to interannual variability
(Racherla and Adams, 2008).
The general results of the Carnegie Mellon effort are broadly consistent with those of the
Harvard research effort, although there are some important differences. In contrast to the
regional pattern of future U.S. O3 change found by the Harvard University group, the
Carnegie Mellon research group found a relatively smaller response in the Northeast and
Midwest but a strong increase in the Southeast, using some similar models and
assumptions as the Harvard project (although with a different IPCC greenhouse gas
scenario and some key differences in the ocean surface boundary condition). These
differences appear to be largely due to (1) differences in how the chemical mechanisms
regulating the reactions and transformation of biogenic VOC emissions are represented in
the two modeling systems and (2) possible differences in future simulated mid-latitude
storm track changes.
4.3. The EPA NERL Research Effort
A research team at EPA's National Exposure Research Lab (NERL) built a coupled
global-to-regional climate and chemistry modeling system covering the continental
United States. They used the output from a global climate simulation with the GISS II'
model (including a tropospheric 03 chemistry model) for 1950-2055, following the Alb
IPCC SRES greenhouse gas emissions scenario for the future simulation years (i.e., the
same simulation described in Mickley et al., 2004) as climate and chemical boundary
conditions for the regional climate and air quality simulations. The Penn State/ National
4-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Center for Atmospheric Research (NCAR) Mesoscale Model Version 5 (MM5) was used
at the Department of Energy's Pacific Northwest National Laboratory (PNNL) to create
downscaled fields from this GCM simulation for the periods 1996-2005 and 2045-2055
(Leung and Gustafson, 2005). The NERL group used this regionally downscaled
meteorology to simulate air quality for 5-year-long subsets of these present and future
time periods with the CMAQ model. Multiple years were simulated to examine the role
of interannual variability in the results.
A key element of this project was extensive evaluations of the simulated meteorological
variables, not just for long-term climate statistics (e.g., monthly and seasonal means), but
for synoptic-scale patterns that can be linked more directly to air quality episodes (Cooter
et al., 2005; Gilliam et al., 2006; Gustafson and Leung, 2007). One important finding was
that the subtropical Bermuda High pressure system off the southeastern United States
coast, a critical component of eastern United States warm season weather patterns, was
not well simulated in the downscaled model runs, a result that is likely attributable to
biases in the GCM. Another key finding was that the reduction in cyclones tracking
across the northern United States found in Mickley et al. (2004) was not as clearly
present when this global model output was downscaled using MM5 (Leung and
Gustafson, 2005).
In a set of future simulations with this global-to-regional climate and air quality modeling
system, for which anthropogenic emissions of precursor pollutants were held constant
while climate changed, the NERL group found increases in future summertime maximum
daily 8-hour (MDA8) 03 concentrations of roughly 2-5 ppb in some areas (e.g.,
Northeast, Mid-Atlantic, and Gulf Coast) compared to the present-day, though with
strong regional variability and even decreases in some regions (Nolte et al., 2008). This
regional variability in future O3 concentration changes was associated primarily with
changes in temperature, the amount of solar radiation reaching the surface, and, to a
lesser extent, climate-induced changes in biogenic emissions. The increases in peak O3
concentrations tended to be greater and cover larger areas than those in mean MDA8 O3.
The NERL team also found significant O3 increases in September and October over large
portions of the country, suggesting a possible extension of the O3 season into the fall in
the future.
4.4. The University of Illinois Research Effort
The University of Illinois group focused on exploring and evaluating, as comprehensively
as possible, the capabilities and sensitivities of the tools and techniques underlying the
full, global-to-regional model-based approach to the problem. They concentrated on
building a system that accounts for global chemistry and climate, and regional
meteorology and air quality, capable of simulating effects of climate changes, emissions
changes, and long-range transport changes on regional air quality for the continental
United States (Huang et al., 2007; 2008). To capture a wider range of sensitivities, they
built different versions of this system, which combines multiple GCMs (Parallel Climate
Model (PCM) and the Hadley Centre Model, HadCM3), SRES scenarios (AlFi, A2, Bl,
4-4
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
B2), and convective parameterizations (the Grell and Kain-Fritsch schemes) with the
Model for OZone And Related chemical Tracers (MOZART) GCTM, a modified version
of the MM5 RCM (referred to as CMM5), and the SARMAP11 Air Quality Model
(SAQM). They also made considerable efforts to evaluate both climate and air quality
variables with respect to historical observations and to understand the implications of
these evaluations for simulations of future changes.
Several important findings emerge from this group's model evaluation efforts. First, they
demonstrated that any individual GCM will likely have significant biases in temperature,
precipitation, and circulation patterns, as a result of both parameterizations and internal
model variability, so multi-model ensemble means will tend to be more accurate than
individual models (Kunkel and Liang, 2005). With proper attention, RCM downscaling
can improve on these GCM biases in climate variables over different temporal scales
(e.g., diurnal, seasonal, interannual), due to higher resolution and more comprehensive
physics, and that furthermore the RCM can produce future simulations of temperature
and precipitation patterns that differ significantly from those of the driving GCM (e.g.,
Liang et al., 2006). They found that the improvements in present-day climate simulation
generally led directly to improvements in simulated air quality endpoints, though they
also found that the performance of their modeling system tended to be better for monthly
and seasonal average 03 concentrations than for multi-day high-03 episodes, reflecting
the primary use for which the driving climate models have been designed (Huang et al.,
2007). In addition, they found a high sensitivity of downscaled climate (and downscaling
skill) to the convective scheme chosen, with different parameterizations working better in
different regions/regimes (Liang et al., 2007). This sensitivity strongly affects simulated
air quality, for example by altering meteorology and hence also biogenic emissions (Tao
et al., 2008).
Notably, the Illinois team also found that the different patterns of GCM biases with
respect to present-day observations in different simulations, as well as the way the RCM
downscaling altered these biases, were consistently reflected in the future GCM and
GCM-RCM differences as well. This suggests a strong link between the ability of a
GCM or GCM-RCM downscaling system to accurately reproduce present-day climate
and the type of future climate it simulates (Liang et al., 2008).
In future simulations with their coupled global-to-regional modeling system completed to
date, based on PCM GCM simulations following both the AlFi and B1 SRES greenhouse
gas scenarios, the Illinois group found changes in O3 due to climate change alone (i.e.,
with anthropogenic pollutant emissions held constant at present-day levels) that were of
comparable magnitude to those seen by the NERL and others,11 though with differences
in regional spatial patterns (Tao et al., 2007). The larger greenhouse gas concentrations,
and hence greater simulated climate change, associated with the AlFi scenario generally
resulted in larger future O3 increases than for the climate change simulation driven by the
B1 scenario.
11 This includes a research group at Columbia University which was not included here because it focused
on only a particular region of the U.S. rather than the entire country.
4-5
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
As noted above, the University of Illinois research group produced two simulations,
denoted as Illinois 1 and Illinois 2. These simulations are identical except for the
greenhouse gas emissions scenario used in the GCM simulation of future global climate,
with Illinois 1 using the IPCC SRES AlFi and Illinois 2 using Bl.
4.5. The Washington State University Research Effort
Similar to the NERL and University of Illinois groups, the Washington State team
developed a combined global and regional climate and air quality modeling system to
investigate changes in O3 (and PM) (Chen et al., 2009; Avise et al., 2009). They used the
PCM, MM5, and CMAQ models, and they focused on the IPCC A2 scenario for future
greenhouse gases. With this system, the Washington State group investigated climate
and air quality changes for the continental United States as a whole, and in addition
focused in more detail on two specific regions: the Pacific Northwest and the northern
Midwest. A key distinguishing feature of their effort is the attention to biogenic
emissions and the consideration of land cover changes (both vegetation cover and urban
distributions), as well as changes in the frequency of wildfires in their simulations.
Evaluations of their coupled system against observations indicated reasonable agreement
with observed climatology and 03 concentrations in their two focus regions.
In five years of simulated summertime 03 under both present-day and future climate
conditions (with constant anthropogenic precursor pollutants), the Washington State
group found future 03 increases in certain regions, most notably in the Northeast and
Southwest, with smaller increases or slight decreases in other regions (Avise et al., 2009).
These climate change effects were most pronounced when considering the extreme high
end of the O3 concentration distribution. The magnitude of the O3 increases found by the
Washington State group (i.e., a few to several ppb) were roughly comparable to those
found by the other regional modeling groups already discussed, though again with
differences in the specific regional spatial patterns of the future changes, linked to
differences in the spatial patterns of key O3 drivers, discussed in more detail in the
EPA/ORD report.
In addition, by accounting for plausible future changes in land-use distribution, they
simulated both net decreases and increases in biogenic emission capacity, depending on
region. They found that reductions in forested area in the Southeast and West due to
increases in development more than offset potential increased biogenic emissions due to
climate change, leading to reduction in MDA8 03 levels, while enhanced use of poplar
plantations for carbon sequestration significantly increased isoprene emissions in the
Midwest and eastern United States, leading to 03 increases.
4-6
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
4.6.
The GIT-NESCAUM-MIT Research Effort
Similar to the NERL, Washington State, and Illinois groups discussed above, the GIT-
NESCAUM-MIT group constructed a linked global-to-regional climate and air quality
modeling system to investigate the impacts of global change on regional U.S. O3 and PM
concentrations (Tagaris et al., 2007; Liao et al., 2007). Specifically, they used CMAQ,
driven by present-day and future climate simulations with the GISS II' GCM downscaled
using MM5 (the same MM5-downscaled GISS II' GCM simulations developed for the
NERL project described above). However, compared to these other groups, they had a
unique focus on understanding the climate sensitivity of regional air quality in the context
of expected future pollutant emissions under the implementation of current and future
control strategies.
Their work to date attempts to determine if climate change will have significant impacts
on the efficacy of O3 and PM emissions control strategies currently being considered in
the United States by focusing on (1) comparing the sensitivity of future regional U.S. air
quality to changes in emissions around present-day and projected future climate and
emissions baselines and (2) accounting for the effects of uncertainties in future climate on
simulated future air quality to evaluate the robustness of these results (see Liao et al.,
2009).
To address these issues, the GIT-NESCAUM-MIT team developed a detailed, spatially
resolved U.S. future air pollutant emissions inventory to understand the relative impacts
of climate change on future air quality in different emissions and control strategy
regimes. They used the latest projection data available for the near future (to about
2020), such as the EPA CAIR Inventory, and they extended point source emissions to
2050 using the IMAGE12 model combined with the IPCC Alb emissions scenario (the
same scenario used in the GISS II' future climate simulations) and mobile source
emissions from Mobile Source Emission Factor Model version 6 (MOBILE6), projecting
reductions of more than 50% in NOx and S02 emissions (Woo et al., 2007).
A key finding from the GIT-NESCAUM-MIT work is that, overall, existing control
strategies should continue to be effective in an altered future climate, though with
regional variations in relative benefit (Tagaris et al., 2007). The magnitude of the
"climate change penalty" for controlling O3 (as defined by the Harvard group) is found to
be consistent with the work of Wu et al. (2008a). The spatial distribution and annual
variation in the contribution of precursors to O3 and PM formation under the combined
future scenario of climate change and emission controls remain similar to the baseline
case, implying the continued effectiveness of current control strategies. The findings
further suggest, however, that compliance with air quality standards in areas at or near the
NAAQS in the future would be sensitive to the amount of future climate change.
4-7
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
4.7. A Summary of Climate Change/Air Quality Research
Efforts
Table 4-1 (taken from Table 3-1 in the EPA/ORD report (U.S. EPA, 2009a)) summarizes
key features of the regional climate and O3 modeling efforts discussed above. These
simulations were carried out with linked systems consisting of a GCM/GCTM, dynamical
downscaling with an RCM, and regional-scale air quality calculations with an RAQM. In
aggregate, they cover a range of models, IPCC SRES scenarios of future greenhouse gas
emissions, climate and meteorological model physical parameterizations, and chemical
mechanisms. Figure 4-1 shows an overview of the different regional modeling results.
Table 4-1. Summary of Regional Climate and P3 Modeling Systems*
I NERL | Illinois 1 I Illinois 2
'V-' wo
Simulation
5 JJAs
4 JJAs
4 JJAs
5 Julys
3 JJAs
Period
GCM
GISS III
PCM
PCM
PCM
GISS III
Global
4° x 5°
2.8° x 2.8°
2.8° x 2.8°
2.8° x 2.8°
4° x 5°
Resolution
GHG
Alb
AlFi
B1
A2
Alb
Scenario
RCM
MM5
CMM5
CMM5
MM5
MM5
Regional
36 km
90/30 km
90/30 km
36 km
36 km
Resolution
Convection
Grell
Grell
Grell
Kain-Fritsch
Grell
Scheme
RAQM
CMAQ
AQM
AQM
CMAQ
CMAQ
Chemical
SAPRC99
RADM21
RADM2
SAPRC99
SAPRC99
Mechanism
Climate
BVOCs;
BVOCs;
BVOCs;
BVOCs;
BVOCs;
Sensitive
Emissions
Evaporative
Evaporative
Evaporative
Evaporative
Evaporative
*More details are given in the EPA/ORD report from which this table was taken (U.S. EPA, 2009a).
There are several key similarities between the results from the different groups:
• For all the present/future simulation pairs, some substantial regions of the country
show future increases in O3 concentrations of roughly 2-8 ppb under a future
climate.
• Other regions show little change in O3 concentrations, or even decreases, though
the decreases tend to be less pronounced than the increases.
• These patterns of O3 differences are accentuated in the 95th percentile MDA8 O3.
The basic result of larger climate sensitivity of O3 concentrations for high- O3
conditions (e.g., 95th percentile MDA8 O3) is one of the most robust findings of
this synthesis—it holds across all the modeling groups and appears in many
different analyses carried out by these groups.
4-8
DRAFT: Do Not Quote or Cite
-------
Figure 4-1. 2050s-Minus-Present Differences in Simulated Summer Mean MDA8 O3 Concentrations
(in ppb) for the (a) NERL; (b) GNM; (c) Illinois 1; (d) Illinois 2; and (e) WSU experiments
(a) NERL
(b) GNM
now 100W 9GW BOW
11CIW
10PW
(e) WSU
51QW 1MW 90W ROW
1 1 1 ~r
-7.5 -6 J -5.5 -4.5 -3.5 -2.5 -1.5 -0.5 0.5 1.5 2.5 3J5 4.5 55 6.5 7.5
(C) Illinois 1
(d) lllnols 2
4-9
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Some pronounced differences in the broad spatial patterns of change across these
research groups emerge as well. For example, the NERL and GNM simulations show
increases in O3 concentration in the Mid-Atlantic and parts of the Northeast, Gulf Coast,
and parts of the West. They also show decreases in the upper Midwest and Northwest and
little change elsewhere, including the Southeast. By contrast, the Illinois 1 experiment
shows the strongest increases in the Southeast, the Northwest, and the Mississippi Valley
(as well as the Gulf Coast, in agreement with NERL), with weaker increases in the upper
Midwest. In addition, these changes tend to be larger than those from the NERL
experiment. The WSU experiment shows the largest increases in the Northeast, parts of
the Midwest, and desert Southwest, with decreases in some parts of the West, the
Southeast, the Northwest, the Plains states, and the Gulf Coast. As is to be expected, the
NERL and GNM patterns are quite similar, with differences primarily reflecting the
averaging over five vs. three summers, respectively. This highlights the potential
importance of interannual variability in driving differences between modeling groups.
There are important differences in the simulated future regional climate changes across
the research groups that seem to drive the differences in the regional patterns of O3
increases (and decreases). The differences in modeling systems among the groups, as
documented in Table 4-1, provide some indication of a number of possible contributing
factors that might be responsible for these differences in simulated future regional climate
patterns, including
• Differences in the driving GCM
• Differences in the SRES greenhouse gas scenario
• Differences in the RCM (and/or model physical parameterizations) used to
simulate regional meteorology
• Differences in the RAQM (and/or chemical mechanisms)
• Differences in the amount of interannual variability captured
Table 4-2 (taken from Table 3-2 in the EPA/ORD report (U.S. EPA, 2009a)) summarizes
key features of the global models used in this analysis.
Table 4-2. Summary of Global Climate and P3 Modeling Systems Used in This Analysis*
Harvard
CM U
Simulation Period
5 summer/falls
10 summers/falls
GCM
GISS III
GISS II'
Resolution
4° x 5°
4° x 5°
GHG Scenario
Alb
A2
GCTM
GEOS-Chem
GISS II'
Climate Sensitive
Emissions
BVOCs; Lightning and soil NOx
BVOCs; Lightning and soil NOx
*More details are given in the EPA/ORD report from which this table was taken (U.S. EPA, 2009a).
In the Harvard experiment, the largest 03 increases are mostly in a sweeping pattern from
the central United States, across the Plains states and the Midwest, and extending into the
Northeast. In contrast to the regional model results discussed above, there is not as
obvious a spatial correlation between the changes in O3 and those of any one of the driver
4-10
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
variables. In the CMU experiment, a different regional pattern of change emerges. Here,
the major increases in future 03 concentrations are instead centered on the Gulf Coast and
eastern seaboard, with minimal O3 changes in the upper Midwest and northern Plains
states.
It is important to reiterate that the differences in IPCC SRES scenarios for the simulations
listed in Tables 4-1 and 4-2 refer only to greenhouse gas concentrations, and not
precursor pollutants. As emphasized above, all of the results used in this analysis are
from simulations that held anthropogenic emissions of precursor pollutants, as well as
other relevant chemical species (e.g., CH4) constant at present-day levels. Climate-
sensitive natural emissions, such as biogenic VOCs, evaporative emissions, and lightning
NOx (depending on the modeling system used), were allowed to change in response to the
simulated climate change, with the biogenic VOCs being the dominant impact. Land use
and land cover also remained constant. Finally, potential impacts of changes in 03
concentrations on plant productivity and carbon uptake were not included (e.g., see Sitch
et al., 2007).
4-11
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
5. POPULATION PROJECTIONS
The size and geographic distribution of the U.S. population in 2050 are key inputs to the
estimation of the 03-related human health impacts of climate change in the U.S. in that
year. The greater the proportion of the population living in areas of significant change in
03 concentrations, the greater the population health impacts will be. Because population
size and distribution in 2050, a year that is well in the future, depend on a number of
factors that are difficult to predict, there is substantial uncertainty about these population
inputs to our analysis. We have therefore selected five population projections for our
analysis to illustrate how the estimated 03-related human health impact of climate change
in 2050 is affected by our estimate of population size and geographic distribution in that
year.
One of our "projected" populations is just the 2000 Census population (i.e., we assumed
no change from the 2000 Census population by 2050). For another of our population
projections we extrapolated from the Woods and Poole population projections for the
year 2030 already in BenMAP. The remaining three population projections included in
our analysis come from the ICLUS project (USEPA 2009b). We describe the Woods and
Poole projection and the ICLUS population projections below.
5.1. Extrapolation of Woods and Poole Population Projections
to 2050
BenMAP uses Woods and Poole population growth projections to model populations in a
future year. Woods and Poole population growth projections incorporate the assumptions
from the U.S. Census Bureau population growth model into a comprehensive model of
economic and demographic changes over time. These projections are available at the
county level for several population sub-groups, defined by age, gender, race, and
ethnicity. BenMAP contains a series of population growth projections, based on Woods
and Poole data, for each population sub-group in each county. There are 3,109 counties
and 304 different population sub-groups per county.12
Woods and Poole population growth projections are available only through 2030,
however, whereas our analysis year is 2050. Therefore, it was necessary to extrapolate
Woods and Poole population growth projections to 2050. Given the large number of
population growth series, we used automatic forecasting algorithms that have been
implemented in the forecast package for R (Hyndman (2009), R Development Core Team
(2009)).
In order to generate our forecasts, we used a set of models that belong to the class of
exponential smoothing (ES) forecasting methods. (See Gardner (2006) and Hyndman
12 For detailed information about subgroup definitions, see the BenMAP User Manual (Abt Associates Inc.,
2008), available at: http://www.epa.gov/air/benmap/docs.html..
5-1
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
(2009) for the theoretical background of exponential smoothing models.) We evaluated
the following three ES models: simple exponential smoothing, linear exponential
smoothing, and damped-trend exponential smoothing. These models are categorized by
their trend component: none, additive, and damped, respectively. We estimated all three
models for each population growth series and then chose the best-fitting model based on
the Bayesian Information Criterion (BIC), a standard measure of goodness of fit of a
model to the underlying data. The best model was used to forecast each series out to
2050.
These ES forecasting methods try to extrapolate trends seen in a given set of years
beyond the final year of the dataset. Thus the set of years on which the extrapolation is
based could affect the resulting extrapolation. We applied the method described above to
each of the following three series of years: 2000 - 2030; 2010 - 2030; and 2020 - 2030.
We then averaged the results. This gives somewhat more weight to the latter years,
which is appropriate, since time trends may change over the longer course of years
beginning in 2000 or 2010.
The resulting 2050 population forecast was adjusted to match the Census national
population projection for 2050.13 For each of the 304 population sub-groups we
calculated the 2050 national total, as implied by the extrapolated Woods and Poole
growth projections. We then calculated percent differences between these population
totals and the population totals projected by the Census Bureau. Finally, we adjusted
each county- and population subgroup-specific extrapolated Woods and Poole projection
using corresponding percent differences. This method allowed us to match the Census
Bureau national population projection as well as preserve some of the county-specific
demographic patterns and trends.
5.2. ICLUS Population Projections to 2050
As noted above, the ICLUS project developed land-use outputs based on the social,
economic, and demographic storylines in the IPCC SRES, and adapted these to the
United States. ICLUS outputs are derived from a pair of models: a demographic model
that generates population projections and a spatial allocation model that distributes
projected population into housing units across the landscape. The models were run for
the conterminous United States and output is available for each scenario by decade to
2100. A detailed description of the methods used can be found in the second GCRP
report on which the current project builds (U.S. EPA, 2009b).
Population projections were developed for the four main SRES storylines and a base
case. The base case population projection uses the standard Census projection method;
we refer to this as the Census projection.
13 The Census national population projections for 2050 can be obtained from
http://www.census.gov/population/www/proiections/downloadablefiles.html.
5-2
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
The ICLUS project uses the SRES storylines because these storylines are direct inputs
into general circulation models developed by the climate change science community.
These storylines were selected to facilitate future, more integrated assessments of climate
and land use at national or regional scales, because the broad underlying assumptions are
the same.
The SRES describes storylines along two major axes: economic versus environmentally-
driven development (A-B) and global versus regional development (1-2); the four
quadrants defined by these axes comprise the four storylines, Al, A2, Bl, and B2. GCRP
adapted these storylines to the United States. Table 5-1 below (Table 3-1 in the GCRP
report (U.S. EPA, 2009b)) provides a qualitative description of the global storylines
modified for the United States.
Table 5-1. ICLUS Population Projection: Demographic Characteristics*
Monline
Demographic Model m
l-'erlililt
Domestic Migration
Net International
Migration
Al
Low
High
High
Bl
Low
Low
High
A2
High
High
Medium
B2
Medium
Low
Medium
Baseline ("Census")
Medium
Medium
Medium
* Source: U.S. EPA, 2009b
The SRES storylines do not provide a clear blueprint for downscaling to the local or even
the national level. In incorporating the SRES storylines into county-level projections for
the United States, an effort was made to be consistent in qualitative terms with the global
SRES storylines. Given the wide range of potential interpretations, this consistency was
understood to imply that the qualitative trends do not contradict established theory,
historical precedent, or current thinking. It was also a goal to model a wide a range of
assumptions, while remaining consistent with the SRES and U.S. demographic patterns.
For each of the storylines adapted to the United States, the fertility assumptions are
exactly consistent with the global assumptions, while domestic and international
migration patterns leave more room for interpretation and are more specifically adapted
to the United States. The low U.S. Census scenario for mortality was chosen for all
storylines used in the modeling. These model inputs were varied to develop the different
scenarios rather than to investigate the relative importance of each of the inputs.
Considering the projected trajectory of the total U.S. population under each of the five
ICLUS scenarios, scenarios Al and Bl have the same relatively low population
trajectories, while A2 has a relatively high population trajectory; scenario B2 and the
base case have the same medium population trajectory (see Figure 3-3 in U.S. EPA,
2009b).
For the current project, we selected three of the ICLUS population projections - Al, A2,
and the base case, BC (referred to as the baseline in Table 5-1) - to provide the lower and
5-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
upper bound ICLUS total population projections as well as a "middle" case. Rationales
connected to the selected SRES storylines are discussed briefly below for scenarios A1
and A2.
5.2.1. ICLUS population projection A1
A1 represents a world of fast economic development, low population growth, and high
global integration. In this storyline fertility is assumed to decline and remain low in a
manner similar to recent and current experience in many European countries (Sardon,
2004). A plausible rationale would be that the rapid economic growth in this storyline
leads to continuing high participation of women in the workforce, but it becomes
increasingly difficult to combine work with childbearing due to inflexibilities in labor
markets. At the same time, social changes in family structures lead to increasing
individuation, a rise in divorce rates, a further shift toward cohabitation rather than
marriage, later marriages and delayed childbearing, all of which contribute to low
fertility. Substantial aging resulting from the combination of low birth rates and
continued low death rates raises the demand for immigration. Meanwhile, economic
growth throughout the world and an increasingly unified global economy encourage the
free movement of people across borders. Domestic migration is anticipated to be
relatively high as well, as economic development encourages a flexible and mobile
workforce.
5.2.2. ICLUS population projection A2
The A2 storyline represents a world of continued economic development, yet with a more
regional focus and slower economic convergence between regions. Fertility is assumed to
be higher than in A1 and B1 due to slower economic growth, and with it, a slower decline
in fertility rates. International migration is assumed to be low because a regionally-
oriented world would result in more restricted movements across borders. Domestic
migration is high because, like in Al, the continued focus on economic development is
likely to encourage movement within the United States.
5-4
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
6. MODELING HUMAN HEALTH IMPACTS OF PREDICTED
CHANGES IN AMBIENT OZONE CONCENTRATIONS
The meteorology under each of two different scenarios (with and without climate change)
predicted by a climate change model to occur by a future year are input to an air quality
model, as described above. The air quality model in turn predicts the corresponding
ambient O3 concentrations, under each of the two scenarios, in each 30 km x 30 km cell
of a grid covering the contiguous United States (see Section 3). Both the size and
demography of the U.S. population by 2050 are similarly predicted within each of these
grid cells. These scenario-specific 03 concentrations and projected 2050 populations at
the grid cell level are key inputs to BenMAP, which contains within it the remaining
components of the analysis necessary to estimate the human health impacts in the U.S.
population in 2050 due to climate change under different scenarios. The flow of
modeling inputs to the analysis is illustrated in Figure 3-1 above. A brief description of
BenMAP is given below in Section 6.1.
Once the O3 concentrations in the "with climate change" and "without climate change"
scenarios have been modeled, and the grid cell-specific populations have been projected
to 2050, the estimation of the human health impacts of predicted changes in ambient O3
concentrations due to climate change follows a structure that is identical to the structure
of a typical air pollutant benefit analysis. We describe that structure below in Section
6.2. The specific methods we used to estimate the 03-related human health impacts of
climate change in 2050 are described in Section 6.3.
6.1. An Overview of BenMAP 14
BenMAP is a powerful, yet easy-to-use tool that helps analysts estimate human health
benefits resulting from changes in air quality. BenMAP was originally developed to
analyze national-scale air quality regulations, including, for example, the National
Ambient Air Quality Standards for Particulate Matter (2006) and Ozone (2008) as well as
the Locomotive Marine Engine Rule (2008).
BenMAP is primarily intended as a tool for estimating the human health effects and
economic benefits associated with changes in ambient air pollution. The improvements in
human health as a result of air pollution control regulations are typically referred to as the
benefits of the regulations. As part of the process of developing new regulations,
government agencies are typically required to assess the benefits and the costs of that
regulation. Essentially, benefit analysis develops monetary values to inform the policy
making process and allows decision makers to directly compare costs and benefits using
the same measure (i.e., dollars). BenMAP is a tool that was developed to support these
types of benefit analyses.
14 This section is adapted from Chapter 1 ("Welcome to BenMAP") of the BenMAP User Manual (Abt
Associates Inc., 2008).
6-1
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
BenMAP estimates benefits from improvements in human health, such as reductions in
premature mortality, heart attacks, chronic respiratory illnesses, and other adverse health
effects. Other benefits of reducing air pollution (i.e., visibility and ecosystem effects) are
not quantified in the current version of BenMAP. After estimating the reductions in
adverse health effects, BenMAP calculates the monetary benefits associated with those
reductions, although this final step may be omitted.
First BenMAP determines the change in the ambient air pollutant from a baseline
scenario to a control scenario within each grid cell of an air quality model grid.15 Because
BenMAP does not include an air quality model, this data must be input into BenMAP as
modeling data or generated from air pollution monitoring and/or modeling data pre-
loaded into BenMAP. BenMAP has several options for generating grid cell-specific
changes in ambient air pollutant concentrations. A more detailed description of a
commonly-used method is given below in Section 6.3.1.
