Climate Change arid Children's Health and Well-Being in the United States

Appendix A: Approach for Detailed
Analyses

This appendix details the overarching analytic approach, methods, and uncertainties associated with
the five detailed analyses included in the report. The report relies on a standardized approach to
estimating and presenting risks of climate change on children's health and well being, and to
assessing the geographies and demographic groups that may experience these risks most acutely or
disproportionately. The approach relies in part on methods and information developed previously
and published in Appendix C of a recent EPA report, Climate Change and Social Vulnerability in the
United States.1 The below information is general to each of the detailed analyses; additionally,
specific information related to implementing each analysis is highlighted in separate appendices.

CLIMATE STRESSOR AND IMPACT SELECTION

A key step in developing this report was considering different types of climate stressors, and,
subsequently, impacts that children are likely to experience. Figure 1 presents a framework found in
Hellden et al., a recent synthesis literature review on the impacts of climate change on children's
health, which maps climate change to its direct and indirect effects and ultimate health impacts.2
This report examines five key aspects of climate change (referred to throughout as "climate
stressors"), consistent with topics in the "direct effects" and "indirect effects" boxes in Figure 1, and
provides an overview of how those stressors affect children. Each of the detailed analyses focus on
one impact per stressor for which existing quantitative evidence is sufficiently available to support a
detailed and spatially resolved projection of future conditions for children under different climate
change scenarios. Each chapter closes with considerations for other important potential pathways of
harm, and previews insights from new literature about those potential future impacts on children.

Figure i. Relationship Between Climate Stressors and impacts on Children

Source: Figure 3 of Hellden et al. (2021).

April 2023

1


-------
Climate Change and Children's Health and Well-Being in the United States

How is mental health addressed in this report?

As described in Figure 1, mental health can be affected by various climate stressors. To the extent
possible, this report describes any literature linking the climate stressors with mental health effects on
children. As described in Chapter 8, quantitative evidence on the mental health impacts of climate change
on children is generally lacking and remains a key area for future research.

An example of climate change impacts on child mental health is through the concept of "climate anxiety."
Essentially, this refers to generalized concern and worry pertaining to climate change and its effects on
the future natural environment and human quality of life.3 While the terms "eco-anxiety" and "climate
change anxiety" have been used for nearly a decade,4 the effects of climate change on mental health have
been well-known for far longer.5 Often, these sentiments arise in children upon experiencing poor
environmental conditions, a severe weather event, or a series of such, but also upon having feelings of
futility or despair about the future and the state of changing global conditions.6 Even though adults also
can experience climate anxiety, research demonstrates that children may be more predisposed to
experiencing this specific type of anxiety and may experience it more intensely. For instance, older
children, including pre-teenagers, adolescents, and young adults, understand the likelihood of
experiencing climate change effects for the duration of their lives, which has been linked to feelings of
hopelessness and trauma.7 Additionally, these experiences are occurring at times of important
psychological development when trauma may have longer-term mental health effects, as children are
likely to maintain those memories with greater clarity.8,9 Further, older children, including adolescents,
are more likely to experience generalized depression or anxiety, irrespective of extrinsic factors, which can
be compounded by these same bleak feelings.10 11

The sad fact remains that access to mental healthcare in the United States is not a given, and often is
prohibitively expensive,1213 predominantly offered in English, or hard to access due to geographic location
(e.g., difficult to reach without reliable access to a car or easy public transportation).14 15 Many
practitioners simply do not accept any type of insurance coverage, including private insurance or
Medicaid, which would improve access to services.16 Despite its relative effectiveness in helping to
mitigate poor mental health,171819 child psychotherapy uptake in the United States is relatively limited,20
especially among populations that are BIPOC, immigrants or children of immigrants, and individuals
without private insurance.21-22

Finally, but no less significantly, there is limited research on how children who either identify as LGBTQIA+,
or whose caregivers or family members are LGBTQIA+, are affected by climate change. What does exist,
however, suggests that these individuals are at greater risk of experiencing more severe mental health
outcomes due to climatic factors as well as implicit biases.23,24 LGBTQIA+ youth statistically have higher
rates of suicidality, depression, and homelessness, relative to their peers, and often face insecurity in their
support from caregivers.25 Therefore, in the face of climate change-related hazards, these children are left
with a distinct predisposition for experiencing significant mental health outcomes.

April 2023

2


-------
Climate Change and Children's Health and Well-Being in the United States

IMPACTS BY DEGREE APPROACH

As described in Chapter 2, this report conveys climate risk information using an "impacts-by-degree"
framework that presents impacts to the health and well-being of children in the contiguous U.S.
under different levels of future global temperature change. The impacts-by-degree framework builds
on approaches employed in numerous published studies to produce physical and economic
estimates of climate change impacts in the contiguous U.S. (CONUS), for a broad range of the most
economically important impact sectors. The overall framework is based on a recently published
conceptual paper and demonstration of the method by Sarofim et al.26 The main objective of the
framework is to provide estimates of the physical and economic impacts in the U.S. from 21st
century trajectories of temperature and sea level rise. The methods adopt the same mainstream
scenarios and projections used in the climate science community, but instead of estimating an
impact at a specific period of time under an explicit greenhouse gas (GHG) emissions scenario,
impacts are simulated during the years when future warming thresholds are reached. The framework
is implemented using a set of underlying published studies, referred to as sectoral impact models
and analyses, which relates climate change projections to:

1.	Related environmental stressors (e.g., extreme temperatures, precipitation, floods, air
quality) to assess exposure to vulnerable individuals and physical assets;

2.	Physical impacts of climate-driven environmental stressors, such as property damage,
health effects, or damaged infrastructure; and

3.	Economic processes that are important to understand the relationship between physical
impacts and economic outcomes, such as reduced economic welfare.

Using an impacts-by-degree approach aids in communicating risk information as it can provide a
range of estimates expected for a given temperature change. The general steps and components in
this approach are outlined below, with reference to more detailed information in this and other
technical appendices that support this report.

CLIMATE DATA

Consistent with guidance for the development of the Fourth National Climate Assessment,27 this
report uses representative concentration pathway 8.5 (RCP8.5) as a higher emission scenario and
RCP4.5 as a lower emission scenario.* This selection is not an endorsement of either RCP8.5 or
RCP4.5 and does not indicate any judgment regarding the likelihood of either scenario. Because this
report estimates impacts under increasing degrees of future warming, the use of RCP8.5 allows for
analysis of the widest potential temperature range in the modeling approaches, while limiting the
number of total scenarios necessary for running through sectoral impact models. RCP8.5 provides

* RCP8.5 and a lower emissions scenario (RCP4.5) were recommended for use in NCA4. The Sixth Assessment of the
Intergovernmental Panel on Climate Change (IPCC; Working Group I), which was released in summer 2021, provided
updated scenarios and temperature projections based on the Coupled Model Intercomparison Project Phase 6
(CMIP6). However, downscaled climate projections for the U.S. were not available in time for the development of
this report.

April 2023

3


-------
Climate Change and Children's Health and Well-Being in the United States

projections for the full range of plausible 21st century temperatures, obviating the need to run
multiple scenarios to address low, medium, and high impacts. Using multiple scenarios could provide
insights into how a specific impact by degree warming level for RCP8.5 might differ from the same
level of warming, with different timing (e.g., for RCP4.5), but these differences have been shown to
be minimal once controls for socioeconomic inputs such as population and GDP are incorporated, as
was done here.28

The analyses in this report use climate projections from the fifth phase of the Coupled Model
Intercomparison Project (CMIP5).29 For most sectors, six climate models are used: the Geophysical
Fluid Dynamics Laboratory coupled general circulation model (GFDL_CM3), the Canadian Earth
System Model (CanESM2), the Community Climate System Model (CCSM4), the Goddard Institute for
Space Studies model (GISS_E2_R), the Hadley Centre Global Environmental Model (FladGEM2_ES),
and the Model for Interdisciplinary Research on Climate (MIROC5). These six GCMs are listed in Table
1. In the case of the air quality analysis (see Chapter 4), only two of the six GCMs (GFDL_CM3 and
CCSM4) were used due to computational constraints of the dynamic downscaling and atmospheric
chemistry modeling steps.

