&EPA
United States
Envirofunmlal Protection
Agnncy
Health Risk and Exposure Assessment
for Ozone
Final Report
Chapter 5 Appendices
-------
This page left intentionally blank
-------
EPA-452/R-14-004c
August 2014
Health Risk and Exposure Assessment for Ozone
Final Report
Chapter 5 Appendices
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Risk and Benefits Group
Research Triangle Park, North Carolina 27711
-------
DISCLAIMER
This final document has been prepared by staff from the Risk and Benefits Group, Health
and Environmental Impacts Division, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Any findings and conclusions are those of the authors and do
not necessarily reflect the views of the Agency.
Questions related to this document should be addressed to Dr. Bryan Hubbell, U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, C539-07,
Research Triangle Park, North Carolina 27711 (email: hubbell.bryan@epa.gov).
-------
CHAPTER 5 APPENDICES
APPENDIX 5A:
APPENDIX 5B:
APPENDIX 5C:
APPENDIX 5D:
APPENDIX 5E:
APPENDIX 5F:
APPENDIX 5G:
Title
Description of the Air Pollutants Exposure Model
(APEX)
Inputs to the APEX Exposure Model
Generation of Adult and Child Census-tract
Level Asthma Prevalence using NHIS (2006-
2010) and US Census (2000) Data
Variability Analysis and Uncertainty
Characterization
Updated Analysis of Air Exchange Rate Data:
Memorandum from ICF International
Detailed Exposure Results
Targeted Evaluation of Exposure Model Input
and Output Data
Pages
(5A-1 to 5A-22)
(5B-1 to 5B-38)
(5C-1 to 5C-80)
(5D-1 to 5D-12)
(5E-1 to 5E-25)
(5F-1 to 5F-55)
(5G-1 to 5G-62)
-------
This page left intentionally blank
-------
APPENDIX 5A
Description of the Air Pollutants Exposure Model (APEX)
Table of Contents
5A-1. OVERVIEW 5A-1
5A-2. MODEL INPUTS 5A-3
5A-3. DEMOGRAPHIC CHARACTERISTICS 5A-4
5A-4. ATTRIBUTES OF INDIVIDUALS 5A-4
5A-5. CONSTRUCTION OF LONGITUDINAL DIARY SEQUENCE 5A-5
5A-6. KEY PHYSIOLOGICAL PROCESSES MODELED 5A-7
5A-7. ESTIMATING MICROENVIRONMENTAL CONCENTRATIONS 5A-9
5A-7.1. Mass Balance Model 5A-9
5A-7.2. Factors Model 5A-16
5A-8. EXPOSURE AND DOSE TIME SERIES CALCULATIONS 5A-17
5A-9. MODEL OUTPUT 5A-19
5A-10. REFERENCES 5A-21
5A-i
-------
List of Tables
Table 5A-l. Ventilation coefficient parameter estimates (hi) and residuals distributions (ei)
from Graham and McCurdy (2005) 5A-9
List of Figures
Figure 5A-1. Illustration of the mass balance model used by APEX 5A-10
Figure 5A-2. Example of microenvironmental and exposure concentrations for a simulated
individual over a 48 hours simulation. (H: home, A: automobile, S: school, P:
playground, O: outdoors at home) 5A-l8
Figure 5A-3. The percent of simulated children (ages 5-18) at or above 8-hour average Os
exposures while at moderate or greater exertion 5A-20
5A-ii
-------
This Appendix briefly describes the EPAs Air Pollutants Exposure (APEX) model.
5A-1. OVERVIEW
APEX is the human inhalation exposure model within the Total Risk Integrated
Methodology (TRIM) framework (US EPA 2012a,b). APEX is conceptually based on the
probabilistic NAAQS Exposure Model (pNEM) that was used to estimate population exposures
for the 1996 Os NAAQS review (Johnson et al., 1996a; 1996b; 1996c). Since that time the
model has been restructured, improved, and expanded to reflect conceptual advances in the
science of exposure modeling and newer input data available for the model. Key improvements
to algorithms include replacement of the cohort approach with a probabilistic sampling approach
focused on individuals, accounting for fatigue and oxygen debt after exercise in the calculation
of ventilation rates (Isaacs et al., 2008), and new approaches for construction of longitudinal
activity patterns for simulated persons (Glen et al., 2008; Rosenbaum et al., 2008). Major
improvements to data input to the model include updated air exchange rates (AERs), population
census and commuting data, and the daily time-location-activities database. These
improvements are described later in this and other Chapter 5 Appendices.
APEX estimates human exposure to criteria and toxic air pollutants at local, urban, or
regional scales using a stochastic, microenvironmental approach. That is, the model randomly
selects data on a sample of hypothetical individuals in an actual population database and
simulates each individual's movements through time and space (e.g., at home, in vehicles) to
estimate their exposure to the pollutant. APEX can assume people live and work in the same
general area (i.e., that the ambient air quality is the same at home and at work) or optionally can
model commuting and thus exposure at the work location for individuals who work.
The APEX model is a microenvironmental, longitudinal human exposure model for
airborne pollutants. It is applied to a specified study area, which is typically a metropolitan area.
The time period of the simulation is typically one year, but can easily be made either longer or
shorter. APEX uses census data, such as gender and age, to generate the demographic
characteristics of simulated individuals. It then assembles a composite activity diary to represent
the sequence of activities and microenvironments that the individual experiences. Each
microenvironment has a user-specified method for determining air quality. The inhalation
exposure in each microenvironment is simply equal to the air concentration in that
microenvironment. When coupled with breathing rate information and a physiological model,
various measures of dose can also be calculated.
The term microenvironment is intended to represent the immediate surroundings of an
individual, in which the pollutant of interest is assumed to be well-mixed. Time is modeled as a
sequence of discrete time steps called events. In APEX, the concentration in a microenvironment
5A-1
-------
may change between events. For each microenvironment, the user specifies the method of
concentration calculation (either mass balance or regression factors, described later in this
paper), the relationship of the microenvironment to the ambient air, and the strength of any
pollutant sources specific to that microenvironment. Because the microenvironments that are
relevant to exposure depend on the nature of the target chemical and APEX is designed to be
applied to a wide range of chemicals, both the total number of microenvironments and the
properties of each are free to be specified by the user.
The ambient air data are provided as input to the model in the form of time series at a
list of specified locations. Typically, hourly air concentrations are used, although temporal
resolutions as small as one minute may be used. The spatial range of applicability of a given
ambient location is called an air district. Any number of air districts can be accommodated in a
model run, subject only to computer hardware limitations. In principle, any microenvironment
could be found within a given air district. Therefore, to estimate exposures as an individual
engages in activities throughout the period it is necessary to determine both the
microenvironment and the air district that apply for each event.
An exposure event is determined by the time reported in the activity diary; during any
event the district, microenvironment, ambient air quality, and breathing rate are assumed to
remain fixed. Since the ambient air data change every hour, the maximum duration of an event
is limited to one hour. The event duration may be less than this (as short as one minute) if the
activity diary indicates that the individual changes microenvironments or activities performed
within the hour.
An APEX simulation includes the following steps:
1. Characterize the study area - APEX selects sectors (e.g., census tracts) within a study area
based on user-defined criteria and thus identifies the potentially exposed population and
defines the air quality and weather input data required for the area.
2. Generate simulated individuals - APEX stochastically generates a sample of simulated
individuals based on the census data for the study area and human profile distribution data
(such as age-specific employment probabilities). The user must specify the size of the
sample. The larger the sample, the more representative it is of the population in the study
area and the more stable the model results are (but also the longer the computing time).
3. Construct a long-term sequence of activity events and determine breathing rates - APEX
constructs an event sequence (activity pattern) spanning the period of simulation for each
simulated person. The model then stochastically assigns breathing rates to each event, based
on the type of activity and the physical characteristics of the simulated person.
4. Calculate pollutant concentrations in microenvironments - APEX enables the user to define
any microenvironment that individuals in a study area would visit. The model then
calculates concentrations of each pollutant in each of the microenvironments.
5A-2
-------
5. Calculate pollutant exposures for each simulated individual - Microenvironmental
concentrations are time weighted based on individuals' events (i.e., time spent in the
microenvironment) to produce a sequence of time-averaged exposures (or minute by minute
time series) spanning the simulation period.
6. Estimate dose - APEX can also calculate the dose time series for each of the simulated
individuals based on the exposures and breathing rates for each event. For Os, the adverse
health metric of interest is decrement in forced expiratory volume occurring in one second
(FEVi). This algorithm responsible for combining the time series of APEX estimated
exposure and breathing rates for individuals is discussed in greater detail in the main body of
the HRE A, Chapters.
The model simulation continues until exposures are determined for the user-specified
number of simulated individuals. APEX then calculates population exposure statistics (such as
the number of exposures exceeding user-specified levels) for the entire simulation and writes out
tables of distributions of these statistics.
5A-2. MODEL INPUTS
APEX requires certain inputs from the user. The user specifies the geographic area and
the range of ages and age groups to be used for the simulation. Hourly (or shorter) ambient air
quality and hourly temperature data must be furnished for the entire simulation period. Other
hourly meteorological data (humidity, wind speed, wind direction, precipitation) can be used by
the model to estimate microenvironmental concentrations, but are optional.
In addition, most variables used in the model algorithms are represented by user-specified
probability distributions which capture population variability. APEX provides great flexibility in
defining model inputs and parameters, including options for the frequency of selecting new
values from the probability distributions. The model also allows different distributions to be
used at different times of day or on different days, and the distribution can depend conditionally
on values of other parameters. The probability distributions available in APEX include beta,
binary, Cauchy, discrete, exponential, extreme value, gamma, logistic, lognormal, loguniform,
normal, off/on, Pareto, point (constant), triangle, uniform, Weibull, and nonparametric
distributions. Minimum and maximum bounds can be specified for each distribution if a
truncated distribution is appropriate. There are two options for handling truncation. The
generated samples outside the truncation points can be set to the truncation limit; in this case,
samples "stack up" at the truncation points. Alternatively, new random values can be selected, in
which case the probability outside the limits is spread over the specified range, and thus the
probabilities inside the truncation limits will be higher than the theoretical untruncated
distribution.
5A-3
-------
5A-3. DEMOGRAPHIC CHARACTERISTICS
The starting point for constructing a simulated individual is the population census
database; this contains population counts for each combination of age, gender, race, and sector.
The user may decide what spatial area is represented by a sector, but the default input file defines
a sector as a census tract. Census tracts are variable in both geographic size and population
number, though usually have between 1,500 and 8,000 persons. Currently, the default file
contains population counts from the 2000 census for every census tract in the United States, thus
the default file should be sufficient for most exposure modeling purposes. The combination of
age, gender, race, and sector are selected first. The sector becomes the home sector for the
individual, and the corresponding air district becomes the home district. The probabilistic
selection of individuals is based on the sector population and demographic composition, and
taken collectively, the set of simulated individuals constitutes a random sample from the study
area.
The second step in constructing a simulated individual is to determine their employment
status. This is determined by a probability which is a function of age, gender, and home sector.
An input file is provided which contains employment probabilities from the 2000 census for
every combination of age (16 and over), gender, and census tract. APEX assumes that persons
under age 16 do not commute. For persons who are determined to be workers, APEX then
randomly selects a work sector, based on probabilities determined from the commuting matrix.
The work sector is used to assign a work district for the individual that may differ from the home
district, and thus different ambient air quality may be used when the individual is at work.
The commuting matrix contains data on flows (number of individuals) traveling from a
given home sector to a given work sector. Based on commuting data from the 2000 census, a
commuting data base for the entire United States has been prepared. This permits the entire list
of non-zero flows to be specified on one input file. Given a home sector, the number of
destinations to which people commute varies anywhere from one to several hundred other tracts.
5A-4. ATTRIBUTES OF INDIVIDUALS
In addition to the above demographic information, each individual is assigned status and
physiological attributes. The status variables are factors deemed important in estimating
microenvironmental concentrations, and are specified by the user. Status variables can include,
but are not limited to, people's housing type, whether their home has air conditioning, whether
they use a gas stove at home, whether the stove has a gas pilot light, and whether their car has air
conditioning. Physiological variables are important when estimating pollutant specific dose.
These variables could include height, weight, blood volume, pulmonary diffusion rate, resting
metabolic rate, energy conversion factor (liters of oxygen per kilocalorie energy expended),
5A-4
-------
hemoglobin density in blood, maximum limit on metabolic equivalents of work (MET) ratios
(see below), and endogenous CO production rate. All of these variables are treated
probabilistically taking into account interdependences where possible, and reflecting variability
in the population.
Two key personal attributes determined for each individual in this assessment are body
mass (BM) and body surface area (BSA). Each simulated individual's body mass was randomly
sampled from age- and gender-specific body mass distributions generated from National Health
and Nutrition Examination Survey (NHANES) data for the years 1999-2004.l Details in their
development and the parameter values are provided by Isaacs and Smith (2005). Then age- and
gender-specific body surface area can be estimated for each simulated individual. Briefly, the
BSA calculation is based on logarithmic relationships developed by Burmaster (1998) that use
body mass as an independent variable as follows:
BSA=e-2'27Sl BM°'6S21 Equation (5A-1)
where,
BSA = body surface area (m2)
BM = body mass (kg)
5A-5. CONSTRUCTION OF LONGITUDINAL DIARY SEQUENCE
The activity diary determines the sequence of microenvironments visited by the
simulated person. A longitudinal sequence of daily diaries must be constructed for each
simulated individual to cover the entire simulation period. The default activity diaries in APEX
are derived from those in the EPA's Consolidated Human Activity Database (CHAD) (US EPA,
2000; 2002), although the user could provide area specific diaries if available. There are over
53,000 CHAD diaries, each covering a 24 hour period, that have been compiled from several
studies. CHAD is essentially a cross-sectional database that, for the most part, only has one
diary per person. Therefore, APEX must assemble each longitudinal diary sequence for a
simulated individual from many single-day diaries selected from a pool of similar people.
APEX selects diaries from CHAD by matching gender and employment status, and by
requiring that age falls within a user-specified range on either side of the age of the simulated
individual. For example, if the user specifies plus or minus 20%, then for a 40 year old
simulated individual, the available CHAD diaries are those from persons aged 32 to 48. Each
1 Demographic (Demo) and Body Measurement (BMX) datasets for each of the NHANES studies were obtained
fromhttp://www.cdc.gov/nchs/nhanes/nhanes_questionnaires.htm.
5A-5
-------
simulated individual therefore has an age window of acceptable diaries; these windows can
partially overlap those for other simulated individuals. This differs from a cohort-based
approach, where the age windows are fixed and non-overlapping. The user may optionally
request that APEX allow a decreased probability for selecting diaries from ages outside the
primary age window, and also for selecting diaries from persons of missing gender, age, or
employment status. These options allow the model to continue the simulation when diaries are
not available within the primary window.
The available CHAD diaries are classified into diary pools, based on the temperature and
day of the week. The model will select diaries from the appropriate pool for days in the
simulation having matching temperature and day type characteristics. The rules for defining
these pools are specified by the user. For example, the user could request that all diaries from
Monday to Friday be classified together, and Saturday and Sunday diaries in another class.
Alternatively, the user could instead create more than two classes of weekdays, combine all
seven days into one class, or split all seven days into separate classes.
The temperature classification can be based either on daily maximum temperature, daily
average temperature, or both. The user specifies both the ranges and numbers of temperatures
classes. For example, the user might wish to create four temperature classes and set their ranges
to below 50 °F, 50-69 °F, 70-84 °F, and above a daily maximum of 84 °F. Then day type and
temperature classes are combined to create the diary pools. For example, if there are four
temperature classes and two day type classes, then there will be eight diary pools.