Next, BenMAP applies health impact functions to the exposed population. Health impact
functions are derived from concentration-response (C-R) functions estimated in
epidemiology studies. A C-R function describes the relationship between ambient
concentrations of a pollutant and the corresponding population levels of an adverse health
effect. A health impact function describes the relationship between changes in air
pollutant concentrations and the corresponding changes in the health effect. The basic
structure of a typical air pollutant benefit analysis that BenMAP is used to carry out is
described in Section 6.2 below.
6.2. The Structure of an Air Pollutant Benefit Analysis
The analysis of the impacts of climate change on 03-related health effects is structured
like most air pollution benefits analyses carried out by EPA using BenMAP. The key
components of a BenMAP benefits analysis are:
• Ambient concentrations (at the air quality model grid cell level) of a criteria air
pollutant in a specified year under two scenarios:
o a baseline scenario, and
o a control scenario;
• Concentration-response functions relating ambient concentrations of the
pollutant to the incidences of adverse health effects in the population;
• Baseline incidence rates (numbers of cases per unit population per year) for the
adverse health effects included; and
• Population (at the air quality model grid cell level) in the specified year.
15 The baseline scenario is the scenario for which we have baseline incidence rates, usually obtained from
vital statistics sources. It is therefore the scenario that either represents current air pollutant levels or is the
closer of the two scenarios to current levels. In the typical air pollutant benefits analysis, the control
scenario is a scenario in which an air pollutant rule or regulation has been implemented in a future year.
Air pollutant levels are therefore lower than baseline levels in the typical air pollutant benefits analysis.
6-2
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
As we show below, to calculate the change in incidence of a health effect attributable to
implementation of an air pollutant rule or regulation in a typical benefits analysis we need
the baseline incidence rate for the health effect - i.e., the number of cases of the health
effect per unit population (e.g., per 100,000 population) per year. Because such
incidence rates are typically obtained from vital statistics sources or state or local health
departments, they reflect current (baseline) conditions. The baseline scenario is thus the
scenario reflecting current conditions, and the "control" scenario is the scenario reflecting
conditions when controls have been put in place to implement a proposed rule or
regulation.
The C-R functions used in an O3 benefits analysis are empirically estimated relationships,
reported by epidemiological studies, between ambient concentrations of 03 and the
incidence of specified health effects in a population. Below we describe the basic
method used to estimate the changes in the incidence of a health endpoint associated with
specified changes in O3, using a "generic" C-R function of the most common functional
form.
Although some epidemiological studies have estimated linear C-R functions and some
have estimated logistic functions, most of the studies in the air pollution epidemiological
literature have used a method referred to as "Poisson regression" to estimate exponential
(or log-linear) C-R functions in which the natural logarithm of the health endpoint is a
linear function of the air pollutant (e.g., 03):
where x is the ambient O3 level, y is the incidence of the health endpoint of interest at O3
level x, p is the coefficient of ambient 03 concentration, and B is the incidence at x=0,
i.e., when there is no ambient O3. The relationship between a specified ambient O3 level,
xo, for example, and the incidence of a given health endpoint associated with that level
(denoted as yo) is then
Because the log-linear form of C-R function (equation (1)) is by far the most common
form, we use this form to illustrate the "health impact function" - the relationship
between a change in the pollutant concentration and the corresponding change in
incidence of the health effect in the population.
If we let x0 denote the baseline O3 level, and x/ denote the control scenario O3 level, and
y0 and >'/ denote the corresponding incidences of the health effect, we can derive the
following relationship between the change in x, Ax= (x0 - X/), and the corresponding
change my, Ay, from equation (l):16
16 In a typical benefits analysis, in which the baseline represents air pollutant concentrations before
implementation of a proposed rule or regulation and the control scenario represents air pollutant
concentrations after implementation, the baseline concentration is higher than the corresponding control
y = Befk
(1)
y0=Befk° .
(2)
6-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
4y = (j0 - Ji) = Jo[l-^Ax]
(3)
Alternatively, the difference in health effects incidence can be calculated indirectly using
relative risk. Relative risk (RR) is a measure commonly used by epidemiologists to
characterize the comparative health effects associated with a particular air quality
comparison. The risk of mortality at ambient O3 level x0 relative to the risk of mortality
at ambient 03 level xh for example, may be characterized by the ratio of the two
mortality rates: the mortality rate among individuals when the ambient O3 level is x0 and
the mortality rate among (otherwise identical) individuals when the ambient 03 level is
X], This is the RR for mortality associated with the difference between the two ambient
03 levels, x0 and x}. Given a C-R function of the form shown in equation (1) and a
particular difference in ambient O3 levels, Ax, the RR associated with that difference in
ambient 03, denoted as RR Vo is equal to epAx. The difference in health effects incidence,
Ay, corresponding to a given difference in ambient O3 levels, Ax, can then be calculated
based on this RRax as
Ay = Oo -^i) = ^0[l-(l/^Ax)]. (4)
Equations (3) and (4) are simply alternative ways of expressing the relationship between
a given difference in ambient 03 levels, Ax, and the corresponding difference in health
effects incidence, Ay. These health impact equations are the key equations that combine
air quality information, C-R function information, and baseline health effects incidence
information to estimate ambient O3 health risk.17
Changes in adverse health effects are calculated in BenMAP within each grid cell of the
air quality grid by applying each health impact function, described in equations (3) and
(4), to the exposed population in the grid cell. While BenMAP applies the same
"national" health impact function to all grid cells, population estimates and baseline
incidence rates are as location-specific as possible. The grid cell-specific changes in
health effects are then summed across grid cells to produce county-level, state-level,
and/or national estimates of health impacts.
6.3. Estimation of the 03-Related Human Health Impacts of
Climate Change
This project is analogous to a typical air pollutant benefits analysis. However, instead of
asking about the human health benefits of a proposed rule or regulation that will affect
scenario concentration. This does not have to be the case, however. If the baseline concentration is lower
than the corresponding control scenario concentration, Ax will be negative.
17 Note that yO in equations (3) and (4) is the baseline incidence, not the baseline incidence rate. We
typically can obtain baseline incidence rates. To derive the baseline incidence, we multiply the incidence
rate by the appropriate population.
6-4
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
ambient O3 concentrations, we are asking about the human health effects of climate
change that will affect ambient 03 concentrations. The pollutant of interest is 03 and the
specified year is 2050. Each linked pair of climate change and air quality models
produces two 03 scenarios, with 03 concentrations at the air quality model grid cell level:
a "with climate change" scenario and a "without climate change" scenario, both in the
future year 2050.
Because both the with- and without-climate-change scenarios are in 2050, neither
scenario is really a "baseline" scenario - that is, we cannot obtain baseline incidence rates
from vital statistics sources for either scenario. Instead, we projected current baseline
incidence rates to the year 2050, as described in Section 6.3.3 below. The names
"baseline" and "control" scenario, used in a typical air pollution benefits analysis, thus
don't really fit here. What matters, however, is that we are comparing two different
scenarios. We refer to these as the "without climate change" scenario and the "with
climate change" scenario. Both scenarios are hypothesized "states of the world" in 2050.
The analysis allows population to change from the present to 2050 in both scenarios. In
the "with climate change" scenario, it also allows climate change-related meteorology, as
well as the corresponding concentrations of ambient O3 to change from the present to
2050, but keeps everything else (e.g., economic activity and anthropogenic emissions)
constant. Thus any change in O3 concentrations, and corresponding changes in human
health effects, between the with- and without-climate change scenarios can be attributed
to climate change.
6.3.1. Ambient O3 concentrations: Adjustment of modeled with- and
without-climate-change O3 concentrations in 2050
Each climate change/air quality model combination described above in Section 4
produced a pair of modeled summer average of daily 8-hour maximum O3 concentrations
in each 30 km by 30 km grid cell: the estimated "with climate change" and "without
climate change" concentrations. Air pollution benefits analysts generally acknowledge,
however, that they have more confidence in monitored air pollutant concentrations than
modeled concentrations, since monitor values are actual measurements. However, unlike
modeled values, monitors do not exist in all grid cells of an air quality model grid. In a
typical BenMAP analysis, then, both modeled and monitor values are used to produce
grid cell-specific air pollutant estimates in the baseline and control scenarios that are
considered superior to either the monitor values or the modeled values alone.
There are several options in BenMAP for estimating baseline and control scenario grid
cell-specific air pollutant concentrations using monitor and modeled values. The option
that EPA typically uses for a future-year analysis first applies a spatial interpolation of
monitor values to grid cell centers and then applies a temporal adjustment using the ratios
of modeled values. These spatial and temporal adjustment procedures are described in
6-5
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
detail in Appendix C ("Air Pollution Exposure Estimation Algorithms") of the BenMAP
User Manual (Abt Associates Inc., 2008). We describe them briefly here.18
Although there are several ways to spatially interpolate monitor values to grid cell
centers, EPA benefits analyses typically use a method called Voronoi Neighbor
Averaging (VNA). The VNA algorithm interpolates air quality at every grid cell by first
identifying the set of monitors that best "surround" the center of the grid-cell. In
particular, BenMAP identifies the nearest monitors, or "neighbors," by drawing a
polygon, or "Voronoi" cell, around the center of each BenMAP grid cell. The polygons
have the special property that the boundaries are the same distance from the two closest
points. An example of a grid cell and "neighboring" monitors is shown in Figure 6-1; the
corresponding Voronoi cells are shown in Figure 6-2.
Figure 6-1. Example of Grid Cells and Monitors
¦k
*
*
#
¦k
•k
k
*
# = Center Grid-Cell "E"
"k
= Air Pollution Monitor
18 The brief description given here is a condensed version of the more detailed description given in Section
C.3 of Appendix C of the BenMAP User Manual.
DRAFT: Do Not Quote or Cite
6-6
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Figure 6-2. Illustration of Voronoi Cells Around the Same Grid Cell Center
# = Center Grid-Cell "E'
= Air Pollution Monitor
To estimate the air pollutant level in each grid cell, BenMAP calculates the metric (e.g.,
the summer average of daily 8-hour maxima) for each of the neighboring monitors, and
then calculates an inverse-distance weighted average of the metrics. The further the
monitor is from the BenMAP grid-cell, the smaller the weight.
However, monitors are more likely to be located in areas with higher population than in
rural areas. If a grid cell is in a rural area, a weighted average of "neighboring" monitor
values may still not give a good approximation to the pollutant concentration in the grid
cell. Because of this, EPA typically also uses a spatial scaling technique in which, for
each of the neighboring monitors, BenMAP multiplies the monitoring data by the ratio of
the baseline modeling data for the destination grid cell to the baseline modeling data for
the grid cell containing the monitor. For example, suppose the destination grid cell
(without a monitor) is in a rural area, and the modeled baseline pollutant value is half the
modeled baseline pollutant value in the grid cell containing a neighbor monitor. That
monitor value would be multiplied by one half (and similarly the other neighboring
monitors would be multiplied by the appropriate ratios) before the inverse-distance
weighted average of monitor-specific metrics is calculated.
This first step of spatial interpolation of monitor values and spatial scaling using "without
climate change" modeling values produced "without climate change" scenario estimates
of summer average daily 8-hour maxima for each grid cell. Year 2007 monitor values
were used.
Grid cell-specific "with climate change" scenario O3 concentrations were estimated by
combining both spatial and temporal scaling. After the first step of spatial scaling
described above, BenMAP applied the ratio of the modeled future-year value to the
6-7
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
modeled baseline value in the destination cell.19 The future-year ("with climate change")
03 concentration in a grid cell is thus estimated by (1) spatially interpolating a present-
year value, using both interpolation of (year 2007) monitor values and spatial scaling
using present-year modeled values, and then (2) temporally scaling the resulting value by
the ratio of future-year to present-year modeled values. This produces future-year ("with
climate change") estimates that take advantage of both the ratios of future to recent-year
modeled values and information we have from actual recent-year monitor measurements.
6.3.2. Concentration-response functions
There are often several epidemiological studies reporting multiple concentration-response
(C-R) functions for the same pollutant/health endpoint combination, and substantial
thought goes into the selection of appropriate health endpoints, studies, and C-R
functions. For this project, we followed the selection of health endpoints, studies, and C-
R functions EPA used in the benefits analysis for the O3 National Ambient Air Quality
Standards (NAAQS) Regulatory Impact Analysis (RIA) completed in 2008.20 The 03-
related health endpoints and studies EPA used in the O3 NAAQS RIA are listed in Table
6-2 of the RIA.21 A more detailed summary of health endpoints, epidemiological studies,
and C-R functions used - including the estimated coefficient ("beta") of O3 in the
function and the standard error of the estimate, the location(s) and age range covered, and
the O3 metric used - is given in Table 6-1 below. As can be seen in Table 6-1, we
included the following adverse health effects in our analysis: mortality from all causes;
non-accidental mortality;22 hospital admissions for respiratory illnesses; hospital
admissions for chronic obstructive pulmonary disease (COPD), with and without asthma;
hospital admissions for pneumonia; emergency room (ER) visits for asthma; school loss
days from all causes; and minor restricted activity days.
In all cases, the metric used was the daily 8-hour maximum.23 In several cases, however,
the original C-R function used a different metric (e.g., the 24-hour mean), and these
coefficients were converted to coefficients for the daily 8-hour maximum.24
For several health endpoints, two or more C-R functions were pooled. In particular, for
respiratory hospital admissions we undertook the following pooling procedure:
1. Moolgavkar et al. (1997) estimated C-R functions in Minneapolis for hospital
admissions (HA), pneumonia (ICD-9 codes 480-487) and HA, COPD (ICD 490-
496). We summed the results from these two non-overlapping subcategories.
19 BenMAP actually combined the two steps into one by multiplying the monitor value by the ratio of the
future-year value in the destination cell to the baseline value in the cell containing the monitor.
20 The 03 NAAQS RIA is available online at: http://www.epa.g0v/ttn/naaas/standards/0z0ne/s o3 cr.html.
21 Available online at: http://www.epa.gov/ttn/ecas/regdata/RIAs/6a-ozoneriachapter6appendixa.pdf.
22 This typically excludes accidents, homicides, and suicides.
23 The measure of 03 concentration input to BenMAP from the climate change/air quality models is the 03
season average of the daily 8-hour maxima. The C-R functions are daily functions, so this 03 season
average of daily 8-hour maxima would be applied to each day.
24 The process of converting C-R function coefficients is described in Appendix G ("Ozone Health Impact
Functions in U.S. Setup") of the BenMAP User Manual. See, in particular, Section G.5 ("Converting
Functions to 8-Hour Daily Maximum Metric").
6-8
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
Table 6-1. Summary of Concentration-Response Functions Used to Estimate Climate Change-Related Impacts of 03 on Human Health
Health Endpoint
Study
Location
Age
Range
Metric
Beta
Std. Err.
Notes
Mortality, All Cause
Bell et al. (2005)
US & non-US cities
All ages
Daily 8-hour max.1
0.000795
0.000212
Warm season
Mortality, All Cause
Levy et al. (2005)
US & non-US cities
All ages
Daily 8-hour max.2
0.001119
0.000179
Warm season
Mortality, Non-Accidental
Bell et al. (2004)
95 US cities
All ages
Daily 8-hour max.1
0.000261
0.000089
Warm season
Mortality, Non-Accidental
Ito et al. (2005)
Meta-analysis 7
All ages
Daily 8-hour max.1
0.001173
0.000239
Warm season
Meta-analysis
All ages
Daily 8-hour max.2
0.000532
0.000088
Hospital admission (HA), All
Respiratory
Burnett et al. (2001)
Toronto, CAN
0-1
Daily 8-hour max.2
0.008177
0.002377
Warm season
HA,COPD 4
Moolgavkar et al. (1997)
Minneapolis, MN
65+
Daily 8-hour max.1
0.00196
0.001238
All year
HA, Pneumonia4
Moolgavkar et al. (1997)
Minneapolis, MN
65+
Daily 8-hour max.1
0.00266
0.000762
All year
HA, Pneumonia4
Schwartz (1994a)
Minneapolis, MN
65+
Daily 8-hour max.1
0.002784
0.001305
All year
HA , COPD (less asthma)4
Schwartz (1994b)
Detroit, MI
65+
Daily 8-hour max.1
0.003424
0.001293
All year
HA, Pneumonia4
Schwartz (1994b)
Detroit, MI
65+
Daily 8-hour max.1
0.003230
0.000806
All year
HA , All respiratory 4
Schwartz (1995)
New Haven, CT
65+
Daily 8-hour max.1
0.001777
0.000936
Warm season
HA , All Respiratory 4
Schwartz (1995)
Tacoma, WA
65+
Daily 8-hour max.1
0.004931
0.001770
Warm season
ER, Asthma5
Peel et al. (2005)
Atlanta, GA
All ages
Daily 8-hour max.
0.000870
0.000529
ER, Asthma5
Wilson etal. (2005)
Portland, ME
All ages
Daily 8-hour max.
0.003000
0.001000
ER, Asthma5
Wilson etal. (2005)
Manchester, NH
All ages
Daily 8-hour max.
-0.001000
0.002000
School Loss Days, All Cause6
Chen et al. (2000)
Wachoe Co, NV
5-17
Daily 8-hour max.2
0.015763
0.004985
All year
School Loss Days, All Cause6
Gilliland et al. (2001)
Southern California
5-17
Daily 8-hour max.3
0.007824
0.004445
All year
Minor Restricted Activity
Days
Ostro and Rothschild
(1989)
Nationwide
18-64
Daily 8-hour max.2
0.002596
0.000776
1 Converted from 24-hour mean.
2 Converted from daily 1-hour maximum
3 Converted from 8-hour mean
4 These studies were pooled in BenMAP to generate pooled incidence estimates for respiratory hospital admissions.
5 These studies were pooled in BenMAP to generate pooled incidence estimates for asthma-related ER visits. Note: Jaffe et al. (2003) is listed in Table 6-2 of
EPA's 03 NAAQS RIA as being among those studies included in the pooled analysis for asthma-related ER visits. However, we were informed via personal
communication with Neal Fann (EPA/OAQPS) that this study was ultimately not included because it covered a substantially different age range (ages 5 - 34)
from the other studies.
6 These studies were pooled in BenMAP to generate pooled incidence estimates for school loss days.
7 This was a meta-analysis of 43 U.S. and non-U.S. studies.
6-9
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
2. Schwartz (1994a) also estimated C-R functions in Minneapolis for the same two
subcategories. However, this study found a significant effect only for HA,
pneumonia. So the estimate of "PM-related HA for respiratory illness" in
Minneapolis based on Schwartz (1994a) was taken to be just PM-related HA,
pneumonia.
3. The estimates of "PM-related HA for respiratory illness" in Minneapolis from (1)
and (2) above were pooled using a fixed effects pooling method.25
4. Schwartz (1994b) estimated C-R functions for the same two non-overlapping
subcategories in Detroit. We similarly summed these results.
5. Finally, Schwartz (1995) estimated C-R functions for "HA, all respiratory" in
New Haven, CT and Tacoma, WA. We pooled the HA, All respiratory results
from these C-R functions with the results from steps (3) and (4).26
To obtain the asthma ER visits results, we pooled Peel et al. (2005) and Wilson et al.
(2005) using the random/fixed effects method. To obtain the results for school absence
days, we pooled Gilliland et al. (2001) and Chen et al. (2000) also using the random/fixed
effects method.
6.3.3. Baseline incidence rates
This section describes the development of baseline incidence rates for mortality and
morbidity health endpoints examined in our analyses. First, we describe the source of
2004-2006 individual-level mortality data and the calculation of county-level mortality
rates. Second, we describe how we use national-level Census mortality rate projections to
develop 2050 county-level mortality rate projections, which are used as baseline
mortality incidence rates. We then describe the baseline morbidity incidence rates,
including hospitalization rates and emergency room (ER) visit rates.
Mortality
We obtained individual-level mortality data, including residence county FIPS, age at
death, month of death, and underlying causes (ICD-10 codes), for years 2004-2006 for
the entire United States from the Centers for Disease Control (CDC), National Center for
Health Statistics (NCHS). The detailed mortality data allowed us to generate cause-
specific death counts at the county level for selected age groups. The county-level death
counts are then divided by the corresponding county-level population to obtain the
mortality rates. To provide more stable estimates, we used three years (2004-2006) of
mortality and population data,27 i.e.,
25 When choosing fixed effects as the pooling method, pooling weights are generated automatically based
on the inverse variance of each input result, with the weights normalized to sum to one. Results with a
larger absolute variance get smaller weights. (For more details, see Section J.2.1.3 in the BenMAP User
Manual, Appendix J).
26 For more details on the pooling method, see Section J.2.1.4 in BenMAP User Manual Appendix J.
27 The population data for 2004-2006 were Woods and Poole estimates based on the 2000 Census.
6-10
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
2006
Y,deathvk
Mortality Rate{2004 - 2006)(//. = 20062004 ,
populationijk
2004
where / represents the specific cause of mortality (e.g., non-accidental mortality), j
represents a specific county, and k represents a specific age group.
Mortality rates based on 20 or fewer deaths were considered unreliable.28 If the rate for a
given cause of death was unreliable in certain counties in a state, we summed up the
deaths attributed to that cause in those counties, as well as the populations in those
counties and created an aggregate rate for that cause of death in those counties. If that
aggregate "state-level" rate was unreliable, we aggregated to the region level,29 and if the
region-level rate was still unreliable, we aggregated to the national level.30
To estimate age- and county-specific mortality rates in the year 2050, we calculated
adjustment factors, based on a series of Census Bureau projected national mortality rates,
to adjust the above age- and county-specific mortality rates in 2004-2006 to
corresponding rates for 2050. The procedure we used was as follows:
• For each age group, we calculated the ratio of the Census Bureau national
mortality rate projection in year 2050 to the national mortality rate in 2005. Note
that the Census Bureau projected mortality rates were derived from crude death
rates.31
• To estimate mortality rates in 2050 that are both age-group-specific and county-
specific, we multiplied the county- and age-group-specific mortality rates for
2004-2006 by the appropriate ratio calculated in the previous step. For example,
to estimate the projected mortality rate in 2050 among ages 18-24 in Wayne
County, MI, we multiplied the mortality rate for ages 18-24 in Wayne County in
2004-2006 by the ratio of Census Bureau projected national mortality rate in 2050
for ages 18-24 to Census Bureau national mortality rate in 2005 for ages 18-24.
28 Refer to http://www.health.state.nv.us/diseases/chronic/ratesmall.htm for an explanation of why rates
based on fewer than 20 cases are marked as unreliable.
29 We used the four regions defined by the U.S. Bureau of the Census. See the definitions on the next page.
30 At each level of aggregation, only those counties with unreliable rates for the specified cause of death
were included. So, for example, if 5 counties in a given state had unreliable rates for a specific cause of
death, a "state-level" rate was created by summing the deaths from that cause across those counties and
dividing by the sum of the populations in those counties. If this "state-level" rate was still unreliable, we
repeated the process at the region level.
31 The following formula, given by Chiang (1967, p.2 equation 7), was used: M = Q/(1-(1-A)*Q), where M
denotes the projected mortality rate, Q denotes the crude death rate, and A denotes the fraction of the
interval (one year) lived by individuals who die in the interval. A=0.1 if age < 1, and A=0.5 otherwise.
6-11
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Hospitalizations
Regional hospitalization counts were obtained from the National Center for Health
Statistics' (NCHS) National Hospital Discharge Survey (NHDS). NHDS is a sample-
based survey of non-Federal, short-stay hospitals (<30 days), and is the principal source
of nationwide hospitalization data.32 The survey collects data on patient characteristics,
diagnoses, and medical procedures. Public use data files for the year 1999 survey were
downloaded (from ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHDS/) and
processed to estimate hospitalization counts by region. NCHS groups states into four
regions using the following groupings defined by the U.S. Bureau of the Census:
• Northeast - Maine, New Hampshire, Vermont, Massachusetts, Rhode Island,
Connecticut, New York, New Jersey, Pennsylvania
• Midwest - Ohio, Indiana, Illinois, Michigan, Wisconsin, Minnesota, Iowa,
Missouri, North Dakota, South Dakota, Nebraska, Kansas
• South - Delaware, Maryland, District of Columbia, Virginia, West Virginia,
North Carolina, South Carolina, Georgia, Florida, Kentucky, Tennessee,
Alabama, Mississippi, Arkansas, Louisiana, Oklahoma, Texas
• West - Montana, Idaho, Wyoming, Colorado, New Mexico, Arizona, Utah,
Nevada, Washington, Oregon, California, Alaska, Hawaii
We used the 2000 Census to obtain more age specificity, and then corrected the 2000
Census figures so that the total population equaled the total for 1999 forecasted by
NHDS. In particular, for each type of hospital admission (ICD code or codes) we: (1)
calculated the count of hospital admissions by region in 1999 for the age groups of
interest, (2) calculated the 2000 regional populations corresponding to these age groups,
(3) calculated regional correction factors that equal the regional total population in 1999
divided by the regional total population in 2000, (4) multiplied the 2000 population
estimates by these correction factors, (5) divided the 1999 regional count of hospital
admissions by the estimated 1999 population, and (6) applied the regional rates to every
county in that region.
Similar to mortality rates, the hospitalization rates are also cause-specific and the hospital
admissions endpoints are defined by different combinations of ICD codes that are used in
the selected epidemiological studies.
Emergency Room Visits for Asthma
Regional asthma emergency room visit counts were obtained from the National Hospital
Ambulatory Medical Care Survey (NHAMCS). NHAMCS is a sample-based survey,
conducted by NCHS. The target universe of the NHAMCS is in-person visits made in the
United States to emergency and outpatient departments of non-Federal, short-stay
hospitals (hospitals with an average stay of less than 30 days) or those whose specialty is
32 Note that the following hospital types are excluded from the survey: hospitals with an average patient
length of stay of greater than 30 days, federal, military, Department of Veterans Affairs hospitals,
institutional hospitals (e.g. prisons), and hospitals with fewer than six beds.
6-12
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
general (medical or surgical) or children's general. Public use data files for the year 2000
survey were downloaded (from:
ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHAMCS/) and processed to
estimate hospitalization counts by region. We obtained population estimates from the
2000 U.S. Census. The NCHS regional groupings described above were used to estimate
regional emergency room visit rates.
6.3.4. Populations
The extent of the impacts of climate change-related changes in ambient 03 concentrations
will depend in part on the extent to which areas of large O3 changes coincide with areas
of high population density. Because the target year for this analysis is 2050, we must rely
on population projections. We have used four different population projections in our
analysis; these are described in detail in Section 5 above. Regardless of the population
projection being used, BenMAP derives grid cell-specific population estimates.
6.3.5. Summary of key features of the analysis
In summary, for a given climate change/air quality model combination and a given
population projection, we used BenMAP to estimate the changes in incidence of O3-
related health effects that are estimated to occur in a future year as a result of climate
change-induced changes in O3 concentrations. The basic features of the analysis are as
follows:
• The target year for the analysis: 2050;
• The O3 metric: the average of daily 8-hour maxima over the O3 season;
• Air quality model grid (for all air quality models used): 30 km x 30 km grid cell
of an air quality grid over the coterminous U.S.
• Adjustment of modeled without-climate-change and with-climate-change scenario
O3 metrics:
o Without climate change scenario: spatial adjustment in BenMAP, using
both monitor and modeled values (described in Section 6.3.1);
o With-climate-change scenario: spatial and temporal adjustment in
BenMAP, using both monitor and modeled values (described in Section
6.3.1);
• Selection of health endpoints, epidemiological studies, and C-R functions: chosen
to match the suite used in EPA's recently completed benefits analysis for the O3
NAAQS RIA.
• Results calculated at the grid cell level and then aggregated to
o The regional level33; and
o The national level.
33 The country was divided into three broad regions: The Northeast, the Southeast, and the West. The
definitions of these regions are given in Section 3.
6-13
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
6.3.6. Assessing and characterizing uncertainty
As noted above in Section 3, there is substantial uncertainty surrounding each of the
inputs to our analysis. While some of this uncertainty - in particular, the statistical
uncertainty surrounding estimated coefficients in C-R functions - can easily be
quantified, much of it cannot. Each climate change model is an attempt to approximate a
future reality, just as each air quality model is an attempt to approximate a future reality,
contingent on the future reality approximated by the linked climate change model. Each
population projection is an attempt to approximate the size, geographic distribution, and
composition of the U.S. population over forty years into the future.
We do not have ways to quantitatively assess much of this uncertainty. Even assigning
probabilities to the different models (representing our subjective assessments about the
relative accuracy with which each approximates a future reality) is premature. Because
of this, we have chosen to present our analysis as a series of "sensitivity analyses" or
"what if' scenarios designed to assess the impact of the various assumptions and
modeling approaches on the results of the analysis. The goal of the analysis, then, is to
present a range of predicted (Vrelated human health impact levels, and illustrate how
different (uncertain) inputs to the analysis affect the output.
We carried out the analysis using all combinations of the seven climate change/air quality
models described in Section 4, the five different population projections described in
Section 5, and the two definitions of "O3 season" - June, July, and August, used in most
of the climate change models, as well as an expanded O3 season from May through
September. We also used more than one C-R function for a given 03-related health
endpoint if more than one was used in EPA's O3 NAAQS RIA benefits analysis. No one
set of input characterizations was given any more weight than any other set when we
interpreted the results.