Table i. CMIP5 Global Climate Models (GCMs) Used in the Analyses of this Technical Report

Center (modelling group)

Model Acronym

References

Canadian Centre for Climate Modeling and Analysis

CanESM2

Von Salzen et al. 201330

Geophysical Fluid Dynamics Laboratory

GFDL-CM3

Donner et al. 201131

National Center for Atmospheric Research

CCSM4

Gent et al. 201132
Neale et al. 201333

NASA Goddard Institute for Space Studies

GISS-E2-R

Schmidt et al. 200634

Met Office Hadley Centre

HadGEM2-ES

Collins et al. 201135
Davies et al. 200536

Atmosphere and Ocean Research Institute, National
Institute for Environmental Studies, and Japan Agency
for Marine-Earth Science and Technology

MIROC5

Watanabe et al. 201037

Five of the six GCMs (all but GFDL_CM3) were used in the second modeling phase of the impact-by-
degree framework development, and the overall Climate Change Impact and Risk Assessment (CIRA)
project.38 These five GCMS were chosen based on a consideration of independence and skill at
matching historical observed U.S. climate, and coverage of a wide range of future precipitation and
temperature outcomes. GFDL_CM3 was added to that set with the most important criteria being the
inclusion of an additional high temperature model that was different from other models already
included, as evaluated by estimates of inter-model distance.39 Other warm models considered
included CESM1_CAM5, which was excluded based on similarity to CCSM4; ACCESS1_3, which has
similarities to FladGEM2_ES; and CNRM_CM5, which was slightly cooler and slightly less skillful by
the empirical metrics than GFDL_CM3.40 GFDL_CM3 was added to the suite of climate models to
include better coverage of the impacts-by-degree approach for higher levels of warming in the U.S.,
and that model's results are also considered in some of the impacts analyzed in this report. Sarofim
et al. provides further justification for this rationale.41

April 2023

4


-------
Climate Change and Children's Health and Well-Being in the United States

Most sectoral analyses of this report require downscaled climate projections to reduce model bias
and provide finer resolution. The approach presented here relies primarily on the LOCA (Localized
Constructed Analog)42,43 approach to produce daily temperature (maximum and minimum) and
precipitation data at a 1/16-degree scale (approximately 6.25 km). The only detailed analysis in this
report that did not use LOCA data is the coastal flooding analysis (Chapter 6), which relies on sea
level rise and storm surge projections described below. Moreover, the air quality analysis utilizes
dynamically downscaled climate projections (see Chapter 4).

To aid in the selection of specific GCMs, the LASSO44 tool was used to produce scatter plots showing
the variability across the CMIP5 ensemble for projected changes (2085-2095 compared to the 1986-
2005 reference period) in annual and summertime temperature and precipitation. Figure 2 shows
the range of temperature and precipitation outcomes across the CMIP5 ensemble. The GCMs used in
the climate projections for this report are displayed with blue circles around them to highlight their
location within the scatter plots. The model identified as the double median across
temperature/precipitation outcomes is shown in a red rectangle.

April 2023

5


-------
Climate Change arid Children's Health and Well-Being in the United States

Figure 2. Variability of Projected Annual (top) and Summertime (bottom) Temperature and
Precipitation Change across the CMiPs Ensemble forthe Continental U.S.

32 realizations of climate change from CMIP5 (LOCA)

Emissions scenario: RCP85 Study area: CONUS

s 2
^-88

ro OJ

1	=ACCESS1-0

2	= ACCESS1-3

3	= bcc-csm1-1

4	= bcc-csml

6 = CCSM4

m

9	= CMCC-CM

10	= CMCC-CMS

17	= GFDL-CM3

18	- GFDL-ESM2G

19	= GFDL-ESM2M

20	= GISS-E2-H
22 = GISS-E2-R

24	= HadGEM2-AO

25	= HadGEM2-CC

28	= IPSL-CM5A-LR

29	= IPSL-CM5A-MR

34	= MPI-ESM-LR

35	= MPI-ESM-MR

36	= MRI-CGCM3

37	= NorESM1-M

A mean annual temperature (°C): 2085 to 2095
1986 to 2005 baseline

32 realizations of climate change from CMIP5 (LOCA)

Emissions scenario: RCP85 Study area: CONUS

0	LO

1	l

.9- o

1	1	1	1	1	1	r

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

A mean JJA temperature (°C): 2085 to 2095
1986 to 2005 baseline

1	= ACCESS 1-0

2	= ACCESS1-3

3	= bcc-csm1-1

4	= bcc-csm1

6 = CCSM4

m

9	= CMCC-CM

10	= CMCC-CMS

17	= GFDL-CM3

18	= GFDL-ESM2G

19	= GFDL-ESM2M

20	= GISS-E2-H
22 = GISS-E2-R

24	= HadGEM2-AO

25	= HadGEM2-CC

26	= HadGEM2-ES

28	= IPSL-CM5A-LR

29	= IPSL-CM5A-MR

34	= MPI-ESM-LR

35	= MPI-ESM-MR

36	= MRI-CGCM3

37	= NorESM1-M

Source: U.S. EPA (2017). Notes: Application of the LASSO tool (see text for reference) to produce scatter plots
showing the variability across the CMIP5 ensemble for projected changes (2085-2095 compared to the 1986-2005
reference period) in annual (top panel) and summertime (bottom panel) temperature and precipitation. Numerals
show individual GCM temperature and precipitation outcomes across the CMIP5 ensemble. GCMs used in this
report are displayed with blue circles around them to highlight their location within the scatter plots. The red
rectangle shows the model identified as the double median across temperature/precipitation outcomes.

April 2023

6


-------
Climate Change arid Children's Health and Well-Being in the United States

ARRIVAL TIMES BY INTEGER WARMING

As part of the impacts-by-degree framework, the arrival times of global average temperature
increases compared to the 1986-2005 baseline were identified from the GCMs described above.
These arrival times represent the first 11-year period to have an average temperature equal to that
of the warming degree. Figure 3 shows the year at which the 11-year moving average for when each
of the GCMs first reached each degree above the baseline, and the 11-year window around that year
(e.g., CanESM2 has an initial arrival year of 2018, and its 11-year "window" encompasses 2011-
2022).

Figure 3. Arrival Years of Global Increases in Temperature

2010	2020	2030	2040	2050	2060	2070	2080	2090 2100

CanESM2

2018







2042

2062

2080



CCSM4



2026





2057

2077



GISS-E2-R



2034

2074

HadGEM2-ES



2020





2047

2059

2078



MIROC5





2028



2059

2077



GFDL-CM3



2020





2043

2061

2077

2095

Degrees of Warming

¦ 1 2 3 4 5

Notes: This graphic shows the 11-year windows assigned to each integer temperature by GCM under a higher
emission scenario (RCP8.5). Values are calculated using a 1986-2005 baseline. Arrival years, or the year at which the
11-year moving average reaches the given integer, are listed in each bin. Source: Sarofim et al. (2021)

Temperature change is not uniform across the globe, and the projected global average temperature
changes shown in Figure 4 manifest differently in the U.S. Figure 4 shows the projected county-level
temperature changes that correspond to global warming of 2°C and 4°C. As shown, changes in global
temperatures generally result in higher changes in average annual temperatures in the U.S.

Figure 4. Projected Changes in Average Annual Temperatures Across the U.S.

2°C Global Warming	4°C Global Warming

Northeast

Northern
Great
Plains

lorthwest

iM id west

Southwest

Southeast

—Southern
\Great Plains

12.6°F 14.4°F

5.4°F

Degrees

4°C	5°C

Source: EPA (2021). These maps show the county-level average annual temperature changes associated with global
average temperature changes of2°C and 4°C relative to the 1986-2005 baseline.

April 2023

7


-------
Climate Change and Children's Health and Well-Being in the United States

It is important to note that the 1986-2005 baseline is 0.61°C warmer than preindustrial (1850-1900)
temperatures at the global scale.45 Throughout the main report, results are presented for 2°C and
4°C of global warming. Impacts could be experienced between 2042-2074 (2056 average) and
between 2077 and sometime after 2100 (2097 average) depending on the specific climate model.
However, the risk and impact estimates are only available for 21st century years. As a result,
measurements of the potential risks to children's health presented in this report for 4°C only include
the three models with arrival years before 2100 (CanESM2, HadGEM2-ES, and GFDL-CM3). 2°C and
4°C of global warming were chosen for this report to provide a range of results that might be realized
in the 21st century.