APEX then determines the day-type and the applicable temperature for each person's
simulated day. APEX allows multiple temperature stations to be used; the sectors are
automatically mapped to the nearest temperature station. This may be important for study areas
such as the greater Los Angeles area, where the inland desert sectors may have very different
temperatures from the coastal sectors. For selected diaries, the temperature in the home sector of
the simulated person is used. For each day of the simulation, the appropriate diary pool is
identified and a CHAD dairy is randomly drawn. When a diary for every day in the simulation
period has been selected, they are concatenated into a single longitudinal diary covering the
entire simulation for that individual. APEX contains three algorithms for stochastically selecting
diaries from the pools to create the longitudinal diary. The first method selects diaries at random
after stratification by age, gender, and diary pool; the second method selects diaries based on
metrics related to exposure (e.g., time spent outdoors) with the goal of creating longitudinal
diaries with variance properties designated by the user (Glen et al., 2008); and the third method
uses a clustering algorithm to obtain more realistic recurring behavioral patterns (Rosenbaum
2008).
5A-6
-------
The final step in processing the activity diary is to map the CHAD location codes into the
set of APEX microenvironments, supplied by the user as an input file. The user may define the
number of microenvironments, from one up to the number of different CHAD location codes
(which is currently 115).
5A-6. KEY PHYSIOLOGICAL PROCESSES MODELED
Ventilation is a general term describing the movement of air into and out of the lungs.
The rate of ventilation is determined by the type of activity an individual performs which in turn
is related to the amount of oxygen required to perform the activity. Minute or total ventilation
rate is used to describe the volume of air moved in or out of the lungs per minute. Quantitatively,
•
the volume of air breathed in per minute (VI ) is slightly greater than the volume expired per
•
minute (VE ). Clinically, however, this difference is not important, and by convention, the
ventilation rate is always measured by the expired volume.
•
The rate of oxygen consumption (V 01} is related to the rate of energy usage in
performing activities as follows:
Vo2=EEx ECF Equation (5 A-2)
where,
•
V02 = Oxygen consumption rate (liters O2/minute)
EE = Energy expenditure (kcal/minute)
ECF = Energy conversion factor (liters O2/kcal).
The ECF shows little variation and typically, commonly a value between 0.20 and 0.21 is
used to represent the conversion from energy units to oxygen consumption. APEX can randomly
sample from a uniform distribution defined by these lower and upper bounds to estimate an ECF
for each simulated individual. The activity-specific energy expenditure is highly variable and
can be estimated using metabolic equivalents (METs), or the ratios of the rate of energy
consumption for non-rest activities to the resting rate of energy consumption, as follows
EE=MET xRMR Equation (5A-3)
where,
EE = Energy expenditure (kcal/minute)
5A-7
-------
MET = Metabolic equivalent of work (unitless)
RMR = Resting metabolic rate (kcal/minute)
APEX contains distributions of METs for all activities that might be performed by
simulated individuals. APEX randomly samples from the various METs distributions to obtain
values for every activity performed by each individual. Age- and gender-specific RMR are
estimated once for each simulated individual using a linear regression model (see Johnson et al.,
2002)2 as follows
RMR = [b0 + \ (BM} + e]F Equation (5 A-4)
where,
RMR = Resting metabolic rate (kcal/min)
bo = Regression intercept (MJ/day)
hi = Regression slope (MJ/day/kg)
BM = body mass (kg)
e = randomly sampled error term, N{0, se)3 (MJ/day)
F = Factor for converting MJ/day to kcal/min (0.166)
• •
Finally, Graham and McCurdy (2005) describe an approach to estimate VE using V01.
In that report, a series of age- and gender-specific multiple linear regression equations were
derived from data generated in 32 clinical exercise studies. The algorithm accounts for
variability in ventilation rate due to variation in oxygen consumption, the variability within age
groups, and both inter- and intra-personal and variability. The basic algorithm is
• •
\n(VE/BM) = b0 +bl \n(VoilBM) + b2 \n(\ + age) + b3 gender+ eh +ew Equation (5A-5)
where,
In = natural logarithm of variable
•
V ElBM = activity specific ventilation rate, body mass normalized (liter air/kg)
bt = see below
2 The regression equations were adapted by Johnson (2002) using data reported by Schofield (1985). The regression
coefficients and error terms used by APEX are provided in the APEX physiology input file.
3 The value used for each individual is sampled from a normal distribution (N) having a mean of zero (0) and
variability described by the standard error (se)
5A-8
-------
VoilBM = activity specific oxygen consumption rate, body mass normalized
(liter/O2/kg)
age = the age of the individual (years)
gender = gender value (-1 for males and +1 for females)
eb = randomly sampled error term for between persons N{0, se), (liter
air/kg)
ew = randomly sampled error term for within persons N{0, se), (liter
air/kg)
As indicated above, the random error (e) is allocated to two variance components used to
estimate the between-person (inter-individual variability) residuals distribution (eb) and within-
person (intra-individual variability) residuals distribution (ew). The regression parameters bo, bi,
b2, and bs are assumed to be constant over time for all simulated persons, eb is sampled once per
person, while whereas ew varies from event to event. Point estimates of the regression
coefficients and standard errors of the residuals distributions are given in Table 5A-1.
Table 5A-1. Ventilation coefficient parameter estimates (bi) and residuals distributions (e\)
from Graham and McCurdy (2005).
Age
jgroupj
<20
20-<34
34-<61
en-
Regression Coefficients1
bo
4.3675
3.7603
3.2440
2.5826
bi
1.0751
1.2491
1.1464
1.0840
b2
-0.2714
0.1416
0.1856
0.2766
b3
0.0479
0.0533
0.0380
-0.0208
Random Error1
6b
0.0955
0.1217
0.1260
0.1064
ew
0.1117
0.1296
0.1152
0.0676
1 The values of the coefficients and residuals distributions described by Equation (5A-5).
5A-7. ESTIMATING MICROENVIRONMENTAL CONCENTRATIONS
The user provides rules for determining the pollutant concentration in each
microenvironment. There are two available models for calculating microenvironmental
concentrations: mass balance and regression factors. Any indoor microenvironment may use
either model; for each microenvironment, the user specifies whether the mass balance or factors
model will be used.
5A-7.1. Mass Balance Model
The mass balance method assumes that an enclosed microenvironment (e.g., a room
within a home) is a single well-mixed volume in which the air concentration is approximately
spatially uniform. The concentration of an air pollutant in such a microenvironment is estimated
using the following four processes (and illustrated in Figure 5A-1):
5A-9
-------
Inflow of air into the microenvironment;
Outflow of air from the microenvironment;
Removal of a pollutant from the microenvironment due to deposition, filtration, and
chemical degradation; and
Emissions from sources of a pollutant inside the microenvironment.
Microenvironment
Air
outflow
Air
inflow
Indoorsources
V
Removal due to:
•Chemical reactions
•Deposition
•Filtration
Figure 5A-1. Illustration of the mass balance model used by APEX.
Considering the microenvironment as a well-mixed fixed volume of air, the mass balance
equation for a pollutant in the microenvironment can be written in terms of concentration:
• .
out
"removal
Equation (5A-6)
where,
C(t) = Concentration in the microenvironment at time t
C in = Rate of change in C(t) due to air entering the microenvironment
C out = Rate of change in C(t) due to air leaving the microenvironment
C removal = Rate of change in C(t) due to all internal removal processes
C source = Rate of change in C(t) due to all internal source terms
5 A-10
-------
Concentrations are calculated in the same units as the ambient air quality data, e.g., ppm,
ppb, ppt, or |ig/m3. In the following equations concentration is shown only in |ig/m3 for brevity.
The change in microenvironmental concentration due to influx of air, Cm, is given by:
C/n = ^outdoor x ?penetration x ^airexchange Equation (5A-7)
where,
Coutdoor = Ambient concentration at an outdoor microenvironment or outside an
indoor microenvironment (|ig/m3)
/penetration = Penetration factor (unitless)
exchange = Air ex change rate (hr"1)
Since the air pressure is approximately constant in microenvironments that are modeled
in practice, the flow of outside air into the microenvironment is equal to that flowing out of the
microenvironment, and this flow rate is given by the air exchange rate. The air exchange rate
(hr"1) can be loosely interpreted as the number of times per hour the entire volume of air in the
microenvironment is replaced. For some pollutants (especially particulate matter), the process of
infiltration may remove a fraction of the pollutant from the outside air. The fraction that is
retained in the air is given by the penetration f actor /penetration.
A proximity factor (/proximity) and a local outdoor source term are used to account for
differences in ambient concentrations between the geographic location represented by the
ambient air quality data (e.g., a regional fixed-site monitor) and the geographic location of the
microenvironment. That is, the outdoor air at a particular location may differ systematically
from the concentration input to the model representing the air quality district. For example, a
playground or house might be located next to a busy road in which case the air at the playground
or outside the house would have elevated levels for mobile source pollutants such as carbon
monoxide and benzene. The concentration in the air at an outdoor location or directly outside an
indoor microenvironment (Coutdoor) is calculated as:
^outdoor = 'proximity*-1 ambient + ^LocalOutdoorSources Equation (5A-8)
where,
C ambient = Ambient air district concentration (|ig/m3)
5 A-11
-------
/proximity = Proximity factor (unitless)
CiocaiOutdoorSources = the contribution to the concentration at this location from local
sources not represented by the ambient air district
concentration (|ig/m3)
During exploratory analyses, the user may examine how a microenvironment affects
overall exposure by setting the microenvironment's proximity or penetration factor to zero, thus
effectively eliminating the specified microenvironment.
Change in microenvironmental concentration due to outflux of air is calculated as the
concentration in the microenvironment C(t) multiplied by the air exchange rate:
C(t) Equation (5 A-9)
The third term (C removal) in the mass balance calculation (Equation 5A-6) represents
removal processes within the microenvironment. There are three such processes in general:
chemical reaction, deposition, and filtration. Chemical reactions are significant for Os, for
example, but not for carbon monoxide. The amount lost to chemical reactions will generally be
proportional to the amount present, which in the absence of any other factors would result in an
exponential decay in the concentration with time. Similarly, deposition rates are usually given
by the product of a (constant) deposition velocity and a (time-varying) concentration, also
resulting in an exponential decay. The third removal process is filtration, usually as part of a
forced air circulation or HVAC system. Filtration will normally be more effective at removing
particles than gases. In any case, filtration rates are also approximately proportional to
concentration. Change in concentration due to deposition, filtration, and chemical degradation in
a microenvironment is simulated based on the first-order equation:
^removal = [^deposition + ^ filtration + ^chemical
^ .. /c A i r>\
Equation (5 A- 10)
where,
C removal = Change in microenvironmental concentration due to removal
processes (|ig/m3/hr)
= Removal rate of a pollutant from a microenvironment due to
deposition (hr"1)
5 A-12
-------
= Removal rate of a pollutant from a microenvironment due to
filtration (hr1)
= Removal rate of a pollutant from a microenvironment due to
chemical degradation (hr"1)
Removal = Removal rate of a pollutant from a microenvironment due to the
combined effects of deposition, filtration, and chemical
degradation (hr1)
The fourth term in the mass balance calculation represents pollutant sources within the
microenvironment. This is the most complicated term, in part because several sources may be
present. APEX allows two methods of specifying source strengths: emission sources and
concentration sources. Either may be used for mass balance microenvironments, and both can be
used within the same microenvironment. The source strength values are used to calculate the
term C source (|ig/m3/hr).
Emission sources are expressed as emission rates in units of |ig/hr, irrespective of the
units of concentration. To determine the rate of change of concentration associated with an
emission source SE, it is divided by the volume of the microenvironment:
Csource,SE = -TT Equation (5A-11)
where,
C source.SE = Rate of change in C(t) due to the emission source SE (|ig/m3/hr)
SE = The emission rate (|ig/hr)
V = The volume of the microenvironment (m3)
Concentration sources (Sc) however, are expressed in units of concentration. These must
be the same units as used for the ambient concentration (e.g., |ig/m3). Concentration sources are
normally used as additive terms for microenvironments using the factors model. Strictly
speaking, they are somewhat inconsistent with the mass balance method, since concentrations
should not be inputs but should be consequences of the dynamics of the system. Nevertheless, a
suitable meaning can be found by determining the rate of change of concentration (C source) that
would result in a mean increase of Sc in the concentration, given constant parameters and
equilibrium conditions, in this way:
Assume that a microenvironment is always in contact with clean air (ambient = zero), and
it contains one constant concentration source. Then the mean concentration over time in this
5 A-13
-------
microenvironment from this source should be equal to Sc. The mean source strength expressed
in ppm/hr or |ig/m3/hr is the rate of change in concentration ( C source.sc). In equilibrium,
CQ = Csource,sc Equation (5A-12)
p , P
"air exchange ^ "removal
where, Cs is the mean increase in concentration over time in the microenvironment due to the
source C source.sc . Thus, C source.sc can be expressed as
Csource, sc=Csx Rmean Equation (5A-13)
where Rmean is the chemical removal rate. From Equation (5A-13), Rmean is the sum of the air
exchange rate and the removal rate (Rair exchange + Rremovai) under equilibrium conditions. In
general, however, the microenvironment will not be in equilibrium, but in such conditions there
is no clear meaning to attach to C source.se since there is no fixed emission rate that will lead to a
fixed increase in concentration. The simplest solution is to use Rmean = Rair exchange + Rremovai.
However, the user is given the option of specifically specifying Rmean (see discussion below).
This may be used to generate a truly constant source strength C source.se by making Sc and Rmean
both constant in time. If this is not done, then Rmean is simply set to the sum of (Rair exchange +
Rremovai). If these parameters change over time, then C source.se also changes. Physically, the
reason for this is that in order to maintain a fixed elevation of concentration over the base
conditions, then the source emission rate would have to rise if the air exchange rate were to rise.
Multiple emission and concentration sources within a single microenvironment are
combined into the final total source term by combining Equations (5A-1 1) and (5A-13):
1 ne nc
^source = Csource,SE + Csource,SC = \7^ESi + Rmean^CSi Equation (5A-14)
where,
SEI = Emission source strength for emission source /' (|ig/hr, irrespective of
the concentration units)
Sa = Emission source strength for concentration source /' (|ig/m3)
ne = Number of emission sources in the microenvironment
nc = Number of concentration sources in the microenvironment
5 A-14
-------
In Equations (5A-11) and (5A-14), if the units of air quality are ppm rather than |ig/m3,
7/Fis replaced byf/V, where/= ppm / |ig/m3 = gram molecular weight / 24.45. (24.45 is the
volume (liters) of a mole of the gas at 25°C and 1 atmosphere pressure.)
Equations (5A-7), (5A-9), (5 A-10), and (5 A-14) can now be combined with Equation (5A-6) to
form the differential equation for the microenvironmental concentration C(t). Within the time
period of a time step (at most 1 hour), C source and C in are assumed to be constant. Using
C- combined C source < C in leads tO!
dC(t) _
it ~ vornuirieu ~ ~mr exvnanqe ~ \~ / " "removal ~ \~ / -,-, • / /- *. i /-\
dt Equation (5 A- 15)
= ^
^combined air exchange^ V / remova/^V /
combined
Solving this differential equation leads to:
r ( r ^
C(f)= comb/"ed+ C(t0)- combined \eRmean(t^} Equation (5 A-16)
mean \ mean )
where,
C(to) = Concentration of a pollutant in a microenvironment at the beginning of
a time step (|ig/m3)
C(t) = Concentration of a pollutant in a microenvironment at time t within the
time step (|ig/m3).