The entire set of analyses, using all different combinations of input characterizations, thus
creates a potentially wide range of results that serves to illustrate
• the breadth of uncertainty surrounding estimates of 03-related health impacts that
may be attributable to climate change in a future year (2050), and
• which uncertain inputs "matter most."
An uncertain input to an analysis can be important in different ways:
• It can be important because the value of the outcome of the analysis is sensitive to
the value of the (uncertain) input - i.e., a relatively small change in the input
value results in a relatively large change in the outcome of the analysis.
• It can be important because it contributes a relatively large share of the
uncertainty about the outcome of the analysis, so that if we could reduce the
uncertainty about the input we would disproportionately reduce the uncertainty
about the outcome.
6-14
DRAFT: Do Not Quote or Cite
-------
1 • It can be important because it has the potential to affect the decision that a
2 decision-maker would make based on the analysis.34
3
4
5
6
7
34 These types of uncertainty importance are not mutually exclusive. An uncertain input can be important
in all three ways, or in one or two ways. It is possible, for example, that several uncertain inputs could be
important in the first two ways but not in the third way, if the decision-maker would make the same
decision regardless of the values of the uncertain inputs used in the analysis - e.g., that the outcome of the
analysis depends on the values of the uncertain inputs but that, given any of the possible values, the
decision-maker would make the same decision.
6-15
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
7. RESULTS AND DISCUSSION
We produced 7 (climate change/air quality models) x 5 (population projections) x 2 (O3
season definitions) = 70 potential "answers" to the question: How many 03-related cases
of a given health effect (e.g., premature mortality) may be attributable to climate change in
the conterminous United States in the year 2050? For some health effects for which we
have more than one C-R function, we have produced some multiple of 70 results. In the
case of all-cause mortality, for example, for which we have two different C-R functions,
we produced twice this number, or 140 potential "answers." In addition, we've considered
several different health endpoints - all the health endpoints included in the 2008 03
NAAQS RIA benefits analysis. This includes all-cause mortality, non-accidental
mortality, hospital admissions for respiratory illnesses, emergency room visits for asthma,
school loss days, and minor restricted activity days. Results for all combinations of these
health endpoints, C-R functions, climate change/air quality models, population projections,
and O3 season definitions were aggregated to the national level as well as to each of the
three regions considered in this analysis. We present all of these results in tables in
Appendix A. Below we summarize and discuss the most salient features of these results.
7.1. National Results
There is a wide range of possible "answers" to the question posed above - i.e., there is, not
surprisingly, a large amount of uncertainty about the impact of climate change on future
(2050) 03-related human health effects. This is evident both for 03-related mortality,
discussed in Section 7.1.1, and morbidity, discussed in Section 7.1.2.
7.1.1. Mortality
Figure 7-1 shows the impact of climate change/air quality model, population projection,
and C-R function on estimates of the national incidence of 03-related all-cause premature
mortality attributable to climate change, when the 03 season is defined as June, July, and
August (as it was in most of the climate change/air quality modeling efforts). The top
panel of Figure 7-1 shows estimates based on Bell et al. (2005); the bottom panel shows
estimates based on Levy et al. (2005). The indicators for the climate change models and
population projections on the x-axis of Figure 7-1, as well as the abbreviations for the
models and population projections used in subsequent figures and tables, are given in
Figure 7-2.
Looking across all combinations of climate change/air quality models, population
projections and C-R functions for all-cause mortality considered in our analysis, based on
the 03 season defined as June, July, and August (shown in Figure 7-1), estimates of
national 03-related all-cause premature mortality in 2050 attributable to climate change
range from -657 to 2,550 - that is, from over 600 cases of 03-related premature mortality
avoided because of climate change to over 2,500 cases attributable to climate change.
7-1
DRAFT: Do Not Quote or Cite
-------
1 Figure 7-1. Estimated National 03-Related Cases of All-Cause Mortality in 2050 03 Season (Defined
2 as June, July, and August) Due to Climate Change: Impact of Climate Change/Air Quality
3 Model, Population Projection, and C-R Function
Bell et al. (2005)
3000
2500
2000
1500
1000
500
0
-500
-1000
G1 G2 G3 G4 F4 F5 B5 F3 G5 F2 B4 F1 B3 B2 B1 C5 A5 E5 D5 C4 C2 C3 C1 A4 E4 A3 A2 D4 E3 A1 E2 D3 D2 E1 D1
Climate Change Model/Population Projection
Levy et al. (2005)
3000
2500
2000
1500
1000
500
0
-500
-1000
7-2 DRAFT: Do Not Quote or Cite
U ' u
n n
_ - n n n n
llBB
G1 G2 G3 G4 F4 F5 B5 F3 G5 F2 B4 F1 B3 B2 B1 C5 A5 E5 D5 C4 C2 C3 C1 A4 E4 A3 A2 D4 E3 A1 E2 D3 D2 E1 D1
Climate Change Model/Population Projection
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Figure 7-2. Indicators and Abbreviations for Climate Change/Air Quality Models and Population
Projections in Figures
Indicator in
Figures
Climate Change/Air Quality Model
Abbreviation Used in Figures
A
Carnegie-Mellon University
CMU
B
GIT-NESCAUM-MIT
GNM
C
Harvard University
Harvard
D
University of Illinois - using A1 Fi GHG scenario
lllinois-1
E
University of Illinois - using B1 GHG scenario
lllinois-2
F
EPA's National Exposure Research Lab
NERL
G
Washington State University
WSU
Indicator in
Figures
Population Projection
Abbreviation Used in Figures
1
Integrated Climate and Land-Use Scenarios project - A1 scenario
ICLUS A1
2
Integrated Climate and Land-Use Scenarios project - A2 scenario
ICLUS A2
3
Integrated Climate and Land-Use Scenarios project - base case
Exponential smoothing projections of Woods & Poole 2030
ICLUS_BC
4
populations in BenMAP To 2050
Woods & Poole
5
Year 2000 Census Population
Census 2000
If we use the expanded definition of the 03 season, from May through September, the
range of results expands accordingly - from -1,092 to 4,241. Moreover, this range does
not reflect the full extent of uncertainty, because it does not incorporate the uncertainty
surrounding each individual input to the analysis.35 However, while the wide ranges of
estimates for both definitions of the 03 season include some that are negative, the
preponderance of estimates are positive, suggesting that, all else being equal, we would
expect climate change to increase the incidence of 03-related all-cause premature mortality
in 2050.
The results for non-accidental mortality follow a similar pattern to what is shown in Figure
7-1 for all-cause mortality. Because the coefficient of 03 in the C-R function reported in
Bell et al. (2004) is smaller than the coefficients in Bell et al. (2005) and Levy et al.
(2005), however, the magnitudes of the estimates based on Bell et al. (2004) are
substantially smaller. Looking across all combinations of climate change/air quality
models and population projections, for an 03 season defined as June, July, and August,
estimates of 03-related non-accidental premature mortality due to climate change based on
Bell et al. (2004) range from -147 to 570. Once again, the great preponderance of the
estimates is positive. This is broadly consistent with what other researchers have reported
on the 03-related human health impacts of climate change (see, e.g., Knowlton et al., 2004;
Bell 2007; Hwang et al., 2004; Tagaris et al.; 2009).
35 For example, the figures shown in this section are based on the point estimates of the C-R functions from
Bell et al. (2005) and from Levy et al. (2005), but there is statistical uncertainty surrounding each of these
point estimates. Another set of inputs to the analysis for which we did not incorporate uncertainty are the
baseline incidence rates. While current rates are relatively uncertain, we used rates projected to the year
2050. Like the population projections, these projected baseline incidence rates similarly have substantial
uncertainty surrounding them.
7-3
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
7.1.2. Morbidity
Matrices of point estimates of 03-related morbidity in 2050 attributable to climate change,
for the different combinations of climate change/air quality model and population
projection are shown for each of the morbidity endpoints in Table 7-1. For more complete
estimates, including 95 percent confidence or credible intervals, see Appendix A.
While the general magnitudes of the estimates of 03-related morbidity in 2050 attributable
to climate change will differ from those for mortality - and will vary from one morbidity
endpoint to another - the broad pattern of results seen for mortality across the different
climate change/air quality models, for each population projection, is largely mirrored for
the morbidity endpoints we included in our analysis. In particular, the order of the models
in terms of predicted results for mortality (for any given population projection) - Illinois-1,
Illinois-2, CMU, Harvard, GNM, NERL, and WSU - is followed for the morbidity
endpoints as well.
However, because several of the morbidity endpoints focus on specific age subgroups of
the population, and the population projections differ to some extent in their predicted age
distributions, there are some notable differences in patterns across the population
projections, given a climate change/air quality model - both between mortality and some
of the morbidity endpoints, and between different morbidity endpoints. The influence of
age distribution in the projected population is discussed in more detail in Section 7.2.2
below.
Like the results for 03-related mortality, the preponderance of results for 03-related
morbidity is positive - i.e., overall, the models predict that climate change will increase the
incidence of 03-related morbidity. This is broadly consistent with the few morbidity results
reported by other researchers (see, e.g., Hwang et al., 2004; Tagaris et al.; 2009).
7-4
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
Table 7-1. Estimated National 03-Related Incidence of Morbidity Attributable to Climate Change in
2050 During the 03 Season, Taken to be June, July, and August, Based on Different
Combinations of Climate Change/Air Quality Model and Population Projection*
Population Projection
Climate Change/Air
Quality Model
ICLUS_A1
ICLUS_A2
ICLUS_BC
Woods &
Poole
Census_2000
Hospital Admissions for Respiratory Illness (Ages <1)
lllinois-1
1570
2650
1990
2350
1600
lllinois-2
1610
2740
2060
2350
1610
CMU
1250
2060
1550
1830
1290
Harvard
710
1230
940
1100
820
GNM
190
310
200
170
10
NERL
-40
-100
-100
-100
-160
WSU
-430
-770
-540
-510
-190
Hospital Admissions for Respiratory Illness (Ages 65+)
lllinois-1
6050
5500
5410
4850
1940
lllinois-2
5650
5120
5110
4630
1780
CMU
5190
4630
4580
3880
1670
Harvard
2530
2320
2410
2130
940
GNM
300
220
80
10
-250
NERL
70
10
-140
-620
-310
WSU
-1480
-1420
-1050
-650
30
Emergenc
/ Room Visits for Asthma (All Ages)
lllinois-1
1370
1710
1490
1760
1290
lllinois-2
1330
1670
1460
1720
1240
CMU
1230
1500
1300
1490
1130
Harvard
700
870
770
900
730
GNM
-80
-130
-130
-180
-220
NERL
-90
-130
-130
-170
-200
WSU
0
-60
0
-60
190
School Loss Days (Ages 5-17)
lllinois-1
633000
925000
743000
880000
659000
lllinois-2
638000
937000
755000
893000
650000
CMU
522000
745000
599000
679000
545000
Harvard
299000
445000
362000
422000
347000
GNM
50000
67000
44000
35000
-29000
NERL
-25000
-50000
-50000
-67000
-84000
WSU
-134000
-212000
-153000
-197000
-27000
Minor Restricted Activity Days (Ages 18 - 64)
lllinois-1
1959000
2063000
1934000
2333000
1681000
lllinois-2
1941000
2049000
1927000
2362000
1612000
CMU
1637000
1688000
1582000
1818000
1436000
Harvard
926000
990000
941000
1131000
872000
GNM
120000
108000
73000
58000
-78000
NERL
-76000
-109000
-130000
-202000
-213000
WSU
-333000
-375000
-301000
-460000
2000
'Respiratory hospital admissions and emergency room visits for asthma are rounded to the nearest 10; school loss
days and minor restricted activity days are rounded to the nearest 1000.
7-5
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
7.2. Uncertainty
As noted above, there is substantial uncertainty surrounding each of the inputs to our
analysis, particularly because the analysis focuses so far in the future. In particular, there
is uncertainty surrounding
• the meteorological conditions that will result from the accumulation of greenhouse
gases in the atmosphere (as predicted by the different climate change models);
• the corresponding changes in O3 concentrations (as simulated by the different air
quality models);
• the size, as well as the age and geographic distributions of the population that will
be affected (as represented in the different population projections); and
• the relationships between adverse health effects in that population and (future) O3
concentrations (as embodied in the different C-R functions).
We do not have ways to quantitatively assess much of this uncertainty. Even assigning
probabilities to the different models (representing our subjective assessments about the
relative accuracy with which each approximates a future reality) is premature. Instead, we
present our analysis as a series of "sensitivity analyses" or "what if' scenarios designed to
assess the impact of the various assumptions and modeling approaches on the results of the
analysis.
We discuss two of the most important sources of uncertainty below. Uncertainty due to
different C-R functions for the same health endpoint, as well as the standard uncertainty
surrounding individual C-R functions due to the statistical estimation of their coefficients,
can be seen in the tables of results in Appendix A.
7.2.1. Influence of O3 Changes from the Climate Change/Air Quality Models
The source of the greatest uncertainty appears to be the climate change/air quality models.
Figure 7-3 illustrates the influence of climate change/air quality model on estimated non-
accidental deaths attributable to climate change in 2050, using the C-R function for non-
accidental mortality from Bell et al. (2004). Figure 7-4 provides the legend for Figure 7-3.
The range of results across climate change/air quality models is the largest when combined
with the ICLUSAl population projection. Using the C-R function from Bell et al. (2004)
for non-accidental deaths and the June, July, August O3 season definition, the combination
of the Illinois-1 modeling system and the ICLUSAl population projection predicted 570
03-related non-accidental deaths attributable to climate change in 2050; at the other
extreme, the combination of WSU and ICLUS Al predicted almost 150 03-related deaths
avoided because of climate change in 2050. The difference between the two estimates is
over 700 deaths.
7-6
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Figure 7-3. Estimated National 03-Related Non-Accidental Mortality (Based on Bell et al., 2004) in
2050 03 Season (Defined as June, July, and August) Due to Climate Change: Impact of
Climate Change/Air Quality Model, for Each Population Projection
600
500
400
300
200
100
0
-100
-200
ICLUS_A1 ICLUS_A2 ICLUS_BC Woods & Census_2000
Poole
I
1
Figure 7-4. Legend for Climate Change/Air Quality Models in Figures
~ lllinois-1
~ lllinois-2
~ CMU
~ Harvard
~ GNM
~ NERL
¦ WSU
7-7 DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Similarly, using the C-R function for all-cause mortality from Levy et al. (2005) and the
June, July, August O3 season definition, the combination of the Illinois-1 modeling system
and ICLUSAl population projection predicted over 2,500 03-related deaths attributable
to climate change in 2050 nationally; at the other extreme, the combination of the WSU
modeling system and ICLUS Al predicted over 650 03-related deaths avoided because of
climate change in 2050. The difference between these two estimates is over 3,200 deaths.
The results from the individual climate-Ch simulations input to BenMAP are in agreement
on a number of fundamental points (U.S. EPA, 2009a). For example, they all found that
climate change caused increases in summertime O3 concentrations over substantial regions
of the country, with these increases in the range of 2-8 ppb. They also found a greater
sensitivity of peak O3 events to climate change than mean summer O3. However, there are
also clear differences across the simulations in the spatial distributions of 03 changes
across the country, with areas of little O3, or even decreases, interspersed throughout the
areas of increases with different patterns from simulation to simulation (see Figure 4-1
above). There seems to be (very generally) more agreement on uniform climate-induced O3
increases for the eastern half of the country than for the West, though parts of the
Southeast also show some of the strongest disagreements across the modeling groups.
These differences across simulations will be discussed in more detail in Section 7.3 below.
The wide range of predicted 03-related mortality incidence attributable to climate change -
including a fundamental difference in the message about whether climate change will
increase or decrease 03-related mortality - highlights the need to use an ensemble
approach, rather than relying on any one modeling system to predict the 03-related human
health effects attributable to climate change in a future year. This is perhaps the most
important "take away" message of our analysis.
However, while there is a very wide range of results, including those that suggest that
climate change would decrease the incidence of 03-related mortality, the large
preponderance of the results across the different climate change/air quality simulations
show positive values, thereby suggesting that, all else being equal, climate change would
lead to an increase in 03-related non-accidental deaths in 2050.
7.2.2. Influence of Projected Population Changes
The population projection also made a significant difference, although a smaller difference
than the climate change/air quality model. Figure 7-5 illustrates the influence of
population projection on estimated non-accidental deaths attributable to climate change in
2050, using the C-R function for non-accidental mortality from Bell et al. (2004). Figure
7-6 provides the legend for Figure 7-5.
7-8
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Figure 7-5. Estimated National 03-Related Non-Accidental Mortality (Based on Bell et al., 2004) in
2050 03 Season (Defined as June, July, and August) Due to Climate Change: Impact of
Population Projection, for Each Climate Change/Air Quality Model
600
500
400
300
200
100
-100
-200
lllinois-1 lllinois-2 CMU Harvard GNM NERL WSU
Figure 7-6. Legend for Population Projections in Figures
Qj&jQUS_A1
2000
1500
¦WUS_A2
Tiiiu-
U i i r^r'imr up
-500
~ll(^QlJS_BC
1 H S
~ Woo® &jj^oole
~ Census_2000
7-9
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
The spread in results across population projections is the largest when combined with the
Illinois-1 climate change/air quality model. Using the C-R function from Bell et al. (2004)
and the June, July, August O3 season definition, the combination of the ICLUSAl
population projection and the Illinois-1 resulted in about 570 03-related deaths attributable
to climate change in 2050, as noted above; at the other extreme, the combination of
Census_2000 and Illinois-1 resulted in only 175 03-related deaths due to climate change in
2050. The difference between the two estimates is almost 400 deaths (as compared to a
difference of over 700 deaths across climate change/air quality models).
Our analysis is one of the first to try to project population growth increases as well as
changes in age and geographic distributions by a future year, and we find that this affects
the estimates of health impacts substantially. The impact of projecting the size of a future
population is clearly illustrated by comparing the results based on the Census_2000
population projection to those based on any of the other population projections. The
Census_2000 population "projection" isn't really a projection - i.e., it assumes that the
population in 2050 will be exactly what it was in the year 2000. This is unrealistic in a
way that will produce a known (downward) bias in results. In fact, of the almost 400-case
difference in results produced by the two population projection extremes, noted above, 67
percent (or 265 deaths) is due to the difference between the result produced by the
Census_2000 population "projection" (175 deaths) and the next highest result, produced by
the Woods & Poole projection (439 deaths).36
Our results illustrate, then, how important it is to take into account that the population is
likely to have grown between the present and a year as far in the future as 2050. Public
health strategies to reduce adverse health consequences will need to account for both the
changes in risks from climate change and population changes.
Not only is the total population exposed to O3 in a future year important, but the age and
geographic distributions of that population can also make a substantial difference in the
impact of climate change on 03-related adverse health effects. The ICLUS A2 population
projection, for example, is, in total, greater than the ICLUS Al population projection
(424.8 million vs. 386.7 million). However, the ICLUS Al population projection is
skewed more towards the older ages than the ICLUS A2 population projection, as shown
below in Figure 7-7. In particular, about 26 percent of the ICLUS Al population
projection is 65 or older, versus only about 21 percent of the ICLUSA2 population
projection.37 Since older people have substantially higher baseline incidence rates for
mortality (and other adverse health effects) than younger people, the same increase in O3
concentration will result in more deaths among an older population than a younger one.
This is reflected in the slightly higher numbers of 03-related deaths attributable to climate
change in 2050 when the ICLUS Al population projection is used (as compared with the
ICLUS A2 population projection) despite the overall smaller population.
36 These numbers are still based on Bell et al. (2004) and the June, July, and August definition of the 03
season.
37 This results in a greater number of people ages 65 and up in the ICLUS Al population projection (over
100 million) than in the ICLUSA2 population projection (under 91 million), even though the latter total
population is somewhat larger.
7-10
DRAFT: Do Not Quote or Cite
-------
1 Figure 7-7. Age Distributions of ICLUS Al and ICLUS A2 Population Projections to the Year 2050
Age Distribution of ICLUS_A1 Population Projection
25%
20%
Percent c^f Population
15%
10%
5%
0%
< 1
1-17 18-24 25-34
35-44 45-54
Age Range
55-64 65-74 75-84
85+
Age Distribution of ICLUS_A2 Population Projection
25%
20%
Percent c^f Populatio
15%
10%
5%
0% -M 1 1 1 1 —^1 1 1 1
<1 1-17 18-24 25-34 35-44 45-54 55-64 65-74 75-84 85+
Age Range
2
3
7-11
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
The importance of the age distribution of the affected population is particularly apparent
when we consider the impact of climate change on 03-related morbidity endpoints,
because several of these health endpoints focus on specific age subgroups in the
population. The impact of age distribution can be seen, for example, if we compare the
mortality results to the results for hospital admissions for respiratory illness among infants.
For every climate change/air quality model, the ICLUSAl population projection predicts
a greater magnitude of 03-related non-accidental mortality than does the ICLUS A2
population projection (as shown above in Figure 7-5). In contrast, estimates of 03-related
respiratory hospital admissions among infants attributable to climate change in 2050 based
on the ICLUS Al population projection are uniformly smaller in magnitude than the
corresponding estimates based on the ICLUSA2 population projection, regardless of the
climate change/air quality model (as shown in Table 7-1 above). This is because the
ICLUSA2 population projection has a greater percentage of the population (and a larger
total population) under 1 year of age relative to the ICLUS Al population projection, and
a smaller percentage of the population in the 65 and older categories relative to
ICLUS Al. Thus the ICLUS Al population projection predicts fewer infant respiratory
hospitalizations but more deaths (the preponderance of which occur among those 65 and
older).
7.3. Regional Results
The three broad regions into which we divided the country for this analysis are shown in
Figure 3-2 above. The particular regional divisions used were chosen to roughly match the
major divisions in the climate-03 results (see U.S. EPA, 2009a; also Figure 4-1 above for
the regional modeling results): i.e., a relatively more uniformly positive O3 sensitivity to
climate in the Northeast across the simulations; large amplitudes of climate-induced 03
changes in the Southeast but with large disagreements across the simulations; and a fairly
mixed picture west of the Mississippi. Clearly the regions chosen are very broad, thereby
limiting the spatial specificity of the discussion. The high degree of variability across the
simulations, and the generally high level of uncertainty in regional climate modeling,
suggest that it is probably not particularly useful to look in detail at smaller-scale areas
(e.g., an individual state). Note, however, that the basic regional results shown here are not
particularly sensitive to the choice of averaging domains - for example, averaging over
only the state of California as opposed to the entire West.
As discussed above in Section 7.2.1, there are significant differences across the seven
climate-03 simulations in the spatial patterns of O3 changes they simulate, resulting in
particular in relatively large differences in the Southeast, and to a certain extent the West.
Nationally, these inter-simulation differences in O3 response to global climate change are
due largely to differences in how the modeling systems simulate the following key factors
(in roughly the following order of importance; U.S. EPA, 2009a):
• Regional patterns of incoming solar radiation at the surface (which strongly affects
03 photochemistry) driven primarily by differences in regional cloud cover patterns
across the simulations;
• Regional patterns of simulated temperature change;
7-12
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
• How the different models respond to changes in climate-induced VOC emissions
from natural sources (e.g., vegetation).
These differences, in turn, stem from differences in how the models represent the
following:
• Large-scale circulation patterns that strongly affect regional meteorology, such as
the extra-tropical storm tracks and the subtropical high pressure systems over the
adjacent oceans;
• Small-scale physical processes that must be parameterized in the models, in
particular those related to the production of clouds and precipitation;
• Key chemical pathways that control the control the interaction between NOx,
VOCs, and O3 in regions of large increases in biogenic VOC emissions.
The national estimates of (Vrelated human health effects attributable to climate change are
the sums of the regional estimates, and they can mask very different regional scenarios, as
illustrated below in Figures 7-8 and 7-9. (The legends for these figures are given in
Figures 7-4 and 7-6, respectively.) The WSU climate change/air quality model offers a
particularly striking example of this. At the national level, the WSU model predicts
modest overall decreases in (Vrelated premature mortality as a result of climate change.
These modest national decreases are the sums of much more substantial decreases in the
Southeast, small decreases in the West, and substantial increases in the Northeast. Using
the ICLUSAl population projection, the June, July, August definition of the O3 season,
and Bell et al. (2004), for example, the WSU model predicts about 280 (Vrelated deaths in
the Northeast attributable to climate change and about 370 (Vrelated deaths in the
Southeast and 50 (Vrelated deaths in the West avoided as a result of climate change. The
national total, then, is 280 + (-370) + (-50) = -150. The modest national result obscures
the more substantial impacts of climate change on (Vrelated deaths in opposite directions
in the Northeast and Southeast.
As discussed above, and more extensively in U.S. EPA (2009a), the climate change/air
quality models differ substantially in the regional patterns of climate-induced O3 changes
they simulate. With the exception of Illinios-1 and Illinios-2, none of the models shows
regional impacts uniformly in one direction - i.e., increases in O3 concentrations
attributable to climate change in some regions are accompanied by decreases in other
regions. While the WSU model shows large decreases in (Vrelated deaths in the Southeast
and large increases in the Northeast, two of the other models - GNM and NERL - show
just the opposite regional effects, although neither of these models show effects of the
same magnitude as the WSU model.
7-13
DRAFT: Do Not Quote or Cite
-------
Figure 7-8. Estimated National and Regional 03-Related Non-Accidental Mortality (Based on Bell et al., 2004) in 2050 03 Season (Defined as June,
July, and August) Due to Climate Change: Impact of Climate Change/Air Quality Model, for Each Population Projection
300
200
100
-100
I
I I
in
ICLUS_A1 ICLUS_A2 ICLUS_BC Woods & Census_2000
Poole
National
~ 1
ICLUS_A1 ICLUS_A2 ICLUS_BC
Southeast
Woods & Census_2000
Poole
600
500
400
300
200
100
-100
-200
-300
-400
600
500
400
300
200
100
0
-100
-200
-300
-400
BWMfe
ICLUS_A1 ICLUS_A2 ICLUS_BC Woods & Census_2000
Poole
Northeast
ICLUS_A1 ICLUS_A2 ICLUS_BC
West
Woods & Census_2000
Poole
7-14
DRAFT: Do Not Quote or Cite
-------
Figure 7-9. Estimated National and Regional 03-Related Non-Accidental Mortality (Based on Bell et al., 2004) in 2050 03 Season (Defined as June,
July, and August) Due to Climate Change: Impact of Population Projection, for Each Climate Change/Air Quality Model
600
500
400
300
200
100
0
-100
-200
-300
-400
600
500
400
300
200
100
0
-100
-200
-300
-400
I
U
III jnois-1 III j nois-2 CMU Harvard GNM NERL WSU
National
I 1 T [ 1 T
III lnols-1 MM nols-2 CMU Harvard GNM NERL
Southeast
600
500
400
300
200
100
-100
-200
-300
-400
600
500
400
300
200
100
0
-100
-200
-300
-400
iMfai
1
lllinois-1 III lnols-2 CMU Harvard GNM
Northeast
NERL WSU
mr
mjiL
LB-LT'ULLh1
WSU
lllinois-1 lllinois-2 CMU Harvard GNM
West
NERL
WSU
7-15
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
All of the models show relatively small effects in the West. This is, in part, because the
population in the West is smaller than the populations in the Northeast and Southeast. Using the
ICLUS Al population projection, for example, the population in the West is only 58 percent of
the population in the Northeast in 2050 and only 75 percent of the population in the Southeast.
For some of the climate change/air quality models, the smaller magnitude of effects in the West is
also due to smaller magnitude 03 changes in the West. The CMU model, for example, simulated
average increases in O3 concentration in the Northeast and Southeast of about 2.3 and 2.2 ppb,
respectively, whereas it simulated an average decrease in the West of about 0.8 ppb. In contrast,
however, the GNM model simulates greater 03-related adverse health effects in the West than in
the Southeast, even given the smaller population - using the Woods & Poole population
projection, Bell et al. (2004), and an O3 season of June, July, and August, for example, the GNM
model simulates 57 03-related deaths in the west due to climate change and only 13 in the
southeast.
7.4. Extension of the 03 Season
While the climate change/air quality models used in this analysis generally defined the O3 season
as June, July, and August, i.e., climatological summer in the Northern Hemisphere, most air
pollution epidemiology studies focusing on O3 have defined the season more broadly. The
"expanded" 03 season of May through September is actually more consistent with the current
understanding of the O3 season in most locations in the United States.38 By including only the
three summer months in their modeling of climate change-induced 03 changes, the climate
change/air quality modeling efforts considered here may thus have understated the potential 03-
related human health impacts of climate change. A few of the modeling groups have also
investigated climate-induced changes in the spring and fall and found a response similar to that in
the summer (e.g., see Nolte et al., 2008; Chen et al., 2009; Racherla and Adams, 2008). Therefore,
here we estimate the sensitivity of the results to an alternate O3 season definition.
We expand the O3 season from June, July, and August to May through September simply by
increasing all results by 66 percent, reflecting the facts that there are 92 days in June, July, and
August and 153 days in May through September and that we applied a seasonal average (based on
June, July, and August) of daily 8-hour maxima to each day in the 03 season. The contrast in
estimated 03-related human health impacts is illustrated for all-cause mortality in Figure 7-10.