SEA LEVEL RISE PROJECTIONS

This report projects impacts using future increases in global mean sea level (GMSL) in increments of
25 cm, up to 150 cm, relative to GMSL in 2000. Results in the main report convey impacts under 50
cm and 100 cm of global sea level rise. The underlying economic impact literature provides results
for each year up to 2100, using six GMSL trajectories developed for the USGCRP's Fourth National
Climate Assessment. The scenarios are categorized according to the future change in GMSL in 2100,
relative to the year 2000 (e.g., 100 cm, 200 cm). Projections of location-specific differences in
relative (or local) sea level change46 account for land uplift or subsidence, oceanographic effects, and
responses of the geoid and the lithosphere to shrinking land ice. Mean values for each tide gauge
location are used. A distance weighting procedure for interpolating between tide gauge locations is
employed to attribute tide gauge-level results to each coastal county. This procedure allows us to
connect changes in GMSL with a) county-scale relative sea level rise (SLR) that considers these local
scale factors, and b) data on the economic impacts of each increment in SLR for those localities.

Figure 5 shows the specific 11-year bins used to connect the underlying economic impact literature
to GMSL increments in the NCA4 SLR trajectories. The SLR bins are based on the published NCA SLR
trajectories and calculated using the temperature binning "arrival time" method adopted in
supporting literature, adapted for GMSL arrival timing.47

Figure 5. Arrival Years of Global Mean Sea Level (GMSL) Rise

2020	2030	2040	2050	2060	2070	2080	2090	2100

30 cm

2080

50 cm

2052

100 cm

2040

2065

2083



150 cm

2036

2054



2068

2080

| 2092



200 cm

2033

2048

2060 | 2070

2079

2085



250 cm

2031

2045 2055



2064

2079



Increments of GMSL Rise

25 ¦ 50 ¦ 75 ¦ 100 ¦ 125 ¦ 150

Notes: This graphic shows the 11-year windows assigned to each 25 cm increment for results from each of the
National Climate Scenario GMSL scenarios. Values are calculated using a year-2000 baseline. Arrival years, or the
year at which the 11-year moving average reaches the given integer, are listed in each bin.

April 2023

8


-------
Climate Change and Children's Health and Well-Being in the United States

POPULATION CONSIDERATIONS AND DATA

Ultimately, this report conveys the risk of the climate stressors and associated impacts to all children
in the contiguous U.S. To do so, the detailed analyses incorporate projections for the future
population of children. The analysis relies on U.S. Census data for 2010 as well as future projections
published in EPA's Integrated Climate and Land Use Scenarios version 2 (ICLUSv2)48 model through
2100. Given U.S. Census data for 2020 are available, the analysis evaluates how ICLUS compares with
the U.S. Census data for that year and found only small differences. For consistency with other CIRA
analyses, ICLUS is used for all years post-2010. The analysis takes the following approach:

•	Step 1: Establish the baseline population of children by county using the 2010 U.S. Census.

•	Step 2: Apply the growth rates at the county level calculated from ICLUS populations of
children for each year between 2010 and 2100 to the 2010 population from Step 1.

•	Step 3: Calculate percentage of children within each county by census tract and block group
using 2015-2019 American Community Survey data (e.g., within county x, 10% of children live
in census tract y).

•	Step 4: Allocate population of children by census tract and block group for each year in ICLUS
by multiplying percentages from Step 3 by total county population in Step 2.

While some of the detailed analyses focus on impacts across all children aged 0-17, some of the
impact measures are specific to age ranges of children (i.e., 0-5 years, 14-17, etc.). The four steps
above are performed for each relevant age category. Figure 6 below describes how the population of
children is expected to change over the 21st century. Relative to 2010 levels, the total population of
children is projected to increase by 22% by 2050 and 35% by 2090. To provide a better
representation of the influence of climate specifically on the impacts measured in the report, the
technical appendix for each detailed analysis also presents results assuming constant 2010
population, removing the influence of population growth on the results.

Figure 6. Future Projections of Children's Population in the Contiguous U.S.

120,000,000

S 100,000,000

£ 80,000,000
u

^ 60,000,000

I—

^ 40,000,000
£

= 20,000,000

2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
¦ 0 to 4 ¦ 5 to 13 ¦ 14 to 17

Notes: This figure presents the growth in population of children across the 21st century (x-axis shows 21st century
years) using the 2010 U.S. Census data and projections from ICLUS. Specific age sub-groups are described using
colors: aged 0-4 using orange, aged 5-13 using brown, and aged 14-17 using blue.

April 2023

9


-------
Climate Change and Children's Health and Well-Being in the United States

SOCIAL VULNERABILITY APPROACH

This report considers ways in which climate change impacts may be experienced disproportionately
by overburdened populations, leveraging a previously published analytic approach to assess the
likelihood of this occurring.49 Across each relevant detailed analysis, this report uses a standard set of
demographics: Black, Indigenous, and People of Color (BIPOC), low income, limited English speaking,
and no health insurance. The specific data used to define each of these populations is identified
below. For each analysis, children belonging to any of these four populations are first identified and
are located within the spatial domain considered to be vulnerable to impacts for the analysis. For
example, the coastal flooding analysis only considers children that live in coastal areas.

Climate change impacts are modeled using the methods developed for each analysis to identify high
impact areas. "High impact" is defined as areas in the highest tercile of impacts per capita. Note that
the spatial resolution of analysis varies by sector (e.g., county, census tract, census block group), but
is consistent within each analysis. Once high impact areas are identified, the number of socially
vulnerable children, and the "reference" non-socially vulnerable children in those areas, are
tabulated (see details below on definitions of socially vulnerable and reference populations for each
specific population). From this, the likelihood of living in a high impact location is calculated for both
populations, relative to the reference domain. The relative likelihoods described in this report are the
result of comparing likelihoods of living in high impact areas for populations that are and are not
socially vulnerable. This standardized approach allows us to present relative likelihoods of high
impacts at regional and national scales, in which regional-level relative likelihoods are based on
regional spatial domains and populations.

In standardized form, the difference in risk is calculated as:

A D _ (^Pvh\ / (YjPrh
IaK —

IPvJf VE3

-1

where AR is the risk difference, expressed as a percent; Pvh is the sum of the socially vulnerable
population in all "high impact" areas; Pv is the total socially vulnerable population; Prh is the sum of
the reference population in "high impact" areas; and Pr is the total reference population. As an
example, the details of an illustrative calculation for the impact of air quality on new cases of asthma
for BIPOC children are presented in Table 2.

Table 2. Example Calculation of Disproportionate Impacts on BIPOC Children - New Asthma Cases
Associated with Air Quality

Variables

Calculation and interpretation

Pv= 12 million BIPOC children across CONUS
Pr= 11 million non-BIPOC children across CONUS
Pvh= 4 million BIPOC children in high impact census
tracts

Prh= 2 million non-BIPOC children in high impact census
tracts

BIPOC children are 83% more likely than non-BIPOC
children to live in areas with the highest rates per
capita of new asthma diagnoses linked with
degrading air quality.

April 2023

10


-------
Climate Change and Children's Health and Well-Being in the United States

DEMOGRAPHIC CONSIDERATIONS AND DATA

This report relies on a contemporaneous picture of demographics across CONUS to assess the
likelihood of disproportionate impacts on particular demographic groups of children. This is a
departure from the health risk assessment analyses that project impacts to children while taking into
consideration future changes in population. Projecting the future distribution of children across
demographic groups is less certain, especially for specific variables such as those who are uninsured.
Thus, these distributions generally are unavailable at the level of detail necessary for the social
vulnerability analyses described in this report. However, shifting demographics and socioeconomic
change will affect the spatial distribution and magnitude of vulnerability to climate change. Multi-
sector assessments have demonstrated compounding effects of population growth and climate
change impacts, particularly with regards to health-related effects.50 Therefore, the results of this
report should be interpreted with this limitation in mind, as actual impacts could be larger or smaller
based on potentially changing demographics.