Based on Equation (5 A- 16), the following three concentrations in a microenvironment
are calculated:
r~(t *. ~,"\ ^combined ^source ~*~ ^in -a +• fc * i n\
il = C(t -^ oo ) = — - = — - - - Equation (5 A- 1 7)
^ air exchange ~*~ ^removal
C(t0 + 7) = Ceujl + (C(t0 ) - Cequil )e-R™*"T Equation (5 A- 1 8)
-1 M) ~^ -^ —. ~ ^mpsr?^1
Cmean=~ \C(t)dt = Cequll + (C(t0}-Cequ^ ~ T Equation (5 A-19)
equ equi,
^
5 A-15
-------
where,
= Concentration in a microenvironment (|ig/m3) if t — » oo (equilibrium
state).
C(to) = Concentration in a microenvironment at the beginning of the time step
C(to+T) = Concentration in a microenvironment at the end of the time step
C mean = Mean concentration over the time step in a microenvironment (|ig/m3)
t\mean = t\car exchange ~r Kremoval (flr )
At each time step of the simulation period, APEX uses Equations (5 A- 17), (5 A- 18), and
(5 A- 19) to calculate the equilibrium, ending, and mean concentrations, respectively. The
calculation continues to the next time step by using Cfto+T) for the previous hour as Cfto).
5A-7.2. Factors Model
The factors model is simpler than the mass balance model. In this method, the value of
the concentration in a microenvironment is not dependent on the concentration during the
previous time step. Rather, this model uses the following equation to calculate the concentration
in a microenvironment from the user-provided hourly air quality data:
Cmeen = C ambient fproximity fpenetration + SC/ Equation (5 A-20)
i=1
where,
Cmean = Mean concentration over the time step in a microenvironment (|ig/m3)
Cambient= The concentration in the ambient (outdoor) environment (|ig/m3)
/proximity = Proximity factor (unitless)
/penetration = Penetration factor (unitless)
Sa = Mean air concentration resulting from source i (|ig/m3)
nc = Number of concentration sources in the microenvironment
The user may specify distributions for proximity, penetration, and any concentration
source terms. All of the parameters in Equation (5 A-20) are evaluated for each time step,
5 A-16
-------
although these values might remain constant for several time steps or even for the entire
simulation.
The ambient air quality data are supplied as time series over the simulation period at
several locations across the modeled region. The other variables in the factors and mass balance
equations are randomly drawn from user-specified distributions. The user also controls the
frequency and pattern of these random draws. Within a single day, the user selects the number
of random draws to be made and the hours to which they apply. Over the simulation, the same
set of 24 hourly values may either be reused on a regular basis (for example, each winter
weekday), or a new set of values may be drawn. The usage patterns may depend on day of the
week, on month, or both. It is also possible to define different distributions that apply if specific
conditions are met. The air exchange rate is typically modeled with one set of distributions for
buildings with air conditioning and another set of distributions for those which do not. The
choice of a distribution within a set typically depends on the outdoor temperature and possibly
other variables. In total there are eleven such conditional variables which can be used to select
the appropriate distributions for the variables in the mass balance or factors equations.
For example, the hourly emissions of CO from a gas stove may be given by the product
of three random variables: a binary on/off variable that indicates if the stove is used at all during
that hour, a usage duration sampled from a continuous distribution, and an emission rate per
minute of usage. The binary on/off variable may have a probability for on that varies by time of
day and season of the year. The usage duration could be taken from a truncated normal or
lognormal distribution that is resampled for each cooking event, while the emission rate could be
sampled just once per stove.
5A-8. EXPOSURE AND DOSE TIME SERIES CALCULATIONS
The activity diaries provide the time sequence of microenvironments visited by the
simulated individual and the activities performed by each individual. The pollutant
concentration in the air in each microenvironment is assumed to be spatially uniform throughout
the microenvironment and unchanging within each diary event and is calculated by either the
factors or the mass balance method, as specified by the user. The exposure of the individual is
given by the time sequence of airborne pollutant concentrations that are encountered in the
microenvironments visited. Figure 5A-2 illustrates the exposures for one simulated 12-year old
child over a 2-day period. On both days the child travels to and from school in an automobile,
goes outside to a playground in the afternoon while at school, and spends time outside at home in
the evening.
5 A-17
-------
ppm
0.14:
0.12:
0.10:
o.osi
o.oe;
0.04:
0.02;
nnn
0
p o
A
O
Jfl hk ,
.up ' 'Hn,
O| 1 LJ
CTJ HuWH
c**^^ [•
HH H
P
0
A 0
,HH SSSSSS^HH ^
00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00
time of day
Figure 5A-2. Example of microenvironmental and exposure concentrations for a simulated
individual over a 48 hours simulation. (H: home, A: automobile, S: school, P: playground,
O: outdoors at home).
In addition to exposure, APEX models breathing rates based on the physiology of each
individual and the exertion levels associated with the activities performed. For each activity type
in CHAD, a distribution is provided for a corresponding normalized metabolic equivalent of
work or METs (McCurdy, 2000). METs are derived by dividing the metabolic energy
requirements for the specific activity by a person's resting, or basal, metabolic rate. The MET
ratios have less interpersonal variation than do the absolute energy expenditures. Based on age
and gender, the resting metabolic rate, along with other physiological variables is determined for
each individual as part of their anthropometric characteristics. Because the MET ratios are
sampled independently from distributions for each diary event, it would be possible to produce
time-series of MET ratios that are physiologically unrealistic. APEX employs a MET
adjustment algorithm based on a modeled oxygen deficit to prevent such overestimation of MET
and breathing rates (Isaacs et al., 2008). The relationship between the oxygen deficit and the
applied limits on MET ratios are nonlinear and are derived from published data on work capacity
and oxygen consumption. The resulting combination of microenvironmental concentration and
breathing ventilation rates provides a time series of inhalation intake dose for most pollutants.
5 A-18
-------
5A-9. MODEL OUTPUT
APEX calculates the exposure and dose time series based on the events as listed on the
activity diary with a minimum of one event per hour but usually more during waking hours.
APEX can aggregate the event level exposure and dose time series to output hourly, daily,
monthly, and annual averages. The types of output files are selected by the user, and can be as
detailed as event-level data for each simulated individual (note, Figure 5A-2 was produced from
the event output file). A set of summary tables are produced for a variety of exposure and dose
measures. These include tables of person-minutes at various exposure levels, by
microenvironment, a table of person-days at or above each average daily exposure level, and
tables describing the distributions of exposures for different groups. An example of how APEX
results can be depicted is given in
OJOB
0.1
0.12
Oi«n* fecesurt Lwtfl (ppm4hr)
, which shows the percent of children with at least one 8-hour average exposure at or above
different exposure levels, concomitant with moderate or greater exertion. These are results from
a simulation of Os exposures for the greater Washington, D.C. metropolitan area for the year
2002. From this graph ones sees, for example, that APEX estimates 30 percent of the children in
this area experience exposures above 0.08 ppm-8hr while exercising, at least once during the
year.
5 A-19
-------
OJOB
0.1
0.12
Oi«n* Bceesur* Lwtl (ppm4hr)
Figure 5A-3. The percent of simulated children (ages 5-18) at or above 8-hour average
exposures while at moderate or greater exertion.
5A-20
-------
5A-10. REFERENCES
Burmaster, D.E. (1998). LogNormal distributions for skin area as a function of body weight.
Risk Analysis. 18(l):27-32.
Glen, G., Smith, L., Isaacs, K., McCurdy, T., Langstaff, J. (2008). A new method of
longitudinal diary assembly for human exposure modeling. J Expos Sci Environ Epidem.
18:299-311.
Graham, S.E., McCurdy, T. (2005). Revised ventilation rate (VE) equations for use in
inhalation-oriented exposure models. Report no. EPA/600/X-05/008 is Appendix A of US
EPA (2009). Metabolically Derived Human Ventilation Rates: A Revised Approach Based
Upon Oxygen Consumption Rates (Final Report). Report no. EPA/600/R-06/129F.
Available at: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=202543.
Isaacs, K., Glen, G., McCurdy, T., Smith, L. (2008). Modeling energy expenditure and oxygen
consumption in human exposure models: accounting for fatigue and EPOC. J Expos Sci
Environ Epidemiol. 18:289-298.
Isaacs, K., Smith, L. (2005). New Values for Physiological Parameters for the Exposure Model
Input File Physiology.txt. Memorandum submitted to the U.S. Environmental Protection
Agency under EPA Contract EP-D-05-065. NERL WA 10. Alion Science and Technology.
Found in US EPA. (2009). Risk and Exposure Assessment to Support the Review of the
SO2 Primary National Ambient Air Quality Standard. EPA-452/R-09-007. August 2009.
Available at
http://www.epa.gov/ttn/naaqs/standards/so2/data/200908SO2REAFinalReport.pdf.
Johnson, T., Capel, J., McCoy, M. (1996a). Estimation of Ozone Exposures Experienced by
Urban Residents Using a Probabilistic Version of NEM and 1990 Population Data.
Prepared by IT Air Quality Services for the Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina,
September.
Johnson, T., Capel, J., Mozier, J., McCoy, M. (1996b). Estimation of Ozone Exposures
Experienced by Outdoor Children in Nine Urban Areas Using a Probabilistic Version of
NEM. Prepared for the Air Quality Management Division under Contract No. 68-DO-
30094, April.
Johnson, T., Capel, J., McCoy, M., Mozier, J. (1996c). Estimation of Ozone Exposures
Experienced by Outdoor Workers in Nine Urban Areas Using a Probabilistic Version of
NEM. Prepared for the Air Quality Management Division under Contract No. 68-DO-
30094, April.
Johnson, T. (2002). A Guide to Selected Algorithms, Distributions, and Databases Used in
Exposure Models Developed By the Office of Air Quality Planning and Standards. Revised
Draft. Prepared for U.S. Environmental Protection Agency under EPA Grant No.
CR827033.
McCurdy, T. (2000). Conceptual basis for multi-route intake dose modeling using an energy
expenditure approach. J Expo Anal Environ Epidemiol. 10:1-12.
5A-21
-------
McCurdy, T., Glen, G., Smith, L., Lakkadi, Y. (2000). The National Exposure Research
Laboratory' s Consolidated Human Activity Database. JExp Anal Environ Epidemiol.
10:566-578.
Rosenbaum, A. S. (2008). The Cluster-Markov algorithm in APEX. Memorandum prepared for
Stephen Graham, John Langstaff. USEPA OAQPS by ICF International.
Schofield, W. N. (1985). Predicting basal metabolic rate, new standards, and review of previous
work. HumNutrClinNutr. 39C(S1):5-41.
US EPA. (2002). Consolidated Human Activities Database (CHAD) Users Guide. Database and
documentation available at: http://www.epa.gov/chadnetl/.
US EPA. (2012a). Total Risk Integrated Methodology (TRIM) - Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4.4) Volume I: User's Guide. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park,NC. EPA-452/B-12-00la. Available at:
http://www.epa.gov/ttn/fera/human_apex.html
US EPA. (2012b). Total Risk Integrated Methodology (TRIM) - Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4.4) Volume II: Technical Support
Document. Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-452/B-12-001b. Available at:
http://www.epa.gov/ttn/fera/human_apex.html
5A-22
-------
APPENDIX 5B
Inputs to the APEX Exposure Model
Table of Contents
5B-1 POPULATION DEMOGRAPHICS 5B-1
5B-2 POPULATION COMMUTING PATTERNS 5B-2
5B-3 ASTHMA PREVALENCE RATES 5B-3
5B-4 HUMAN ACTIVITY DATA 5B-4
5B-4.1 CHAD Updates Since the 2007 Ozone NAAQS Review 5B-6
5B-4.2 Longitudinal Activity Pattern Methodology 5B-8
5B-5 PHYSIOLOGICAL AND METABOLIC EQUIVALENTS DATA 5B-11
5B-6 MICROENVIRONMENTS MODELED 5B-12
5B-6.1 Air Exchange Rates for Indoor Residential Microenvironments 5B-15
5B-6.2 Air Conditioning Prevalence for Indoor Residential MicroEnvironments 5B-17
5B-6.3 AER Distributions for Other Indoor Microenvironments 5B-19
5B-6.4 Proximity and Penetration Factors for In-vehicle and Near-Road
Microenvironments 5B-20
5B-6.5 Proximity and Penetration Factors for Outdoor Microenvironments 5B-21
5B-6.6 Ozone Decay and Deposition Rates 5B-21
5B-7 AMBIENT OZONE CONCENTRATIONS 5B-23
5B-8 METEOROLOGICAL DATA 5B-32
5B-9 CONDITIONAL VARIABLES 5B-35
5B-10 REFERENCES 5B-36
5B-i
-------
List of Tables
Table 5B-1. Consolidated Human Activity Database (CHAD) study information and diary-
days used by APEX 5B-9
Table 5B-2. Microenvironments modeled and calculation method used 5B-13
Table 5B-3. AERs for indoor residential microenvironments (ME-1) with A/C by study area
and temperature 5B-16
Table 5B-4. AERs for indoor residential microenvironments (ME-1) without A/C by study
area and temperature 5B-17
Table 5B-5. American Housing Survey A/C prevalence from Current Housing Reports (Table
1-4) for selected urban areas 5B-18
Table 5B-6. Parameter values for distributions of penetration and proximity factors used for
estimating in-vehicle microenvironmental concentrations 5B-20
Table 5B-7. VMT fractions of interstate, urban, and local roads in the study areas used to
select in-vehicle proximity factor distributions 5B-22
Table 5B-8. Identification of U.S. counties and the number of APEX air districts included each
study area 5B-25
Table 5B-9. Ambient monitors used to define exposure modeling domain and the population
modeled in each study area 5B-26
Table 5B-10. Study area meteorological stations, locations, and hours of missing data 5B-33
List of Figures
Figure 5B-1. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Atlanta, Boston, Baltimore and Chicago study areas 5B-28
Figure 5B-2. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Cleveland, Dallas, Denver and Detroit study areas 5B-29
Figure 5B-3. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Houston, Los Angeles, New York and Philadelphia study areas...
5B-30
Figure 5B-4. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Sacramento, St. Louis and Washington DC study areas 5B-31
5B-ii
-------
The APEX model inputs require extensive analysis and preparation to ensure the model
outputs are appropriate as intended, reasonable, and relevant. This Appendix describes the
preparation and the sources of data for the APEX input files.
5B-1 POPULATION DEMOGRAPHICS
APEX accounts for important population characteristics in representing study area
demographics. Population counts and employment probabilities by age and gender are used to
develop representative profiles of hypothetical individuals for the simulation. For the main-body
results of the Os Health Risk and Exposure Assessment (HREA), we estimated population-based
exposures using US Census tract-level population counts stratified by age in one-year
increments, from birth to 99 years, and were obtained from the 2000 Census of Population and
Housing Summary File 1 (SF1).1 The SF1 contains the 100-percent data, which is the
information compiled from the questions asked of all people and about every housing unit.
Three standard APEX input files are used for the current Os assessment:
• pop geo2000011403.txt: census tract ID's, their latitudes and longitudes
• pop^fall2000 043003.txt: tract-level population counts for females by age
• pop^fall2000 043003.txt: tract-level population counts for males by age
Census tract employment rates were developed using the Employment Status: 2000-
Supplemental Tables.2 The file input to APEX is stratified by gender and age group, so that each
gender/age group combination is given an employment probability fraction (ranging from 0 to 1)
within each census tract. The age groupings in this employment file are: 16-19, 20-21, 22-24,
25-29, 30-34, 35-44, 45-54, 55-59, 60-61, 62-64, 65-69, 70-74, and >75. Children under 16
years of age are assumed to not be employed.3
One standard APEX input file is used for the current Os assessment:
• Employment2000 043003.txt: census tract employment probabilities by age
groups
1 http://www.census.gov/census2000/sumfilel.html.