This result shows that a longer 03 season has the potential to significantly increase the incidence
of adverse health outcomes associated with climate change. Whether climate change, in general,
has the potential to increase the 03 season beyond is an area that warrants further research.
38 The 03 season actually varies somewhat geographically. In some locations - e.g., Los Angeles and Houston - it is
considered to be all year.
7-16 DRAFT: Do Not Quote or Cite
-------
Figure 7-10. Estimated National 03-Related Cases of All-Cause Mortality in 2050 03 Season (Bell et al., 2005)
Due to Climate Change: Impact of 03 Season Definition
03 Season: June, July, and August
3500
3000
2500
2000
1500
1000
500
I I
n n
-500
-1 ooo
G1 G2 G3 G4 F4 F5 B5 F3 G5 F2 B4 F1 B3 B2 B1 C5 A5 E5 D5 C4 C2 C3 C1 A4 E4 A3 A2 D4 E3 A1 E2 D3 D2 E1 D1
Climate Change Model/Population Projection
03 Season: May through September
3500
3000
2500
2000
1500
1000
500
0
-500
-1000
. n n
G1 G2 G3 G4 F4 F5 B5 F3 G5 F2 B4 F1 B3 B2 B1 C5 A5 E5 D5 C4 C2 C3 C1 A4 E4 A3 A2 D4 E3 A1 E2 D3 D2 E1 D1
Climate Change Model/Population Projection
7-17
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
7.5. Conclusions
How important are the projected (Vrelated adverse health effects of climate change? Are more
03-related premature deaths likely to be caused by climate change than would be avoided as a
result of implementing more stringent O3 NAAQS? A comparison of our results to the results
obtained in the benefit analysis for the most recent 03 NAAQS RIA (U.S. EPA, 2008b) may help
put our results in perspective.39 Because we used the suite of C-R functions used in the O3
NAAQS RIA benefit analysis, such a comparison is relatively straightforward, although there are
two notable differences between our analysis and that analysis to keep in mind. First, our analysis
defined the 03 season as either June, July, and August or May through September, whereas the
benefit analysis for the O3 NAAQS RIA defined the O3 season as May through September.40
Second, the change in 03 concentrations in our analysis is from a baseline of 03 concentrations in
the absence of climate change, making no assumption about whether or not the current O3
NAAQS have been attained. In the benefit analysis for the most recent 03 NAAQS RIA, the
results that are reported if alternative standards are just met are incremental to the current standard
- i.e., the baseline is a scenario in which the current standard is just met.
With those differences in mind, we used Bell et al. (2004) to compare our results to those in the 03
NAAQS RIA benefit analysis. Because that analysis defined the O3 season as May through
September, we used our results based on the "expanded" 03 season rather than the shorter season
of June, July, and August for the comparison. Using Bell et al. (2004), the O3 NAAQS RIA
benefit analysis estimated that national full attainment of standards set at 0.079, 0.075, 0.070, and
0.065 ppm would result in 24, 71, 250, and 450 premature non-accidental deaths avoided,
respectively (see U.S. EPA, 2008b, Tables 6-18, 6-14, 6-10, and 6-6). All of these estimates are
contained within the broader range of results, based on the same study, produced in our analysis.
Three of the models (Illinois-1, Illinois-2, and CMU) produced results (using Bell et al., 2004) all
of which were greater in magnitude than the estimates of premature deaths avoided if each of the
alternative 03 NAAQS were just met - results from these models range from about 950 to about
580 premature deaths attributable to climate change.41 Other models (NERL) produced results
under 10 in absolute value. It is not surprising that the results produced in our analysis encompass
and extend beyond those produced in the O3 NAAQS RIA benefit analysis, since there is
substantial additional uncertainty introduced into our analysis by the climate change models.
What does all of this mean? Recalling the purpose of this report, we have attempted to assess the
sensitivity of modeled human health impacts to assumptions about the following key inputs:
• Climate-induced changes in meteorological conditions;
39 In the 03 NAAQS RIA benefit analysis 03 concentrations are largely decreasing as a result of alternative (more
stringent) standards being met, so the estimated changes in health effects are numbers of cases avoided as a result of
these alternative standards being just met. In our analysis, in contrast, 03 concentrations are increasing in most (but
not all) of the scenarios considered as a result of climate change, so the estimated changes in health effects are largely
numbers of cases attributable to climate change. However, it is reasonable to compare the magnitudes of estimates -
i.e., we can compare the absolute values of our estimates to those of the 03 NAAQS RIA benefit analysis.
40 Personal communication with Neal Fann, EPA/OAQPS, on September 24, 2009.
41 Estimates based on Census_2000 were included in our analysis largely to illustrate the importance of projecting the
population to a future year. These estimates thus have a substantial downward bias, and are therefore omitted from
the comparisons discussed here.
7-18
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
• Corresponding changes in O3 concentrations;
• The size and geographic distribution of the affected population;
• The relationships that link O3 levels to specific health outcomes;
• The fraction of the year over which O3 is assumed to affect health.
Given this context, we can draw the following conclusions:
• Looking across all combinations of climate change/air quality models, population
projections, O3 season definitions, and C-R functions for all-cause mortality considered in
our analysis, estimates of national 03-related all-cause mortality around 2050 attributable
to climate change span a range of over 5000, i.e., from roughly -1000 to +4200. Despite
this range, the large preponderance of the estimates are positive, suggesting that, all else
being equal, climate change would be likely to increase the incidence of 03-related all-
cause premature mortality in 2050.
• The source of the greatest uncertainty at the national level appears to be the particular
climate change/air quality scenario used.
• The choice of population projection also made a significant difference, although only about
half that of climate change/air quality scenario at the national level.
• It is important to take into account that the size of the population exposed to O3 will
increase by a future year. Failing to do so may result in substantially downward biased
estimates of future 03-related adverse health impacts of climate change.
• Not only is the total population exposed to O3 in a future year important, but the age (and
geographic) distribution of that population can also make a significant difference in the
estimated impact of climate change on 03-related adverse health effects (e.g., the
difference in ICLUS Al and ICLUS A2 results discussed above).
• The national results can mask important regional differences. The Northeast showed the
greatest agreement (of generally adverse health impacts associated with climate-induced
O3 increases) across the seven climate/air quality scenarios used in this study, while the
Southeast showed large disagreements in health impacts across the different scenarios. The
West generally showed the smallest impacts, largely due to the smaller projected
populations compared to the Northeast and Southeast.
• A climate-induced extension of the O3 season later into the fall and earlier into the spring
has the potential to significantly increase the incidence of negative health outcomes.
At this stage in the development of our scientific understanding of climate change and its potential
impact on air pollution-related human health, it would be unwise to rely on any one model or any
one population projection. This may be the most important "take away" message of our analysis.
The different model combinations can produce widely varying results, particularly at the regional
level, in some cases leading to fundamentally different conclusions about the overall impact of
climate change on 03-related health effects. This has a number of implications for the
development of meaningful analyses to assess the range of benefits associated with responses to
climate change.
45
7-19
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
REFERENCES
Abt Associates Inc. 2008. BenMAP: Environmental Benefits Mapping and Analysis
Program, User's Manual. Prepared for Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC. September 2008.
Avise, J; Chen, J; Lamb, B; et al. (2009) Attribution of projected changes in summertime
U.S. ozone and PM2.5 concentrations to global changes. Atm Chem and Phys
9:1111-1124.
Bell, ML. (2007) Climate change, ambient ozone, and health in 50 US cities. Climatic
Change 82(1-2): 61-76.
Bell, ML; et al. (2004) Ozone and short-term mortality in 95 US urban communities,
1987-2000. JAMA 292(19): p. 2372-2378.
Bell, ML; Dominici, F; Samet, JM. (2005) A meta-analysis of time-series studies of
ozone and mortality with comparison to the national morbidity, mortality, and air
pollution study. Epidemiology 16(4): p. 436-445.
Bell, ML; Peng, RD; et al. (2006) The exposure response curve for ozone and risk of
mortality and the adequacy of current ozone regulations. Environ Health Perspect
114(4): 532-536.
Bell, ML; Davis, DL; et al. (2008) Ancillary human health benefits of improved air
quality resulting from climate change mitigation. Environ Health 7: 41.
Bernard, SM; Samet, JM; et al. (2001) The potential impacts of climate variability and
change on air pollution-related health effects in the United States. Environ Health
Perspect 109 Suppl 2: 199-209.
Bloomfield, P; Royle, JA; Steinberg, LJ; et al. (1996) Accounting for meteorological
effects in measuring urban ozone levels and trends. Atmos Environ 30:3067-3077.
Burnett, RT; Smith-Doiron, M; Stieb, D; Raizenne, ME; Brook, JR; Dales, RE; et al.
(2001) Association between ozone and hospitalization for acute respiratory diseases in
children less than 2 years of age. Am J Epidemiol 153(5):444-452.
Camalier, L; Cox, W; Dolwick, P. (2007) The effects of meteorology on ozone in urban
areas and their use in assessing ozone trends. Atmos Environ 4:7127-7137.
Chen, L; Jennison, BL; Yang, W; Omaye, ST. (2000) Elementary school absenteeism and
air pollution. Inhal Toxicol 12(11):997-1016.
7-20
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Chen, J; Avise, J; Lamb, B; et al. (2009) The effects of global changes upon regional
ozone pollution in the United States. Atm Chem and Phys 9:1125-1141.
Chiang, P.L. 1967. Variance and Covariance of Life Table Functions Estimated from a
Sample of Deaths. National Center for Health Statistics. Washington, DC. March.
Cifuentes, L; Borja-Aburto, VH; et al. (2001) Climate change. Hidden health benefits of
greenhouse gas mitigation. Science 293(5533): 1257-1259.
Cooter, E; Gilliam, R; Gilliland, A; et al. (2005) Examining the impact of climate change
and variability on regional air quality over the United States. Presented at the U.S.
Climate Change Science Program, Climate Science in Support of Decision Making
Workshop, Arlington, VA, 14-16 November.
Cox, WM; Chu, S-H. (1993) Meteorologically adjusted ozone trends in urban areas: A
probabilistic approach. Atmos Environ 27B:425-434.
D'Amato, G; Cecchi, L. (2008) Effects of climate change on environmental factors in
Respiratory allergic diseases. Clin Exp Allergy 38(8): 1264-1274.
Deschenes, O; Greenstone, M. (2007) Climate Change, Mortality and Adaptation:
Evidence from Annual Fluctuations in Weather in the U.S. Massachusetts Institute of
Technology Department of Economics Working Paper Series. Cambridge, MA,
Massachusetts Insistute of Technology: Working Paper 07-19, Revised Dec. 2, 2008.
Ebi, KL; Mills, DM; et al. (2006) Climate change and human health impacts in the
United States: an update on the results of the U.S. national assessment. Environ Health
Perspect 114(9): 1318-1324.
Ebi, KL; McGregor, G. (2008) Climate change, tropospheric ozone and particulate
matter, and health impacts. Environ Health Perspect 116(11): 1449-1455.
Gardner Jr., ES. (2006) Exponential smoothing: The state of the art—Part II,
International Journal of Forecasting 22:637- 666.
Gilliam, RC; Hogrefe, C; Rao, ST. (2006) New methods for evaluating meteorological
models used in air quality applications. Atmos Environ 40:5073-5086,
doi:10.1016/j.atmosenv.2006.01.023.
Gilliland, FD; Berhane, K; Rappaport, EB; Thomas, DC; Avol, E; Gauderman, WJ; et al.
(2001) The effects of ambient air pollution on school absenteeism due to respiratory
illnesses. Epidemiology 12(l):43-54.
Gustafson, WI; Leung, RL. (2007) Regional downscaling for air quality assessment: A
reasonable proposition? Bull Am Meteorol Soc 88:1215-1227.
Hayhoe, KD; Cayan; et al. (2004) Emissions pathways, climate change, and impacts on
7-21
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
California. ProcNatl Acad Sci USA 101(34): 12422-12427.
Huang, H-C; Liang, X-Z; Kunkel, KE; et al. (2007) Seasonal simulation of tropospheric
ozone over the Midwestern and Northeastern United States: An application of a coupled
regional climate and air quality modeling system. J Appl Meteor Clim 46:945-960.
Huang, H-C; Lin, J; Tao, Z; et al. (2008) Impacts of long-range transport of global
pollutants and precursor gases on U.S. air quality under future climatic conditions. J
GeophysRes 113:D19307, doi:10.1029/2007JD009469.
Hwang R, Burer MJ, Bell M. 2004. Smog in the Forecast: Global Warming, Ozone
Pollution and Health in California. San Francisco: National Resources Defense Council.
Hyndman, RJ (2009) forecast: Forecasting functions for time series, R package version
1.24. URL: http://vvvvvv.robhvndman.info/Rlibrarv/forecast/.
IPCC (Intergovernmental Panel on Climate Change). (2007) Climate change 2007: The
physical science basis. Contribution of Working Group I to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change [Solomon, S; Qin, D;
Manning, M; et al.; (eds.)]. Cambridge, United Kingdom and New York, NY: Cambridge
University Press, 996 pp.
IPCC. 2007. Climate Change 2007: Impacts, Adaptation and Vulnerability. Cambridge,
UK.
Ito, K; De Leon, SF; Lippmann, M. (2005) Associations between ozone and daily
mortality: analysis and meta-analysis. Epidemiology 16(4): p. 446-457.
Jerrett, M; Burnett, RT; Pope, CA. (2009) Long-term ozone exposure and mortality. N
Engl J Med 360:1085-1095.
Kalkstein, LS; Greene, JS. (1997) Evaluation of climate/mortality relationships in
large U.S. cities and the possible impacts of a climate change. Environ Health Perspect.
105(1): 84-93.
Kalkstein, LS; Greene, JS. (2007) An analysis of potential heat-related mortality
increases in U.S. cities under a business-as-usual climate change scenario. Environment
America Global Warming Solutions Reports. Retrieved Feb 5, 2009.
Kinney, P; O'Neill, MS; et al. (2008) Approaches for estimating effects of climate
change on heat related deaths: Challenges and opportunities. Environmental Science and
Policy II: 87-96.
Kinney, PL. (2008) Climate change, air quality, and human health. Am J Prev Med .
35(5): 459-467.
Knowlton, K; Rosenthal, JE; et al. (2004) Assessing ozone-related health impacts under a
7-22
DRAFT: Do Not Quote or Cite
-------
1 Changing climate. Environ Health Perspect 112(15): 1557-1563.
2
3 Kunkel, KE; Liang, X-Z. (2005) GCM simulations of the climate in the central United
4 States. J Climate 18:1016-1031.
5
6 Leibensperger, EM; Mickley, LJ; Jacob, DJ. (2008) Sensitivity of U.S. air quality to mid-
7 latitude cyclone frequency and implications of 1980-2006 climate change. Atm Chem
8 Phys 8:7075-7086.
9 Leung, LR; Gustafson, WI. (2005) Potential regional climate change and implications to
10 U.S. air quality. Geophys Res Lett 32:L16711, doi:10.1029/2005GL022911.
11 Levy, JI; Chemerynski, SM; Sarnat, JA. (2005) Ozone exposure and mortality: an
12 empiric bayes metaregression analysis. Epidemiology 16(4): p. 458-468.
13
14 Liang, X-Z; Pan, J; Zhu, J; et al. (2006) Regional climate model downscaling of the U.S.
15 summer climate and future change. J Geophys Res 111 :D10108,
16 doi: 10.1029/2005JD006685.
17 Liang, X-Z; Xu, M; Kunkel, KE; et al. (2007) Regional climate model simulation of
18 U.S.-Mexico summer precipitation using the optimal ensemble of two cumulus
19 parameterizations. J Climate 20:5201-5207.
20 Liang, X-Z; Kunkel, KE; Meehl, GA; et al. (2008) Regional climate models downscaling
21 analysis of general circulation models present climate biases propagation into future
22 change projections. Geophys Res Lett 35:L08709, doi:10.1029/2007GL032849.
23 Liao, K-J; Tagaris, E; Manomaiphiboon, K; et al. (2007) Sensitivities of ozone and fine
24 particulate matter formation to emissions under the impact of potential future climate
25 change. Environ Sci Technol 41: 8355-8361, doi:10.1021/es070998z.
26 Liao, K-J; Tagaris, E; Manomaiphiboon, K; et al. (2009) Quantification of the impact of
27 climate uncertainty on regional air quality. Atmos Chem Phys 9:865-878.
28 Mickley, LJ; Jacob, DJ; Field, BD; et al. (2004) Effects of future climate change on
29 regional air pollution episodes in the United States. Geophys Res Lett 30:L24103,
30 doi: 10.1029/2004GL021216.
31
32 Moolgavkar, SH; Luebeck, EG; Anderson, EL. (1997) Air pollution and hospital
33 admissions for respiratory causes in Minneapolis St. Paul and Birmingham.
34 Epidemiology 8(4):364-370.
35
36 Morris, RE; Guthrie, PD; Knopes, CA. (1995) Photochemical modeling analysis under
37 global warming conditions. Proceedings of the 88th air & waste management association
38 annual meeting and exhibition: Pittsburgh, PA: Paper No. 95—WP 74B.02.
7-23
DRAFT: Do Not Quote or Cite
-------
1 Nolte, CG; Gilliland, AB; Hogrefe, C. (2008) Linking global to regional models to assess
2 future climate impacts on surface ozone levels in the United States. J Geophys Res
3 113:D14307, doi:10.1029/2007JD008497.
4 NAS (National Academy of Sciences). (2002) Estimating the public health benefits of
5 proposed air pollution regulations. The National Academies Press. Washington, DC: 192
6 pp.
7 NRC (National Research Council). (1991) Rethinking the ozone problem in urban and
8 regional air pollution. National Academy Press, Washington, DC; 489 pp.
9 NRC (National Research Council). (2001) Global air quality. National Academy Press,
10 Washington, DC; 41 pp.
11 NRC (National Research Council). (2004) Air quality management in the United States.
12 Committee on Air Quality Management in the United States, National Research Council.
13 National Academy Press, Washington, DC: 426 pp.
14 Patz, JA.; McGeehin; et al. (2000) The potential health impacts of climate variability and
15 change for the United States: executive summary of the report of the health sector of the
16 U.S. National Assessment. Environ Health Perspect 108(4): 367-76.
17
18 Peel, JL; Tolbert, PE; Klein, M; et al. (2005) Ambient air pollution and respiratory
19 emergency department visits. Epidemiology Vol. 16 (2): 164-174.
20
21 R Development Core Team. (2009) R: A Language and Environment for Statistical
22 Computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-
23 project.org.
24
25 Racherla, PN; Adams, PJ. (2006) Sensitivity of global ozone and fine particulate matter
26 concentrations to climate change. J Geophys Res 111 :D24103,
27 doi: 10.1029/2005JD006939.
28 Racherla, PN; Adams, PJ. (2008) The response of surface ozone to climate change over
29 the Eastern United States. Atmos Chem Phys 8:871-885.
30 Sardon, JP; Robertson, DA. (2004) Recent demographic trends in the developed
31 countries. Population 59(2):263-314.
32 Schwartz, J. (1994a) PM(10) ozone, and hospital admissions for the elderly in
33 Minneapolis St Paul, Minnesota. Arch Environ Health 49(5):366-374.
34
35 Schwartz, J. (1994b) Air pollution and hospital admissions for the elderly in Detroit,
36 Michigan. Am J Respir Crit Care Med 150(3):648-655.
37
38 Schwartz, J. (1995) Short term fluctuations in air pollution and hospital admissions of the
39 elderly for respiratory disease. Thorax 50(5):531-538.
40
7-24
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Shea, KM. (2008) Climate change and allergic disease. Journal of Allergy and Clinical
Immunology 122(3): 443-453.
Sillman, S; Samson, PJ. (1995) Impact of temperature on oxidant photochemistry in
urban, polluted rural and remote environments. J Geophys Res 100:11497-11508.
Sitch, S; Cox, PM; Collins, WJ; et al. (2007) Indirect radiative forcing of climate change
through ozone effects on the land-carbon sink. Nature 448:791-794.
Tagaris, E; Manomaiphiboon, K; Liao, K-J; et al. (2007) Impacts of global climate
change and emissions on regional ozone and fine particulate matter concentrations over
the United States. J Geophys Res 112:D14312, doi:10.1029/2006JD008262.
Tagaris, E; Liao, K; DeLucia, AJ; Deck, L; Amar, P; Russell AG. (2009) Potential
Impact of Climate Change on Air Pollution-Related Human Health Effects.
Environmental Science & Technology 43(13): 189-194.
Tao, Z; Williams, A; Huang, H-C; et al. (2007) Sensitivity of U.S. surface ozone to future
emissions and climate changes. Geophys Res Lett 34:L08811,
doi: 10.1029/2007GL02945 5.
Tao, Z; Williams, A; Huang, H-C; et al. (2008) Sensitivity of surface ozone simulation to
cumulus parameterization. J Appl Meteor Clim:47, 1456-1466.
Thompson, ML; Reynolds, J; Cox, LH; et al. (2001) Review of statistical methods for the
meteorological adjustment of tropospheric ozone. Atmos Environ 35:617-630.
U.S. EPA (U.S. Environmental Protection Agency). (1989) The Potential Effects of
Global Climate Change on the United States. Appendix F: Air Quality. Report to
Congress, EPA Office of Policy, Planning, and Evaluation. U.S. Environmental
Protection Agency, Washington, DC, EPA-230-05-89-056.
U.S. EPA (U.S. Environmental Protection Agency). (1999) Guideline for Developing an
Ozone Forecasting System. Report prepared by the Office of Air Quality Planning and
Standards. U.S. Environmental Protection Agency, Research Triangle Park, NC, EPA
454/R-99-009.
U.S. EPA (U.S. Environmental Protection Agency). (2008a) National Air Quality: Status
and Trends through 2007. U.S. Environmental Protection Agency, Washington, DC,
EPA-454/R-08-06.
U.S. EPA (U.S. Environmental Protection Agency). (2008b) Final Ozone NAAQS
Regulatory Impact Analysis. Report prepared by the Office of Air Quality Planning and
Standards. U.S. Environmental Protection Agency, Research Triangle Park, NC, EPA
452/R-08-003.
7-25
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
U.S. EPA. 2009a. Assessment of the Impacts of Global Change on Regional U.S. Air
Quality: A Synthesis of Climate Change Impacts on Ground-Level Ozone An Interim
Report of the U.S. EPA Global Change Research Program
U.S. EPA. 2009b. Land-Use Scenarios: National-Scale Housing-Density Scenarios
Consistent with Climate Change Storylines. Global Change Research Program, National
Center for Environmental Assessment, Washington, DC; EPA/600/R-08/076F. Available
from: National Technical Information Service, Springfield, VA, and online at
http://www.epa.gov/ncea.
West JJ, Szopa S, Hauglustaine DA. (2007) Human mortality effects of future
concentrations of tropospheric ozone. CR Geoscience. 339:775-783.
Wilson, AM; Wake, CP; Kelly, T; et al. (2005) Air pollution, weather, and respiratory
emergency room visits in two northern New England cities: an ecological time-series
study. Environ Res. Vol. 97 (3): 312-321.
Woo, J-H; He, S; Amar, P; et al. (2007) Development of North American emission
inventories for air quality modeling under climate change. J Air Waste Manage Assoc
58:1483-1494.
Wu, S; Mickley, LJ; Jacob, DJ; et al. (2007) Why are there large differences between
models in global budgets of tropospheric ozone? J Geophys Res 112:D05302.
doi: 10.1029/2006JD007801.
Wu, S; Mickley, LJ; Leibensperger, EM; et al. (2008a) Effects of 2000-2050 global
change on ozone air quality in the United States. J Geophys Res 113:D06302,
doi: 10.1029/2007JD008917.
Wu, S; Mickley, LJ; Jacob, DJ; et al. (2008b) Effects of 2000-2050 changes in climate
and emissions on global tropospheric ozone and the policy relevant surface background
ozone in the United States. J Geophys Res 113:D18312, doi:10.1029/2007JD009639.