The report examines disproportionate impacts of climate change on children of various demographic
groups by relying on ACS data, averaged across 2015-2019. Where available, data are collected at the
block group level, or if necessary, at the census tract level. This analysis relied on the IPUMS+
platform to download ACS data through its National Historical Geographic Information System
(NHGIS). The NGHIS codes for data this report relies upon are provided in Table 3.51

SOCIALLY VULNERABLE GROUPS CONSIDERED

This analysis considers four groups of socially vulnerable children. These variables were chosen
primarily because the literature suggests children in these categories are disproportionately
vulnerable to the specific climate stressors and impacts analyzed.

• BIPOC: The term BIPOC used in this report refers to individuals identifying as Black or African
American; American Indian or Alaska Native; Asian; Native Hawaiian or Other Pacific Islander;
and/or Hispanic or Latino. The BIPOC children included herein also only include those living in
the CONUS. The ACS provides race and ethnicity data at the census block group level. This
report relies on total population as well as White, non-Hispanic population to calculate the
BIPOC population at the census block group and census tract spatial scales. Age-stratified
demographic information is available at the census tract level, so these estimates are specific
to children aged 0-17. Age-stratified race and ethnicity information is not available from the
ACS at the block group level employed in the coastal flooding health risk analysis, so all-age
demographic information is combined with block group-level age distributions to estimate
the distribution of children for racial components within the BIPOC category. For calculations
of disproportionate effects on socially vulnerable BIPOC populations, the White non-Hispanic
population is used as the reference population. Across CONUS, the ACS 2015-2019 identifies
31 million BIPOC children and 42 million non-BIPOC children.

f IPUMS had previously been an acronym for Integrated Public Use Microdata Series, but not all of the data it
accesses is public, or is microdata, so since 2016 it has been known only by its acronym.

April 2023

11


-------
Climate Change and Children's Health and Well-Being in the United States

•	Low income: "Low income" at the individual level is defined as children living in households
that have an aggregate income that is no more than twice the poverty threshold. This
variable is not age-stratified, so this report relies on an all-ages poverty estimate. Additional
information on the definition of poverty thresholds can be found on the U.S. Census
website.52 In this report, the estimates of households that fall into income-to-poverty
threshold ratios below two times the poverty threshold are aggregated. The reference
population is individuals living in households with income greater than two times the poverty
threshold. Across CONUS, the ACS 2015-2019 identifies 24 million low-income children and
49 million children living above two times the poverty threshold.

•	Limited English speaking: Children are considered "limited English speaking" in the ACS if
those 14 years old and older have at least some difficulties with speaking English. This
variable is not age-stratified in the ACS, so this report relies on an all-ages estimate of
language isolation and assumes the proportion of children who are linguistically isolated is
consistent with that of the rest of the population. In this report, the estimates of populations
who live in limited English-speaking households are aggregated by primary language spoken
at home. The reference population is individuals who do not live in a limited English-speaking
household. Across CONUS, the ACS 2015-2019 identifies 4 million children who fall into the
U.S. Census-defined category of "Limited English speaking" and 69 million children who are
not considered to be in this category.

•	No health insurance: The ACS provides age-stratified estimates of children with and without
health insurance at the block group level. In this report, children of both sexes younger than
6 years old and between the ages of 6 and 18 years old who have no health insurance are
aggregated, so the estimate is specific to all children aged 0-18 years. The reference
population is children who have health insurance. Across CONUS, the ACS 2015-2019
identifies 4 million children lacking health insurance and 69 million children with insurance.

Introductory sections of each chapter summarize the literature and/or the conceptual links between
impacts and vulnerability of these populations. There are additional dimensions of social vulnerability
not considered in this report (e.g., disability, household composition, and others), warranting further
analysis. Additional disproportionate risks may be present when evaluating the interconnections
between social vulnerability measures, connections that are not explored in this report.

As illustrated in Figure 7, the demographic groups described above are spatially correlated with each
other. The key disproportionality results, however, do not necessarily exhibit the same degree of
correlation nationally or by region that are shown in the full ACS dataset, as each impact examines a
different spatial domain based on the specific locations of the high impact terciles. Many individuals
also may meet the ACS definition for inclusion in multiple categories from among the four chosen.
Supplemental analyses were considered regarding disproportionate effects for individuals included in
multiple categories of social vulnerability; however, ACS data support only limited versions of
analyses that aggregate characteristics and are not stratified by age. For example, available low
income cross-tabulations are focused on individuals with income below the poverty line, rather than
below twice the poverty line, but do not reflect age-related effects.

April 2023

12


-------
Climate Change arid Children's Health and Well-Being in the United States

Figure 7. Spatial Distribution of Select Groups of Overburdened Children (Aged 0-17)

Low Income	BIPOC

Limited English Speaking	No Health Insurance

0% - 10% 10% - 25% 25% - 50% S 50% - 75% H 75% - 80% H 80% - 100%

Notes: This graphic shows the spatial distribution of four groups of overburdened children by census tracts based on the U.S. Census Bureau's
American Community Survey 2015-2019 (specific data tables are documented in Table 3). The percentages convey the portion of children living in
each census tract that meet the definition. NCA regions are outlined in black.

April 2023


-------
Climate Change and Children's Health and Well-Being in the United States

Table 3. Underlying Demographic Data from U.S. Census Bureau's American Community Survey
2015-2019



NHGIS Field

ACS



Data table

Spatial Scale

Code

Source



Table



Description	Use

Sex by Age

ALTOE003; 004;
005; 006; 027;
028; 029; 030

B01001

Block Group;
Census Tract

Total population of
individuals by age (<5, 5-
9,10-14,15-17) and sex
(male, female)

BIPOC; Low
Income; No
health
insurance

Race

ALUCE002

B02001

Block Group

White Alone

BIPOC

Race

ALUCE003

B02001

Block Group

Black or African-
American Alone

BIPOC

Race

ALUCE004

B02001

Block Group

American Indian and
Alaska Native Alone

BIPOC

Race

ALUCE005

B02001

Block Group

Asian Alone

BIPOC

Race

ALUCE006

B02001

Block Group

Native Hawaiian and
Other Pacific Islander
Alone

BIPOC

Race

ALUCE007

B02001

Block Group

Some Other Race Alone

BIPOC

Race

ALUCE008

B02001

Block Group

Two or More Races

BIPOC

Hispanic or Latino
Origin by Race

ALUKE003

B03002

Block Group

White Alone, Not
Hispanic or Latino

BIPOC

Hispanic or Latino
Origin by Race

ALUKE012

B03002

Block Group

Hispanic or Latino (all
races)

BIPOC

Sex by Age (White
Alone)

AL4FE003; 004;
005; 006; 018;
019; 020; 021

B01001A

Census Tract

Population of white
individuals by age (<5, 5-
9,10-14,15-17 years)
and sex (male, female)

BIPOC

Sex by Age (Black or
African American
Alone)

AL4GE003;
004; 005; 006;
018; 019; 020;
021

B01001B

Census Tract

Population of Black or
African American
individuals by age (<5, 5-
9,10-14,15-17 years)
and sex (male, female)

BIPOC

Sex by Age (American
Indian and Alaska
Native Alone)

AL4HE003;
004; 005; 006;
018; 019; 020;
021

B01001C

Census Tract

Population of American
Indian and Alaska native
individuals by age (<5, 5-
9,10-14,15-17 years)
and sex (male, female)

BIPOC

Sex by Age (Asian
Alone)

AL4IE003;
004; 005; 006;
018; 019; 020;
021

B01001D

Census Tract

Population of Asian
individuals by age (<5, 5-
9,10-14,15-17 years)
and sex (male, female)

BIPOC

Sex by Age (Native
Hawaiian and Other
Pacific Islander Alone)

AL4JE003;
004; 005; 006;
018; 019; 020;
021

B01001E

Census Tract

Population of Native
Hawaiian and other
Pacific Islander
individuals by age (<5, 5-
9,10-14,15-17 years)
and sex (male, female)

BIPOC

Sex by Age (Some
Other Race Alone)