2 http://www.census.gov/population/www/cen2000/phc -t28.html.
3 While children can be employed at ages <16, staff feel that when modeling population-based exposures for these
young children regardless of whether or not they have been designated as being employed, it is likely the overall
study group exposure results would not be significantly affected given the small fraction of the population that
may be employed at these ages and that the principal factor influencing high O3 exposure concentrations is
afternoon time spent outdoors. In such a simulation that included employed children <16, only their home tract to
work tract commuting would be affected and unless they were employed as an outdoor worker (also a further
subdivision of those employed), substantial time spent outdoors is unlikely to occur at work.
5B-1
-------
5B-2 POPULATION COMMUTING PATTERNS
To more realistically simulate human behavior, APEX incorporates workplace patterns
into the assessment by use of home-to-work commuting data. By design, commuting is only
used for those simulated individuals who are employed (i.e., > 16 years old). The commuting
data were derived from the 2000 Census Transportation Planning Package (CTPP) Part 3-
Journey-to-Work (JTW) files.4 These files contain counts of individuals commuting from home
to work locations at varying geographic scales. These data were processed to calculate fractions
for each tract-to-tract flow to create a national commuting flow file distributed with APEX. This
database contains commuting data for each of the 50 states and Washington, D.C. Important
processing and application assumptions include the following:
• Commuting within the Home Tract: the APEX commuting database does not
differentiate people that work at home from those that commute within their home tract.
• Commuting Distance Cutoff: all persons in home-work flows up to 120 km are daily
commuters and no persons in more widely separated flows commute daily, thus the list of
destinations for each home tract was restricted to only those work tracts that are within
120 km of the home tract.5
• Eliminated Records: tract-to-tract pairs that represented workers who either worked
outside of the U.S. (9,631 tract pairs with 107,595 workers) or worked in an unknown
location (120,830 tract pairs with 8,940,163 workers) were eliminated. An additional 515
workers in the commuting database whose data were missing from the original files,
possibly due to privacy concerns or errors, were also deleted.
• Simulation of Leavers: we restricted the simulated population to those who do not
commute to destinations outside the study area because we have not estimated ambient
concentrations of Os in counties outside of the modeled areas.
4 Files downloaded from http://transtats.bts.gov/.
5 Plotting log(flows) versus log(distance) indicates a near-constant slope out to a distance of approximately 120
kilometers. Beyond that distance, the relationship also had a fairly constant, though less inclined, slope. A simple
interpretation is that for distances up to 120 km, the majority of the flow was due to persons traveling between
home and work tracts daily, with the numbers of such persons decreasing rapidly with increasing distance.
Beyond 120 km, the majority of the flow is made up of persons who stay at the workplace for extended times, in
which case the separation distance is not as crucial in determining the flow.
5B-2
-------
An additional commuting input file was recently developed as a companion to the APEX
commuting flow file. Also derived from the 2000 census are tract-level population counts of
one-way commute times, and given in 13 time bins (in minutes): < 5, 5 to 9, 10 to 14, 15 to 19,
20 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 44, 45 to 59, 60 to 89, 90-120, works at home (0
minutes commuting time). APEX uses these time bins to create a cumulative probability
distribution of commuting times for each tract, which it then uses in conjunction with the
distribution of commuting distances to assign a profile-level one-way commuting time variable
to each employed person in the population. This commuting time profile variable is then used to
select for CHAD diaries having appropriate commute times in their daily activity pattern (i.e., a
total time spent in travel locations or activities before and after work activities) to represent the
simulated individual.
Two standard APEX input files are used for the current Os assessment:
• Commuting2000 010505. txt: home/work census tract ID's, cumulative
probabilities of commuting to work tract from home tract, distances of home to
work tract (km)
• CommutingTimes2000 050610.txt: tract-level counts of all workers, commuters,
and commute time bins
5B-3 ASTHMA PREVALENCE RATES
One of the important study group in the exposure assessment is asthmatic school-age
children (ages 5-18). Modeling exposures for this study group with APEX requires the
estimation of children's asthma prevalence rates. The estimates are based on children's asthma
prevalence data from the 2006-2010 National Health Interview Survey (NHIS). Briefly, 2000
US census tract level asthma prevalence was estimated for children (by single age years) and
adults (by age groups), also stratified by gender and family income/poverty ratio (i.e., whether
the family income was considered below or at/above the US Census estimate of poverty level for
the given year). Given the significant differences in asthma prevalence by age, gender, region,
and poverty status, the variability in the spatial distribution of poverty status across census tracts
(and also stratified by age), and the spatial variability in local scale ambient concentrations of
many air pollutants, the goal was to better represent the variability in population-based exposures
when accounting for and modeling these newly refined attributes of this study group. A detailed
description of how the NHIS data were processed for input to APEX is provided in Appendix
5C.
One standard APEX input file is used for the current Os assessment:
• AsthmaPrevalence053112.txt: tract-level asthma prevalence by age (for ages <18)
and age groups (for ages > 17)
5B-3
-------
5B-4 HUMAN ACTIVITY DATA
Exposure models use human activity pattern data to predict and estimate exposure to
pollutants. Different human activities, such as outdoor exercise, indoor reading, or driving,
would lead to varying pollutant exposures. In addition, different human activities require
different energy expenditures, and thus, higher exposure media consumption rates lead to higher
doses received. To accurately model individuals and their exposure to pollutants, it is critical to
have a firm understanding of the locations where people spend time and the activities performed
in such locations.
The Consolidated Human Activity Database (CHAD) provides time series data on human
activities through a database system of collected human diaries, or daily time location activity
logs (US EPA, 2002). The purpose of CHAD is to provide a basis for conducting multi-route,
multi-media exposure assessments (McCurdy et al., 2000). The data contained within CHAD
come from multiple surveys with somewhat variable study-specific structure (e.g., minute-by-
minute versus time-block averaged sequence of diary events), though common to all studies
included, individuals provided information on their locations visited and activities performed for
each survey day. Personal attribute data for these surveyed individuals, such as age and gender,
are included in CHAD as well. The latest version of CHAD master (071113) contains data for
54,373 person-days.
The CHAD served as the primary source of time location activity pattern data and was
processed to retain appropriate diary data for use by APEX. Diaries with missing personal
attribute data (i.e., age, gender), missing diary day information (i.e., either daily mean/ maximum
temperature, day-of-week), or having 3-hours or more of missing location and/or activity
information are not used by APEX. For the latter case, CHAD diaries were evaluated for
instances where a diary may contain enough information for the purposes of this exposure
assessment allowing it to be adjusted to reduce the missing information to less than 3 hours on a
given day. For example, the diary structure of the ozone averting behavior (OAB) study resulted
in nearly all of the diary days (n=2,776) having no diary information between the hours of 8PM
and midnight. In processing the CHAD data for this subset of diaries, the location was assumed
by staff to be indoors at their residence and persons were engaged in a sleep activity. This
substitution was judged by staff as a reasonable approximation based on the limited likelihood of
a person's highest Os exposures occurring at this time of day, while still retaining the relevant
activity pattern data of interest (e.g., locations visited and activities performed during the
daytime hours).
The following is a list of adjustments made to CHAD diary data where study specific
structure was a factor in missing data or diary information was present in either CHAD location
or activity codes to infer specific information where data were missing.
5B-4
-------
• OAB (a children's study) missing location and activity events from 8PM - 12AM
were set to 'indoor residence' and 'sleep';
• BAL missing activity events at SAM occurring indoors were set to 'personal
care';
• ISR missing activity events occurring when attending school were set to either
'attend K-12' (ages 5-18) or 'attend day-care' (ages <5);
• NSA (an adults study) missing activity events at 8PM - 12AM occurring indoor
residences were set to 'leisure, general';
• Locations missing for a number of staff judged outdoor activities6 were set to
'outdoor, general';
• Locations missing for a number of staff judged indoor residential activities7 were
set to "indoor, residence"; and
• Locations missing for a number of staff judged general indoor activities8 were set
to "indoor, other".
Three standard APEX input files are used for the current Os assessment:
• CHADQuest 013013B.txt: personal (e.g., age, gender, employment status, county
of residence, etc.) and day (e.g., daily maximum temperature, day-of-week)
attribute meta data for each diary day
• CHADEvents 013013B. txt: time sequence of locations visited and activities
performed by individuals for each diary day
• CHADSTATSOutdoor 013013B.txt: time spent outdoors for each diary day
Table 5B-1 summarizes the studies and number of diary days used by APEX in this
modeling analysis, providing over 41,000 diary-days of activity data (nearly 18,000 diary-days
for ages 4-18) collected between 1982 and 2010.
b For CHAD activity codes (US EPA, 2002) "11300", "11630", "17100", "17110", "17112", "17120", "17131" or
"17170".
7 For CHAD activity codes (US EPA, 2002) "11100", "11110", "11200", "11210", "11220", "14000", "14100",
"14110", "14120", "14300", "14400", "14500", "14600", or "17223".
8 For CHAD activity codes (US EPA, 2002) "13300", "13400", "15400", "16300", "16400", or "16500".
5B-5
-------
5B-4.1 CHAD Updates Since the 2007 Ozone NAAQS Review
Since the time of the prior Os NAAQS review conducted in 2007, there have been
a number new data sets incorporated into CHAD and used in our current exposure assessment,
most of which were from recently conducted studies. The data from these eight additional
studies incorporated in CHAD and available for use by APEX have more than doubled the total
activity pattern data used for Os exposure modeling in 2007 and has increased the number of
children diaries by a factor of five. The studies from which these new data were derived are
briefly described below.
• DEA. The diaries are from 2 seasons of the 6-season sampling period (2004-2007) used
by EPA in the Detroit Exposure and Aerosol Research Study (DEARS) (Williams et al.,
2008). The intent was to obtain environmental samples and time use data for 10 days—5
in each of 2 seasons per participant located in 6 areas in Wayne County, Michigan (in and
around Detroit). A 15-minute block diary approach was used to collect activity data.
Participants were all adults and activity data was collected from Tuesday through
Saturday. Just over 300 diary-days from DEARS are used by APEX.
• EPA. The diaries were collected as part of an ongoing longitudinal internal EPA study
by EPA scientists, and in some cases, their families. This dataset contains two long-term
longitudinal diaries: one by a 60 year-old-male in 1999-2000 (McCurdy and Graham,
2003), and one by a 35 year old male in 2002. Additional longitudinal diaries were kept
for a 35-year-old female and her infant daughter in 2008 (though the infant data are not
used here). The remaining diaries are from a study of a group of 9 adults (Isaacs et al.
2012). In this portion of the study, all subjects were studied for approximately 17
consecutive days in each of 4 seasons in 2006 and 2007. Approximately of 1,400 diary-
days are used by APEX.
• ISR. The diaries are from phase I (1997), phase II (2002-03), and phase III (2007-08) of
the University of Michigan's Panel Study of Income Dynamics (PSID), respectively
(University of Michigan, 2012). Nationally representative activity pattern data from
nearly 11,000 children ages 0-13 (phase I), ages 5-19 (phase II), and ages 10-19 (phase
III) were added to the APEX activity pattern data. For each child, time use data were
reported by primary care-givers, school teachers, and/or the children themselves on two
nonconsecutive days in a single week, in no particular season, though mostly occurring
during the spring and fall (phase I), winter (phase II), and spring, fall and winter (phase
III) months.
• NSA. The diaries were collected as part of the National-Scale Activity Survey (NSAS),
an EPA-funded study of averting behavior related to air quality alerts (Knowledge
Networks, 2009). Data were collected from about 1,200 adults aged 35-92 in seven
metropolitan areas (Atlanta, St. Louis, Sacramento, Washington DC, Dallas, Houston,
and Philadelphia). Data were collected over 1-15 (partially consecutive) days across the
2009 ozone season, providing approximately 7,000 person days of data for use by APEX.
• OAB. The diaries were collected in a study of children's activities on high and low ozone
days during the 2002 ozone season (Mansfield et al., 2009). Children ages 2-12 from 35
U.S. metropolitan areas having the worst Os pollution were studied, and of whom, about
5B-6
-------
half of were asthmatics. Activity data were collected on 6 nonconsecutive days from
each subject, with some subjects providing fewer days, providing nearly 2,200 persons
days of data to APEX.
• SEA. The diaries are from a particulate matter (PM) exposure study of susceptible study
groups living in Seattle, WA between 1999 and 2002 (Liu et al., 2003). Two cohorts
were studied: an older adult group with either chronic obstructive pulmonary disease
(COPD) or coronary heart disease and a children's group (ages 6-13) with asthma.
Activity data were collected on 10 consecutive days from each subject, with some
subjects providing fewer days. Over 1,600 adult diaries and more than 300 children
diaries were included in the APEX activity pattern file.
• SUP. The diaries are from the SUPERB study (Study of Use of Products and Exposure-
Related Behaviors) undertaken by researchers from the University of California at Davis
Bennett et al., 2012a; Hertz-Picciotto et al., 2012). The study focused on the use of
household and personal care products from 47 California households, 30 with children
(ages 1-18) living in 22 counties in northern California, and 17 with an older adult (>55
y) living in 3 central California counties. Two days of activity data were obtained via the
internet for each participant—a weekday and a weekend day. Approximately 2,500
diary-days from SUPERB met appropriate criteria for use in APEX.
• RTF. The diaries were collected in a panel study of PM exposure in the Research
Triangle Park (RTF), NC area (Williams et al., 2003a, b). Two older adult cohorts (ages
55-85) were studied: a cohort having implanted cardiac defibrillators living in Chapel
Hill, NC and a second group of 30 people having controlled hypertension and residing in
a low-to-moderate SES neighborhood in Raleigh, NC. Data were collected on
approximately 8 consecutive days in 4 consecutive calendar seasons in 2000-2001.
Approximately 900 diary-days were included from this study.
5B-7
-------
5B-4.2 Longitudinal Activity Pattern Methodology
An important issue in this assessment is the approach used for creating an Os-season or
year-long activity sequence for each simulated individual based on a largely cross-sectional
activity database of 24-hour records. The typical subject in the time location activity studies in
CHAD provided about two days of diary data. For this reason, the construction of a season-long
activity sequence for each individual requires some combination of repeating the same data from
one subject and using data from multiple subjects. The best approach would reasonably account
for the day-to-day and week-to-week repetition of activities common to individuals (though
recognizing even these diary sequences are not entirely correlated) while maintaining realistic
variability among individuals comprising each study group.
The method currently used in APEX for creating longitudinal diaries was designed to
capture the tendency of individuals to repeat activities, based on reproducing realistic variation in
a key diary variable, which is a user selected function of diary variables. For this Cb analysis,
the key variable selected is the amount of time an individual spends outdoors each day, one of
the most important determinants of exposure to high levels of Os. The actual diary construction
method targets two statistics, a population diversity statistic (D) and a within-person
autocorrelation statistic (A). The D statistic reflects the relative importance of within- and
between-person variance in the key variable. The^4 statistic quantifies the lag-one (day-to-day)
key variable autocorrelation. Further details regarding the longitudinal methodology can be
found in US EPA (2013a, b).