7-26
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
APPENDIX A:
TABLES OF RESULTS
A-l
DRAFT: Do Not Quote or Cite
-------
Table A-l. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al.(2005)
CMU
ICLUS_A1
1530
(530 - 2540)
930
(440- 1420)
790
(380 - 1200)
-190
(-290 - -80)
Bell et al.(2005)
CMU
ICLUS_A2
1380
(470 - 2290)
830
(400- 1270)
720
(340 - 1100)
-180
(-270 - -80)
Bell et al.(2005)
CMU
ICLUS_BC
1360
(470 - 2250)
850
(400 - 1300)
670
(320 - 1030)
-170
(-260 - -70)
Bell et al.(2005)
CMU
Woods & Poole
1120
(340 - 1890)
820
(390 - 1250)
480
(230 - 730)
-180
(-270 - -80)
Bell et al.(2005)
CMU
Census_2000
500
(190 - 800)
340
(160 - 530)
190
(90 - 300)
-40
(-60 - -20)
Bell et al.(2005)
GNM
ICLUS_A1
120
(-380 - 630)
-150
(-380 - 80)
110
(-40 - 260)
160
(40 - 280)
Bell et al.(2005)
GNM
ICLUS_A2
100
(-370 - 570)
-150
(-370 - 60)
90
(-50 - 230)
160
(40 - 270)
Bell et al.(2005)
GNM
ICLUS_BC
60
(-410 - 520)
-180
(-400 - 40)
80
(-50-210)
160
(40 - 270)
Bell et al.(2005)
GNM
Woods & Poole
30
(-410-470)
-190
(-410-30)
40
(-60- 140)
180
(50-310)
Bell et al.(2005)
GNM
Census_2000
-60
(-220 - 100)
-110
(-200 --10)
0
(-30 - 40)
40
(10-80)
Bell et al.(2005)
Harvard
ICLUS_A1
770
(110- 1440)
680
(320 - 1050)
-80
(-250 - 100)
160
(40 - 280)
Bell et al.(2005)
Harvard
ICLUS_A2
710
(110 - 1320)
620
(290 - 950)
-70
(-230 - 90)
160
(40 - 270)
Bell et al.(2005)
Harvard
ICLUS_BC
730
(130 - 1320)
640
(290 - 980)
-50
(-200 - 90)
140
(40 - 250)
Bell et al.(2005)
Harvard
Woods & Poole
630
(70 - 1190)
600
(270 - 920)
-100
(-230 - 40)
130
(20 - 240)
Bell et al.(2005)
Harvard
Census_2000
280
(60 - 490)
260
(120 -400)
-10
(-60 - 30)
30
(0 - 60)
A-2
DRAFT: Do Not Quote or Cite
-------
Table A-l cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al.(2005)
lllinois-1
ICLUS_A1
1810
(800 - 2820)
820
(390 - 1250)
780
(370 - 1200)
200
(30 - 370)
Bell et al.(2005)
lllinois-1
ICLUS_A2
1660
(730 - 2590)
760
(360 - 1160)
710
(340 - 1090)
180
(30 - 340)
Bell et al.(2005)
lllinois-1
ICLUS_BC
1620
(710-2520)
790
(380 - 1200)
660
(310-1010)
170
(20 - 320)
Bell et al.(2005)
lllinois-1
Woods & Poole
1410
(600 - 2220)
770
(370 - 1180)
500
(240 - 770)
130
(-10-270)
Bell et al.(2005)
lllinois-1
Census_2000
570
(250 - 890)
340
(160 - 530)
190
(90 - 290)
O
O Is"-
1
0
Bell et al.(2005)
lllinois-2
ICLUS_A1
1690
(770-2610)
670
(310 - 1030)
660
(300 - 1010)
360
(160 - 560)
Bell et al.(2005)
lllinois-2
ICLUS_A2
1540
(710 -2380)
630
(290 - 960)
580
(270 - 900)
330
(150 - 520)
Bell et al.(2005)
lllinois-2
ICLUS_BC
1530
(700 - 2360)
660
(310-1010)
550
(250 - 850)
310
(140 -490)
Bell et al.(2005)
lllinois-2
Woods & Poole
1340
(610-2070)
640
(300 - 980)
410
(180 -630)
290
(120-460)
Bell et al.(2005)
lllinois-2
Census_2000
520
(240 - 800)
290
(140 -450)
150
(70 - 230)
80
(30 - 120)
Bell et al.(2005)
NERL
ICLUS_A1
40
(-510 - 580)
-120
(-330 - 100)
320
(110 - 530)
-160
(-290 - -40)
Bell et al.(2005)
NERL
ICLUS_A2
20
(-480 - 510)
-120
(-310-80)
290
(100 -470)
-150
(-270 - -40)
Bell et al.(2005)
NERL
ICLUS_BC
-30
(-520 - 460)
-150
(-350 - 50)
260
(90 - 440)
-150
(-260 - -40)
Bell et al.(2005)
NERL
Woods & Poole
-170
(-610-270)
-170
(-370 - 30)
160
(40 - 280)
-160
(-280 - -40)
Bell et al.(2005)
NERL
Census_2000
-80
(-230 - 80)
-90
(-180--10)
60
(20 - 100)
O
O 1
1
1 0
00
1
A-3
DRAFT: Do Not Quote or Cite
-------
Table A-l cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al.(2005)
wsu
ICLUS_A1
-470
(-1990 - 1060)
870
(350 - 1390)
-1170
(-1930 --410)
-170
(-410-80)
Bell et al.(2005)
wsu
ICLUS_A2
-450
(-1830 - 930)
770
(300 - 1230)
-1060
(-1750 --370)
-160
(-380 - 70)
Bell et al.(2005)
wsu
ICLUS_BC
-350
(-1680 - 990)
790
(320- 1260)
-970
(-1610 --340)
-160
(-390 - 60)
Bell et al.(2005)
wsu
Woods & Poole
-180
(-1400 - 1030)
790
(320- 1260)
-800
(-1310 --290)
-170
(-410-60)
Bell et al.(2005)
wsu
Census_2000
-10
(-440 - 430)
330
(130 - 520)
-280
(-460 --100)
-50
(-120-10)
Levy et al.(2005)
CMU
ICLUS_A1
2160
(1310-3010)
1310
(900 - 1730)
1110
(760- 1460)
-260
(-350 --180)
Levy et al.(2005)
CMU
ICLUS_A2
1940
(1170-2720)
1170
(800 - 1540)
1020
(700 - 1340)
-250
(-330 --170)
Levy et al.(2005)
CMU
ICLUS_BC
1910
(1160-2670)
1200
(820 - 1580)
950
(650 - 1250)
-230
(-310--160)
Levy et al.(2005)
CMU
Woods & Poole
1570
(920 - 2230)
1150
(790 - 1520)
670
(460 - 880)
-250
(-330 --170)
Levy et al.(2005)
CMU
Census_2000
700
(440 - 960)
490
(330 - 640)
270
(190 - 360)
-60
(-80 - -40)
Levy et al.(2005)
GNM
ICLUS_A1
170
(-260 - 600)
-210
(-410--10)
150
(20 - 280)
230
(130 - 330)
Levy et al.(2005)
GNM
ICLUS_A2
140
(-260 - 530)
-210
(-390 - -30)
130
(10-240)
220
(130 - 320)
Levy et al.(2005)
GNM
ICLUS_BC
80
(-310 -470)
-250
(-440 - -60)
110
(0 - 220)
220
(130 - 320)
Levy et al.(2005)
GNM
Woods & Poole
40
(-330 - 420)
-270
(-450 - -90)
60
(-30 - 140)
260
(150 - 360)
Levy et al.(2005)
GNM
Census_2000
-80
(-220 - 50)
-150
(-230 - -70)
10
(-30 - 40)
60
(30 - 90)
2
A-4
DRAFT: Do Not Quote or Cite
-------
Table A-l cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Levy et al.(2005)
Harvard
ICLUS_A1
1090
(530 - 1650)
960
(650- 1270)
-110
(-260 - 40)
230
(130 - 330)
Levy et al.(2005)
Harvard
ICLUS_A2
1000
(490 - 1520)
880
(590 - 1160)
-90
(-230 - 40)
220
(130 - 320)
Levy et al.(2005)
Harvard
ICLUS_BC
1020
(520 - 1530)
900
(610-1190)
-80
(-200 - 50)
200
(110 -290)
Levy et al.(2005)
Harvard
Woods & Poole
890
(410 - 1370)
840
(570 - 1110)
-130
(-240 - -20)
180
(90 - 270)
Levy et al.(2005)
Harvard
Census_2000
390
(210 - 570)
370
(250 - 490)
-20
(-60 - 20)
40
(20 - 70)
Levy et al.(2005)
lllinois-1
ICLUS_A1
2550
(1690 - 3410)
1160
(800 - 1530)
1100
(750 - 1450)
290
(140 -430)
Levy et al.(2005)
lllinois-1
ICLUS_A2
2340
(1550 - 3120)
1070
(730 - 1410)
1010
(690 - 1320)
260
(130 - 390)
Levy et al.(2005)
lllinois-1
ICLUS_BC
2280
(1510 - 3050)
1110
(760- 1460)
930
(640- 1220)
240
(110 - 360)
Levy et al.(2005)
lllinois-1
Woods & Poole
1990
(1300 -2670)
1090
(750 - 1430)
710
(490 - 940)
190
(70 - 300)
Levy et al.(2005)
lllinois-1
Census_2000
810
(540 - 1080)
490
(330 - 640)
270
(180 - 350)
50
(20 - 80)
Levy et al.(2005)
lllinois-2
ICLUS_A1
2380
(1610-3160)
950
(650 - 1250)
920
(620- 1230)
510
(340 - 680)
Levy et al.(2005)
lllinois-2
ICLUS_A2
2180
(1470 -2890)
890
(600 - 1170)
820
(550 - 1090)
470
(310 -630)
Levy et al.(2005)
lllinois-2
ICLUS_BC
2150
(1450 -2860)
930
(640- 1230)
780
(520 - 1030)
440
(290 - 590)
Levy et al.(2005)
lllinois-2
Woods & Poole
1890
(1270 -2500)
900
(610-1190)
570
(380 - 760)
410
(270 - 560)
Levy et al.(2005)
lllinois-2
Census_2000
730
(500 - 970)
410
(280 - 550)
210
(140 -280)
110
(70 - 150)
2
A-5
DRAFT: Do Not Quote or Cite
-------
Table A-l cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Levy et al.(2005)
NERL
ICLUS_A1
50
(-410-510)
-160
(-340 - 20)
450
(270 - 620)
-230
(-340 --130)
Levy et al.(2005)
NERL
ICLUS_A2
20
(-400 - 440)
-170
(-330 - 0)
400
(250 - 560)
-220
(-310--120)
Levy et al.(2005)
NERL
ICLUS_BC
-40
(-460 - 370)
-210
(-380 - -40)
370
(230 - 520)
-210
(-310--110)
Levy et al.(2005)
NERL
Woods & Poole
-240
(-610-130)
-240
(-410--70)
220
(120 - 320)
-220
(-320 --120)
Levy et al.(2005)
NERL
Census_2000
-110
(-240 - 30)
-130
(-200 - -60)
80
(50 - 120)
-60
(-90 - -30)
Levy et al.(2005)
WSU
ICLUS_A1
-660
(-1950 -630)
1220
(780- 1670)
-1640
(-2290 --1000)
-240
(-440 - -30)
Levy et al.(2005)
WSU
ICLUS_A2
-640
(-1800 - 530)
1080
(690- 1470)
-1500
(-2080 --910)
-220
(-410--30)
Levy et al.(2005)
WSU
ICLUS_BC
-490
(-1610-640)
1120
(710 - 1520)
-1370
(-1910 --830)
-230
(-420 - -40)
Levy et al.(2005)
WSU
Woods & Poole
-260
(-1280-770)
1110
(710-1510)
-1130
(-1560 --690)
-240
(-440 - -50)
Levy et al.(2005)
WSU
Census_2000
-10
(-380 - 360)
460
(300 - 620)
-390
(-550 - -240)
-80
(-130 --20)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
\ concentration-response function. Incidences are rounded to the nearest ten.
2
A-6
DRAFT: Do Not Quote or Cite
-------
Table A-2. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al. (2004)
CMU
ICLUS_A1
480
(80 - 890)
290
(100 -490)
250
(80-410)
-60
(-100 --20)
Bell et al. (2004)
CMU
ICLUS_A2
430
(70 - 800)
260
(90 - 440)
230
(80 - 380)
-60
(-90 - -20)
Bell et al. (2004)
CMU
ICLUS_BC
430
(70 - 780)
270
(90 - 450)
210
(70 - 350)
-50
(-90 - -20)
Bell et al. (2004)
CMU
Woods & Poole
350
(40 - 660)
260
(80 - 430)
150
(50 - 250)
-50
(-90 - -20)
Bell et al. (2004)
CMU
Census_2000
150
(30 - 270)
110
(40 - 180)
60
(20 - 100)
-10
(-20 - 0)
Bell et al. (2004)
GNM
ICLUS_A1
40
(-160-240)
-50
(-140 - 50)
30
(-30 - 100)
50
(0-100)
Bell et al. (2004)
GNM
ICLUS_A2
30
(-160-220)
-50
(-130 -40)
30
(-30 - 80)
50
(0-100)
Bell et al. (2004)
GNM
ICLUS_BC
20
(-170-210)
-60
(-150 - 30)
20
(-30 - 80)
50
(0-100)
Bell et al. (2004)
GNM
Woods & Poole
10
(-170 - 190)
-60
(-150 - 30)
10
(-30 - 50)
60
(10-110)
Bell et al. (2004)
GNM
Census_2000
-20
(-80 - 50)
-30
(-70 - 0)
0
(-10-20)
10
(0 - 30)
Bell et al. (2004)
Harvard
ICLUS_A1
240
(-20-510)
220
(70 - 360)
-20
(-90 - 50)
50
(0-100)
Bell et al. (2004)
Harvard
ICLUS_A2
220
(-20 - 470)
200
(60 - 330)
-20
(-80 - 40)
50
(0 - 90)
Bell et al. (2004)
Harvard
ICLUS_BC
230
(-10-470)
200
(60 - 340)
-20
(-80 - 40)
50
(0 - 90)
Bell et al. (2004)
Harvard
Woods & Poole
200
(-30 - 420)
190
(60 - 320)
-30
(-80 - 20)
40
(0 - 80)
Bell et al. (2004)
Harvard
Census_2000
80
(0-170)
80
(30 - 140)
0
(-20-10)
10
(0 - 20)
A-7
DRAFT: Do Not Quote or Cite
-------
Table A-2 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al. (2004)
lllinois-1
ICLUS_A1
570
(160 - 980)
260
(90 - 430)
250
(80-410)
60
(0-130)
Bell et al. (2004)
lllinois-1
ICLUS_A2
520
(150 - 890)
240
(80 - 400)
220
(70 - 370)
60
(0-120)
Bell et al. (2004)
lllinois-1
ICLUS_BC
510
(150 - 870)
250
(80-410)
210
(70 - 350)
50
(-10-110)
Bell et al. (2004)
lllinois-1
Woods & Poole
440
(120-760)
240
(80 - 400)
160
(50 - 260)
40
(-10-100)
Bell et al. (2004)
lllinois-1
Census_2000
170
(50 - 300)
110
(40 - 180)
60
(20 - 100)
10
(0 - 30)
Bell et al. (2004)
lllinois-2
ICLUS_A1
530
(160 - 900)
210
(70 - 360)
210
(60 - 350)
110
(30 - 190)
Bell et al. (2004)
lllinois-2
ICLUS_A2
480
(150 - 820)
200
(60 - 330)
180
(60-310)
100
(30 - 180)
Bell et al. (2004)
lllinois-2
ICLUS_BC
480
(150-810)
210
(70 - 350)
170
(50 - 290)
100
(30 - 170)
Bell et al. (2004)
lllinois-2
Woods & Poole
420
(130-710)
200
(60 - 340)
130
(40-210)
90
(20- 160)
Bell et al. (2004)
lllinois-2
Census_2000
160
(50 - 270)
90
(30 - 150)
50
(10-80)
20
(10-40)
Bell et al. (2004)
NERL
ICLUS_A1
10
(-210 -230)
-40
(-120 - 50)
100
(20 - 180)
-50
(-100-0)
Bell et al. (2004)
NERL
ICLUS_A2
10
(-190 -200)
-40
(-120-40)
90
(20- 160)
-50
(-90 - 0)
Bell et al. (2004)
NERL
ICLUS_BC
-10
(-210-190)
-50
(-130 -40)
80
(10-150)
-50
(-90 - 0)
Bell et al. (2004)
NERL
Woods & Poole
-50
(-230 - 120)
-50
(-130 - 30)
50
(0-100)
-50
(-100-0)
Bell et al. (2004)
NERL
Census_2000
-20
(-90 - 40)
-30
(-60-10)
20
(0 - 30)
-10
(-30 - 0)
A-8
DRAFT: Do Not Quote or Cite
-------
Table A-2 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al. (2004)
wsu
ICLUS_A1
-150
(-760 - 470)
280
(70 - 490)
-370
(-670 - -60)
-50
(-150 -40)
Bell et al. (2004)
wsu
ICLUS_A2
-140
(-700 -410)
240
(60 - 430)
-330
(-610--60)
-50
(-140-40)
Bell et al. (2004)
wsu
ICLUS_BC
-110
(-640 - 430)
250
(60 - 440)
-310
(-560 - -50)
-50
(-140-40)
Bell et al. (2004)
wsu
Woods & Poole
-60
(-540 - 430)
250
(60 - 440)
-250
(-450 - -50)
-60
(-150 -40)
Bell et al. (2004)
wsu
Census_2000
0
(-170- 170)
100
(30 - 180)
-80
(-150--10)
-20
(-40-10)
Ito et al. (2005)
CMU
ICLUS_A1
2180
(1090 - 3270)
1320
(790 - 1860)
1120
(670- 1560)
-260
(-370 --150)
Ito et al. (2005)
CMU
ICLUS_A2
1950
(970 - 2940)
1180
(710 - 1650)
1020
(610 - 1430)
-250
(-350 --140)
Ito et al. (2005)
CMU
ICLUS_BC
1920
(960 - 2890)
1210
(720- 1690)
950
(570 - 1330)
-230
(-330 --130)
Ito et al. (2005)
CMU
Woods & Poole
1570
(740-2410)
1160
(690- 1620)
660
(400 - 930)
-250
(-350 --140)
Ito et al. (2005)
CMU
Census_2000
690
(370 - 1010)
480
(290 - 670)
260
(160 - 370)
-50
(-80 - -30)
Ito et al. (2005)
GNM
ICLUS_A1
180
(-370 - 720)
-210
(-460 - 40)
150
(-10 - 320)
230
(100 - 360)
Ito et al. (2005)
GNM
ICLUS_A2
140
(-360 - 650)
-210
(-440 - 20)
130
(-20 - 280)
220
(100 - 350)
Ito et al. (2005)
GNM
ICLUS_BC
80
(-420 - 590)
-250
(-490 --10)
110
(-30 - 250)
220
(100 - 350)
Ito et al. (2005)
GNM
Woods & Poole
50
(-430 - 520)
-270
(-500 - -30)
60
(-50 - 160)
260
(120 - 390)
Ito et al. (2005)
GNM
Census_2000
-80
(-250 - 90)
-150
(-250 - -50)
0
(-30 - 40)
60
(30 - 90)
A-9
DRAFT: Do Not Quote or Cite
-------
Table A-2 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ito et al. (2005)
Harvard
ICLUS_A1
1090
(370 - 1810)
970
(570 - 1370)
-110
(-300 - 80)
230
(100 - 360)
Ito et al. (2005)
Harvard
ICLUS_A2
1000
(350 - 1660)
880
(520- 1240)
-90
(-260 - 80)
220
(100 - 340)
Ito et al. (2005)
Harvard
ICLUS_BC
1030
(380 - 1670)
900
(530 - 1270)
-80
(-230 - 80)
200
(90 - 320)
Ito et al. (2005)
Harvard
Woods & Poole
890
(280 - 1490)
840
(490 - 1190)
-130
(-270- 10)
180
(70 - 290)
Ito et al. (2005)
Harvard
Census_2000
380
(160-610)
360
(210-510)
-20
(-70 - 30)
40
(10-70)
Ito et al. (2005)
lllinois-1
ICLUS_A1
2560
(1470-3660)
1170
(700 - 1630)
1110
(660 - 1550)
290
(100 -470)
Ito et al. (2005)
lllinois-1
ICLUS_A2
2340
(1340 - 3340)
1070
(640 - 1500)
1010
(600 - 1410)
260
(90 - 430)
Ito et al. (2005)
lllinois-1
ICLUS_BC
2280
(1310 - 3260)
1120
(670- 1570)
930
(560 - 1300)
240
(80 - 390)
Ito et al. (2005)
lllinois-1
Woods & Poole
1970
(1110 -2840)
1090
(650 - 1520)
700
(420 - 990)
180
(40 - 330)
Ito et al. (2005)
lllinois-1
Census_2000
780
(450 - 1120)
480
(290 - 670)
260
(150 - 360)
50
(10-90)
Ito et al. (2005)
lllinois-2
ICLUS_A1
2390
(1400 - 3390)
950
(560 - 1340)
930
(540 - 1310)
510
(290 - 730)
Ito et al. (2005)
lllinois-2
ICLUS_A2
2180
(1270 - 3080)
890
(520 - 1250)
820
(480 - 1170)
470
(270 - 670)
Ito et al. (2005)
lllinois-2
ICLUS_BC
2160
(1260 - 3050)
940
(550 - 1320)
780
(460 - 1100)
440
(250 - 630)
Ito et al. (2005)
lllinois-2
Woods & Poole
1870
(1090 -2650)
900
(530 - 1270)
560
(330 - 800)
410
(230 - 590)
Ito et al. (2005)
lllinois-2
Census_2000
710
(420 - 1010)
410
(240 - 570)
200
(120 -290)
100
(60 - 150)
A-10
DRAFT: Do Not Quote or Cite
-------
Table A-2 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ito et al. (2005)
NERL
ICLUS_A1
50
(-540 - 650)
-160
(-390 - 70)
450
(230 - 680)
-230
(-370 --100)
Ito et al. (2005)
NERL
ICLUS_A2
20
(-510 - 560)
-160
(-380 - 50)
400
(200 -610)
-220
(-340 - -90)
Ito et al. (2005)
NERL
ICLUS_BC
-40
(-570 - 480)
-210
(-420- 10)
370
(190 - 560)
-210
(-330 - -90)
Ito et al. (2005)
NERL
Woods & Poole
-240
(-710 -230)
-240
(-460 - -20)
220
(90 - 350)
-220
(-350 - -90)
Ito et al. (2005)
NERL
Census_2000
-100
(-270 - 60)
-120
(-210--40)
80
(40- 120)
-60
(-90 - -30)
Ito et al. (2005)
WSU
ICLUS_A1
-650
(-2300 - 1000)
1240
(670 - 1810)
-1650
(-2470 - -830)
-240
(-500 - 20)
Ito et al. (2005)
WSU
ICLUS_A2
-630
(-2120-870)
1090
(590 - 1590)
-1500
(-2240 - -750)
-220
(-470 - 20)
Ito et al. (2005)
WSU
ICLUS_BC
-480
(-1920 - 960)
1130
(610- 1640)
-1370
(-2050 - -680)
-230
(-470- 10)
Ito et al. (2005)
WSU
Woods & Poole
-240
(-1540 - 1060)
1120
(610 - 1630)
-1110
(-1650 --570)
-250
(-500 - 0)
Ito et al. (2005)
WSU
Census_2000
0
(-460 - 460)
450
(250 - 660)
-380
(-570 --190)
-80
(-140--10)
Schwartz (2005)
CMU
ICLUS_A1
730
(100 - 1370)
450
(140-760)
370
(120 -630)
-90
(-150 --20)
Schwartz (2005)
CMU
ICLUS_A2
660
(80 - 1230)
400
(120-670)
340
(110 - 580)
-80
(-150 --20)
Schwartz (2005)
CMU
ICLUS_BC
650
(90-1210)
410
(130 -690)
320
(100 - 540)
-80
(-140--20)
Schwartz (2005)
CMU
Woods & Poole
540
(50 - 1030)
390
(120-670)
230
(70 - 380)
-80
(-150 --20)
Schwartz (2005)
CMU
Census_2000
240
(50 - 430)
160
(50 - 280)
90
(30 - 150)
-20
(-30 - 0)
A-ll
DRAFT: Do Not Quote or Cite
-------
Table A-2 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (2005)
GNM
ICLUS_A1
60
(-260 - 380)
-70
(-220 - 80)
50
(-40 - 150)
80
(0-160)
Schwartz (2005)
GNM
ICLUS_A2
50
(-250 - 340)
-70
(-210-60)
40
(-40 - 130)
80
(0-150)
Schwartz (2005)
GNM
ICLUS_BC
30
(-270 - 320)
-90
(-230 - 50)
40
(-40- 120)
80
(0-150)
Schwartz (2005)
GNM
Woods & Poole
20
(-260 - 300)
-90
(-230 - 50)
20
(-40 - 80)
90
(10-170)
Schwartz (2005)
GNM
Census_2000
-30
(-130 -70)
-50
(-110-10)
0
(-20 - 20)
20
(0 - 40)
Schwartz (2005)
Harvard
ICLUS_A1
370
(-50 - 790)
330
(100 - 560)
-40
(-150 -70)
80
(0-150)
Schwartz (2005)
Harvard
ICLUS_A2
340
(-40 - 720)
300
(90-510)
-30
(-130 -70)
70
(0-150)
Schwartz (2005)
Harvard
ICLUS_BC
350
(-30 - 720)
300
(90 - 520)
-30
(-120-70)
70
(0-140)
Schwartz (2005)
Harvard
Woods & Poole
300
(-50 - 660)
290
(80 - 490)
-50
(-130 -40)
60
(-10-130)
Schwartz (2005)
Harvard
Census_2000
130
(0 - 270)
120
(40-210)
-10
(-40 - 20)
10
(0 - 30)
Schwartz (2005)
lllinois-1
ICLUS_A1
860
(230 - 1500)
390
(120-670)
370
(110 -630)
100
(-10 -200)
Schwartz (2005)
lllinois-1
ICLUS_A2
790
(210 - 1370)
360
(110-610)
340
(100 - 570)
90
(-10-180)
Schwartz (2005)
lllinois-1
ICLUS_BC
770
(200 - 1340)
380
(120-640)
310
(100 - 530)
80
(-10-170)
Schwartz (2005)
lllinois-1
Woods & Poole
670
(160 - 1180)
370
(120 -630)
240
(70-410)
60
(-20 - 150)
Schwartz (2005)
lllinois-1
Census_2000
270
(70 - 470)
160
(50 - 280)
90
(30 - 150)
20
(-10-40)
A-12
DRAFT: Do Not Quote or Cite
-------
Table A-2 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (2005)
lllinois-2
ICLUS_A1
810
(230-1380)
320
(100 - 550)
310
(90 - 530)
170
(50 - 300)
Schwartz (2005)
lllinois-2
ICLUS_A2
730
(210-1260)
300
(90-510)
280
(80 - 470)
160
(40 - 270)
Schwartz (2005)
lllinois-2
ICLUS_BC
730
(210-1250)
320
(90 - 540)
260
(70 - 450)
150
(40 - 260)
Schwartz (2005)
lllinois-2
Woods & Poole
640
(180-1100)
310
(90 - 520)
190
(50 - 330)
140
(30 - 240)
Schwartz (2005)
lllinois-2
Census_2000
250
(70 - 420)
140
(40 - 240)
70
(20- 120)
40
(10-60)
Schwartz (2005)
NERL
ICLUS_A1
20
(-330 - 360)
-60
(-190-80)
150
(20 - 280)
-80
(-160 - 0)
Schwartz (2005)
NERL
ICLUS_A2
10
(-310-320)
-60
(-180-70)
140
(20 - 250)
-70
(-150 - 0)
Schwartz (2005)
NERL
ICLUS_BC
-20
(-320 - 290)
-70
(-200 - 60)
130
(20 - 230)
-70
(-140 - 0)
Schwartz (2005)
NERL
Woods & Poole
-80
(-360 - 200)
-80
(-210 - 40)
80
(0-150)
-70
(-150 - 0)
Schwartz (2005)
NERL
Census_2000
-40
(-140 - 60)
-40
(-100-10)
30
(0 - 50)
-20
(-40 - 0)
Schwartz (2005)
WSU
ICLUS_A1
-220
(-1180 - 740)
420
(90 - 750)
-560
(-1030 - -80)
-80
(-240 - 70)
Schwartz (2005)
WSU
ICLUS_A2
-210
(-1080-660)
370
(80 - 660)
-510
(-940 - -70)
-80
(-220 - 70)
Schwartz (2005)
WSU
ICLUS_BC
-160
(-1000 - 680)
380
(80 - 680)
-460
(-860 - -60)
-80
(-220 - 60)
Schwartz (2005)
WSU
Woods & Poole
-90
(-850 - 680)
380
(80 - 680)
-380
(-700 - -60)
-80
(-230 - 60)
Schwartz (2005)
WSU
Census_2000
0
(-270 - 270)
160
(40 - 280)
-130
(-240 - -20)
-30
(-60-10)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
\ concentration-response function. Incidences are rounded to the nearest ten.
A-13
DRAFT: Do Not Quote or Cite
-------
Table A-3. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Ages 65 and Up) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be June, July, and August*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (1995); Schwartz
(1994a); Moolgavkar et al.
(1997); Schwartz (1994b);
Moolgavkar et al. (1997)
CMU
ICLUS_A1
5190
(-400 - 13460)
3220
(160- 8160)
2510
(130-6460)
-530
(-1340 --10)
CMU
ICLUS_A2
4630
(-390 - 12010)
2850
(140-7220)
2280
(120 - 5870)
-490
(-1260--10)
CMU
ICLUS_BC
4580
(-350- 11860)
2920
(150-7400)
2130
(110 - 5490)
-460
(-1180--10)
CMU
Woods & Poole
3880
(-430- 10150)
2850
(140-7280)
1520
(80 - 3950)
-490
(-1240--10)
CMU
Census_2000
1670
(-60 - 4360)
1200
(60 - 3090)
580
(30-1510)
-110
(-290 - 0)
GNM
ICLUS_A1
300
(-1760 -2390)
-570
(-1690- 550)
390
(-310 - 1070)
470
(-90 - 970)
GNM
ICLUS_A2
220
(-1680 -2130)
-570
(-1610-470)
330
(-290 - 930)
460
(-20 - 940)
GNM
ICLUS_BC
80
(-1840 - 1970)
-670
(-1760-440)
290
(-290 - 850)
460
(-80 - 970)
GNM
Woods & Poole
10
(-2100-2410)
-710
(-2140-400)
170
(-270 - 730)
550
(-70 - 1590)
GNM
Census_2000
-250
(-1120 -490)
-400
(-1070- 120)
20
(-130 -220)
130
(-20 - 350)
Harvard
ICLUS_A1
2530
(-920 - 6560)
2440
(90 - 6040)
-370
(-1370-460)
460
(-90 - 1200)
Harvard
ICLUS_A2
2320
(-820 - 5990)
2200
(80 - 5430)
-320
(-1220-420)
440
(-80 - 1140)
Harvard
ICLUS_BC
2410
(-740-6180)
2270
(90 - 5600)
-270
(-1100-410)
410
(-80 - 1070)
Harvard
Woods & Poole
2130
(-830 - 5540)
2180
(70 - 5410)
-420
(-1180 -250)
380
(-110 - 990)
Harvard
Census_2000
940
(-220 - 2440)
950
(30 - 2370)
-90
(-360 - 120)
90
(-30 - 230)
A-14
DRAFT: Do Not Quote or Cite
-------
Table A-3 cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Ages 65 and Up)
Attributable to Climate Change in 2050 During the 03 Season, Taken to be June, July, and August*
Climate
Change Model
Population
Estimated 03-
Related Incidence Attributable to Climate Change in 2050
Study**
Projection to
2050
National
Northeast
Southeast
West
lllinois-1
ICLUS_A1
6050
(90 - 15300)
2910
(150-7200)
2580
(120-6630)
560
(-180 - 1470)
lllinois-1
ICLUS_A2
5500
(80 - 13910)
2660
(140-6590)
2330
(110 - 5980)
510
(-170 - 1340)
lllinois-1
ICLUS_BC
5410
(70- 13660)
2780
(150-6880)
2150
(100 - 5540)
470
(-170- 1240)
lllinois-1
Woods & Poole
4850
(0- 12320)
2780
(140-6920)
1710
(80 -4410)
360
(-220 - 990)
lllinois-1
Census_2000
1940
(40 - 4950)
1240
(60-3110)
600
(30 - 1560)
100
(-50 - 280)
lllinois-2
ICLUS_A1
5650
(180 - 14170)
2450
(100- 5960)
2180
(60 - 5610)
1020
(10 -2600)
Schwartz (1995); Schwartz
lllinois-2
ICLUS_A2
5120
(160- 12840)
2270
(100- 5520)
1920
(50 - 4940)
930
(10 -2380)
(1994a); Moolgavkar et al.
(1997); Schwartz (1994b);
lllinois-2
ICLUS_BC
5110
(160- 12790)
2410
(110- 5840)
1820
(50 - 4690)
890
(10-2260)
Moolgavkar et al. (1997)
lllinois-2
Woods & Poole
4630
(140-11610)
2400
(110- 5860)
1410
(40 - 3640)
830
(-10-2110)
lllinois-2
Census_2000
1780
(60 - 4490)
1090
(50 -2710)
470
(10 - 1230)
220
(0 - 560)
NERL
ICLUS_A1
70
(-2150 -2340)
-440
(-1490- 550)
950
(-130-2420)
-450
(-1120 - 100)
NERL
ICLUS_A2
10
(-2040 - 2060)
-470
(-1480- 480)
930
(10 -2350)
-450
(-1140 - 30)
NERL
ICLUS_BC
-140
(-2770-2610)
-600
(-2250 - 740)
940
(10 -2880)
-480
(-1520 -40)
NERL
Woods & Poole
-620
(-2960- 1350)
-670
(-1820 - 350)
510
(-90 - 1420)
-460
(-1250 - 30)
NERL
Census_2000
-310
(-1140-410)
-350
(-940 - 110)
170
(-40 - 470)
-130
(-340 - 20)
A-15
DRAFT: Do Not Quote or Cite
-------
Table A-3 cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Ages 65 and Up)
Attributable to Climate Change in 2050 During the 03 Season, Taken to be June, July, and August*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (1995); Schwartz
(1994a); Moolgavkar et al.