AL4KE003;
004; 005; 006;
018; 019; 020;
021

B01001F

Census Tract

Population of individuals
of some other race by
age (<5, 5-9, 10-14,15-
17 years) and sex (male,
female)

BIPOC

April 2023

14


-------
Climate Change and Children's Health and Well-Being in the United States

Data table

NHGIS Field
Code

ACS
Source
Table

Spatial Scale

Description

Use

Sex by Age (Two or
More Races)

AL4LE003;
004; 005; 006;
018; 019; 020;
021

B01001G

Census Tract

Population of individuals
of two or more races by
age (<5, 5-9, 10-14,15-
17 years) and sex (male,
female)

BIPOC

Sex by Age (White
Alone, Not Hispanic or
Latino)

AL4ME003;
004; 005; 006;
018; 019; 020;
021

B01001H

Census Tract

Population of white non-
Hispanic individuals by
age (<5, 5-9, 10-14,15-
17 years) and sex (male,
female)

BIPOC

Sex by Age (Hispanic or
Latino)

AL4NE003;
004; 005; 006;
018; 019; 020;
021

B01001I

Census Tract

Population of Hispanic or
Latino individuals by age
(<5, 5-9,10-14,15-17
years) and sex (male,
female)

BIPOC

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE001

C17002

Block Group;
Census Tract

Total Population

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE002

C17002

Block Group;
Census Tract

Under .50

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE003

C17002

Block Group;
Census Tract

.50 to .99

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE004

C17002

Block Group;
Census Tract

1.00 to 1.24

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE005

C17002

Block Group;
Census Tract

1.25 to 1.49

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE006

C17002

Block Group;
Census Tract

1.50 to 1.84

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE007

C17002

Block Group;
Census Tract

1.85 to 1.99

Low Income

Ratio of Income to
Poverty Level in the
Past 12 Months

ALWVE008

C17002

Block Group;
Census Tract

2.00 and over

Low Income

Health Insurance
Coverage Status by Sex
and Age

AMLLE004

B27001

Census Tract

Male: Under 6 years:
With health insurance
coverage

No health
insurance

Health Insurance
Coverage Status by Sex
and Age

AMLLE005

B27001

Census Tract

Male: Under 6 years: No
health insurance
coverage

No health
insurance

Health Insurance
Coverage Status by Sex
and Age

AMLLE007

B27001

Census Tract

Male: 6 to 18 years: With
health insurance
coverage

No health
insurance

Health Insurance
Coverage Status by Sex
and Age

AMLLE008

B27001

Census Tract

Male: 6 to 18 years: No
health insurance
coverage

No health
insurance

Health Insurance
Coverage Status by Sex
and Age

AMLLE032

B27001

Census Tract

Female: Under 6 years:
With health insurance
coverage

No health
insurance

April 2023

15


-------
Climate Change and Children's Health and Well-Being in the United States

Data table

NHGIS Field
Code

ACS
Source
Table

Spatial Scale

Description

Use

Health Insurance
Coverage Status by Sex
and Age

AMLLE033

B27001

Census Tract

Female: Under 6 years:
No health insurance
coverage

No health
insurance

Health Insurance
Coverage Status by Sex
and Age

AMLLE035

B27001

Census Tract

Female: 6 to 18 years:
With health insurance
coverage

No health
insurance

Health Insurance
Coverage Status by Sex
and Age

AMLLE036

B27001

Census Tract

Female: 6 to 18 years:
No health insurance
coverage

No health
insurance

Household Language
by Household Limited
English Speaking Status

ALWTE001

C16002

Block Group;
Census Tract

Total Households

Limited English
speaking

Household Language
by Household Limited
English Speaking Status

ALWTE004

C16002

Block Group;
Census Tract

Spanish: Limited English-
Speaking Household

Limited English
speaking

Household Language
by Household Limited
English Speaking Status

ALWTE007

C16002

Block Group;
Census Tract

Other Indo-European
Languages: Limited
English-Speaking
Household

Limited English
speaking

Household Language
by Household Limited
English Speaking Status

ALWTE010

C16002

Block Group;
Census Tract

Asian and Pacific Island
Languages: Limited
English-Speaking
Household

Limited English
speaking

Household Language
by Household Limited
English Speaking Status

ALWTE013

C16002

Block Group;
Census Tract

Other Languages:

Limited English-Speaking
Household

Limited English
speaking

April 2023

16


-------
Climate Change and Children's Health and Well-Being in the United States

SOURCES OF UNCERTAINTY AND LIMITATIONS

This section describes some of the main sources of uncertainty inherent across the detailed analyses.
Limitations specific to each individual detailed analysis are described in those sections of this report
and appendices.

PROJECTIONS OF FUTURE CLIMATE

With the goal of presenting a consistent set of climate change impact analyses across sectors, this
report presents results using an impacts-by-degree approach. Arrival windows for integral levels of
future warming were identified from each climate model, and these years were used in the
simulations for each sectoral impact analysis. Due to the level of effort necessary to run each
scenario through the sectoral models of this report, only six climate models were chosen. While
these models were chosen to capture a large range of the variability observed across the entire
ensemble, this subset is not a perfect representation of climate models. However, even the full set of
GCMs is not likely to span the entire range of potential physical responses of the climate system to
changes in the concentration of atmospheric GHGs. Previous literature demonstrates the importance
of climate sensitivity assumptions in understanding a wide range of potential changes to the climate
system,53,54 as well as the effect of natural variability on timing and magnitude of impacts.55,56The
Sixth Assessment of the IPCC provides updated scenarios and temperature projections based on the
CMIP6 project. However, these newer projections and the widely accepted downscaled and bias-
corrected projections of the results of CMIP6 GCMs were not available in time for use in this report.

COVERAGE OF CLIMATE STRESSORS AND IMPACTS

The analyses presented in this report cover just a handful of potential impacts of climate change in
the U.S. The five stressors included were chosen because of the availability of robust methods and
data for analysis that offered information specific to children or were easily extrapolated to younger
populations. There are a number of additional impacts of climate change that likely will affect
children, but which are not included in this report. The literature reviews that open each chapter
provide some perspective on the broad range of possible impacts on all children and those
disproportionately affected.

COMPARISON ACROSS IMPACT MEASURES

Unlike previous CIRA reports that primarily focused on the presentation of economic results across
sectors, this report contains limited monetization of impacts. The one exception is the analysis that
projects lost future income associated with heat-induced learning losses, which aggregates impacts
across students graduating each year. While the lack of a common economic metric makes
comparisons across impacts more challenging, a focus on physical impacts (e.g., cases of asthma,
number of children affected by flooding) is more appropriate in this context because the results are
not dependent on the details of a specific economic valuation approach. In addition, some metrics
used in this report have not yet been valued in economic terms, such as the mental stress of children
losing a home to coastal flooding. It should be noted that even the physical measures used here
cannot convey how climate effects and health outcomes experienced in childhood may prevail

April 2023

17


-------
Climate Change and Children's Health and Well-Being in the United States

throughout an individual's life, including leading to future serious health effects. In either
circumstance, these are likely to differ across the impacts considered.

However, to provide perspective on the costs associated with the physical impacts projected in this
report, the report conveys direct medical costs and indirect productivity losses provided by available
research. A major research gap is that these costs are infrequently described specific to children, and
research (where it does exist) clearly shows that the costs differ between children and adults.57 The
unit costs offered in the report are merely to provide perspective on order of magnitude.

IMPACT MODELING

The impact estimates presented in this report were developed using discrete impact models. These
models are complex analytical tools, and choices regarding the structure and parameter values of the
model can create important assumptions that affect the estimation of impacts. Ongoing studies such
as the Inter-Sectoral Impact Model Intercomparison Project (ISI-MIP) are investigating the influence
of structural uncertainties across sectoral impact models.58 The use of additional models for each
analysis of this report would help improve the understanding of potential impacts in the future.

JOINT IMPACTS ACROSS CLIMATE STRESSORS

The results presented for each detailed analysis primarily were developed independently of one
another. As a result, the estimated impacts may omit important cumulative or interactive effects or
outcomes. For example, the air quality and heat analyses do not examine the compounding health
risks that individuals could suffer during heat waves with high ozone concentrations in the air. First-
order connectivity was achieved in limited cases, such as with the coastal flooding analysis, which
includes projected installation of coastal defenses and provides information on location and timing to
inform where coastal properties may receive ancillary protection; however, improved connectivity
between models would aid in gaining a more complete understanding of climate change impacts on
children in the U.S.