Desired D and A values for the key variable are selected by the user and set in the APEX
parameters file, and the method algorithm constructs longitudinal diaries that preserve these
parameters. Longitudinal diary data from a limited field study of children ages 7-12 (Geyh et al.,
2000; Xue et al., 2004) estimated values of approximately 0.2 for/) and 0.2 for A In the
absence of data for estimating these statistics for younger children and others outside the study
age range, and since APEX appears to underestimate repeated activities, values of 0.5 for/) and
0.2 for^4 are used for all ages.
5B-8
-------
Table 5B-1. Consolidated Human Activity Database (CHAD) study information and diary-days used by APEX.
Study Name (CHAD
Abbreviation)
Baltimore Retirement
Home Study (BAL)
California Youth
Activity Patterns Study
(CAY)
California Adults
Activity Patterns Study
(CAA)
California Children
Activity Patterns Study
(CAC)
Cincinnati Activity
Patterns Study (CIN)
Detroit Exposure and
Aerosol Research
Study (DEA)
Denver CO Personal
Exposure Study (DEN)
EPA Longitudinal
Studies (EPA)
Los Angeles Ozone
Exposure Study:
Elementary School
(LAE)
Los Angeles Ozone
Exposure Study: High
School (LAH)
Geographic
Coverage
One building
in Baltimore,
MD
California
California
California
Cincinnati, OH
metro, area
Detroit, Ml
metro, area
Denver, CO
metro, area
RTP, NC
Los Angeles,
CA
Los Angeles,
CA
Study Dates
1/1 997 to 2/1 997;
7/1 998 to 8/1 998
10/1 987 to 9/1 988
10/1 987 to 9/1 988
4/1 989 to 2/1 990
3/1 985 to 4/1 985;
8/1985
7/2005 to 8/2005;
7/2006 to 8/2006
11/1 982 to
2/1983
2/1 999 to 2/2000;
2/2002 to 8/2002;
7/2006 to 6/2008
10/1989
9/1 990 to 10/1 990
Study
Subject
Ages
72-93
12- 17
18-94
<1 -11
<1 -86
18-74
18-70
<1 -60
10- 12
13-17
APEX
Diary-days
(ages 4-94)
304
182
1,555
1,195
2,449
331
714
1,417
50
42
APEX
Diary-days
(ages 4-18)
0
182
36
771
727
5
7
0
50
42
Diary Type,
Time Format,
Survey Design
Diary, 15 Minute
Block, Panel
Recall, Event,
Random
Recall, Event,
Random
Recall, Event,
Random
Diary, Event,
Random
Recall, 15 Minute
Block, Panel
Diary, Event,
Random
Diary, Event,
Panel
Diary, Event,
Panel
Diary, Event,
Panel
Study Reference
Williams et al. (2000)
Robinson et al. (1989),
Wiley etal. (1991 a)
Robinson etal. (1989),
Wiley etal. (1991 a)
Wiley etal. (1991b)
Johnson (1989)
Williams et al. (2008)
Johnson (1984), Akland
etal. (1985)
Isaacs etal. (2012)
Spier etal. (1992)
Spier etal. (1992)
5B-9
-------
Study Name (CHAD
Abbreviation)
National Human
Activity Pattern Study:
Air (NHA)
National Human
Activity Pattern Study:
Water (NHW)
National-Scale Activity
Survey (NSA)
Population Study of
Income Dynamics
PSID 1 (ISR)
Population Study of
Income Dynamics
PSID II (ISR)
Population Study of
Income Dynamics
PSID III (ISR)
RTI Ozone Averting
Behavior (OAB)
RTP Panel (RTP)
Seattle (SEA)
Study of Use of
Products and
Exposure Related
Behavior (SUP)
Washington, D.C.
(WAS)
Totals
Geographic
Coverage
National
National
7 US metro.
areas
National
National
National
35 US metro.
areas
RTP, NC
Seattle, WA
Broader
Sacramento &
San
Francisco, CA
Counties
Wash. DC
metro, area
Study Dates
9/1 992 to 10/1 994
9/1 992 to 10/1994
6/2009 to 9/2009
2/1 997 to 12/1 997
1/2002 to 12/2003
10/2007 to 4/2008
7/2002 to 8/2003
6/2000 to 5/2001
10/1 999 to 3/2002
7/2006 to 3/20 10
11/1 982 to 2/1 983
1982-2010
Study
Subject
Ages
<1 -93
<1 -93
35-92
<1 -13
5-19
10-19
2-12
55-85
6-91
1 -88
18-71
<1 -94
APEX
Diary-days
(ages 4-94)
4,329
4,329
6,825
4,978
4,800
2,650
2,872
871
1,624
3,456
686
41,474
APEX
Diary-days
(ages 4-18)
693
745
0
3,507
4,793
2,614
2,187
0
317
994
10
17,680
Diary Type,
Time Format,
Survey Design
Recall, Event,
Random
Recall, Event,
Random
Recall, 15 Minute
Block, Random
Recall, 15 Minute
Block,
Random/Panel
Recall, 15 Minute
Block,
Random/Panel
Recall, 15 Minute
Block,
Random/Panel
Recall, 15 Minute
Block, Random
Diary, 15 Minute
Block, Panel
Diary, 15 Minute
Block, Panel
Recall, 15 Minute
Block, Panel
Diary, Event,
Random
Study Reference
Klepeisetal. (1996),
Tsang and Klepeis
(1996)
Klepeisetal. (1996),
Tsang and Klepeis
(1996)
Knowledge Networks
(2009)
University of Michigan
(2012)
University of Michigan
(2012)
University of Michigan
(2012)
Mansfield et al. (2006,
2009)
Williams et al. (2003a,b)
Liu et al. (2003)
Bennett et al. (201 2a),
Hertz-Picciotto et al.
(2010)
Hartwell et al. (1984),
Aklandetal. (1985)
5B-10
-------
5B-5 PHYSIOLOGICAL AND METABOLIC EQUIVALENTS DATA
APEX requires several physiological parameters to accurately model processes
that affect pollutant intake rate for individuals. This is because differences in physiology may
cause people with the same exposure and activity scenarios to have different pollutant intake
levels. The physiological parameters file used by APEX contains individual data or data
distributions stratified by age and gender for maximum ventilatory capacity (in terms of age- and
gender-specific maximum oxygen consumption potential, NVChmax), body mass (BM), resting
metabolic rate (RMR), body surface area (BSA), maximum oxygen deficits (MOXD) and
associated recovery time (RECTIME), height, and oxygen consumption-to-ventilation rate
relationships (ECF), among a few others not used for estimating Os exposure and dose).
APEX also uses an input file containing the metabolic equivalents for work (METS) to
estimate the specific energy expended for each activity listed in the diary file. These METS
values are commonly in the form of distributions and were originally derived as relative to an
individual's RMR. Some activities are specified as a single point value (for instance, sleep),
while others, such as athletic endeavors or manual labor, are normally, lognormally, or otherwise
statistically distributed. APEX samples from these distributions and calculates values to
simulate the variable nature of activity levels among different people. These personal- and
activity-level physiological variables are ultimately used to estimate ventilation rate (VE) and
decrements in forced expiratory volume, in one second (dFEVi).
Three standard APEX input files are used for the current Os assessment:
• PhysiologyO 10213 threshold, txt: NVO2max, BM, RMR, B SA, MOXD,
RECTIME, height, ECF, and dFEVi distributions and equation coefficients, by
sex and age groups
• MET Distributions 030612. txt: statistical form and parameters for METS
distributions associated with each activity performed, some by age groups
• Ventilation 121106. txt: distributions and equation coefficients to estimate
individual activity- specific VE by sex and age groups
5B-11
-------
5B-6 MICROENVIRONMENTS MODELED
In APEX, exposure for simulated individuals occurs in microenvironments. For
exposures to be accurately estimated, it is important maintain the spatial and temporal sequence
of microenvironments persons inhabit and appropriately represent the time series of
concentrations that occur within them. As discussed in Appendix 5A, the two methods available
in APEX for calculating pollutant concentrations within microenvironments are a mass balance
model and a transfer factor approach, each of which uses an appropriate ambient pollutant
concentration to estimate the microenvironmental concentration. Table 5B-2 lists the 28
microenvironments selected for this analysis and the exposure calculation method for each. The
variables used and their associated parameters to calculate microenvironmental concentrations
are described in subsequent subsections below.
The CHAD database has 115 locations codes, many of which go beyond the scale
of the microenvironmental modeling (e.g., inside at residence in a bedroom). Therefore these
more specific locations are aggregated by mapping these 115 location codes to the 28 modeled
microenvironments. Further, all microenvironmental concentrations in this exposure assessment
are estimated using an ambient concentration (section 5B-7), though these concentrations not
only vary temporally but spatially, depending on the particular microenvironment. The mapping
of locations to the 28 microenvironments also includes an identifier that designates what ambient
concentration is used in the calculation of the microenvironmental concentration for each event.
For this assessment, we used ambient concentration for each individual based on either their
home (H), work (W), near work (NW), near home (NH), last (L, either NH or NW), other (O,
average of all), or unknown (U, last ME determined) tracts.
Multiple APEX ME input files are used for the current Os assessment, varying by study
area though given in one form. Only one ME mapping file is used:
• ME descriptions 28MEsO3 CSAfstudyarea.JJdateJ.txt: defines calculation
method, variables and their parameters used to estimate all microenvironmental
concentrations
• MicroEnv Mapping CHAD Jo APEX 28MEs 022613.txt: maps 115 CHAD
locations to 28 APEX microenvironments and defines tract-level ambient
concentrations to use for each location
5B-12
-------
Table 5B-2. Microenvironments modeled and calculation method used.
Microenvironment (ME)
Indoor- Residence
Indoor -Community
Center or Auditorium
Indoor- Restaurant
Indoor- Hotel, Motel
Indoor -Office building,
Bank, Post office
Indoor - Bar, Night club,
Cafe
Indoor -School
Indoor- Shopping mall,
Non-grocery store
Indoor - Grocery store,
Convenience store
Indoor - Metro-Subway-
Train station
Indoor- Hospital, Medical
care facility
Indoor- Industrial,
factory, warehouse
Indoor -Other indoor
Outdoor- Residential
Outdoor -Park or Golf
course
Outdoor- Restaurant or
Cafe
Outdoor- School
grounds
Outdoor- Boat
Outdoor- Other outdoor
non-residential
AEPX
ME
Number
1
2
3
4
5
6
7
8
9
10
11
26
27
12
14
15
16
25
13
Not Used In CHAD Study
SEA
BAL, CIN, DEN, LAE, LAH,
OAB, SEA, WAS
SEA
BAL, CAC, CAY, DEN, LAE,
LAH, OAB, SEA, WAS
BAL,
LAE, LAH, SEA
BAL, ISR, SEA
BAL, CAA, CAC, CAY, CIN,
DEA, DEN, LAE, LAH, NSA,
OAB, RTP, SUP, WAS
LAH
BAL, DEA, ISR, LAE, LAH,
OAB, SEA, WAS
BAL,
BAL, CAA, CAY, SEA, WAS
CAA, CAC, CAY, CIN, DEA,
DEN, LAE, LAH, NSA, OAB,
SEA, SUP, WAS
BAL, CAC
BAL, CAA, CAC, CAY, DEA,
DEN, ISR, LAE, LAH, RTP,
SEA, WAS
Calculation
Method
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Mass
balance
Factors
Factors
Factors
Factors
Factors
Factors
Variables1
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
AER & DE
None
None
None
None
None
None
5B-13
-------
Microenvironment (ME)
Near-road - Metro-
Subway-Train stop
Near-road - Within 1 0
yards of street
Near-road - Parking
garage (covered or below
ground)
Near-road - Parking lot
(open), Street parking
Near-road - Service
station
Vehicle - Cars and Light
Duty Trucks
Vehicle - Heavy Duty
Trucks
Vehicle - Bus
Vehicle - Train, Subway
AEPX
ME
Number
17
18
19
20
21
22
28
23
24
Not Used In CHAD Study
BAL, CIN, DEA, DEN, ISR,
LAE, LAH, WAS
CAC, OAB, SEA, SUP
BAL, CAA, CAC, CAY, DEA,
ISR, LAE, NSA, OAB, SEA
CAA, CAC, CAY, ISR, OAB,
SEA, SUP
BAL, LAH, OAB, SEA
BAL, CIN, DEA, DEN, EPA,
ISR, LAE, LAH, NSA, OAB,
RTP, SEA, SUP, WAS
ISR
BAL, CAC, DEA, DEN, ISR,
LAE, LAH, RTP, SEA
Calculation
Method
Factors
Factors
Factors
Factors
Factors
Factors
Factors
Factors
Factors
Variables1
PR
PR
PR
PR
PR
PE&PR
PE&PR
PE&PR
PE&PR
1AER = air exchange rate, DE = decay-deposition rate, PR = proximity factor, PE = penetration factor.
5B-14
-------
5B-6.1 Air Exchange Rates for Indoor Residential Microenvironments
Distributions of air exchange rates (AERs) for the indoor residential microenvironments
(ME-1) were developed using data from several studies. The analysis of these data and the
development of most of the distributions used in the modeling were originally described in detail
in US EPA (2007) Appendix A, though recently updated by Cohen et al. (2012) and provided in
Appendix 5E.
The analyses indicated that the AER distributions for the residential microenvironments
depend on the type of air conditioning (A/C) and on the outdoor temperature, among other
variables for which we do not have sufficient data to estimate. These analyses demonstrate that
the AER distributions vary greatly across cities, A/C types, and temperatures, so that the selected
AER distributions for the modeled cities should also depend on these attributes. For example,
the mean AER for residences with A/C ranges from 0.38 in Research Triangle Park, NC at
temperatures > 25 °C upwards to 1.244 in New York, NY considering the same temperature bin.
For each combination of A/C type, city, and temperature with a minimum of 11 AER
values, exponential, lognormal, normal, and Weibull distributions were fit to the AER values and
compared. Generally, the lognormal distribution was the best-fitting of the four distributions,
and so, for consistency, the fitted lognormal distributions are used for all the cases. Los Angeles
had an adequate number of samples and identifiers to distinguish the estimated AER
distributions by central A/C and room unit A/C for the homes with A/C.
There were a number of limitations in generating study-area specific AER stratified by
temperature and A/C type. For example, AER data and derived distributions were available only
for selected cities, and yet the summary statistics and comparisons demonstrate that the AER
distributions depend upon the city as well as the temperature range and A/C type. As a result,
city-specific AER distributions were used where possible; otherwise staff selected AER data
from a similar city. Another important limitation of the analysis was that distributions were not
able to be fitted to all of the temperature ranges due to limited number of available measurement
data in these ranges. A description of how these limitations were addressed can be found in
Appendix 5E. The AER distributions used for the exposure modeling are given in Table 5B-3
(Residences with A/C) and Table 5B-4 (Residences without A/C).
5B-15
-------
Table 5B-3. AERs for indoor residential microenvironments (ME-1) with A/C by study
area and temperature.