(1997); Schwartz (1994b);
Moolgavkar et al. (1997)
wsu
ICLUS_A1
-1480
(-8380 - 5060)
3050
(-110-7740)
-4030
(-10240- 350)
-500
(-1610 -490)
wsu
ICLUS_A2
-1420
(-7750 - 4460)
2670
(-110-6770)
-3630
(-9230 - 320)
-460
(-1490-460)
wsu
ICLUS_BC
-1050
(-6720 - 4590)
2770
(-100- 7000)
-3340
(-8460 - 300)
-490
(-1500 -430)
wsu
Woods & Poole
-650
(-6570 - 4570)
2810
(-100-7440)
-2930
(-7760- 180)
-540
(-1590 -440)
wsu
Census_2000
30
(-1810 - 1920)
990
(-30-2170)
-820
(-1870-70)
-140
(-410-100)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest ten.
\ "These studies were pooled to estimate respiratory hospital admissions for ages 65 and up.
2
3
4
A-16
DRAFT: Do Not Quote or Cite
-------
Table A-4. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Age < 1) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Burnett et al. (2001)
CMU
ICLUS_A1
1250
(290 - 2220)
830
(350- 1310)
630
(270 - 1000)
-210
(-330 - -80)
CMU
ICLUS_A2
2060
(450 - 3700)
1370
(580-2160)
1070
(460- 1690)
-370
(-590 --150)
CMU
ICLUS_BC
1550
(340 - 2770)
1040
(440- 1640)
780
(340- 1240)
-270
(-440--110)
CMU
Woods & Poole
1830
(390 - 3280)
1110
(470- 1750)
1050
(450- 1670)
-340
(-530 --140)
CMU
Census_2000
1290
(330 - 2260)
880
(370- 1390)
600
(250 - 940)
-190
(-300 - -70)
GNM
ICLUS_A1
190
(-350 - 730)
-160
(-380 - 70)
110
(-30 - 260)
230
(60 - 400)
GNM
ICLUS_A2
310
(-630 - 1250)
-290
(-680- 100)
180
(-70 - 430)
420
(120-720)
GNM
ICLUS_BC
200
(-500 - 910)
-230
(-530 - 60)
130
(-60-310)
310
(90 - 540)
GNM
Woods & Poole
170
(-630 - 980)
-280
(-610-40)
130
(-100 - 360)
330
(80 - 570)
GNM
Census_2000
10
(-540 - 570)
-250
(-520- 10)
50
(-80 - 180)
220
(60 - 380)
Harvard
ICLUS_A1
710
(50- 1390)
620
(260 - 980)
-80
(-250 - 80)
180
(40 - 320)
Harvard
ICLUS_A2
1230
(100-2380)
1040
(430- 1660)
-130
(-400 - 140)
320
(70 - 580)
Harvard
ICLUS_BC
940
(90- 1790)
790
(330 - 1260)
-90
(-290 - 110)
240
(50 - 430)
Harvard
Woods & Poole
1100
(110 -2090)
890
(370- 1410)
-60
(-310-190)
270
(40 - 490)
Harvard
Census_2000
820
(160 - 1490)
680
(280- 1090)
-20
(-140-110)
160
(30 - 290)
A-17
DRAFT: Do Not Quote or Cite
-------
Table A-4 cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Age < 1) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Burnett et al. (2001)
lllinois-1
ICLUS_A1
1570
(580 - 2560)
740
(320- 1170)
640
(270- 1020)
180
(-10 - 380)
lllinois-1
ICLUS_A2
2650
(980 - 4350)
1270
(540-2010)
1070
(450 - 1690)
310
(-20 - 640)
lllinois-1
ICLUS_BC
1990
(730 - 3260)
970
(420- 1540)
780
(330 - 1240)
230
(-10 -480)
lllinois-1
Woods & Poole
2350
(860 - 3860)
1100
(470- 1740)
990
(420- 1570)
260
(-30 - 560)
lllinois-1
Census_2000
1600
(600-2610)
880
(370- 1380)
560
(240 - 880)
160
(-10 - 340)
lllinois-2
ICLUS_A1
1610
(650 - 2580)
630
(260- 1000)
600
(240 - 960)
380
(150 -620)
lllinois-2
ICLUS_A2
2740
(1110 -4380)
1090
(460- 1740)
990
(400 - 1580)
650
(250 - 1060)
lllinois-2
ICLUS_BC
2060
(840 - 3300)
840
(350- 1340)
730
(300 - 1170)
490
(190 -790)
lllinois-2
Woods & Poole
2350
(940 - 3770)
950
(400- 1520)
820
(330 - 1320)
570
(210 - 930)
lllinois-2
Census_2000
1610
(650 - 2580)
760
(320- 1220)
500
(200 - 800)
350
(130 - 560)
NERL
ICLUS_A1
-40
(-600 - 520)
-130
(-340 - 80)
270
(70 - 470)
-180
(-330 - -20)
NERL
ICLUS_A2
-100
(-1050- 850)
-250
(-600 - 110)
450
(130 -780)
-310
(-570 - -40)
NERL
ICLUS_BC
-100
(-810-610)
-200
(-470 - 70)
330
(90 - 570)
-230
(-430 - -30)
NERL
Woods & Poole
-100
(-920 - 720)
-230
(-520 - 60)
410
(120 -700)
-280
(-510--50)
NERL
Census_2000
-160
(-700 - 390)
-220
(-450 - 20)
230
(70 - 400)
-180
(-320 - -30)
A-18
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table A-4 cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Age < 1) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Burnett et al. (2001)
wsu
ICLUS_A1
-430
(-1930 - 1100)
790
(280- 1300)
-1010
(-1690--310)
-210
(-520 - 100)
wsu
ICLUS_A2
-770
(-3300 - 1800)
1290
(450-2140)
-1690
(-2830 - -520)
-370
(-910-180)
wsu
ICLUS_BC
-540
(-2410 - 1380)
980
(340 - 1630)
-1230
(-2060 - -370)
-290
(-700 - 130)
wsu
Woods & Poole
-510
(-2610- 1640)
990
(340 - 1650)
-1320
(-2320 - -290)
-180
(-620 - 280)
wsu
Census_2000
-190
(-1570- 1220)
820
(280- 1360)
-800
(-1360--230)
-210
(-500 - 90)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest ten.
A-19
DRAFT: Do Not Quote or Cite
-------
Table A-5. Estimated National and Regional 03-Related Incidence of Emergency Room Visits for Asthma Attributable to Climate
Change in 2050 During the 03 Season, Taken to be June, July, and August**
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Peel et al. (2005); Wilson et
al. (2005)
CMU
ICLUS_A1
1230
(-3370 - 4370)
850
(-2140-2910)
450
(-1130- 1540)
-80
(-280 - 220)
CMU
ICLUS_A2
1500
(-4120 - 5340)
1030
(-2580 - 3510)
570
(-1420- 1940)
-100
(-350 - 270)
CMU
ICLUS_BC
1300
(-3570 - 4620)
910
(-2280 - 3100)
480
(-1190- 1620)
-90
(-300 - 230)
CMU
Woods & Poole
1490
(-4150 - 5350)
1010
(-2530 - 3440)
600
(-1490-2030)
-110
(-390 - 300)
CMU
Census_2000
1130
(-3050 - 3990)
830
(-2080 - 2830)
360
(-900- 1220)
-60
(-200 - 160)
GNM
ICLUS_A1
-80
(-1290 - 1130)
-240
(-1180 - 830)
80
(-320 - 450)
90
(-220 - 320)
GNM
ICLUS_A2
-130
(-1700 - 1450)
-330
(-1480 - 1070)
90
(-390 - 550)
110
(-280 - 400)
GNM
ICLUS_BC
-130
(-1520 - 1280)
-300
(-1300 - 950)
70
(-320 - 450)
100
(-250 - 350)
GNM
Woods & Poole
-180
(-1850 - 1530)
-360
(-1470- 1110)
70
(-390 - 520)
110
(-300 - 430)
GNM
Census_2000
-220
(-1570 - 1190)
-320
(-1250- 930)
30
(-210-260)
70
(-180 -250)
Harvard
ICLUS_A1
700
(-2300 - 2720)
680
(-1740-2340)
-60
(-440-410)
70
(-230 - 280)
Harvard
ICLUS_A2
870
(-2840 - 3380)
850
(-2160-2910)
-70
(-530 - 500)
90
(-290 - 350)
Harvard
ICLUS_BC
770
(-2480 - 2970)
750
(-1900-2560)
-50
(-440-410)
80
(-250 - 300)
Harvard
Woods & Poole
900
(-2920 - 3490)
860
(-2190-2950)
-50
(-500 - 500)
100
(-310 - 370)
Harvard
Census_2000
730
(-1870 -2690)
690
(-1530-2370)
-20
(-250 - 220)
50
(-150-210)
A-20
DRAFT: Do Not Quote or Cite
-------
Table A-5 cont'd. Estimated National and Regional 03-Related Incidence of Emergency Room Visits for Asthma Attributable to Climate
Change in 2050 During the 03 Season, Taken to be June, July, and August**
Climate
Change Model
Population
Estimated 03-
Related Incidence Attributable to Climate Change in 2050
Study**
Projection to
2050
National
Northeast
Southeast
West
lllinois-1
ICLUS_A1
1370
(-3070 - 4740)
830
(-1820-2840)
460
(-1000 - 1560)
80
(-260 - 350)
lllinois-1
ICLUS_A2
1710
(-3840 - 5920)
1050
(-2290 - 3580)
570
(-1240 - 1930)
100
(-310 -430)
lllinois-1
ICLUS_BC
1490
(-3840 - 5140)
930
(-2330- 3160)
470
(-1190-1610)
80
(-320 - 380)
lllinois-1
Woods & Poole
1760
(-4560 - 6090)
1080
(-2710- 3690)
580
(-1450 - 1970)
100
(-400 - 480)
lllinois-1
Census_2000
1290
(-3310 -4450)
890
(-2230 - 3030)
340
(-860 - 1160)
60
(-230 - 270)
lllinois-2
ICLUS_A1
1330
(-3390 - 4580)
760
(-1920-2610)
410
(-1060 - 1430)
160
(-410 - 540)
lllinois-2
ICLUS_A2
1670
(-4260 - 5750)
980
(-2450 - 3330)
510
(-1310 - 1760)
190
(-500 - 660)
Peel et al. (2005); Wilson et
al. (2005)
lllinois-2
ICLUS_BC
1460
(-3700 - 5000)
860
(-2170-2950)
430
(-1100 - 1480)
160
(-440 - 580)
lllinois-2
Woods & Poole
1720
(-4380 - 5910)
1010
(-2550 - 3460)
500
(-1290 - 1730)
210
(-540 - 720)
lllinois-2
Census_2000
1240
(-3150 -4260)
830
(-2090 - 2840)
290
(-750 - 1010)
120
(-310-410)
NERL
ICLUS_A1
-90
(-1330 - 1160)
-210
(-1080 -750)
190
(-490 - 700)
-70
(-280 -210)
NERL
ICLUS_A2
-130
(-1700 - 1450)
-280
(-1330 - 950)
240
(-600 - 870)
-90
(-340 - 250)
NERL
ICLUS_BC
-130
(-1520- 1270)
-260
(-1170 - 850)
200
(-500 - 720)
-80
(-300 - 220)
NERL
Woods & Poole
-170
(-1810 - 1500)
-310
(-1310- 970)
240
(-600 - 860)
-100
(-380 - 280)
NERL
Census_2000
-200
(-1470 - 1130)
-280
(-1100- 820)
140
(-340 - 490)
-60
(-230 - 170)
A-21
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Table A-5 cont'd. Estimated National and Regional 03-Related Incidence of Emergency Room Visits for Asthma Attributable to Climate
Change in 2050 During the 03 Season, Taken to be June, July, and August**
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Peel et al. (2005); Wilson et
al. (2005)
wsu
ICLUS_A1
0
(-1610-1610)
550
(-190- 1270)
-490
(-1190-230)
-60
(-230 - 120)
wsu
ICLUS_A2
-60
(-3800 - 3620)
1250
(-1260-3420)
-1170
(-3270- 1250)
-140
(-640-210)
wsu
ICLUS_BC
0
(-1680 - 1700)
570
(-200- 1490)
-500
(-1290-230)
-60
(-250 - 120)
wsu
Woods & Poole
-60
(-3760 - 3560)
1170
(-1410 - 3250)
-1120
(-3200- 1460)
-110
(-630 - 270)
wsu
Census_2000
190
(-3090 - 3000)
740
(-2120-2760)
-470
(-1820- 1510)
-70
(-440 - 370)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest ten.
"These studies were pooled to estimate respiratory emergency room visits.
A-22
DRAFT: Do Not Quote or Cite
-------
Table A-6. Estimated National and Regional 03-Related Incidence of Minor Restricted Activity Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ostro and Rothschild (1989)
CMU
ICLUS_A1
1637000
(411000 -2868000)
1107000
(456000 - 1762000)
749000
(309000-1191000)
-219000
(-354000 - -85000)
CMU
ICLUS_A2
1688000
(408000 - 2973000)
1142000
(470000- 1817000)
785000
(324000-1248000)
-239000
(-386000 - -92000)
CMU
ICLUS_BC
1582000
(387000 - 2783000)
1084000
(446000-1725000)
719000
(297000 - 1142000)
-220000
(-356000 - -84000)
CMU
Woods & Poole
1818000
(390000 - 3251000)
1230000
(505000-1958000)
886000
(366000-1409000)
-298000
(-481000 --115000)
CMU
Census_2000
1436000
(406000 -2471000)
1018000
(418000-1620000)
571000
(236000 - 908000)
-153000
(-248000 - -57000)
GNM
ICLUS_A1
120000
(-545000 - 787000)
-224000
(-541000 - 94000)
120000
(-54000 - 295000)
224000
(50000 - 398000)
GNM
ICLUS_A2
108000
(-604000 - 823000)
-260000
(-600000 - 80000)
116000
(-65000 - 297000)
252000
(60000 - 445000)
GNM
ICLUS_BC
73000
(-597000 - 745000)
-262000
(-586000 - 63000)
99000
(-66000 - 265000)
236000
(55000 -417000)
GNM
Woods & Poole
58000
(-763000 - 881000)
-343000
(-723000 - 37000)
106000
(-103000 - 314000)
295000
(63000 - 529000)
GNM
Census_2000
-78000
(-651000 -497000)
-291000
(-602000 - 20000)
39000
(-85000 - 163000)
175000
(36000 - 314000)
Harvard
ICLUS_A1
926000
(74000- 1781000)
831000
(331000 - 1332000)
-90000
(-289000 - 109000)
186000
(31000 - 340000)
Harvard
ICLUS_A2
990000
(95000-1887000)
872000
(347000- 1398000)
-87000
(-290000 - 117000)
205000
(38000 - 372000)
Harvard
ICLUS_BC
941000
(103000 - 1782000)
828000
(330000-1329000)
-74000
(-259000 - 111000)
187000
(32000 - 343000)
Harvard
Woods & Poole
1131000
(113000 -2152000)
972000
(387000- 1560000)
-78000
(-305000 - 149000)
237000
(31000 -443000)
Harvard
Census_2000
872000
(156000 - 1591000)
775000
(306000-1246000)
-26000
(-159000 - 108000)
123000
(9000 - 237000)
A-23
DRAFT: Do Not Quote or Cite
-------
Table A-6 cont'd. Estimated National and Regional 03-Related Incidence of Minor Restricted Activity Days Attributable to Climate Change in
2050 During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ostro and Rothschild (1989)
lllinois-1
ICLUS_A1
1959000
(715000 - 3209000)
1007000
(416000 - 1601000)
751000
(305000-1198000)
201000
(-7000 - 409000)
lllinois-1
ICLUS_A2
2063000
(749000 - 3382000)
1074000
(443000- 1707000)
778000
(317000-1242000)
210000
(-12000 -433000)
lllinois-1
ICLUS_BC
1934000
(702000 - 3170000)
1029000
(425000 - 1636000)
711000
(290000- 1133000)
194000
(-12000 -401000)
lllinois-1
Woods & Poole
2333000
(829000 - 3843000)
1227000
(506000-1950000)
862000
(352000- 1374000)
244000
(-29000 - 519000)
lllinois-1
Census_2000
1681000
(622000 - 2745000)
1003000
(414000 - 1595000)
541000
(221000- 862000)
137000
(-14000 -288000)
lllinois-2
ICLUS_A1
1941000
(761000 - 3125000)
862000
(349000 - 1379000)
676000
(265000-1088000)
403000
(148000- 659000)
lllinois-2
ICLUS_A2
2049000
(804000 - 3300000)
931000
(377000 - 1488000)
690000
(270000-1112000)
428000
(157000- 699000)
lllinois-2
ICLUS_BC
1927000
(757000 - 3103000)
895000
(362000 - 1430000)
635000
(249000- 1023000)
397000
(145000- 650000)
lllinois-2
Woods & Poole
2362000
(924000 - 3806000)
1074000
(435000 - 1715000)
766000
(300000-1235000)
522000
(189000- 856000)
lllinois-2
Census_2000
1612000
(632000 - 2596000)
868000
(352000 - 1386000)
457000
(178000 -737000)
287000
(102000- 473000)
NERL
ICLUS_A1
-76000
(-766000 -617000)
-195000
(-488000 - 98000)
310000
(77000 - 543000)
-190000
(-356000 - -25000)
NERL
ICLUS_A2
-109000
(-837000 - 620000)
-226000
(-536000 - 85000)
321000
(83000 - 559000)
-204000
(-384000 - -24000)
NERL
ICLUS_BC
-130000
(-810000 - 551000)
-228000
(-523000 - 68000)
290000
(75000 - 506000)
-193000
(-361000--24000)
NERL
Woods & Poole
-202000
(-1022000 -620000)
-290000
(-630000 - 51000)
346000
(88000 - 604000)
-258000
(-481000 --36000)
NERL
Census_2000
-213000
(-781000 - 356000)
-257000
(-538000 - 25000)
209000
(57000 - 361000)
-164000
(-299000 - -30000)
A-24
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
Table A-6 cont'd. Estimated National and Regional 03-Related Incidence of Minor Restricted Activity Days Attributable to Climate Change in
2050 During the 03 Season, Taken to be June, July, and August*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ostro and Rothschild (1989)
wsu
ICLUS_A1
-333000
(-2213000 - 1560000)
1057000
(353000-1765000)
-1176000
(-2020000 - -325000)
-214000
(-546000 - 119000)
wsu
ICLUS_A2
-375000
(-2336000 - 1598000)
1080000
(355000-1807000)
-1221000
(-2095000 - -339000)
-234000
(-596000 - 130000)
wsu
ICLUS_BC
-301000
(-2119000 - 1530000)
1029000
(342000-1719000)
-1103000
(-1895000--304000)
-226000
(-566000 - 115000)
wsu
Woods & Poole
-460000
(-2622000- 1717000)
1121000
(366000 - 1881000)
-1345000
(-2317000--364000)
-236000
(-671000 -200000)
wsu
Census_2000
2000
(-1476000 - 1490000)
972000
(328000-1619000)
-773000
(-1346000--195000)
-197000
(-458000 - 65000)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest thousand.
A-25
DRAFT: Do Not Quote or Cite
-------
Table A-7. Estimated National and Regional 03-Related Incidence of School Loss Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Chen et al. (2000); Gilliland
et al. (2001)
CMU
ICLUS_A1
522000
(96000 - 1183000)
356000
(126000 - 806000)
241000
(86000 - 544000)
-75000
(-167000 --25000)
CMU
ICLUS_A2
745000
(128000 - 1689000)
509000
(180000- 1151000)
351000
(125000-793000)
-115000
(-255000 - -38000)
CMU
ICLUS_BC
599000
(104000 - 1356000)
414000
(147000 - 935000)
276000
(98000 - 623000)
-91000
(-202000 - -30000)
CMU
Woods & Poole
679000
(100000 - 1540000)
450000
(159000- 1018000)
347000
(124000-785000)
-118000
(-263000 - -39000)
CMU
Census_2000
545000
(113000 - 1235000)
389000
(138000 - 880000)
222000
(79000 - 502000)
-66000
(-147000 --21000)
GNM
ICLUS_A1
50000
(-171000 -265000)
-71000
(-168000 - 34000)
42000
(-18000 - 97000)
79000
(15000 - 138000)
GNM
ICLUS_A2
67000
(-264000 - 389000)
-115000
(-260000 -41000)
57000
(-29000 - 137000)
125000
(25000 -219000)
GNM
ICLUS_BC
44000
(-222000 - 304000)
-98000
(-216000 -29000)
42000
(-25000 - 106000)
100000
(20000 - 175000)
GNM
Woods & Poole
35000
(-277000 - 344000)
-123000
(-257000 - 22000)
45000
(-39000 - 124000)
114000
(19000 -200000)
GNM
Census_2000
-29000
(-282000 -201000)
-123000
(-382000- 10000)
18000
(-33000 - 87000)
76000
(12000 -224000)
Harvard
ICLUS_A1
299000
(4000 - 676000)
267000
(91000- 602000)
-32000
(-95000 - 35000)
64000
(8000 - 144000)
Harvard
ICLUS_A2
445000
(14000- 1006000)
389000
(133000- 876000)
-43000
(-132000 - 52000)
99000
(14000 -224000)
Harvard
ICLUS_BC
362000
(16000 - 818000)
316000
(108000-711000)
-32000
(-102000 -42000)
78000
(10000 - 176000)
Harvard
Woods & Poole
422000
(11000 - 954000)
360000
(123000- 812000)
-31000
(-120000 -61000)
93000
(8000 -210000)
Harvard
Census_2000
347000
(47000 - 783000)
302000
(102000-681000)
-9000
(-59000 - 42000)
54000
(3000 - 121000)
A-26
DRAFT: Do Not Quote or Cite
-------
Table A-7 cont'd. Estimated National and Regional 03-Related Incidence of School Loss Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Chen et al. (2000); Gilliland
et al. (2001)
lllinois-1
ICLUS_A1
633000
(192000 - 1428000)
323000
(115000 -728000)
244000
(85000 - 550000)
66000
(-7000 - 150000)
lllinois-1
ICLUS_A2
925000
(279000 - 2087000)
478000
(170000 - 1078000)
350000
(122000-790000)
97000
(-13000 -219000)
lllinois-1
ICLUS_BC
743000
(224000- 1677000)
391000
(139000 - 882000)
275000
(96000- 621000)
77000
(-11000 - 175000)
lllinois-1
Woods & Poole
880000
(260000 - 1988000)
451000
(160000 - 1018000)
336000
(118000-759000)
93000
(-18000 -211000)
lllinois-1
Census_2000
659000
(205000 - 1488000)
390000
(139000 - 881000)
209000
(74000 - 473000)
59000
(-7000 - 134000)
lllinois-2
ICLUS_A1
638000
(214000 - 1441000)
276000
(96000 - 625000)
226000
(76000- 510000)
136000
(42000 - 306000)
lllinois-2
ICLUS_A2
937000
(314000 -2117000)
415000
(144000 - 938000)
321000
(108000-726000)
201000
(62000 - 454000)
lllinois-2
ICLUS_BC
755000
(253000 - 1706000)
340000
(118000 -769000)
254000
(85000 - 574000)
161000
(50000 - 363000)
lllinois-2
Woods & Poole
893000
(297000 -2019000)
397000
(138000 - 899000)
295000
(97000 - 665000)
201000
(61000 -454000)
lllinois-2
Census_2000
650000
(218000 - 1468000)
343000
(120000 -776000)
184000
(61000-416000)
122000
(37000 - 276000)
NERL
ICLUS_A1
-25000
(-274000 - 202000)
-65000
(-230000 - 34000)
106000
(20000 - 307000)
-66000
(-198000--4000)
NERL
ICLUS_A2
-50000
(-378000 - 284000)
-104000
(-318000 -42000)
151000
(30000 - 409000)
-98000
(-272000 - -5000)
NERL
ICLUS_BC
-50000
(-312000 -219000)
-88000
(-237000 - 30000)
117000
(23000 - 294000)
-79000
(-201000--4000)
NERL
Woods & Poole
-67000
(-372000 - 249000)
-107000
(-284000 - 26000)
139000
(26000 - 350000)
-99000
(-250000 - -6000)
NERL
Census_2000
-84000
(-301000 - 146000)
-105000
(-276000 - 12000)
86000
(20000 -216000)
-64000
(-163000 --5000)
A-27
DRAFT: Do Not Quote or Cite
-------
Table A-7 cont'd. Estimated National and Regional 03-Related Incidence of School Loss Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be June, July, and August*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Chen et al. (2000); Gilliland
et al. (2001)
wsu
ICLUS_A1
-134000
(-1049000 - 555000)
363000
(96000- 1079000)
-416000
(-1212000--89000)
-81000
(-314000 -42000)
wsu
ICLUS_A2
-212000
(-1560000 -767000)
510000
(133000- 1528000)
-599000
(-1751000--130000)
-124000
(-481000 -64000)
wsu
ICLUS_BC
-153000
(-1206000 -647000)
419000
(110000- 1245000)
-468000
(-1362000--101000)
-103000
(-393000 - 49000)
wsu
Woods & Poole
-197000
(-1434000 -695000)
430000
(110000- 1292000)
-543000
(-1611000 --100000)
-84000
(-386000 - 90000)
wsu
Census_2000
-27000
(-1638000 - 1623000)
465000
(81000- 1547000)
-392000
(-1298000--24000)
-101000
(-477000 - 130000)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest thousand.
"These studies were pooled to estimate school loss days.