PROJECTIONS OF FUTURE POPULATION OF CHILDREN

Disaggregated population projections were produced at the county level using EPA's ICLUSv2 model.
The spatial pattern of population change in ICLUS is dependent upon underlying assumptions
regarding fertility, migration rate, and international immigration. These assumptions were
parameterized using the storyline of SSP2, which suggests medium levels of fertility, mortality, and
international immigration. The choice of this population scenario versus another could significantly
influence the estimated impacts across sectors, particularly those most affected by changes in
population and economic growth. Recent demographic trends in the U.S. suggest that population
growth lies closer to the mid-range scenarios consistent with SSP2 (less than 0.5% per year-
contrasted with SSP5, which projects population increases over 1% per year).59 These choices for
future population represent a reasonable central case, but use of other population projections would
affect the results reported. The purpose of this analysis is to focus on understanding the differences
between impacts under multiple climate scenarios. As a result, the exploration of uncertainty

April 2023

18


-------
Climate Change and Children's Health and Well-Being in the United States

surrounding the use of the central case population projection is deferred to future work and the
robust literature exploring the differences amongst scenarios.

SOCIOECONOMIC AND DEMOGRAPHIC CHANGE

This report isolates the effects of climate change on socially vulnerable children by using current
demographics to develop projections. The primary rationale for this approach is that long-term
assumptions and forecasts for national changes in demographics have a high degree of uncertainty,
and therefore are unavailable. Shifting demographics and changes to the socioeconomic statuses and
characteristics of populations will alter the spatial distributions of effects and magnitude of
population sensitivity and vulnerability to climate change. Therefore, the results of the social
vulnerability analyses should be interpreted with this limitation in mind, as actual impacts could be
larger or smaller based on changing demographics.

CONSIDERATION OF ADAPTATION

Populations are likely to adapt to climate change in many ways, with some actions limiting the impact
of climatic exposure, and other actions likely exacerbating impacts. Many of the same factors that
contribute to exposure to climate hazards also influence the ability of individuals and communities to
adapt to climate variability and change. Socioeconomic status, the condition and accessibility of
infrastructure, the accessibility of healthcare, specific demographic characteristics, and other
institutional resources all contribute to the timeliness and effectiveness of adaptive capacity.60

The detailed analyses of this report treat adaptation in unique ways, with some sectors directly
modeling the implications of adaptation responses, and others implicitly incorporating well-
established pathways for adapting to climate stress. For example, most analyses incorporate
empirically-based accounting of individual, community, and infrastructure adaptation in estimating a
climate stressor-response function (i.e., they reflect historical responses to these stressors). As
climate stress worsens and expands geographically, historical adaptation actions implicitly are
incorporated in the estimated response function, and by extension in the estimates presented here,
but do not include new adaptive actions. The heat analysis explicitly holds baseline air conditioning
use constant to underscore the risks associated with no further investment in cooling systems in
schools and homes. The coastal flooding analysis employs a simulation modeling approach that
allows for incorporation of baseline adaptation actions; as an example, continuing and expanding
beach nourishment projects. These simulation modeling approaches also facilitate future adoption of
more complex and extensive adaptive actions, such as changing maintenance practices and
extending seawall protections, which constitute new adaptation scenarios. To the extent that future
adaptation actions beyond those considered are implemented in response to ongoing climate
change, future impacts would likely be lower than estimated in this report.

Adaptation actions that extend beyond historically implemented practices and baseline
infrastructure investments require planning, potentially complex financing, maintenance costs, and
efficacy evaluations with consideration for the specific human and natural environment contexts.
Adaptation plans, therefore, typically are developed and implemented at local scales. The general
adaptation scenarios considered in the analyses of this report do not capture the complex issues

April 2023

19


-------
Climate Change and Children's Health and Well-Being in the United States

driving adaptation decision-making at local and regional scales. For example, the coastal flooding
chapter considers the cost effectiveness of adaptive responses to sea level rise inundation and storm
surge damages by comparing the costs of protection to the value of properties at risk of destruction.
While many factors at the property, community, regional, and national levels will determine adaptive
responses to coastal risks, this sectoral analysis uses the simplistic cost/benefit metric to enable
consistent comparisons for the entire coastline. That said, the adaptation scenarios and estimates
presented in all sections of this report should not be construed as recommending any specific policy
or adaptive action.

GEOGRAPHIC COVERAGE

This report does not examine impacts and damages occurring outside of U.S. borders. Aside from the
inherent value of people and ecosystems around the world, these impacts could affect the U.S.
through changes in migration, impacts on trade, and concerns for conflict and national security. In
addition, the geographic focus of this report is on CONUS, with the detailed analyses excluding
Hawai'i, Alaska, and the U.S. territories (although the District of Columbia is included). The main
reason is that the underlying literature for this report (that is, the sectoral impact models referred to
at the beginning of this Appendix, and in each of the other sector- and climate stressor-specific
Technical Appendices) limits the spatial domain to CONUS. This omission may be particularly
important given the unique climate change vulnerabilities of these locales, socioeconomic
characteristics that often define them, and the subsequent effects on their populations.

SUMMARY

The influence of the sources of uncertainty on the risks of climate change impacting children's health
is difficult to estimate. In theory, a quantitative estimate of the influence of different GCMs in the
climate impact step can be performed to estimate the sensitivity of results to this source of variation
in climate outcomes. Further, the influence of different socioeconomic inputs, sampling margins of
error for the ACS data, or statistical measurement error from certain exposure-response
relationships, or perhaps other sources of uncertainty as well, might be estimated quantitatively.
Many of the underlying peer-reviewed studies relied on within this report perform these types of
analyses to inform readers of the uncertainty associated with each estimate presented. For this
report and the analyses, attempting to combine any quantitative results on uncertainty across
analytic steps would necessarily involve mixing estimates of variability (e.g., across GCMs) with
estimates of statistical uncertainty (e.g., for ACS margins of error, or the impacts that rely on
statistically estimated exposure-response relationships). Moreover, a combined estimate of
uncertainty would ignore other sources of uncertainty that cannot be easily quantified, such as
structural uncertainty associated with the choice of a single sector impacts model, and potential
correlations in sources of uncertainty that may not be fully independent, such as many GCMs sharing
a common structural foundation. Consequently, this report relies on an approach of identifying the
key sources of uncertainty and attempting to qualitatively characterize the potential influence of
each source of uncertainty on the overall results.

April 2023

20


-------
Climate Change and Children's Health and Well-Being in the United States

REFERENCES

1	U.S. Environmental Protection Agency. 2021. "Climate Change and Social Vulnerability in the United States: A

Focus on Six Impacts. U.S. Environmental Protection Agency." EPA 430-R-21-003. Available at:
https://www.epa.gov/cira/social-vulnerability-report

2	Hellden, D., Andersson, C., Nilsson, M., Ebi, K.L, Friberg, P. and Alfven, T., 2021. Climate change and child health:
a scoping review and an expanded conceptual framework. The Lancet Planetary Health, 5(3), pp.el64-el75.

3	Wu, J., Snell, G. and Samji, H., 2020. Climate anxiety in young people: a call to action. The Lancet Planetary

Health, 4(10), pp.e435-e436.

4	Clayton, S., Manning, C.M., Krygsman, K., and Speiser, M., 2017. "Mental health and our changing climate:

Impacts, implications, and guidance." Washington, D.C.: American Psychological Association, and ecoAmerica.

5	Fritze, J.G., Blashki, G.A., Burke, S. and Wiseman, J., 2008. Hope, despair and transformation: Climate change and
the promotion of mental health and wellbeing. International journal of mental health systems, 2(1), pp.1-10.

6	Crandon, T.J., Scott, J.G., Charlson, F.J. and Thomas, H.J., 2022. A social-ecological perspective on climate anxiety

in children and adolescents. Nature Climate Change, 12(2), pp.123-131.