Study Area
Atlanta, Baltimore,
Washington DC
Boston, New York,
Philadelphia
Chicago, Cleveland,
Detroit
Dallas, Houston
Denver, St. Louis
Los Angeles
(Central A/C)
Los Angeles
(Room Unit A/C)
Sacramento
Daily Mean
Temperature
(°C)
< 10
10-20
20-25
>25
<10
10-25
>25
<10
10-20
20-25
>25
<20
20-25
25-30
>30
<10
10-20
20-25
25-30
>30
<20
20-25
>25
<20
20-25
>25
<25
>25
Lognormal Distribution
{GM, GSD, min, max}
{0.962, 1.809,0.1, 10}
{0.562, 1.906,0.1, 10}
{0.397, 1.889,0.1, 10}
{0.380, 1.709,0.1, 10}
{0.711,2.108,0.1, 10}
{1.139,2.677,0.1, 10}
{1.244,2.177,0.1, 10}
{0.744, 1.982,0.1, 10}
{0.811,2.653,0.1, 10}
{0.785,2.817,0.1, 10}
{0.916,2.671,0.1, 10}
{0.407,2.113,0.1, 10}
{0.467, 1.938,0.1, 10}
{0.422,2.258,0.1, 10}
{0.499, 1.717,0.1, 10}
{0.921, 1.854,0.1, 10}
{0.573, 1.990,0.1, 10}
{0.530,2.427,0.1, 10}
{0.527,2.381,0.1, 10}
{0.609,2.369,0.1, 10}
{0.577, 1.897,0.1, 10}
{1.084,2.336,0.1, 10}
{0.861,2.344,0.1, 10}
{0.672, 1.863,0.1, 10}
{1.674,2.223,0.1, 10}
{0.949, 1.644,0.1, 10}
{0.503, 1.921,0.1, 10}
{0.830,2.353,0.1, 10}
Original AER Study Data
Used
Research Triangle Park,
NC
New York, NY
Detroit, Ml and New York,
NY
Houston, TX
All Cities Outside of CA
Los Angeles, CA
Los Angeles, CA
Sacramento, Riverside,
San Bernardino Counties
5B-16
-------
Table 5B-4. AERs for indoor residential microenvironments (ME-1) without A/C by study
area and temperature.
Study Area
Atlanta, Baltimore,
Denver, St. Louis,
Washington DC
Boston, New York,
Philadelphia
Chicago, Cleveland,
Detroit
Dallas, Houston
Los Angeles
Sacramento
Daily Mean
Temperature
(°C)
<10
10-20
>20
<10
10-20
>20
<0
0-10
10-20
20-25
>25
< 10
10-20
>20
<20
20-25
>25
< 10
10-20
20-25
>25
Lognormal Distribution
{GM, GSD, min, max}
{0.923, 1.843,0.1, 10}
{0.951,2.708,0.1, 10}
{1.575,2.454,0.1, 10}
{1.016,2.138,0.1, 10}
{0.791,2.042,0.1, 10}
{1.606,2.119,0.1, 10}
{1.074, 1.772,0.1, 10}
{0.760, 1.747,0.1, 10}
{1.447,2.950,0.1, 10}
{1.531,2.472,0.1, 10}
{1.901,2.524,0.1, 10}
{0.656, 1.679,0.1, 10}
{0.625,2.916,0.1, 10}
{0.916,2.451,0.1, 10}
{0.744,2.057,0.1, 10}
{1.448,2.315,0.1, 10}
{0.856,2.018,0.1, 10}
{0.526,3.192,0.1, 10}
{0.665,2.174,0.1, 10}
{1.054, 1.711,0.1, 10}
{0.827,2.265,0.1, 10}
Original AER Study Data
Used
All Cities Outside of CA
New York, NY
Detroit, Ml and New York,
NY
Houston, TX
Los Angeles, CA
Sacramento, Riverside,
San Bernardino Counties
5B-6.2 Air Conditioning Prevalence for Indoor Residential MicroEnvironments
The selection of an AER distribution is conditioned on the presence or absence of
A/C. We assigned this housing attribute to indoor residential microenvironments (ME-1) using
A/C prevalence data from the American Housing Survey (AHS)9. A/C prevalence is noted as
distinct from usage rate, the latter represented by the AER distribution and dependent on
temperature. The A/C prevalence data were assigned to our study areas where the AHS data best
matched our exposure simulation years (Table 5B-5). Because we were able to stratify the AER
distributions by three A/C types in Los Angeles, both the individual central and room unit values
were used. In all other study areas, the sum of room unit and central A/C prevalence was used.
' Available at: http://www.census.gov/housing/ahs/data/metro.html.
5B-17
-------
Table 5B-5. American Housing Survey A/C prevalence from Current Housing Reports (Table 1-4) for selected urban areas.
Metropolitan
Area
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles3
New York4
Philadelphia
Sacramento
St. Louis
Washington, DC
Area1
MA
MSA
CMSA
PMSA
PMSA
PMSA
MA
PMSA
PMSA
PMSA
PMSA
PMSA
PMSA
MA
MA
Study
Years
2004
2007
2007
2009
2004
2002
2004
2009
2007
2003
2009
2009
2004
2004
2007
Total Occupied
Housing Units
(xlOOO)
1595.8
1012.3
1057.1
3010.7
769.3
1235.3
855.7
1672.5
1872.0
5152.4
4493.3
1916.2
669.4
1139.6
1949.1
Number of Occupied Housing Units (x1 000) with:
Central
A/C
1473.8
785.4
291.5
2050.6
416.4
1146.3
425.1
1194.3
1682.5
2448.4
872.4
1095.9
549.5
974.4
1729.6
>1 Central
A/C
245.9
44.4
20.3
116.2
14.1
171.6
16.8
46.5
153.7
161.6
38.2
52.3
30.7
53.7
145.7
1 Room
Unit
39.8
55.7
259.4
412.0
132.9
25.5
123.8
192.3
46.0
702.1
1036.9
197.9
57.1
65.8
69.2
2 Room
Units
32.5
72.4
198.7
265.1
47.0
29.8
20.3
82.8
59.9
118.6
1184.1
260.8
12.3
43.5
64.8
3+ Room
Units
18.3
65.4
156.0
124.4
17.6
27.8
3.9
29.2
60.6
46.9
812.6
265.8
2.4
16.6
61.2
% of Occupied Housing Units with:2
Central
A/C
92
78
28
68
54
92.8
50
71
90
47.5
19
57
82
86
89
Window
Units
6
19
58
27
26
6.7
17
18
9
16.8
68
38
11
11
10
Central &
Window A/C
98
97
86
95
80
99.5
67
90
99
64.4
87
95
93
97
99
1 MA - metropolitan area; CMSA - consolidated metropolitan statistical area; PMSA - primary metropolitan statistical area.
2 Shaded areas indicate final values used in APEX functions files to select AER distributions used for indoor residential microenvironments (ME-1).
3 Los Angeles includes Los Angeles-Long Beach, Riverside-San Bernardino-Ontario, and Anaheim-Santa Ana MSA's.
4 New York is represented by the NY-Nassau-Suffolk-Orange MSA.
5B-18
-------
5B-6.3 AER Distributions for Other Indoor Microenvironments
To estimate AER distributions for non-residential, indoor environments (e.g., offices,
libraries), we obtained and analyzed two AER data sets: "Turk" (Turk et al., 1989); and "Persily"
(Persily and Gorfain, 2004; Persily et al., 2005). The Turk data set includes 40 AER
measurements from offices (25 values), schools (7 values), libraries (3 values), and multi-
purpose buildings (5 values), each measured using an SFe tracer over two or four hours in
different seasons of the year. The Persily data were derived from the US EPA Building
Assessment Survey and Evaluation (BASE) study, which was conducted to assess indoor air
quality, including ventilation, in a large number of randomly selected office buildings throughout
the US. This data base consists of 390 AER measurements in 96 large, mechanically ventilated
offices. AERs were measured both by a volumetric method and by a CCh ratio method, and
included their uncertainty estimates. For these analyses, we used the recommended "Best
Estimates" defined by the values with the lower estimated uncertainty; in the vast majority of
cases the best estimate was from the volumetric method.
Due to the small sample size of the Turk data, the data were analyzed without
stratification by building type and/or season. For the Persily data, the AER values for each office
space were averaged, rather using the individual measurements, to account for the strong
dependence of the AER measurements for the same office space over a relatively short period.
The mean values are similar for the two studies, but the standard deviations are about twice as
high for the Persily data. We fitted exponential, lognormal, normal, and Weibull distributions to
the 96 office space average AER values from the more recent Persily data, and the best fitting of
these was the lognormal. The fitted parameters for this distribution are a geometric mean of
1.109, geometric standard deviation of 3.015, and bounded by the lower and upper values of the
sample data set (0.07, 13.8}. These are used for AER distributions for several indoor non-
residential microenvironments (ME-2, ME-4, ME-5, ME-8, ME-9, ME-10, ME-11, ME-26)
except for indoor schools (ME-7) and indoor restaurants, bars, night clubs, and cafes (ME-3 and
ME-6).
The AER distribution used for indoor schools (ME-7) is a discrete distribution (0.8, 1.3,
1.8, 2.19, 2.2, 2.21, 3.0, 0.6, 0.1, 0.6, 0.2, 1.8, 1.3, 1.2, 2.9, 0.9, 0.9, 0.9, 0.9, 0.4, 0.4, 0.4, 0.4,
0.9, 0.9, 0.9, 0.9, 0.3, 0.3, 0.3, 0.3} developed using data from Turk et al. (1989) and Shendell et
al. (2004).
5B-19
-------
The AER distribution used for indoor restaurants, bars, night clubs, and cafes (ME-3,
ME-6) is a fitted lognormal distribution, having a geometric mean = 3.712, geometric standard
deviation = 1.855 and bounded by the lower and upper values of the sample data set (1.46,
9.07}. This distribution was developed using data from Bennett et al. (2012b), who measured
these six values in restaurants (details on derivation provided in Appendix 5E).
5B-6.4 Proximity and Penetration Factors for In-vehicle and Near-Road
Microenvironments
For the in-vehicle proximity and penetration factors (ME-22, ME-23, ME-24, ME-28),
we use distributions developed from the Cincinnati Ozone Study (American Petroleum Institute,
1997, Appendix B; Johnson et al., 1995). This field study was conducted in the greater
Cincinnati metropolitan area in August and September, 1994. Vehicle tests were conducted
according to an experimental design specifying the vehicle type, road type, vehicle speed, and
ventilation mode. Vehicle types were defined by the three study vehicles: a minivan, a full-size
car, and a compact car. Road types were interstate highways (interstate), principal urban arterial
roads (urban), and local roads (local). Nominal vehicle speeds (typically met over one minute
intervals within 5 mph) were at 35 mph, 45 mph, or 55 mph. Ozone concentrations were
measured inside the vehicle, outside the vehicle, and at six fixed-site monitors in the Cincinnati
area. Table 5B-6 lists the parameters of the normal distributions developed for penetration and
proximity factors for in-vehicle microenvironments used in this modeling analysis.
Table 5B-6. Parameter values for distributions of penetration and proximity factors used
for estimating in-vehicle microenvironmental concentrations.
Microenvironmental
Factor
Penetration
Proximity
Road type
All
Local
Urban
Interstate
Arithmetic
Mean
0.300
0.755
0.754
0.364
Standard
Deviation
0.232
0.203
0.243
0.165
Lower
Bound1
0.100
0.422
0.355
0.093
Upper
Bound
1.0
1.0
1.0
1.0
A 5th percentile value estimated using a normal approximation as Mean - 1.64 x standard deviation.
5B-20
-------
The Vehicle Miles of Travel (VMT) fractions10 provided by the U.S. Department of
Transportation (DOT) are used to generate daily conditional variables that determine the
selection of which proximity factor distributions are used to estimate in-vehicle
microenvironmental concentrations (Table 5B-7). For local and interstate road types, the VMT
for the same DOT categories were used. For urban roads, the VMT for all other DOT road types
were summed (i.e., other freeways/expressways, other principal arterial, minor arterial, and
collector). At the time of this writing, data were only available for four of our modeled years,
2006-2008 and 2010. Staff assumed that values for 2009 would be best represented by averaging
2008 and 2010.
For all outdoors-near-road microenvironments (ME-17, ME-18, ME-19, ME-20, ME-21)
we employed the distribution for local roads (i.e., a normal distribution (0.755, 0.203}, bounded
by 0.422 and 1.0), based on the assumption that most of the outdoors-near-road ozone exposures
will occur proximal to local roads.
5B-6.5 Proximity and Penetration Factors for Outdoor Microenvironments
All outdoor microenvironments (ME-12, ME-13, ME-14, ME-15, ME-16, ME-25) are
assumed well represented by the census tract level Os concentrations. Therefore, both the
penetration factor and proximity factor for this microenvironment were set to equal 1.
5B-6.6 Ozone Decay and Deposition Rates
A distribution for combined Os decay and deposition rates was obtained from the
analysis of measurements from a study by Lee et al. (1999). This study measured decay rates in
the living rooms of 43 residences in Southern California. Measurements of decay rates in a
second room were made in 24 of these residences. The 67 decay rates range from 0.95 to 8.05
hour"1. A lognormal distribution was fit to the measurements from this study, yielding a
geometric mean of 2.51 and a geometric standard deviation of 1.53. These values are
constrained to lie between 0.95 and 8.05 hour"1. This distribution was used for all indoor
microenvironments.
10 U.S. Department of Transportation, Federal Highway Administration. Annual Highway Statistics, Table HM-71:
Urbanized Areas - Miles And Daily Vehicle Miles Of Travel. For example, 2010 data available at:
www.fhwa.dot. gov/policvinformation/statistics/2010/xls/hm71 .xls
5B-21
-------
Table 5B-7. VMT fractions of interstate, urban, and local roads in the study areas used to select in-vehicle proximity factor
distributions.
Study Area
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Wash., DC
2006
Inter-
state Urban Local
0.34 0.46 0.20
0.34 0.59 0.07
0.32 0.55 0.13
0.30 0.58 0.12
0.40 0.44 0.16
0.30 0.66 0.04
0.23 0.67 0.10
0.26 0.64 0.10
0.24 0.72 0.04
0.29 0.66 0.05
0.19 0.66 0.15
0.23 0.65 0.12
0.25 0.72 0.03
0.36 0.45 0.19
0.31 0.61 0.08
2007
Inter-
state Urban Local
0.34 0.47 0.19
0.34 0.59 0.07
0.32 0.55 0.13
0.30 0.58 0.12
0.40 0.44 0.16
0.30 0.66 0.04
0.24 0.66 0.10
0.25 0.65 0.10
0.24 0.72 0.04
0.29 0.67 0.04
0.19 0.65 0.16
0.24 0.65 0.11
0.25 0.70 0.05
0.37 0.45 0.18
0.31 0.61 0.08
2008
Inter-
state Urban Local
0.32 0.45 0.23
0.34 0.59 0.07
0.32 0.54 0.14
0.31 0.57 0.12
0.39 0.45 0.16
0.30 0.65 0.05
0.25 0.65 0.10
0.24 0.66 0.10
0.24 0.72 0.04
0.28 0.67 0.05
0.19 0.66 0.15
0.24 0.65 0.11
0.23 0.69 0.08
0.38 0.45 0.17
0.30 0.61 0.09
2009
Inter-
state Urban Local
0.31 0.44 0.25
0.34 0.59 0.07
0.32 0.54 0.14
0.30 0.57 0.13
0.39 0.45 0.16
0.30 0.66 0.04
0.25 0.65 0.10
0.25 0.65 0.10
0.24 0.72 0.04
0.29 0.66 0.05
0.19 0.66 0.15
0.20 0.59 0.21
0.23 0.69 0.08
0.37 0.45 0.18
0.30 0.61 0.09
2010
Inter-
state Urban Local
0.30 0.43 0.27
0.34 0.59 0.07
0.32 0.54 0.14
0.31 0.56 0.13
0.38 0.46 0.16
0.29 0.67 0.04
0.25 0.65 0.10
0.26 0.63 0.11
0.24 0.72 0.04
0.29 0.66 0.05
0.19 0.65 0.16
0.18 0.52 0.30
0.23 0.69 0.08
0.36 0.44 0.20
0.30 0.61 0.09
A few individual fractions have been adjusted to yield an annual sum of 1.00.