A-28
DRAFT: Do Not Quote or Cite
-------
Table A-8. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al.(2005)
CMU
ICLUS_A1
2550
(880 - 4220)
1550
(740 - 2360)
1310
(630 - 2000)
-310
(-480 --140)
Bell et al.(2005)
CMU
ICLUS_A2
2290
(780 - 3810)
1380
(660-2110)
1200
(570 - 1830)
-290
(-460 --130)
Bell et al.(2005)
CMU
ICLUS_BC
2260
(780 - 3740)
1410
(670-2150)
1120
(530 - 1710)
-270
(-430 --120)
Bell et al.(2005)
CMU
Woods & Poole
1860
(560 - 3150)
1360
(640 - 2070)
790
(380 - 1210)
-290
(-460 --130)
Bell et al.(2005)
CMU
Census_2000
830
(320 - 1340)
570
(270 - 870)
320
(150 -490)
-70
(-110--30)
Bell et al.(2005)
GNM
ICLUS_A1
200
(-640- 1040)
-250
(-630 - 140)
180
(-70 - 430)
270
(70 - 470)
Bell et al.(2005)
GNM
ICLUS_A2
160
(-620 - 940)
-250
(-610-100)
150
(-80 - 380)
260
(70 - 460)
Bell et al.(2005)
GNM
ICLUS_BC
90
(-680 - 870)
-300
(-670 - 70)
130
(-90 - 340)
260
(70 - 450)
Bell et al.(2005)
GNM
Woods & Poole
50
(-680 - 790)
-320
(-680 - 40)
70
(-100 -230)
300
(90-510)
Bell et al.(2005)
GNM
Census_2000
-100
(-370 - 170)
-180
(-330 - -20)
10
(-60 - 70)
70
(20- 120)
Bell et al.(2005)
Harvard
ICLUS_A1
1280
(180 -2390)
1140
(530 - 1750)
-130
(-420- 170)
270
(70 - 470)
Bell et al.(2005)
Harvard
ICLUS_A2
1180
(180 -2190)
1030
(480 - 1590)
-110
(-380 - 150)
260
(70 - 450)
Bell et al.(2005)
Harvard
ICLUS_BC
1210
(220 - 2200)
1060
(490 - 1630)
-90
(-330 - 150)
240
(60 - 420)
Bell et al.(2005)
Harvard
Woods & Poole
1050
(120 - 1980)
990
(450 - 1530)
-160
(-380 - 60)
220
(40 - 390)
Bell et al.(2005)
Harvard
Census_2000
460
(110 - 820)
430
(200 - 670)
-20
(-100 - 50)
50
(10-100)
2
A-29
DRAFT: Do Not Quote or Cite
-------
Table A-8 cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al.(2005)
lllinois-1
ICLUS_A1
3010
(1330 -4690)
1370
(650 - 2090)
1300
(620 - 1990)
340
(60 - 620)
Bell et al.(2005)
lllinois-1
ICLUS_A2
2750
(1210 -4300)
1260
(600 - 1920)
1190
(560 - 1810)
310
(50 - 560)
Bell et al.(2005)
lllinois-1
ICLUS_BC
2690
(1180 -4200)
1310
(630 - 2000)
1100
(520- 1670)
280
(40 - 520)
Bell et al.(2005)
lllinois-1
Woods & Poole
2340
(1000 - 3690)
1280
(610 - 1960)
840
(400 - 1280)
220
(-10 -450)
Bell et al.(2005)
lllinois-1
Census_2000
950
(420- 1480)
570
(270 - 870)
320
(150 -480)
60
(0-120)
Bell et al.(2005)
lllinois-2
ICLUS_A1
2810
(1280 -4340)
1120
(520- 1720)
1090
(500 - 1680)
600
(260 - 940)
Bell et al.(2005)
lllinois-2
ICLUS_A2
2570
(1170 - 3970)
1040
(490 - 1600)
970
(440 - 1500)
550
(240 - 860)
Bell et al.(2005)
lllinois-2
ICLUS_BC
2540
(1160 - 3920)
1100
(510 - 1690)
920
(420- 1420)
520
(230 - 810)
Bell et al.(2005)
lllinois-2
Woods & Poole
2220
(1010 - 3440)
1060
(500 - 1630)
670
(310 - 1040)
490
(210-770)
Bell et al.(2005)
lllinois-2
Census_2000
860
(400 - 1330)
490
(230 - 750)
250
(110 - 380)
130
(50 - 200)
Bell et al.(2005)
NERL
ICLUS_A1
60
(-850 - 970)
-190
(-550 - 160)
530
(180 - 870)
-270
(-480 - -70)
Bell et al.(2005)
NERL
ICLUS_A2
30
(-800 - 850)
-200
(-520 - 130)
480
(170 -790)
-250
(-450 - -60)
Bell et al.(2005)
NERL
ICLUS_BC
-50
(-860 - 760)
-240
(-580 - 90)
440
(150 -730)
-250
(-440 - -60)
Bell et al.(2005)
NERL
Woods & Poole
-280
(-1010 -450)
-280
(-620 - 50)
260
(60 - 460)
-260
(-460 - -60)
Bell et al.(2005)
NERL
Census_2000
-120
(-390 - 140)
-150
(-290 --10)
100
(30 - 170)
-70
(-130 --20)
2
A-30
DRAFT: Do Not Quote or Cite
-------
Table A-8 cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al.(2005)
wsu
ICLUS_A1
-780
(-3310 - 1760)
1440
(580 -2310)
-1940
(-3210--670)
-280
(-680 - 130)
Bell et al.(2005)
wsu
ICLUS_A2
-750
(-3050 - 1550)
1270
(500 - 2040)
-1770
(-2920 - -620)
-260
(-640- 120)
Bell et al.(2005)
wsu
ICLUS_BC
-580
(-2790- 1640)
1310
(530 -2100)
-1620
(-2680 - -560)
-270
(-640 - 100)
Bell et al.(2005)
wsu
Woods & Poole
-310
(-2320- 1710)
1310
(530 - 2090)
-1330
(-2180 --480)
-290
(-670 - 100)
Bell et al.(2005)
wsu
Census_2000
-10
(-740-710)
540
(220 - 860)
-460
(-770--160)
-90
(-190-10)
Levy et al.(2005)
CMU
ICLUS_A1
3600
(2180 - 5010)
2180
(1500 -2870)
1850
(1270-2430)
-440
(-580 - -290)
Levy et al.(2005)
CMU
ICLUS_A2
3230
(1950 -4520)
1950
(1340 -2570)
1700
(1160-2230)
-410
(-550 - -280)
Levy et al.(2005)
CMU
ICLUS_BC
3180
(1930 -4440)
1990
(1370-2620)
1580
(1080 -2080)
-390
(-520 - -260)
Levy et al.(2005)
CMU
Woods & Poole
2620
(1520 - 3710)
1920
(1310 -2520)
1120
(770- 1470)
-420
(-550 - -280)
Levy et al.(2005)
CMU
Census_2000
1170
(740- 1600)
810
(550 - 1060)
460
(310 -600)
-100
(-130 --60)
Levy et al.(2005)
GNM
ICLUS_A1
290
(-430 - 1000)
-350
(-670 - -20)
250
(40 - 470)
380
(210 - 550)
Levy et al.(2005)
GNM
ICLUS_A2
230
(-430 - 890)
-360
(-660 - -50)
210
(20-410)
370
(210 - 530)
Levy et al.(2005)
GNM
ICLUS_BC
130
(-520 - 790)
-420
(-730 --110)
180
(0 - 360)
370
(210 - 530)
Levy et al.(2005)
GNM
Woods & Poole
70
(-550 - 690)
-450
(-750 --140)
100
(-40 - 240)
420
(250 - 600)
Levy et al.(2005)
GNM
Census_2000
-140
(-370 - 90)
-250
(-380 --120)
10
(-40 - 60)
100
(60 - 150)
2
A-31
DRAFT: Do Not Quote or Cite
-------
Table A-8 cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Levy et al.(2005)
Harvard
ICLUS_A1
1810
(870 - 2740)
1600
(1090 -2120)
-180
(-430 - 70)
380
(210 - 550)
Levy et al.(2005)
Harvard
ICLUS_A2
1670
(820 - 2520)
1460
(990 - 1930)
-160
(-380 - 70)
370
(210 - 530)
Levy et al.(2005)
Harvard
ICLUS_BC
1700
(870 - 2540)
1490
(1010 - 1970)
-130
(-330 - 80)
340
(190 -490)
Levy et al.(2005)
Harvard
Woods & Poole
1480
(690 - 2270)
1400
(940 - 1850)
-220
(-410--40)
300
(150 -450)
Levy et al.(2005)
Harvard
Census_2000
650
(350 - 950)
610
(410-810)
-30
(-100 - 30)
70
(30-110)
Levy et al.(2005)
lllinois-1
ICLUS_A1
4240
(2820 - 5670)
1930
(1320 -2540)
1840
(1260-2420)
470
(240-710)
Levy et al.(2005)
lllinois-1
ICLUS_A2
3890
(2580 - 5190)
1780
(1220-2340)
1670
(1140-2200)
430
(210 -650)
Levy et al.(2005)
lllinois-1
ICLUS_BC
3790
(2520 - 5070)
1850
(1270-2430)
1550
(1060 -2030)
400
(190 -600)
Levy et al.(2005)
lllinois-1
Woods & Poole
3300
(2160-4440)
1810
(1240 -2380)
1180
(810 - 1560)
310
(110 - 500)
Levy et al.(2005)
lllinois-1
Census_2000
1340
(890 - 1790)
810
(550 - 1060)
450
(300 - 590)
90
(40- 140)
Levy et al.(2005)
lllinois-2
ICLUS_A1
3970
(2670 - 5260)
1580
(1070 -2080)
1540
(1040 -2040)
850
(560 - 1130)
Levy et al.(2005)
lllinois-2
ICLUS_A2
3620
(2440-4810)
1470
(1000 - 1950)
1370
(920 - 1820)
780
(520 - 1040)
Levy et al.(2005)
lllinois-2
ICLUS_BC
3580
(2420 - 4750)
1550
(1060 -2050)
1300
(870- 1720)
740
(490 - 980)
Levy et al.(2005)
lllinois-2
Woods & Poole
3140
(2110-4170)
1500
(1020 - 1980)
950
(640- 1260)
690
(450 - 920)
Levy et al.(2005)
lllinois-2
Census_2000
1220
(820- 1620)
690
(470 - 910)
350
(240 - 460)
180
(120-240)
2
A-32
DRAFT: Do Not Quote or Cite
-------
Table A-8 cont'd. Estimated National and Regional 03-Related Incidence of All Cause Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Levy et al.(2005)
NERL
ICLUS_A1
90
(-680 - 850)
-270
(-570 - 30)
750
(450 - 1040)
-390
(-560 --210)
Levy et al.(2005)
NERL
ICLUS_A2
40
(-660 - 740)
-280
(-550 - 0)
670
(410 - 930)
-360
(-520 --190)
Levy et al.(2005)
NERL
ICLUS_BC
-70
(-760-610)
-340
(-630 - -60)
620
(380 - 860)
-350
(-510--190)
Levy et al.(2005)
NERL
Woods & Poole
-400
(-1010 -220)
-400
(-680 --120)
370
(200 - 540)
-370
(-530 - -200)
Levy et al.(2005)
NERL
Census_2000
-180
(-400 - 50)
-210
(-330 - -90)
140
(80 - 200)
-100
(-150 --60)
Levy et al.(2005)
WSU
ICLUS_A1
-1090
(-3230 - 1050)
2040
(1300 -2770)
-2730
(-3800 --1670)
-390
(-740 - -50)
Levy et al.(2005)
WSU
ICLUS_A2
-1060
(-3000 - 890)
1800
(1140-2450)
-2490
(-3460 --1520)
-360
(-680 - -40)
Levy et al.(2005)
WSU
ICLUS_BC
-810
(-2680 - 1070)
1850
(1190 -2520)
-2280
(-3170 --1390)
-380
(-700 - -70)
Levy et al.(2005)
WSU
Woods & Poole
-430
(-2130 - 1280)
1850
(1190 -2510)
-1870
(-2590 --1160)
-410
(-730 - -80)
Levy et al.(2005)
WSU
Census_2000
-20
(-630 - 600)
760
(490 - 1030)
-650
(-910 --400)
-130
(-220 - -40)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
\ concentration-response function. Incidences are rounded to the nearest ten.
2
A-33
DRAFT: Do Not Quote or Cite
-------
Table A-9. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050 During
the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al. (2004)
CMU
ICLUS_A1
810
(130 - 1480)
490
(160 - 820)
410
(140 -690)
-100
(-170 --30)
Bell et al. (2004)
CMU
ICLUS_A2
720
(110 - 1330)
440
(150 -730)
380
(130 -630)
-90
(-160 --30)
Bell et al. (2004)
CMU
ICLUS_BC
710
(120-1310)
450
(150 -740)
350
(120 - 590)
-90
(-150 --30)
Bell et al. (2004)
CMU
Woods & Poole
580
(70 - 1100)
430
(140-710)
250
(80-410)
-90
(-160 --30)
Bell et al. (2004)
CMU
Census_2000
250
(60 - 450)
180
(60 - 290)
100
(30 - 160)
-20
(-40--10)
Bell et al. (2004)
GNM
ICLUS_A1
60
(-270 - 400)
-80
(-230 - 80)
60
(-40- 160)
90
(0-170)
Bell et al. (2004)
GNM
ICLUS_A2
50
(-260 - 360)
-80
(-220 - 60)
50
(-40- 140)
80
(10-160)
Bell et al. (2004)
GNM
ICLUS_BC
30
(-280 - 340)
-90
(-240 - 60)
40
(-50 - 130)
80
(10-160)
Bell et al. (2004)
GNM
Woods & Poole
20
(-280 - 310)
-100
(-240 - 50)
20
(-40 - 90)
90
(10-180)
Bell et al. (2004)
GNM
Census_2000
-30
(-140 - 80)
-50
(-120-10)
0
(-20 - 30)
20
(0 - 40)
Bell et al. (2004)
Harvard
ICLUS_A1
400
(-40 - 850)
360
(110 -600)
-40
(-160 - 80)
90
(10-170)
Bell et al. (2004)
Harvard
ICLUS_A2
370
(-30 - 780)
320
(100 - 550)
-30
(-140-70)
80
(10-160)
Bell et al. (2004)
Harvard
ICLUS_BC
380
(-20 - 780)
330
(100 - 560)
-30
(-130 -70)
80
(0-150)
Bell et al. (2004)
Harvard
Woods & Poole
330
(-40 - 700)
310
(100 - 530)
-50
(-140-40)
70
(0-140)
Bell et al. (2004)
Harvard
Census_2000
140
(0 - 280)
130
(40 - 220)
-10
(-40 - 20)
20
(0 - 30)
A-34
DRAFT: Do Not Quote or Cite
-------
Table A-9 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al. (2004)
lllinois-1
ICLUS_A1
950
(270- 1620)
430
(140-720)
410
(130 -690)
110
(-10-220)
Bell et al. (2004)
lllinois-1
ICLUS_A2
870
(250 - 1480)
400
(130 -660)
370
(120-620)
100
(-10 -200)
Bell et al. (2004)
lllinois-1
ICLUS_BC
850
(240- 1450)
410
(140 -690)
340
(110 - 570)
90
(-10-180)
Bell et al. (2004)
lllinois-1
Woods & Poole
730
(200- 1270)
400
(130 -670)
260
(90 - 430)
70
(-20- 160)
Bell et al. (2004)
lllinois-1
Census_2000
290
(90 - 500)
180
(60 - 290)
100
(30 - 160)
20
(-10-40)
Bell et al. (2004)
lllinois-2
ICLUS_A1
880
(270 - 1500)
350
(110 - 590)
340
(110 - 580)
190
(50 - 320)
Bell et al. (2004)
lllinois-2
ICLUS_A2
810
(250 - 1360)
330
(110 - 550)
300
(90 - 520)
170
(50 - 300)
Bell et al. (2004)
lllinois-2
ICLUS_BC
800
(250 - 1350)
350
(110 - 580)
290
(90 - 490)
160
(50 - 280)
Bell et al. (2004)
lllinois-2
Woods & Poole
690
(210-1170)
330
(110 - 560)
210
(60 - 350)
150
(40 - 260)
Bell et al. (2004)
lllinois-2
Census_2000
260
(80 - 450)
150
(50 - 250)
70
(20 - 130)
40
(10-70)
Bell et al. (2004)
NERL
ICLUS_A1
20
(-340 - 380)
-60
(-200 - 80)
170
(30-310)
-90
(-170-0)
Bell et al. (2004)
NERL
ICLUS_A2
10
(-320 - 340)
-60
(-190 -70)
150
(30 - 270)
-80
(-160-0)
Bell et al. (2004)
NERL
ICLUS_BC
-20
(-340 - 310)
-80
(-210-60)
140
(20 - 250)
-80
(-150-0)
Bell et al. (2004)
NERL
Woods & Poole
-90
(-380 - 200)
-90
(-220 - 40)
80
(0-160)
-80
(-160-0)
Bell et al. (2004)
NERL
Census_2000
-40
(-140-60)
-50
(-100-10)
30
(0 - 60)
-20
(-40 - 0)
A-35
DRAFT: Do Not Quote or Cite
-------
Table A-9 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Bell et al. (2004)
wsu
ICLUS_A1
-240
(-1270-780)
460
(110-810)
-610
(-1120 --100)
-90
(-250 - 70)
Bell et al. (2004)
wsu
ICLUS_A2
-240
(-1160 -690)
400
(90-710)
-560
(-1020 --90)
-80
(-230 - 70)
Bell et al. (2004)
wsu
ICLUS_BC
-180
(-1070 -710)
420
(100 -730)
-510
(-930 - -80)
-90
(-230 - 60)
Bell et al. (2004)
wsu
Woods & Poole
-90
(-900 -710)
410
(100 -730)
-410
(-750 - -80)
-90
(-250 - 60)
Bell et al. (2004)
wsu
Census_2000
0
(-280 - 280)
170
(40 - 290)
-140
(-260 - -20)
-30
(-70-10)
Ito et al. (2005)
CMU
ICLUS_A1
3620
(1810 - 5440)
2200
(1320 - 3090)
1860
(1110 -2600)
-440
(-620 - -250)
Ito et al. (2005)
CMU
ICLUS_A2
3250
(1610 -4890)
1960
(1180 -2750)
1700
(1020 -2380)
-410
(-580 - -240)
Ito et al. (2005)
CMU
ICLUS_BC
3200
(1600 -4800)
2010
(1200 -2810)
1580
(950 -2210)
-390
(-550 - -220)
Ito et al. (2005)
CMU
Woods & Poole
2610
(1230 -4000)
1920
(1150 -2700)
1100
(660- 1540)
-410
(-580 - -240)
Ito et al. (2005)
CMU
Census_2000
1140
(610 - 1680)
790
(470- 1110)
440
(260-610)
-90
(-130 --50)
Ito et al. (2005)
GNM
ICLUS_A1
290
(-620- 1200)
-350
(-760 - 70)
260
(-20 - 530)
380
(160 -600)
Ito et al. (2005)
GNM
ICLUS_A2
240
(-600 - 1080)
-350
(-740 - 30)
210
(-30 - 460)
370
(160 - 580)
Ito et al. (2005)
GNM
ICLUS_BC
140
(-700 - 980)
-420
(-820 - -20)
180
(-50 - 420)
370
(170 - 580)
Ito et al. (2005)
GNM
Woods & Poole
80
(-710 - 870)
-450
(-830 - -60)
100
(-80 - 270)
430
(200 - 650)
Ito et al. (2005)
GNM
Census_2000
-140
(-420 - 150)
-240
(-410--80)
10
(-60 - 70)
100
(40 - 150)
A-36
DRAFT: Do Not Quote or Cite
-------
Table A-9 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ito et al. (2005)
Harvard
ICLUS_A1
1820
(620 - 3010)
1610
(950 - 2280)
-180
(-500 - 140)
380
(170 -600)
Ito et al. (2005)
Harvard
ICLUS_A2
1670
(580 - 2760)
1460
(860 - 2060)
-150
(-440 - 130)
370
(170 - 570)
Ito et al. (2005)
Harvard
ICLUS_BC
1710
(640 - 2780)
1500
(880 -2120)
-130
(-390 - 140)
340
(150 - 530)
Ito et al. (2005)
Harvard
Woods & Poole
1470
(470 - 2480)
1400
(820 - 1980)
-220
(-460- 10)
300
(110 -490)
Ito et al. (2005)
Harvard
Census_2000
630
(260 - 1010)
600
(350 - 850)
-30
(-110-50)
70
(20- 120)
Ito et al. (2005)
lllinois-1
ICLUS_A1
4260
(2440 - 6080)
1940
(1160-2720)
1840
(1100 -2590)
470
(170 -780)
Ito et al. (2005)
lllinois-1
ICLUS_A2
3890
(2220 - 5550)
1780
(1070 -2500)
1670
(1000 -2350)
430
(150-710)
Ito et al. (2005)
lllinois-1
ICLUS_BC
3800
(2170 - 5430)
1860
(1110 -2600)
1550
(920-2170)
390
(130 -660)
Ito et al. (2005)
lllinois-1
Woods & Poole
3280
(1840 -4730)
1810
(1090 -2540)
1170
(700- 1640)
300
(60 - 550)
Ito et al. (2005)
lllinois-1
Census_2000
1300
(750 - 1860)
790
(470- 1110)
430
(260 - 600)
80
(20 - 150)
Ito et al. (2005)
lllinois-2
ICLUS_A1
3980
(2320 - 5630)
1580
(940 - 2230)
1540
(900 -2190)
850
(480 - 1210)
Ito et al. (2005)
lllinois-2
ICLUS_A2
3620
(2120-5120)
1470
(870 - 2080)
1370
(800 - 1940)
780
(440- 1110)
Ito et al. (2005)
lllinois-2
ICLUS_BC
3580
(2100 - 5070)
1560
(920 -2190)
1300
(760- 1840)
730
(420 - 1050)
Ito et al. (2005)
lllinois-2
Woods & Poole
3110
(1820-4410)
1500
(890 -2110)
940
(550 - 1330)
680
(380 - 980)
Ito et al. (2005)
lllinois-2
Census_2000
1180
(690- 1670)
680
(400 - 950)
340
(200 - 480)
170
(100 -250)
A-37
DRAFT: Do Not Quote or Cite
-------
Table A-9 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ito et al. (2005)
NERL
ICLUS_A1
90
(-890 - 1070)
-270
(-650 - 110)
750
(370 - 1120)
-390
(-610--160)
Ito et al. (2005)
NERL
ICLUS_A2
40
(-850 - 930)
-270
(-620 - 80)
670
(340 - 1010)
-360
(-570 --150)
Ito et al. (2005)
NERL
ICLUS_BC
-70
(-950 - 800)
-340
(-700 - 20)
620
(310 - 930)
-350
(-550 --150)
Ito et al. (2005)
NERL
Woods & Poole
-400
(-1180 - 390)
-400
(-760 - -40)
370
(150 - 580)
-360
(-580 --150)
Ito et al. (2005)
NERL
Census_2000
-170
(-450 - 100)
-210
(-360 - -60)
130
(60-210)
-100
(-150 --40)
Ito et al. (2005)
WSU
ICLUS_A1
-1080
(-3820 - 1670)
2060
(1110 - 3000)
-2740
(-4100 --1380)
-400
(-830 - 40)
Ito et al. (2005)
WSU
ICLUS_A2
-1040
(-3520 - 1440)
1810
(970 - 2650)
-2490
(-3720 --1250)
-370
(-770 - 40)
Ito et al. (2005)
WSU
ICLUS_BC
-790
(-3190 - 1600)
1870
(1010 -2730)
-2280
(-3410 --1140)
-390
(-790 - 10)
Ito et al. (2005)
WSU
Woods & Poole
-400
(-2560- 1770)
1860
(1010-2710)
-1850
(-2750 - -950)
-410
(-820 - 0)
Ito et al. (2005)
WSU
Census_2000
0
(-760 - 760)
750
(410 - 1090)
-630
(-940 --310)
-130
(-230 - -20)
Schwartz (2005)
CMU
ICLUS_A1
1220
(170-2270)
740
(230- 1260)
620
(190 - 1050)
-150
(-260 - -40)
Schwartz (2005)
CMU
ICLUS_A2
1090
(140 -2050)
660
(210-1120)
570
(180 - 960)
-140
(-240 - -40)
Schwartz (2005)
CMU
ICLUS_BC
1080
(150 -2010)
680
(210-1150)
530
(160 - 900)
-130
(-230 - -40)
Schwartz (2005)
CMU
Woods & Poole
890
(80 - 1700)
650
(200 - 1110)
380
(120-640)
-140
(-240 - -40)
Schwartz (2005)
CMU
Census_2000
390
(80-710)
270
(80 - 460)
150
(50 - 260)
I
O)
0 ,
1 CO
. O
o
A-38
DRAFT: Do Not Quote or Cite
-------
Table A-9 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (2005)
GNM
ICLUS_A1
100
(-430 - 630)
-120
(-360 - 130)
90
(-70 - 250)
130
(0 - 260)
Schwartz (2005)
GNM
ICLUS_A2
80
(-410 - 570)
-120
(-340 - 110)
70
(-70 - 220)
130
(10 -250)
Schwartz (2005)
GNM
ICLUS_BC
50
(-440 - 540)
-140
(-370 - 90)
60
(-70 - 200)
130
(10 -250)
Schwartz (2005)
GNM
Woods & Poole
30
(-440 - 490)
-150
(-380 - 80)
30
(-70- 140)
150
(10 -280)
Schwartz (2005)
GNM
Census_2000
-50
(-220- 120)
-80
(-180 -20)
0
(-30 - 40)
o
O Is"-
CO 1
o
Schwartz (2005)
Harvard
ICLUS_A1
610
(-80-1310)
550
(160 - 930)
-60
(-250 - 120)
130
(0 - 260)
Schwartz (2005)
Harvard
ICLUS_A2
560
(-70 - 1200)
490
(140 - 840)
-50
(-220- 110)
120
(10-240)
Schwartz (2005)
Harvard
ICLUS_BC
580
(-50 - 1200)
510
(150 - 870)
-40
(-200 - 110)
110
(0 - 230)
Schwartz (2005)
Harvard
Woods & Poole
500
(-90 - 1090)
480
(140 - 820)
-80
(-210-60)
100
(-10-210)
Schwartz (2005)
Harvard
Census_2000
220
(0 - 440)
210
(60 - 350)
-10
(-60 - 40)
20
(0 - 50)
Schwartz (2005)
lllinois-1
ICLUS_A1
1430
(380 - 2490)
660
(200 - 1110)
620
(190 - 1050)
160
(-20 - 330)
Schwartz (2005)
lllinois-1
ICLUS_A2
1310
(340 - 2280)
600
(190 - 1020)
560
(170 - 950)
150
(-20-310)
Schwartz (2005)
lllinois-1
ICLUS_BC
1280
(340 - 2230)
630
(200 - 1060)
520
(160 - 880)
130
(-20 - 290)
Schwartz (2005)
lllinois-1
Woods & Poole
1120
(270- 1970)
620
(190 - 1040)
400
(120 -680)
100
(-40 - 250)
Schwartz (2005)
lllinois-1
Census_2000
450
(120 -780)
270
(80 - 460)
150
(50 - 250)
30
(-10-70)
A-39
DRAFT: Do Not Quote or Cite
-------
Table A-9 cont'd. Estimated National and Regional 03-Related Incidence of Non-Accidental Mortality Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (2005)
lllinois-2
ICLUS_A1
1340
(380 - 2300)
540
(160-910)
520
(150 - 890)
290
(80 - 500)
Schwartz (2005)
lllinois-2
ICLUS_A2
1220
(350 - 2100)
500
(150 - 850)
460
(130 - 790)
260
(70 - 460)
Schwartz (2005)
lllinois-2
ICLUS_BC
1210
(350 - 2080)
530
(160 - 900)
440
(120 - 750)
250
(60 - 430)
Schwartz (2005)
lllinois-2
Woods & Poole
1060
(300-1830)
510
(150 - 870)
320
(90 - 550)
230
(60-410)
Schwartz (2005)
lllinois-2
Census_2000
410
(120-700)
230
(70 - 400)
120
(30 - 200)
60
(10-110)
Schwartz (2005)
NERL
ICLUS_A1
30
(-540 - 600)
-90
(-320 -130)
250
(30 - 470)
-130
(-260 - 0)
Schwartz (2005)
NERL
ICLUS_A2
10
(-510-530)
-90
(-300-110)
230
(30 - 420)
-120
(-240 - 0)
Schwartz (2005)
NERL
ICLUS_BC
-30
(-540 - 480)
-120
(-330 -100)
210
(30 - 390)
-120
(-240 - 0)
Schwartz (2005)
NERL
Woods & Poole
-140
(-600 - 330)
-140
(-350 - 70)
130
(0 - 250)
-120
(-250 - 0)
Schwartz (2005)
NERL
Census_2000
-60
(-230- 110)
-70
(-160-20)
50
(0 - 90)
-30
(-70 - 0)
Schwartz (2005)
WSU
ICLUS_A1
-370
(-1970- 1240)
700
(150- 1250)
-930
(-1720--130)
-140
(-390-120)
Schwartz (2005)
WSU
ICLUS_A2
-350
(-1800- 1100)
610
(130-1100)
-840
(-1560--120)
-130
(-370 -110)
Schwartz (2005)
WSU
ICLUS_BC
-270
(-1670- 1130)
630
(130-1130)
-770
(-1430--110)
-130
(-370-100)
Schwartz (2005)
WSU
Woods & Poole
-150
(-1420- 1130)
630
(140-1130)
-640
(-1170--100)
-140
(-390-100)
Schwartz (2005)
WSU
Census_2000
0
(-460 - 450)
260
(60 - 460)
-220
(-410--30)
-40
(-110-20)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
\ concentration-response function. Incidences are rounded to the nearest ten.
A-40
DRAFT: Do Not Quote or Cite
-------
Table A-10. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Ages 65 and Up) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (1995); Schwartz
(1994a); Moolgavkar et al.
(1997); Schwartz (1994b);
Moolgavkar et al. (1997)
CMU
ICLUS_A1
8640
(-670 - 22390)
5350
(270- 13580)
4170
(210- 10740)
-880
(-2230 - -20)
CMU
ICLUS_A2
7700
(-640 - 19970)
4740
(240- 12010)
3790
(190 - 9770)
-820
(-2090 - -20)
CMU
ICLUS_BC
7620
(-590- 19730)
4860
(240- 12310)
3540
(180 - 9130)
-770
(-1970 --10)
CMU
Woods & Poole
6450
(-710- 16880)
4740
(230 - 12110)
2520
(130 -6560)
-810
(-2070 - -20)
CMU
Census_2000
2780
(-100-7250)
2000
(90 - 5140)
960
(50 - 2520)
-180
(-480 - 0)
GNM
ICLUS_A1
500
(-2920 - 3970)
-950
(-2810-910)
650
(-520 - 1780)
790
(-160-1610)
GNM
ICLUS_A2
360
(-2790 - 3550)
-950
(-2680 - 790)
550
(-480 - 1550)
760
(-30 - 1560)
GNM
ICLUS_BC
130
(-3060 - 3280)
-1110
(-2930 - 740)
480
(-480- 1420)
760
(-130-1610)
GNM
Woods & Poole
10
(-3490-4010)
-1190
(-3560 - 670)
290
(-450 - 1210)
910
(-120-2650)
GNM
Census_2000
-420
(-1860- 810)
-670
(-1770-200)
40
(-210 - 370)
210
(-40 - 580)
Harvard
ICLUS_A1
4210
(-1530- 10900)
4050
(160 - 10040)
-610
(-2290 - 760)
770
(-150 -2000)
Harvard
ICLUS_A2
3860
(-1360- 9960)
3660
(140 - 9040)
-530
(-2030 - 700)
730
(-130 - 1900)
Harvard
ICLUS_BC
4010
(-1220- 10280)
3770
(140 - 9310)
-450
(-1830 -680)
680
(-140- 1770)
Harvard
Woods & Poole
3550
(-1390- 9210)
3620
(120 - 9000)
-700
(-1960-420)
630
(-180 - 1640)
Harvard
Census_2000
1570
(-370 - 4050)
1570
(50 - 3950)
-150
(-590 - 200)
150
(-50 - 390)
A-41
DRAFT: Do Not Quote or Cite
-------
Table A-10 cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Ages 65 and Up)
Attributable to Climate Change in 2050 During the 03 Season, Taken to be May through September*
Climate
Change Model
Population
Estimated 03-
Related Incidence Attributable to Climate Change in 2050
Study**
Projection to
2050
National
Northeast
Southeast
West
lllinois-1
ICLUS_A1
10050
(150 -25450)
4840
(250- 11970)
4290
(200 - 11030)
930
(-300 - 2450)
lllinois-1
ICLUS_A2
9140
(130 -23120)
4430
(230- 10950)
3870
(180 - 9950)
840
(-290 - 2230)
lllinois-1
ICLUS_BC
8990
(120-22720)
4630
(240- 11430)
3580
(170-9220)
780
(-290 - 2070)
lllinois-1
Woods & Poole
8070
(10 -20480)
4630
(240- 11520)
2840
(130 -7330)
600
(-360- 1640)
lllinois-1
Census_2000
3230
(60 - 8240)
2060
(100- 5170)
1000
(50 - 2600)
170
(-90 - 470)
lllinois-2
ICLUS_A1
9400
(290 - 23560)
4080
(170- 9910)
3620
(100 - 9330)
1700
(20 - 4320)
Schwartz (1995); Schwartz
lllinois-2
ICLUS_A2
8520
(270 -21350)
3780
(160- 9180)
3190
(90 - 8220)
1550
(20 - 3950)
(1994a); Moolgavkar et al.