7	Hickman, C., Marks, E., Pihkala, P., Clayton, S., Lewandowski, R.E., Mayall, E.E., Wray, B., Mellor, C. and van
Susteren, L, 2021. Climate anxiety in children and young people and their beliefs about government responses
to climate change: a global survey. The Lancet Planetary Health, 5(12), pp.e863-e873.

8	Felix, E., Rubens, S. and Hambrick, E., 2020. The relationship between physical and mental health outcomes in
children exposed to disasters. Current Psychiatry Reports, 22(7), pp.1-7.

9	Fuchs, R., Glaude, M., Hansel, T., Osofsky, J. and Osofsky, H., 2021. Adolescent risk substance use behavior,

posttraumatic stress, depression, and resilience: Innovative considerations for disaster recovery. Substance
Abuse, 42(3), pp.358-365.

10	Knight, Z.G., 2017. A proposed model of psychodynamic psychotherapy linked to Erik Erikson's eight stages of
psychosocial development. Clinical Psychology & Psychotherapy, 24(5), pp.1047-1058.

11	Dodgen, D., Donato, D., Kelly, N., La Greca, A., Morganstein, J., Reser, J., Ruzek, J., Schweitzer, S., Shimamoto,
M.M., Tart, K.T. and Ursano, R., 2016. Ch. 8: Mental Health and Well-Being (pp. 217-246). US Global Change
Research Program, Washington, DC.

12	Gunter, R.W. and Whittal, M.L., 2010. Dissemination of cognitive-behavioral treatments for anxiety disorders:
Overcoming barriers and improving patient access. Clinical Psychology Review, 30(2), pp. 194-202.

13	Mojtabai, R., 2005. Trends in contacts with mental health professionals and cost barriers to mental health care
among adults with significant psychological distress in the United States: 1997-2002. American Journal of Public
Health, 95(11), pp.2009-2014.

14	Belova, A., Gould, C.A., Munson, K., Howell, M., Trevisan, C., Obradovich, N. and Martinich, J., 2022. Projecting
the suicide burden of climate change in the United States. GeoHealth, p.e2021GH000580.

15	Robinson, L.R., Holbrook, J.R., Bitsko, R.H., Hartwig, S.A., Kaminski, J.W., Ghandour, R.M., Peacock, G., Heggs, A.
and Boyle, C.A., 2017. Differences in health care, family, and community factors associated with mental,
behavioral, and developmental disorders among children aged 2-8 years in rural and urban areas—United
States, 2011-2012. MMWR Surveillance Summaries, 66(8), p.l.

16	Chien, A.T., Leyenaar, J., Tomaino, M., Woloshin, S., Leininger, L., Barnett, E.R., McLaren, J.L and Meara, E.,
2022. Difficulty Obtaining Behavioral Health Services for Children: A National Survey of Multiphysician Practices.
The Annals of Family Medicine, 20(1), pp.42-50.

17	Barnes, T.N., Smith, S.W. and Miller, M.D., 2014. School-based cognitive-behavioral interventions in the
treatment of aggression in the United States: A meta-analysis. Aggression and Violent Behavior, 19(4), pp.311-
321.

18	Dickerson, J.F., Lynch, F.L, Leo, M.C., DeBar, L.L., Pearson, J. and Clarke, G.N., 2018. Cost-effectiveness of
cognitive behavioral therapy for depressed youth declining antidepressants. Pediatrics, 141(2).

19	Ross, E.L., Vijan, S., Miller, E.M., Valenstein, M. and Zivin, K., 2019. The cost-effectiveness of cognitive behavioral
therapy versus second-generation antidepressants for initial treatment of major depressive disorder in the
United States: a decision analytic model. Annals of Internal Medicine, 171(11), pp.785-795.

April 2023

21


-------
Climate Change and Children's Health and Well-Being in the United States

20	Fan, Q., DuPont-Reyes, M.J., Hossain, M.M., Chen, L.S., Lueck, J. and Ma, P., 2022. Racial and ethnic differences
in major depressive episode, severe role impairment, and mental health service utilization in US adolescents.
Journal of Affective Disorders, 306, pp. 190-199.

21	Lindly, O., Eaves, M.C., Xu, Y., Tarazi, C.L., Rao, S.R. and Kuhlthau, K.A., 2022. Therapy use for US school-aged
children with developmental disabilities: State variation and determinants. Disability and Health Journal, 15(1),
p.101198.

22	Chien, A.T., Leyenaar, J., Tomaino, M., Woloshin, S., Leininger, L., Barnett, E.R., McLaren, J.L and Meara, E.,
2022. Difficulty Obtaining Behavioral Health Services for Children: A National Survey of Multiphysician Practices.
The Annals of Family Medicine, 20(1), pp.42-50.

23	Goldsmith, L, Raditz, V. and Mendez, M., 2022. Queer and present danger: understanding the disparate impacts
of disasters on LGBTQ+ communities. Disasters, 46(A), pp.946-973.

24	Simmonds, K.E., Jenkins, J., White, B., Nicholas, P.K. and Bell, J., 2022. Health impacts of climate change on
gender diverse populations: A scoping review. Journal of Nursing Scholarship, 54(1), pp.81-91.

25	Rhoades, H., Rusow, J.A., Bond, D., Lanteigne, A., Fulginiti, A. and Goldbach, J.T., 2018. Homelessness, mental
health and suicidality among LGBTQ youth accessing crisis services. Child Psychiatry & Human Development,
49(A), pp.643-651.

26	Sarofim, M.C., Martinich, J., Neumann, J.E., Willwerth, J., Kerrich, Z., Kolian, M., Fant, C. and Hartin, C., 2021. A
temperature binning approach for multi-sector climate impact analysis. Climatic Change, 165(1), pp.1-18.

27	U.S. Global Change Research Program. 2015. U.S. Global Change Research Program General Decisions
Regarding Climate-Related Scenarios for Framing the Fourth National Climate Assessment. USGCRP Scenarios
and Interpretive Science Coordinating Group. Available online at
https://scenarios.globalchange.gov/accouncement/1158

28	Sarofim, M.C., Martinich, J., Neumann, J.E., Willwerth, J., Kerrich, Z., Kolian, M., Fant, C. and Hartin, C., 2021. A
temperature binning approach for multi-sector climate impact analysis. Climatic Change, 165(1), pp.1-18.

29	Taylor, K.E., Stouffer, R.J. and Meehl, G.A., 2012. An overview of CMIP5 and the experiment design. Bulletin of
the American Meteorological Society, 93(A), pp.485-498.

30	von Salzen, K., Scinocca, J.F., McFarlane, N.A., Li, J., Cole, J.N., Plummer, D., Verseghy, D., Reader, M.C., Ma, X.,
Lazare, M. and Solheim, L, 2013. The Canadian fourth generation atmospheric global climate model (CanAM4).
Part I: representation of physical processes. Atmosphere-Ocean, 51(1), pp.104-125.

31	Donner, L.J., Wyman, B.L., Hemler, R.S., Horowitz, L.W., Ming, Y., Zhao, M., Golaz, J.C., Ginoux, P., Lin, S.J.,
Schwarzkopf, M.D. and Austin, J., 2011. The dynamical core, physical parameterizations, and basic simulation
characteristics of the atmospheric component AM3 of the GFDL global coupled model CM3. Journal of Climate,
24(13), pp.3484-3519.

32	Gent, P.R., Danabasoglu, G., Donner, L.J., Holland, M.M., Hunke, E.C., Jayne, S.R., Lawrence, D.M., Neale, R.B.,
Rasch, P.J., Vertenstein, M. and Worley, P.H., 2011. The community climate system model version 4. Journal of
Climate, 24(19), pp.4973-4991.

33	Neale, R.B., Richter, J., Park, S., Lauritzen, P.H., Vavrus, S.J., Rasch, P.J. and Zhang, M., 2013. The mean climate of
the Community Atmosphere Model (CAM4) in forced SST and fully coupled experiments. Journal of Climate,
26(14), pp.5150-5168.

34	Schmidt, G.A., Ruedy, R., Hansen, J.E., Aleinov, I., Bell, N., Bauer, M., Bauer, S., Cairns, B., Canuto, V., Cheng, Y.
and Del Genio, A., 2006. Present-day atmospheric simulations using GISS ModelE: Comparison to in situ,
satellite, and reanalysis data. Journal of Climate, 19(2), pp. 153-192.