5B-22
-------
5B-7 AMBIENT OZONE CONCENTRATIONS
To estimate exposure in this assessment, APEX requires hourly ambient Os
concentrations at a set of locations (or air districts) within study area. We used hourly ambient
monitoring data along with a statistical approach (VNA) to better approximate spatial
heterogeneity (where such heterogeneity might be present) across each study area (Os HREA,
Chapter 4). General processing steps performed to generate the final APEX ambient
concentration input files that were used were as follows.
After identifying the 15 study areas to be modeled in this assessment, staff defined a
broad air quality modeling domain for each study area, specifically bounding where exposures
were to be estimated. We evaluated 1) counties modeled in the previous 2007 Os NAAQS
review common to current study areas, 2) political/statistical county aggregations (MSA,
PMSA), and 3) if the study area was designated as a non-attainment area (NAA), the counties
that were part of the NAA list. A final list of counties was generated using this information
(Table 5B-8), then hourly Os concentrations were estimated at every census tract within the
counties that comprised each study area (Os HREA, Chapter 4). These data served as the air
quality input to APEX with some exception (see below), though note also, not all of the
estimated hourly concentrations would be used in the exposure simulation even if supplied to
APEX.
A 30 km radius of influence was used for each monitoring site within the above county-
level defined study. All census tracts that fell within the 30 km radius of each ambient monitor
used to estimate the air quality concentration fields were selected, then any tracts/monitor radii
that were largely outside of the urban core were removed, thus defining a final exposure
modeling domain in each study area (Table 5B-8).
Because APEX uses 2000 census population data and the air concentrations were
modeled to 2010 census tracts, some of the air district locations differed slightly from that of the
exposure tracts, resulting in different numbers of air districts when compared with the number of
census tracts used in simulating exposures. This difference is expected to have a negligible
effect on exposure and risk results because APEX always uses the air district nearest to the tract
to be modeled, the distances between any two air district centroids within these urban study areas
(census tract level) is expected to be small, and the concentration gradient across that said
distance is also expected to not be significant.
5B-23
-------
Further, staff had computational difficulty in simulating the large number of tracts and air
districts for the Los Angeles, New York, and Chicago study areas (number and size of arrays
needed in APEX calculations was beyond the standard PC capabilities); based on simulations
that ran to completion, the maximum number of air districts possible using a standard 32-bit PC
was estimated as 1,900 - 2,000. Thus, to make the analysis more tractable for these study areas,
first staff reduced the number of air districts originally modeled (i.e., all year 2010 US tracts in
the broad county domain) to the number needed for the actual year 2000 census tracts in the
exposure model domain (i.e., all tracts within 30 km of ambient monitors in the broad county
domain). Using this approach, the number of air districts was reduced to the following: Chicago
(1,882), Los Angeles (3,268), and New York (4,646). For Los Angeles and New York, the
number of air districts was reduced to 2,000 and 1,900 using simple random sampling of these
tracts using SAS's SURVEYSELECT procedure; the number of air districts for Chicago
remained at 1,882. While we estimated this number of districts would run on a standard PC,
these three study areas would only run on a 64-bit PC.
The final list of year 2000 census tract IDs where exposure was modeled is within the
APEX control files. The final list of 2010 census tract IDs where ambient concentrations were
estimated is within the APEX air districts files. Table 5B-9 contains the final list of counties, the
number of US census tracts where exposures were estimated, the number air districts ultimately
used from the air quality input files, and the population counts represented in each study area.
The final list of year 2000 census tract IDs where exposure was modeled is within the APEX
control files. The final list of 2010 census tract IDs where ambient concentrations were
estimated is within the APEX air districts files. Figure 5B-1 through Figure 5B-4 illustrate the
general exposure modeling domains (i.e., the selected census tract centroids falling within 30 km
of a ambient monitor) for each of the 15 study areas.
Multiple unique APEX input files are used for the current Os assessment, varying by the
air quality scenario, year, and study area, though generally in two forms:
• concsCSA[studyarea]S[scenario]?[std. avg.periodJYfyearJ.txt: hourly
concentrations for each tract, by study area, air quality scenario, standard
averaging period, year
• districtsCSAfstudyareaJYfyearJ.txt: tract ID's, latitudes and longitudes, start and
stop dates of concentrations
5B-24
-------
Table 5B-8. Identification of U.S. counties and the number of APEX air districts included each study area.
Study Area (State Abbreviation: List of Counties1)
Atlanta (GA: Barrow, Bartow, Butts, Carroll, Cherokee, Clayton, Cobb, Coweta, Dawson, De Kalb, Douglas, Fayette, Forsyth,
Fulton, Gwinnett, Hall, Haralson, Heard, Henry, Jasper, Lamar, Meriwether, Newton, Paulding, Pickens, Pike, Polk, Rockdale,
Spalding, Troup, Upson, Walton; AL Chambers)
Baltimore (MD: Anne Arundel, Baltimore, Carroll, Harford, Howard, Queen Anne's, Baltimore (City))
Boston (MA: Barnstable, Bristol, Dukes, Essex, Middlesex, Nantucket, Norfolk, Plymouth, Suffolk, Worcester)
Chicago (IL: Cook, DeKalb, DuPage, Grundy, Kane, Kankakee, Kendall, Lake, McHenry, Will; IN: Jasper, Lake, LaPorte,
Newton, Porter, Kenosha)
Cleveland (OH: Ashtabula, Cuyahoga, Geauga, Lake, Lorain, Medina, Portage, Summit)
Dallas (TX: Collin, Dallas, Denton, Ellis, Hunt, Johnson, Kaufman, Parker, Rockwall, Tarrant, Wise)
Denver (CO: Adams, Arapahoe, Boulder, Broomfield2, Clear Creek, Denver, Douglas, Elbert, Gilpin, Jefferson, Larimer, Park,
Weld)
Detroit (Ml: Genesee, Lapeer, Livingston, Macomb, Monroe, Oakland, St. Clair, Washtenaw, Wayne)
Houston (TX: Austin, Brazoria, Chambers, Fort Bend, Galveston, Harris, Liberty, Montgomery, San Jacinto, Waller)
Los Angeles, (CA: Los Angeles, Orange, Riverside, San Bernardino, Ventura)
New York (CT: Fairfield, Middlesex, New Haven; NJ: Bergen, Essex, Hudson, Hunterdon, Mercer, Middlesex, Monmouth,
Morris, Passaic, Somerset, Sussex, Union, Warren; NY: Bronx, Kings, Nassau, New York, Orange, Putnam, Queens,
Richmond, Rockland, Suffolk, Westchester)
Philadelphia (DE: New Castle, MD: Cecil; NJ: Atlantic, Burlington, Camden, Cape May, Cumberland, Gloucester, Ocean,
Salem; PA: Bucks, Chester, Delaware, Montgomery, Philadelphia)
Sacramento (CA: El Dorado, Nevada, Placer, Sacramento, Solano, Sutter, Yolo)
St. Louis (IL: Bond, Calhoun, Clinton, Jersey, Macoupin, Madison, Monroe, Saint Clair, MO: Crawford, Franklin, Jefferson,
Lincoln, Saint Charles, Saint Louis, Warren, Washington, St. Louis City)
Washington, DC (District of Columbia; MD: Calvert, Charles, Frederick, Montgomery, Prince George's, St. Mary's, VA:
Arlington, Clarke, Culpeper, Fairfax, Fauquier, Frederick, Loudoun, Prince William, Spotsylvania, Stafford, Warren, Alexandria
City, Fairfax City, Falls Church City, Fredericksburg City, Manassas City, Manassas Park City, Winchester City, WV: Jefferson)
All AREAS (Counties: 207 Exposure of 215 Air Quality)
APEX Air Districts
(VNA Total)
664(1,019)
603 (679)
1 ,005 (1 ,276)
1 ,882 (2,267)
802 (830)
1012(1,312)
655 (839)
1,419(1,568)
779 (1 ,074)
2,000 (3,920)
1,900(5,003)
1,452(1,735)
447 (623)
494 (626)
1,013(1,391)
16,127(24,162)
1 italicized: in air quality domain but not in exposure modeling domain; considered outside of urban core or no monitors.
2 this county is newly defined in the 2010 census.
5B-25
-------
Table 5B-9. Ambient monitors used to define exposure modeling domain and the population modeled in each study area.
Study Area (State Abbreviation: List of Monitors1)
Atlanta (GA: 130590002, 130670003, 130770002, 130850001, 130890002, 130893001, 130970004, 131130001,
131210055, 131350002, 131510002, 132230003, 132319991, 132470001)
Baltimore (MD: 240030014, 240051007, 240053001, 240130001, 240251001, 240259001, 240290002, 240313001,
240330030, 240338003, 240339991 , 245100054)
Boston (MA: 250092006, 250094004, 250094005, 250095005, 250170009, 250171102, 250213003, 250250041,
250250042, 250270015, 250270024; NH: 3301 11011; Rl: 440071010)
Chicago (IL: 170310001, 170310032, 170310042, 170310064, 170310072, 170310076, 170311003, 170311601,
170314002, 170314007, 170314201, 170317002, 170436001, 170890005, 170971002, 170971007, 171110001,
171971011; IN: 180890022, 180890030, 180892008, 180910005, 180910010, 181270024, 181270026, Wl:
550590019, 551010017, 551270005)
Cleveland (OH: 390071001, 390350034, 390350060, 390350064, 390355002, 390550004, 390850003, 390850007,
390853002, 390930018, 391030003, 391030004, 391331001, 391510016, 391514005, 391530020)
Dallas (TX: 480850005, 481130069, 481130075, 481130087, 481133003, 481210034, 481211032, 481390015,
481390016, 481391044, 482210001, 482311006, 482510003, 482570005, 483670081, 483970001, 484390075,
484391002, 484392003, 484393009, 48439301 1)
Denver (CO: 080013001, 080050002, 080050006, 080130007, 080130011, 080137001, 080137002, 080190004,
080190005, 080310002, 080310014, 080310025, 080350004, 080590002, 080590005, 080590006, 080590011,
080590013, 080690007, 080690011 , 080691004, 080699991, 080930001, 081190003, 081230009)
Detroit (Ml: 260490021, 260492001, 260910007, 260990009, 260991003, 261250001, 261470005, 261610008,
261619991, 261630001, 261630015, 261630016, 261630019)
Houston (TX: 480391004, 482010024, 482010026, 482010029, 482010046, 482010047, 482010051, 482010055,
482010062, 482010066, 482010070, 482010075, 482010416, 482011015, 482011034, 482011035, 482011039,
482011050,483390078)
Los Angeles (CA: 060370002, 060370016, 060370113, 060371002, 060371103, 060371201, 060371301, 060371302,
060371602, 060371701, 060372005, 060374002, 060374006, 060375005, 060376012, 060379033, 060590007,
060591003, 060592022, 060595001, 060650004, 060650009, 060650012, 060651010, 060651016, 060651999,
060652002, 060655001, 060656001, 060658001, 060658005, 060659001, 060659003, 060710001, 060710005,
060710012, 060710306, 060711004, 060711234, 060712002, 060714001, 060714003, 060719002, 060719004,
061110007, 061110009,061111004, 061112002,061112003,061113001)
Census
Tracts for
Exposure
678
618
1,028
2,055
879
1,036
675
1,454
802
3,352
Population
Represented
3,850,951
2,209,226
4,449,291
8,345,373
2,692,846
4,698,392
2,626,239
4,572,479
3,925,054
14,950,340
5B-26
-------
Study Area (State Abbreviation: List of Monitors1)
New York (CT: 090010017, 090011123, 090013007, 090019003, 090070007, 090090027, 090093002; NJ:
340030005, 340030006, 340130003, 340170006, 340190001, 340210005, 340219991, 340230011, 340250005,
340273001 , 340290006, 340315001 ; NY: 360050083, 360050110, 360050133, 360610135, 360790005, 360810098,
360810124, 360850067, 360870005, 361030002, 361030004, 361030009, 361192004)
Philadelphia (DE: 100031007, 100031010, 100031013, MD: 240150003; NJ: 340070003, 340071001, 340110007,
340150002, 340210005, 340219991 , 340290006; PA: 420170012, 420290100, 420450002, 420910013, 421010004,
421010014, 421010024, 421010136)
Sacramento (CA: 060170010, 060170020, 060570005, 060610002, 060610004, 060610006, 060670002, 060670006,
060670010, 060670011, 060670012, 060670013, 060670014, 060675003, 060953003, 061010003, 061010004,
061131003)
St. Louis (IL: 170831001, 171190008, 171191009, 171193007, 171199991, 171630010; MO: 290990012, 290990019,
291130003, 291831002, 291831004, 291890004, 291890005, 291890014, 295100085, 295100086)
Washington, DC (110010025, 110010041, 110010043; MD: 240030014, 240090011, 240130001, 240170010,
240210037, 240313001, 240330030, 240338003, 240339991, 240430009; VA: 510130020, 510330001, 510590005,
510590018, 510590030, 510591005, 510595001, 510610002, 510690010, 511071005, 511390004, 511530009,
511790001, 515100009; WV: 540030003)
All AREAS (324 ambient monitors)
Census
Tracts for
Exposure
4,889
1,555
461
518
1,037
21,037
Population
Represented
18,520,868
5,506,954
1,926,598
2,340,325
4,498,374
85,113,310
A 30 km radius for monitors operating anytime during 2006-2010 was used to select census tracts in defining the exposure modeling domain.
5B-27
-------
CSA122-ATL
CSA148-BOS
CSA999-BAL
CSA176-CHI
Figure 5B-1. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Atlanta, Boston, Baltimore and Chicago study areas.
5B-28
-------
CSA184-CLE
CSA216-DEN
CSA206-DAL
CSA220-DET
Figure 5B-2. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Cleveland, Dallas, Denver and Detroit study areas.
5B-29
-------
CSA288-HOU
CSA408-NY
CSA348-LA
CSA428-PHI
Figure 5B-3. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Houston, Los Angeles, New York and Philadelphia study areas.
5B-30
-------
CSA472-SAC
CSA548-WAS
CSA476-STL
Figure 5B-4. Illustration of APEX exposure modeling domains (2000 US Census tract
centroids) for Sacramento, St. Louis and Washington DC study areas.
5B-31
-------
5B-8 METEOROLOGICAL DATA
Temperature data are used by APEX in selecting human activity data and in estimating
AERs for indoor residential microenvironments. Hourly surface temperature measurements
were obtained from the National Weather Service ISH data files.11 The weather stations used for
each city are given in Table 5B-10. When developing profiles and selecting for am AER, APEX
uses temperature data from the closest weather station to each Census tract.
Missing temperature data were estimated by the following procedure. Where there were
consecutive strings of missing values (data gaps) of 4 or fewer hours, missing values were
estimated by linear interpolation between the valid values at the ends of the gap. Remaining
missing values at a station were estimated by fitting linear regression models for each hour of the
day, with each of the other monitors, and choosing the model which maximizes R2, for each hour
of the day, subject to the constraints that R2 be greater than 0.50 and the number of regression
data values (days) is at least 60. If there were any remaining missing values at this point, for
gaps of 6 or fewer hours, missing values were estimated by linear interpolation between the valid
values at the ends of the gap. Any remaining missing values were replaced with the value at the
closest station for that hour.