(1997); Schwartz (1994b);
lllinois-2
ICLUS_BC
8510
(270-21280)
4000
(180- 9720)
3030
(90 - 7800)
1470
(10 - 3760)
Moolgavkar et al. (1997)
lllinois-2
Woods & Poole
7700
(230- 19310)
3990
(170- 9750)
2340
(70 - 6050)
1370
(-10-3510)
lllinois-2
Census_2000
2960
(100-7470)
1820
(80 - 4500)
780
(20 - 2040)
360
(0 - 930)
NERL
ICLUS_A1
110
(-3570 - 3890)
-720
(-2490- 910)
1580
(-220 - 4020)
-750
(-1860 - 170)
NERL
ICLUS_A2
10
(-3390 - 3430)
-790
(-2460 - 800)
1550
(20 - 3910)
-750
(-1890 - 50)
NERL
ICLUS_BC
-240
(-4600 - 4340)
-1000
(-3740- 1220)
1560
(10 -4800)
-800
(-2520 - 60)
NERL
Woods & Poole
-1030
(-4930 - 2240)
-1120
(-3030 - 590)
850
(-160-2360)
-770
(-2080 - 60)
NERL
Census_2000
-510
(-1900 -690)
-580
(-1560- 190)
280
(-70 - 780)
-210
(-570 - 40)
A-42
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Table A-10 cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Ages 65 and Up)
Attributable to Climate Change in 2050 During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Schwartz (1995); Schwartz
(1994a); Moolgavkar et al.
(1997); Schwartz (1994b);
Moolgavkar et al. (1997)
wsu
ICLUS_A1
-2470
(-13930 - 8420)
5070
(-190- 12870)
-6700
(-17020- 590)
-840
(-2680 - 820)
wsu
ICLUS_A2
-2370
(-12890-7420)
4430
(-190- 11250)
-6030
(-15340- 520)
-770
(-2480 - 770)
wsu
ICLUS_BC
-1750
(-11180 -7630)
4610
(-170- 11650)
-5550
(-14060- 500)
-810
(-2490 - 720)
wsu
Woods & Poole
-1090
(-10920 -7600)
4680
(-160- 12370)
-4870
(-12900- 300)
-890
(-2640 - 730)
wsu
Census_2000
50
(-3000 - 3190)
1650
(-50 - 3600)
-1370
(-3110-120)
-230
(-680 - 160)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest ten.
"These studies were pooled to estimate respiratory hospital admissions for ages 65 and up.
A-43
DRAFT: Do Not Quote or Cite
-------
Table A-ll. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Age < 1) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Burnett et al. (2001)
CMU
ICLUS_A1
2080
(480 - 3690)
1370
(580-2170)
1050
(450- 1660)
-350
(-550 --140)
CMU
ICLUS_A2
3430
(750-6160)
2270
(970 - 3600)
1780
(760-2810)
-620
(-980 - -250)
CMU
ICLUS_BC
2570
(570-4610)
1730
(740 - 2730)
1310
(560 - 2060)
-460
(-730 --180)
CMU
Woods & Poole
3040
(650 - 5460)
1840
(780-2910)
1750
(750 - 2770)
-560
(-880 - -230)
CMU
Census_2000
2140
(550 - 3760)
1460
(620 - 2320)
990
(420- 1570)
-310
(-500 --120)
GNM
ICLUS_A1
310
(-580 - 1210)
-260
(-630- 120)
190
(-60 - 440)
380
(100 -660)
GNM
ICLUS_A2
510
(-1040 -2080)
-490
(-1130- 160)
300
(-110-720)
690
(190 - 1200)
GNM
ICLUS_BC
340
(-830 - 1520)
-390
(-880- 100)
210
(-90 - 520)
520
(140 - 900)
GNM
Woods & Poole
280
(-1040- 1620)
-470
(-1010-70)
220
(-160 - 590)
540
(130 - 960)
GNM
Census_2000
20
(-890 - 940)
-420
(-860- 10)
90
(-130 - 300)
360
(90 - 630)
Harvard
ICLUS_A1
1190
(70-2310)
1030
(430- 1630)
-140
(-410-140)
300
(60 - 540)
Harvard
ICLUS_A2
2050
(170 - 3950)
1730
(720 - 2750)
-220
(-670 - 240)
540
(120 - 960)
Harvard
ICLUS_BC
1560
(150 -2980)
1310
(540 - 2090)
-150
(-480 - 180)
400
(80-710)
Harvard
Woods & Poole
1820
(180 - 3480)
1480
(610- 2350)
-100
(-510 - 320)
440
(70-810)
Harvard
Census_2000
1370
(270 - 2470)
1140
(470- 1810)
-30
(-240 - 180)
260
(40 - 480)
A-44
DRAFT: Do Not Quote or Cite
-------
Table A-ll cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Age < 1) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Burnett et al. (2001)
lllinois-1
ICLUS_A1
2610
(970 - 4260)
1230
(530- 1950)
1060
(450 - 1690)
310
(-10 -630)
lllinois-1
ICLUS_A2
4410
(1620-7230)
2120
(900 - 3340)
1780
(750 - 2820)
520
(-30 - 1070)
lllinois-1
ICLUS_BC
3310
(1220 - 5430)
1620
(690 - 2560)
1300
(550 - 2070)
390
(-20 - 800)
lllinois-1
Woods & Poole
3910
(1430 -6430)
1830
(780 - 2890)
1650
(700-2610)
430
(-50 - 930)
lllinois-1
Census_2000
2660
(1000 -4330)
1460
(620 - 2300)
930
(390 - 1470)
270
(-10 - 560)
lllinois-2
ICLUS_A1
2670
(1090 -4280)
1050
(440- 1670)
990
(410 - 1590)
630
(240 - 1030)
lllinois-2
ICLUS_A2
4550
(1850 -7290)
1820
(760 - 2890)
1640
(670 - 2630)
1090
(420- 1760)
lllinois-2
ICLUS_BC
3430
(1390 - 5490)
1400
(580 - 2220)
1220
(500 - 1950)
810
(310 - 1320)
lllinois-2
Woods & Poole
3900
(1570-6270)
1590
(660 - 2520)
1360
(540 - 2200)
950
(360 - 1550)
lllinois-2
Census_2000
2680
(1090 -4290)
1270
(530 - 2020)
830
(340 - 1330)
570
(220 - 940)
NERL
ICLUS_A1
-70
(-990 - 860)
-220
(-560 - 130)
450
(120 -780)
-290
(-550 - -40)
NERL
ICLUS_A2
-170
(-1750- 1410)
-410
(-1000 - 180)
750
(210 - 1300)
-510
(-950 - -60)
NERL
ICLUS_BC
-170
(-1350- 1010)
-330
(-780 - 120)
550
(150 - 940)
-390
(-720 - -50)
NERL
Woods & Poole
-170
(-1520- 1190)
-380
(-870 - 110)
680
(190 - 1170)
-470
(-850 - -80)
NERL
Census_2000
-260
(-1160-640)
-360
(-750 - 40)
390
(120-660)
-300
(-540 - -50)
A-45
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table A-ll cont'd. Estimated National and Regional 03-Related Incidence of Hospital Admissions for Respiratory Illness (Age < 1) Attributable
to Climate Change in 2050 During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Burnett et al. (2001)
wsu
ICLUS_A1
-710
(-3210 - 1830)
1310
(460-2170)
-1670
(-2810--510)
-350
(-860 - 170)
wsu
ICLUS_A2
-1290
(-5490 - 3000)
2140
(740 - 3560)
-2810
(-4710--860)
-620
(-1520 - 300)
wsu
ICLUS_BC
-890
(-4020 - 2300)
1630
(570-2710)
-2040
(-3430 - -620)
-480
(-1160-210)
wsu
Woods & Poole
-840
(-4330 - 2720)
1640
(560 - 2740)
-2190
(-3860 - -470)
-300
(-1040-460)
wsu
Census_2000
-310
(-2620 - 2040)
1360
(470 - 2260)
-1330
(-2260 - -380)
-340
(-830 - 150)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest ten.
A-46
DRAFT: Do Not Quote or Cite
-------
Table A-12. Estimated National and Regional 03-Related Incidence of Emergency Room Visits for Asthma Attributable to Climate
Change in 2050 During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Peel et al. (2005); Wilson et
al. (2005)
CMU
ICLUS_A1
2040
(-5600 - 7260)
1420
(-3570 - 4850)
750
(-1880 -2560)
-130
(-470 - 360)
CMU
ICLUS_A2
2490
(-6850 - 8870)
1710
(-4290 - 5840)
940
(-2370 - 3220)
-170
(-580 - 450)
CMU
ICLUS_BC
2160
(-5940 - 7690)
1510
(-3790- 5150)
790
(-1980 -2690)
-140
(-500 - 390)
CMU
Woods & Poole
2480
(-6900 - 8890)
1680
(-4210-5720)
990
(-2480 - 3380)
-190
(-650 - 500)
CMU
Census_2000
1880
(-5070 - 6640)
1380
(-3460 - 4700)
600
(-1490- 2030)
-90
(-330 - 260)
GNM
ICLUS_A1
-130
(-2150 - 1890)
-400
(-1970- 1390)
130
(-530 - 740)
140
(-370 - 530)
GNM
ICLUS_A2
-210
(-2830 - 2420)
-550
(-2460- 1790)
150
(-650 - 910)
180
(-470 - 670)
GNM
ICLUS_BC
-220
(-2530 -2130)
-490
(-2160- 1580)
120
(-540 - 740)
160
(-420 - 590)
GNM
Woods & Poole
-300
(-3080 - 2540)
-610
(-2440- 1850)
120
(-650 - 870)
190
(-500 -710)
GNM
Census_2000
-370
(-2600 - 1970)
-530
(-2070- 1550)
40
(-350 - 440)
110
(-300 - 420)
Harvard
ICLUS_A1
1160
(-3820 - 4520)
1140
(-2900 - 3900)
-100
(-730 - 680)
120
(-390 - 460)
Harvard
ICLUS_A2
1450
(-4730 - 5630)
1410
(-3590 - 4840)
-110
(-890 - 830)
150
(-480 - 580)
Harvard
ICLUS_BC
1280
(-4130 -4940)
1240
(-3160- 4250)
-90
(-720 - 690)
130
(-420 - 500)
Harvard
Woods & Poole
1500
(-4860 - 5810)
1430
(-3640 - 4900)
-80
(-840 - 830)
160
(-520 - 620)
Harvard
Census_2000
1210
(-3110 -4480)
1150
(-2540 - 3930)
-30
(-410 - 360)
90
(-250 - 340)
A-47
DRAFT: Do Not Quote or Cite
-------
Table A-12 cont'd. Estimated National and Regional 03-Related Incidence of Emergency Room Visits for Asthma Attributable to Climate
Change in 2050 During the 03 Season, Taken to be May through September*
Climate
Change Model
Population
Estimated 03-
Related Incidence Attributable to Climate Change in 2050
Study**
Projection to
2050
National
Northeast
Southeast
West
lllinois-1
ICLUS_A1
2280
(-5110 -7880)
1390
(-3020 - 4730)
760
(-1660 -2590)
130
(-430 - 580)
lllinois-1
ICLUS_A2
2850
(-6390 - 9840)
1750
(-3800 - 5950)
940
(-2060 - 3220)
160
(-520 -710)
lllinois-1
ICLUS_BC
2470
(-6390 - 8540)
1550
(-3870 - 5260)
790
(-1980 -2680)
140
(-530 - 630)
lllinois-1
Woods & Poole
2930
(-7590 - 10130)
1800
(-4510-6130)
960
(-2420 - 3270)
170
(-670 - 800)
lllinois-1
Census_2000
2150
(-5510 -7400)
1480
(-3710- 5050)
570
(-1420 - 1930)
100
(-380 - 450)
lllinois-2
ICLUS_A1
2220
(-5640-7610)
1270
(-3190-4340)
690
(-1770-2370)
260
(-680 - 900)
lllinois-2
ICLUS_A2
2780
(-7090 - 9560)
1620
(-4080 - 5540)
850
(-2180 -2920)
310
(-830 - 1090)
Peel et al. (2005); Wilson et
al. (2005)
lllinois-2
ICLUS_BC
2420
(-6160 - 8310)
1430
(-3610-4900)
710
(-1830 -2450)
270
(-720 - 960)
lllinois-2
Woods & Poole
2860
(-7290 - 9830)
1680
(-4230 - 5750)
830
(-2150 -2880)
340
(-910 - 1200)
lllinois-2
Census_2000
2060
(-5240 - 7080)
1380
(-3470 - 4720)
490
(-1250 - 1680)
190
(-520 - 680)
NERL
ICLUS_A1
-150
(-2220- 1920)
-350
(-1790- 1250)
320
(-810-1170)
-120
(-460 - 340)
NERL
ICLUS_A2
-210
(-2830 -2410)
-470
(-2210- 1580)
400
(-1000 - 1450)
-140
(-560 - 420)
NERL
ICLUS_BC
-220
(-2530 -2120)
-430
(-1940- 1410)
330
(-830 - 1200)
-130
(-500 - 370)
NERL
Woods & Poole
-290
(-3020 - 2490)
-510
(-2170- 1610)
390
(-990 - 1420)
-160
(-630 - 470)
NERL
Census_2000
-330
(-2450- 1870)
-460
(-1820 - 1370)
230
(-560 - 810)
-100
(-390 - 280)
A-48
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Table A-12 cont'd. Estimated National and Regional 03-Related Incidence of Emergency Room Visits for Asthma Attributable to Climate
Change in 2050 During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Peel et al. (2005); Wilson et
al. (2005)
wsu
ICLUS_A1
0
(-2670 - 2680)
910
(-320-2110)
-810
(-1970- 380)
-100
(-380 - 190)
wsu
ICLUS_A2
-110
(-6320 -6010)
2070
(-2100- 5690)
-1950
(-5440 - 2070)
-230
(-1070 - 360)
wsu
ICLUS_BC
10
(-2800 - 2820)
940
(-340 - 2480)
-830
(-2140 - 390)
-110
(-420-210)
wsu
Woods & Poole
-110
(-6250 - 5930)
1940
(-2350 - 5400)
-1860
(-5320 - 2420)
-190
(-1050 -450)
wsu
Census_2000
320
(-5140 -4980)
1220
(-3530 - 4590)
-780
(-3020 -2510)
-120
(-730 -610)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest ten.
"These studies were pooled to estimate respiratory emergency room visits.
A-49
DRAFT: Do Not Quote or Cite
-------
Table A-13. Estimated National and Regional 03-Related Incidence of Minor Restricted Activity Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ostro and Rothschild (1989)
CMU
ICLUS_A1
2722000
(684000 - 4770000)
1841000
(758000 - 2929000)
1246000
(514000 - 1981000)
-365000
(-588000 --141000)
CMU
ICLUS_A2
2807000
(679000 - 4944000)
1899000
(782000 - 3022000)
1305000
(539000 - 2075000)
-398000
(-642000 --153000)
CMU
ICLUS_BC
2632000
(643000 - 4629000)
1803000
(742000 - 2869000)
1195000
(493000-1900000)
-366000
(-592000 --140000)
CMU
Woods & Poole
3023000
(649000 - 5407000)
2045000
(841000- 3255000)
1474000
(608000 - 2343000)
-496000
(-800000 --191000)
CMU
Census_2000
2388000
(675000 -4109000)
1693000
(695000 - 2695000)
949000
(392000-1509000)
-254000
(-412000 --95000)
GNM
ICLUS_A1
200000
(-907000 - 1309000)
-372000
(-900000- 157000)
200000
(-91000 -491000)
372000
(84000 - 662000)
GNM
ICLUS_A2
180000
(-1005000 - 1368000)
-432000
(-997000- 134000)
193000
(-108000 -495000)
419000
(100000 -740000)
GNM
ICLUS_BC
122000
(-993000 - 1239000)
-436000
(-975000- 105000)
165000
(-110000 -441000)
392000
(92000 - 693000)
GNM
Woods & Poole
96000
(-1270000 - 1465000)
-571000
(-1203000-62000)
176000
(-171000 - 523000)
491000
(104000 - 880000)
GNM
Census_2000
-129000
(-1082000 - 827000)
-485000
(-1001000- 34000)
64000
(-141000 -270000)
291000
(60000 - 523000)
Harvard
ICLUS_A1
1540000
(123000-2962000)
1382000
(551000 -2215000)
-150000
(-480000 - 181000)
309000
(52000 - 566000)
Harvard
ICLUS_A2
1646000
(157000- 3139000)
1450000
(577000 - 2325000)
-145000
(-482000 - 194000)
341000
(63000 -619000)
Harvard
ICLUS_BC
1565000
(171000-2964000)
1377000
(548000 -2210000)
-123000
(-430000 - 184000)
312000
(53000 - 570000)
Harvard
Woods & Poole
1881000
(187000- 3579000)
1617000
(643000 - 2594000)
-130000
(-507000 - 248000)
394000
(52000 - 737000)
Harvard
Census_2000
1451000
(260000 - 2645000)
1289000
(509000 - 2072000)
-43000
(-265000 - 179000)
204000
(15000 - 394000)
A-50
DRAFT: Do Not Quote or Cite
-------
Table A-13 cont'd. Estimated National and Regional 03-Related Incidence of Minor Restricted Activity Days Attributable to Climate Change in
2050 During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ostro and Rothschild (1989)
lllinois-1
ICLUS_A1
3258000
(1189000 - 5336000)
1675000
(692000 - 2663000)
1249000
(508000-1993000)
334000
(-11000 -680000)
lllinois-1
ICLUS_A2
3430000
(1245000 - 5625000)
1786000
(738000 - 2840000)
1295000
(527000 - 2065000)
349000
(-20000 - 720000)
lllinois-1
ICLUS_BC
3216000
(1168000 - 5272000)
1711000
(706000 - 2720000)
1182000
(481000- 1885000)
323000
(-20000 - 667000)
lllinois-1
Woods & Poole
3880000
(1379000 -6391000)
2040000
(842000 - 3243000)
1434000
(586000 - 2285000)
406000
(-49000 - 862000)
lllinois-1
Census_2000
2796000
(1034000 -4565000)
1668000
(689000 - 2652000)
900000
(368000-1433000)
228000
(-23000 - 480000)
lllinois-2
ICLUS_A1
3227000
(1266000 - 5198000)
1434000
(580000 - 2293000)
1124000
(441000 - 1810000)
670000
(246000- 1095000)
lllinois-2
ICLUS_A2
3408000
(1337000 - 5488000)
1549000
(626000 - 2475000)
1148000
(450000 - 1849000)
711000
(261000- 1163000)
lllinois-2
ICLUS_BC
3205000
(1258000 - 5160000)
1488000
(603000 - 2377000)
1056000
(414000 - 1701000)
661000
(242000- 1081000)
lllinois-2
Woods & Poole
3928000
(1536000 -6330000)
1786000
(724000 - 2853000)
1274000
(498000 - 2054000)
867000
(314000- 1423000)
lllinois-2
Census_2000
2680000
(1051000 -4317000)
1443000
(586000 - 2304000)
760000
(296000 - 1226000)
478000
(170000-786000)
NERL
ICLUS_A1
-126000
(-1274000 - 1025000)
-324000
(-811000 - 163000)
515000
(129000 - 903000)
-317000
(-592000--41000)
NERL
ICLUS_A2
-182000
(-1392000 - 1031000)
-375000
(-891000 - 142000)
533000
(138000 - 930000)
-340000
(-638000 - -40000)
NERL
ICLUS_BC
-216000
(-1346000 - 916000)
-379000
(-870000 - 113000)
483000
(125000- 842000)
-320000
(-601000--39000)
NERL
Woods & Poole
-336000
(-1700000 - 1031000)
-482000
(-1048000 - 85000)
575000
(147000- 1005000)
-430000
(-799000 - -59000)
NERL
Census_2000
-354000
(-1298000 - 593000)
-428000
(-896000 -41000)
347000
(94000-601000)
-273000
(-497000 - -49000)
A-51
DRAFT: Do Not Quote or Cite
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table A-13 cont'd. Estimated National and Regional 03-Related Incidence of Minor Restricted Activity Days Attributable to Climate Change in
2050 During the 03 Season, Taken to be May through September*
Study
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Ostro and Rothschild (1989)
wsu
ICLUS_A1
-554000
(-3681000 -2594000)
1758000
(587000 - 2936000)
-1956000
(-3359000 - -540000)
-356000
(-909000 - 199000)
wsu
ICLUS_A2
-624000
(-3884000 - 2658000)
1795000
(591000 - 3006000)
-2030000
(-3485000 - -563000)
-389000
(-990000 -215000)
wsu
ICLUS_BC
-500000
(-3524000 - 2544000)
1711000
(569000 - 2859000)
-1835000
(-3152000--506000)
-376000
(-941000 - 191000)
wsu
Woods & Poole
-765000
(-4360000 - 2855000)
1865000
(609000 - 3128000)
-2236000
(-3853000 - -606000)
-393000
(-1115000 - 333000)
wsu
Census_2000
4000
(-2455000 - 2478000)
1617000
(545000 - 2693000)
-1285000
(-2238000 - -324000)
-328000
(-762000 - 108000)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest thousand.
A-52
DRAFT: Do Not Quote or Cite
-------
Table A-14. Estimated National and Regional 03-Related Incidence of School Loss Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Chen et al. (2000); Gilliland
et al. (2001)
CMU
ICLUS_A1
868000
(159000 - 1967000)
593000
(210000- 1340000)
401000
(142000- 905000)
-125000
(-278000 --41000)
CMU
ICLUS_A2
1239000
(212000 -2809000)
847000
(300000-1915000)
584000
(208000 - 1319000)
-191000
(-425000 - -63000)
CMU
ICLUS_BC
996000
(173000 -2256000)
688000
(244000-1555000)
459000
(163000 - 1037000)
-151000
(-336000 - -49000)
CMU
Woods & Poole
1130000
(166000 -2561000)
749000
(265000- 1694000)
578000
(205000 - 1306000)
-197000
(-438000 - -65000)
CMU
Census_2000
907000
(189000 -2054000)
647000
(229000-1463000)
370000
(131000- 835000)
-110000
(-245000 - -35000)
GNM
ICLUS_A1
83000
(-285000 -441000)
-118000
(-280000 - 56000)
70000
(-29000 - 162000)
131000
(24000 - 229000)
GNM
ICLUS_A2
111000
(-439000 - 647000)
-192000
(-432000 - 69000)
95000
(-48000 - 228000)
208000
(42000 - 364000)
GNM
ICLUS_BC
73000
(-368000 - 506000)
-163000
(-359000 - 49000)
71000
(-42000 - 175000)
166000
(33000 -291000)
GNM
Woods & Poole
59000
(-460000 - 572000)
-205000
(-428000 - 37000)
74000
(-65000 - 206000)
190000
(32000 - 332000)
GNM
Census_2000
-48000
(-469000 - 333000)
-205000
(-635000- 17000)
29000
(-54000 - 145000)
127000
(20000 - 373000)
Harvard
ICLUS_A1
498000
(7000- 1124000)
444000
(152000 - 1000000)
-53000
(-158000 - 58000)
107000
(13000 -240000)
Harvard
ICLUS_A2
741000
(24000-1673000)
646000
(220000 - 1456000)
-71000
(-220000 - 87000)
165000
(23000 - 372000)
Harvard
ICLUS_BC
602000
(26000- 1360000)
525000
(179000- 1183000)
-53000
(-170000 -70000)
130000
(17000 -293000)
Harvard
Woods & Poole
703000
(19000- 1587000)
599000
(204000- 1350000)
-52000
(-199000 - 102000)
155000
(14000 - 350000)
Harvard
Census_2000
577000
(78000- 1302000)
503000
(170000- 1133000)
-15000
(-98000 - 70000)
90000
(6000 - 202000)
A-53
DRAFT: Do Not Quote or Cite
-------
Table A-14 cont'd. Estimated National and Regional 03-Related Incidence of School Loss Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Chen et al. (2000); Gilliland
et al. (2001)
lllinois-1
ICLUS_A1
1052000
(320000 - 2374000)
536000
(191000 - 1210000)
405000
(141000- 914000)
110000
(-12000 -250000)
lllinois-1
ICLUS_A2
1538000
(464000 - 3471000)
795000
(282000 - 1793000)
583000
(203000- 1315000)
161000
(-22000 - 364000)
lllinois-1
ICLUS_BC
1236000
(373000 - 2789000)
650000
(231000 - 1467000)
458000
(160000- 1032000)
128000
(-18000 -291000)
lllinois-1
Woods & Poole
1464000
(433000 - 3306000)
750000
(267000 - 1694000)
559000
(196000- 1262000)
155000
(-30000 - 351000)
lllinois-1
Census_2000
1096000
(341000 -2474000)
649000
(231000 - 1466000)
348000
(122000 -786000)
98000
(-12000 -222000)
lllinois-2
ICLUS_A1
1061000
(356000 - 2397000)
460000
(160000 - 1039000)
376000
(126000 - 849000)
226000
(70000 - 509000)
lllinois-2
ICLUS_A2
1559000
(523000 - 3521000)
690000
(240000 - 1560000)
534000
(179000 - 1207000)
334000
(104000-755000)
lllinois-2
ICLUS_BC
1256000
(421000 -2837000)
566000
(197000 - 1279000)
423000
(142000 - 954000)
268000
(83000 - 604000)
lllinois-2
Woods & Poole
1486000
(494000 - 3357000)
661000
(230000 - 1495000)
490000
(162000 - 1107000)
335000
(102000-755000)
lllinois-2
Census_2000
1081000
(363000 - 2442000)
571000
(199000 - 1291000)
306000
(102000 -691000)
204000
(62000 - 459000)
NERL
ICLUS_A1
-42000
(-456000 - 336000)
-108000
(-383000 - 57000)
176000
(32000 - 510000)
-110000
(-329000 - -6000)
NERL
ICLUS_A2
-84000
(-628000 - 473000)
-172000
(-530000 - 70000)
252000
(49000 - 680000)
-163000
(-452000 - -8000)
NERL
ICLUS_BC
-84000
(-518000 - 363000)
-147000
(-394000 - 51000)
195000
(38000 - 489000)
-132000
(-334000 - -7000)
NERL
Woods & Poole
-111000
(-619000 -414000)
-178000
(-472000 - 44000)
231000
(44000- 581000)
-164000
(-415000--11000)
NERL
Census_2000
-139000
(-501000 -242000)
-175000
(-458000 - 20000)
143000
(33000 - 359000)
-107000
(-272000 - -9000)
A-54
DRAFT: Do Not Quote or Cite
-------
Table A-14 cont'd. Estimated National and Regional 03-Related Incidence of School Loss Days Attributable to Climate Change in 2050
During the 03 Season, Taken to be May through September*
Study**
Climate
Change Model
Population
Projection to
2050
Estimated 03-Related Incidence Attributable to Climate Change in 2050
National
Northeast
Southeast
West
Chen et al. (2000); Gilliland
et al. (2001)
wsu
ICLUS_A1
-224000
(-1744000 - 923000)
604000
(159000- 1794000)
-692000
(-2015000--149000)
-135000
(-523000 - 70000)
wsu
ICLUS_A2
-353000
(-2595000 - 1276000)
848000
(221000-2542000)
-996000
(-2912000--217000)
-206000
(-800000 - 106000)
wsu
ICLUS_BC
-254000
(-2006000 - 1076000)
697000
(182000-2070000)
-779000
(-2265000--167000)
-171000
(-654000 - 81000)
wsu
Woods & Poole
-328000
(-2384000 - 1156000)
714000
(182000-2149000)
-903000
(-2679000--167000)
-140000
(-641000 - 149000)
wsu
Census_2000
-45000
(-2724000 - 2699000)
774000
(134000-2572000)
-652000
(-2159000--41000)
-168000
(-793000 -216000)
*The 95% confidence or credible intervals shown below the incidence estimates characterize only the uncertainty due to statistical error of the coefficient estimate in the
concentration-response function. Incidences are rounded to the nearest thousand.
"These studies were pooled to estimate school loss days.
A-55
DRAFT: Do Not Quote or Cite
------- |