35	Collins, W.J., Bellouin, N., Doutriaux-Boucher, M., Gedney, N., Halloran, P., Hinton, T., Hughes, J., Jones, C.D.,
Joshi, M., Liddicoat, S. and Martin, G., 2011. Development and evaluation of an Earth-System model-HadGEM2.
Geoscientific Model Development, 4(A), pp. 1051-1075.

36	Davies, T., Cullen, M.J., Malcolm, A.J., Mawson, M.H., Staniforth, A., White, A.A. and Wood, N., 2005. A new
dynamical core for the Met Office's global and regional modelling of the atmosphere. Quarterly Journal of the
Royal Meteorological Society: A Journal of the Atmospheric Sciences, Applied Meteorology and Physical
Oceanography, 131(608), pp.1759-1782.

April 2023

22


-------
Climate Change and Children's Health and Well-Being in the United States

37	Watanabe, M., Suzuki, T., O'ishi, R., Komuro, Y., Watanabe, S., Emori, S., Takemura, T., Chikira, M., Ogura, T.,
Sekiguchi, M. and Takata, K., 2010. Improved climate simulation by MIROC5: mean states, variability, and climate
sensitivity. Journal of Climate, 23(23), pp.6312-6335.

38	U.S. Environmental Protection Agency. 2017. Multi-model framework for quantitative sectoral impacts analysis:
a technical report for the Fourth National Climate Assessment. U.S. Environmental Protection Agency, EPA 430-
R-17-001. Available at: https://www.epa.gov/cira/multi-model-framework-quantitative-sectoral-impacts-
analysis

39	Sanderson, B.M., Wehner, M. and Knutti, R., 2017. Skill and independence weighting for multi-model
assessments. Geoscientific Model Development, 10(6), pp.2379-2395.

40	Sanderson, B.M., Wehner, M. and Knutti, R., 2017. Skill and independence weighting for multi-model
assessments. Geoscientific Model Development, 10(6), pp.2379-2395.

41	Sarofim, M.C., Martinich, J., Neumann, J.E., Willwerth, J., Kerrich, Z., Kolian, M., Fant, C. and Hartin, C., 2021. A
temperature binning approach for multi-sector climate impact analysis. Climatic Change, 165(1), pp.1-18.

42	U.S. Bureau of Reclamation, Climate Analytics Group, Climate Central, Lawrence Livermore National Laboratory,
Santa Clara University, Scripps Institution of Oceanography, U.S. Army Corps of Engineers, and U.S. Geological
Survey, 2016: Downscaled CMIP3 and CMIP5 Climate Projections: Release of Downscaled CMIP5 Climate
Projections, Comparison with Preceding Information, and Summary of User Needs. Available online at
http://gdo-dcp.ucllnl.org/downscaled_cmip_projections/techmemo/downscaled_climate.pdf. Data available at
http://gdo-dcp.ucllnl.org/downscaled_cmip_projections/.

43	University of California San Diego. 2017: LOCA statistical downscaling. Scripps Institution of Oceanography.
Available online at http://loca.ucsd.edu/

44	U.S. Environmental Protection Agency. 2019. Locating and Selecting Scenarios Online, https://lasso.epa.gov/

45	Oppenheimer, M., Campos, M., Warren, R., Birkmann, J., Luber, G., O'Neill, B., Takahashi, K., Brklacich, M.,
Semenov, S., Licker, R. and Hsiang, S., 2015. Emergent risks and key vulnerabilities. In Climate Change 2014
Impacts, Adaptation and Vulnerability: Part A: Global and Sectoral Aspects (pp. 1039-1100). Cambridge
University Press.

46	Sweet, W.V., Kopp, R.E., Weaver, C.P., Obeysekera, J., Horton, R.M., Thieler, E.R. and Zervas, C., 2017. Global and
Regional Sea Level Rise Scenarios for the United States (No. CO-OPS 083).

47	Sarofim, M.C., Martinich, J., Neumann, J.E., Willwerth, J., Kerrich, Z., Kolian, M., Fant, C. and Hartin, C., 2021. A
temperature binning approach for multi-sector climate impact analysis. Climatic Change, 165(1), pp.1-18.

48ICLUS v2 population estimates available at: https://www.epa.gov/gcx/iclus-fourth-national-climate-assessment

49	U.S. Environmental Protection Agency. 2021. Climate Change and Social Vulnerability in the United States: A
Focus on Six Impacts. U.S. Environmental Protection Agency, EPA430-R-21-003.

50	U.S. Environmental Protection Agency. 2017. Multi-model framework for quantitative sectoral impacts analysis:
a technical report for the Fourth National Climate Assessment. U.S. Environmental Protection Agency, EPA 430-
R-17-001. Available at: https://www.epa.gov/cira/multi-model-framework-quantitative-sectoral-impacts-
analysis

51	Manson S, Schroeder J, Van Riper D, KuglerT, and RugglesS. IPUMS National Historical Geographic Information
System: Version 15.0 American Community Survey 2014-2018a. Minneapolis, MN: IPUMS. 2020. Note that the
NHGIS field codes in Table 3 are unique to IPUMS - ACS table numbers differ from the field codes shown here,
but the data are identical.

52	U.S. Census Bureau. "How the Census Bureau Measures Poverty." Available online at:
https://www.census.gov/topics/income-poverty/poverty/guidance/poverty-measures.html

53	Paltsev, S., Monier, E., Scott, J., Sokolov, A. and Reilly, J., 2015. Integrated economic and climate projections for
impact assessment. Climatic Change, 131(1), pp.21-33.

54	Monier, E., Gao, X., Scott, J.R., Sokolov, A.P. and Schlosser, C.A., 2015. A framework for modeling uncertainty in
regional climate change. Climatic Change, 131(1), pp.51-66.

55	Monier, E., Gao, X., Scott, J.R., Sokolov, A.P. and Schlosser, C.A., 2015. A framework for modeling uncertainty in
regional climate change. Climatic Change, 131(1), pp.51-66.

April 2023

23


-------
Climate Change and Children's Health and Well-Being in the United States

56	Mills, D., Jones, R., Carney, K., St Juliana, A., Ready, R., Crimmins, A., Martinich, J., Shouse, K., DeAngelo, B. and
Monier, E., 2015. Quantifying and monetizing potential climate change policy impacts on terrestrial ecosystem
carbon storage and wildfires in the United States. Climatic Change, 131(1), pp.163-178.

57	U.S. Environmental Protection Agency. 2003. "Children's Health Valuation Handbook." Available online at:
https://www.epa.gov/environmental-economics/childrens-health-valuation-handbook-2003

58	Huber, V., Schellnhuber, H.J., Arnell, N.W., Frieler, K., Friend, A.D., Gerten, D., Haddeland, I., Kabat, P., Lotze-
Campen, H., Lucht, W. and Parry, M., 2014. Climate impact research: beyond patchwork. Earth System Dynamics,
5(2), pp.399-408.

59	Wear D.N., Prestemon J.P., 2019. Spatiotemporal downscaling of global population and income scenarios for the
United States. PLoS ONE 14(7): e0219242. https://doi.org/10.1371/journal.pone.0219242 and William H. Frey
analysis of U.S. Census Bureau historical population estimates, including 2020-2022 annual estimates released
December 22, 2022, available at https://www.brookings.edu/research/new-census-estimates-show-a-tepid-rise-
in-u-s-population-growth-buoyed-by-im migration/

60	Gamble J.L., Balbus J., Berger M., Bouye K., Campbell V., Chief K., Conlon K., Crimmins A., Flanagan B., Gonzalez-
Maddux C., Hallisey E., Hutchins S., Jantarasami L, Khoury S., Kiefer M., Kolling J., Lynn K., Manangan A.,
McDonald M., Morello-Frosch R., Redsteer M.H., Sheffield P., Thigpen Tart K., Watson J., Whyte K.P., and Wolkin
A.F., 2016. Ch. 9: Populations of Concern. The Impacts of Climate Change on Human Health in the United States:
A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 247-286.

April 2023

24


-------