There were negligible differences between the statistically filled and the original
temperature data with missing values. On average, daily mean temperatures were approximately
0.02 °C greater in the final data set used by APEX, compared with the data set having missing
temperatures. The greatest positive difference occurred at station '2227013864', where the filled
data had a daily average mean of about 0.72 °C greater than that of the data set with missing
values. The greatest negative difference was associated with station '2403603710', where the
filled data had a daily average mean of about -0.27 °C less than that of the data set with missing
values. Given these small differences, the number of stations used to represent meteorological
conditions in each study area and the range of values used by APEX in creating diary pools (e.g.,
50 - 68 °F) or AER distributions (e.g., 55 - 84 °F) , the impact of the filled values to estimated
exposures is assumed negligible.
Multiple unique APEX input files are used for the current Os assessment, varying by the
year and study area, though generally in two forms:
• METdataCSAfstudyareaJYfyearJ.txt: hourly temperature for each MET station,
by study area and year
• METlocsCSAfstudyareaJYfyearJ.txt: MET station ID's, latitudes and longitudes,
start and stop dates of temperature data
11 http://www.ncdc.noaa.gov/oa/climate/surfaceinventories.html
5B-32
-------
Table 5B-10. Study area meteorological stations, locations, and hours of missing data.
Study Area
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
ISH ID1
2217003813
2219013874
2219503888
2225593842
2227013864
2228403892
2228713871
2311013873
2320093801
2406093721
4594013705
2505464710
2506014704
2506454769
2507014765
2509014739
2509594746
4394514710
2530094846
2530594892
2534014819
2535014848
2543094822
2521014895
2524014820
2524504853
2525014852
2258313960
2259003927
2259613961
2466093037
2466693067
2469523036
2476824051
2476994062
2564024018
2565003017
2537094847
2537514822
2537614853
2537714804
2538404888
2539014836
2539514833
2637014826
2241012917
Latitude
32.683
33.633
33.767
32.517
33.917
32.616
33.583
33.95
34.35
39.167
38.817
41.917
41.65
41.917
41.717
42.367
42.267
42.933
41.983
41.917
41.783
41.7
42.2
40.917
41.4
41.517
41.25
32.85
32.9
32.817
38.817
39.567
39.717
40.433
40.45
41.15
39.833
42.217
42.4
42.233
42.617
42.917
42.783
42.267
42.967
29.95
Longitude
-83.65
-84.433
-84.517
-84.95
-84.517
-85.433
-85.85
-83.333
-85.167
-76.683
-76.867
-71.5
-70.517
-70.733
-71.433
-71.017
-71.883
-71.433
-87.917
-88.25
-87.75
-86.333
-89.1
-81.433
-81.85
-81.683
-80.667
-96.85
-97.017
-97.367
-104.717
-104.85
-104.75
-104.633
-105.017
-104.8
-104.65
-83.35
-83
-83.533
-82.833
-82.533
-84.583
-84.467
-83.75
-94.017
2006
15
0
14
16
2502
469
24
4
14
0
34
275
46
5
0
0
34
3
2
71
0
3
1
7
0
12
0
1
0
6
2
2
32
72
19
1
0
1
17
1
11
40
1
4
0
5
2007
2
0
15
4
1611
114
4
4
30
0
86
71
84
10
1
2
15
10
1
12
1
0
0
1
1
116
0
2
0
22
2
1
41
393
532
7
2
1
110
7
27
140
3
35
0
24
2008
140
101
113
170
266
209
168
271
187
101
173
362
126
259
94
97
128
103
126
170
127
91
96
128
82
144
119
93
88
157
108
104
103
876
314
129
91
106
226
148
236
94
148
329
116
119
2009
113
41
103
52
93
77
210
39
59
54
184
741
54
143
51
41
61
49
44
97
44
46
43
56
38
79
45
51
46
84
110
53
52
134
381
46
44
40
104
107
59
32
70
49
52
67
2010
128
18
29
48
73
540
55
45
68
27
42
315
95
285
45
34
53
55
21
64
23
18
18
44
19
68
28
40
30
138
71
45
32
140
164
66
40
27
122
78
94
55
37
144
74
27
5B-33
-------
Study Area
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
ISH ID1
2242012923
2243012960
2243512918
2244503904
2286023119
2288023152
2295023174
2297023129
2381523161
2381603159
2383023187
2391093111
2392623136
4718703104
4718823158
2408454760
2409614706
2502014734
2502594741
2502964707
2503014732
2503504781
2503614757
2503794745
2503814714
2504094702
2505894793
2514554746
2517014737
4486094789
2407093730
2407513735
2408013739
2408454760
2408594732
2408913781
2409614706
2517014737
4596603726
2483023232
2483993225
2584523225
4516023202
2433813802
2434013994
2434503966
2445493996
Latitude
29.3
30
29.65
30.583
33.9
34.2
33.933
33.833
34.85
34.733
34.75
34.117
34.217
33.633
33.617
40.183
40.017
40.717
40.85
41.483
40.783
40.783
41.633
41.067
41.5
41.183
41.167
41.7
40.65
40.65
39.45
39.367
39.867
40.183
40.083
39.667
40.017
40.65
39
38.5
38.7
39.3
38.267
38.55
38.75
38.65
37.767
Longitude
-94.8
-95.367
-95.283
-96.367
-117.25
-118.35
-118.4
-118.167
-116.8
-118.217
-118.717
-119.117
-119.083
-116.167
-114.717
-74.067
-74.6
-74.183
-74.067
-73.133
-73.883
-73.1
-73.883
-73.717
-74.1
-73.15
-71.583
-74.8
-75.45
-73.8
-74.567
-75.083
-75.233
-74.067
-75.017
-75.6
-74.6
-75.45
-74.917
-121.5
-121.583
-120.717
-121.933
-89.85
-90.367
-90.65
-90.4
2006
126
1
0
5
13
2
0
2
69
126
31
2046
21
99
2
251
66
1
0
554
0
2
14
8
565
61
687
1387
5
4
4
20
1
251
0
22
66
5
945
3
0
1
34
40
1
5
186
2007
19
0
0
3
25
12
0
4
4
11
3
16
47
18
15
475
62
0
2
292
0
0
5
0
851
68
349
356
4
0
3
84
0
475
10
1
62
4
500
5
0
1
110
55
0
33
24
2008
2468
173
160
108
103
152
113
99
196
438
375
265
311
134
400
566
131
105
129
569
73
174
406
119
1144
146
586
159
148
101
142
268
122
566
143
156
131
148
1299
115
116
217
152
129
98
189
74
2009
444
57
99
107
48
86
44
173
741
176
276
77
218
161
54
183
83
47
70
842
38
138
388
93
1159
114
2970
375
74
56
112
73
57
183
60
244
83
74
1562
51
52
124
78
68
46
62
134
2010
186
30
33
40
29
37
19
269
241
411
554
792
139
60
159
794
121
22
31
703
17
18
139
34
1358
173
127
582
51
21
161
74
21
794
38
89
121
51
1855
77
42
350
46
41
27
41
402
5B-34
-------
Study Area
Washington DC
ISH ID1
2403093738
2403303706
2403513773
2403603710
2404013721
2405013743
2405303717
2405503714
2406093721
2417713734
4594013705
Latitude
38.933
38.267
38.5
38.717
38.3
38.867
39.15
39.083
39.167
39.4
38.817
Longitude
-77.45
-77.45
-77.3
-77.517
-76.417
-77.033
-78.15
-77.567
-76.683
-77.983
-76.867
2006
3
41
177
725
1414
0
38
28
0
36
34
2007
2
48
113
698
123
1
7
28
0
42
86
2008
102
4587
1937
943
322
96
78
64
101
217
173
2009
52
145
788
1631
83
46
32
38
54
137
184
2010
28
36
610
97
86
21
35
181
27
118
42
1 From the Federal Climate Complex Integrated Surface Hourly (ISH) global database.
5B-9 CONDITIONAL VARIABLES
APEX has added flexibility in using conditional variables in association with selection of
the distributions used to represent input variables, across several modules (i.e., CHAD diary
selection, microenvironmental concentration calculations). In this Os assessment, a number of
temperature ranges are used in selecting the particular AER distribution (section 5B-6.1),
maximum daily temperature is also used in diary selection to best match the study area MET data
for the simulated individual (<55, 55-83, and >84; based on Graham and McCurdy, 2004), air
conditioning prevalence data (section 5B-6.2), and designation of roadway type travelled based
on VMT miles (section 5B-6.4).
A single unique APEX input files is used for the current Os assessment, varying by the
year and study area:
• Functions O3CSA[studyarea]Y[year]J^dateJ.txt: conditional variables and values used
5B-35
-------
5B-10 REFERENCES
AHS. (2003). U.S. Bureau of the Census and U.S. Department of Housing and Urban
Development. 2003 American Housing Survey (AHS): National Survey Data. Available
at: http://www.census.gov/hhes/www/housing/ahs/ahs.html, and
http://www.huduser.org/datasets/ahs.html
Akland, G. G., Hartwell, T. D., Johnson, T. R., Whitmore, R. W. (1985). Measuring human
exposure to carbon monoxide in Washington, D. C. and Denver, Colorado during the
winter of 1982-83. Environ Sci Technol. 19:911-918.
American Petroleum Institute. (1997). Sensitivity Testing of pNEM/Os Exposure to Changes in
the Model Algorithms. Health and Environmental Sciences Department.
Bennett, D. H., Wu, X., league, C.H., Lee, K., Cassady, D. L., Ritz, B., Hertz-Picciotto, I.
(2012a). Passive sampling methods to determine household and personal care product use.
JExpos Sci Environ Epidem. 22: 148-160.
Bennett, D. H, Fisk, W., Apte, M. G., Wu, X., Trout, A., Faulkner, D., Sulivan D. (2012b).
Ventilation, Temperature, and HVAC Characteristics in Small and Medium Commercial
Buildings (SMCBs) in California. Indoor Air. 22(4): 309-320.
Geyh, A. S., Xue, J., Ozkaynak, H., Spengler, J. D. (2000). The Harvard Southern California
chronic ozone exposure study: assessing ozone exposure of grade-school-age children in
two southern California communities. Environ Health Perspect. 108: 265-270.
Hartwell, T. D., Clayton, C. A., Ritchie, R. M., Whitmore, R. W., Zelon, H. S., Jones, S. M.,
Whitehurst, D. A. (1984). Study of Carbon Monoxide Exposure of Residents of
Washington, DC and Denver, Colorado. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Research and Development, Environmental Monitoring
Systems Laboratory. EPA-600/4-84-031.
Hertz-Picciotto, L, Cassady, D., Lee, K., Bennett, D. H., Ritz, B., Vogt, R. (2010). Study of Use
of Products and Exposure-Related Behaviors (SUPERB): study design, methods, and
demographic characteristics of cohorts. Environ Health. 9:54.
Isaacs, K. K., McCurdy, T., Glen, G., Nysewander, M., Errickson, A., Forbes, S., Graham, S.,
McCurdy, L., Smith, L., Tulve, N., and Vallero, D. (2012). Statistical properties of
longitudinal time-activity data for use in human exposure modeling. J Expos Sci Environ
Epidemiol. 23(3): 328-336.
Johnson, T. (1984). A Study of Personal Exposure to Carbon Monoxide in Denver, Colorado.
Research Triangle Park, NC: U.S. Environmental Protection Agency, Environmental
Monitoring Systems Laboratory. EPA-600/4-84-014.
Johnson, T. (1989). Human Activity Patterns in Cincinnati, Ohio. Palo Alto, CA: Electric
Power Research Institute. EPRI EN-6204.
Johnson, T., Pakrasi, A., Wisbeth, A., Meiners, G., Ollison, W. (1995). Ozone exposures Within
Motor Vehicles - Results of a Field Study in Cincinnati, Ohio. Proceedings 88th annual
meeting and exposition of the Air & Waste Management Association, San Antonio, TX.
June 18-23, 1995. Preprint paper 95-WA84A.02.
Klepeis, N. E., Tsang, A. M., Behar, J. V. (1996). Analysis of the National Human Activity
Pattern Survey (NHAPS) Respondents from a Standpoint of Exposure Assessment.
5B-36
-------
Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Development. EPA/600/R-96/074.
Knowledge Networks. (2009). Field Report: National-Scale Activity Survey (NSAS).
Conducted for Research Triangle Institute. Submitted to Carol Mansfield November 13,
2009.
Lee, K., Vallarino, J., Dumyahn, T., Ozkaynak, H., and Spengler, J. D. (1999). Ozone decay
rates in residences. J Air Waste Manag Assoc. 49: 1238-1244.
Liu, L.-J. S., Box, M., Kalman, D., Kaufman, J., Koenig J., Larson, T., Lumley, T., Sheppard, L.,
Wallace, L. (2003). Exposure assessment of particulate matter for susceptible populations
in Seattle. Environ Health Persp. Ill: 909-918.
Mansfield, C., Houtven, G. V., Johnson, F. R., Yang, J.-C. (2009). Environmental Risks and
Behavior: Do children spend less time outdoors when ozone pollution is high? ASS A
annual meeting, January 5, 2009. Update of Houtven et al. (2003), using the OAB CHAD
data set, and related to Mansfield et al. (2006).
Mansfield, C., Johnson, F. R., Van Houtven, G. (2006). The missing piece: averting behavior for
children's ozone exposures. Resource Energy Econ. 28:215-228.
McCurdy, T., Glen, G., Smith, L., Lakkadi, Y. (2000). The National Exposure Research
Laboratory's Consolidated Human Activity Database. J Expo Anal Environ Epidemiol. 10:
566-578.
McCurdy, T., Graham, S. E. (2003). Using human activity data in exposure models: analysis of
discriminating factors. J Expos Anal Environ Epidemiol. 13(4): 294-317.
Persily, A., Gorfain, J. (2004). Analysis of Ventilation Data from the U.S. Environmental
Protection Agency Building Assessment Survey and Evaluation (BASE) Study. National
Institute of Standards and Technology, NISTIR 7145, December 2004.
Persily, A., Gorfain, J., Brunner, G. (2005). Ventilation Design and Performance in U.S. Office
Buildings. ASHRAE Journal. April 2005, 30-35.
Robinson, J. P., Wiley, J. A., Piazza, T., Garrett, K., and Cirksena, K. (1989). Activity Patterns
of California Residents and their Implications for Potential Exposure to Pollution.
California Air Resources Board, Sacramento, CA. CARB-A6-177-33.
Cohen, J., Mallya, H., Rosenbaum, A. (2012). Updated Analysis of Air Exchange Rate, ICF
International Memo to John Langstaff. Available in Appendix 5E of Os HREA.
Shendell, D. G., Winer, A. M., Weker, R., Colome, S. D. (2004). Evidence of inadequate
ventilation in portable classrooms: results of a pilot study in Los Angeles County. Indoor
Air. 14:154-158.
Spier, C. E., Little, D. E., Trim, S. C., Johnson, T. R., Linn, W. S., Hackney, J. D. (1992).
Activity patterns in elementary and high school students exposed to oxidant pollution. J
Expo Anal Environ Epidemiol. 2: 277-293.
Tsang, A. M., Klepeis, N. E. (1996). Descriptive Statistics Tables from a Detailed Analysis of
the National Human Activity Pattern Survey (NHAPS) Data. U.S. Environmental
Protection Agency. EPA/600/R-96/148.
Turk, B. H., Grimsrud, D. T., Brown, J. T., Geisling-Sobotka, K. L., Harrison, J., Prill, R. J.
(1989). Commercial Building Ventilation Rates and Particle Concentrations. ASHRAE
No. 3248.
5B-37
-------
US EPA. (2002). Consolidated Human Activities Database (CHAD) Users Guide. Database and
documentation available at: http://www.epa.gov/chadnet 1 /
US EPA. (2007). Ozone Population Exposure Analysis for Selected Urban Areas. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, NC. Available at: http://www.epa.gOv/ttn/naaqs/standards/ozone/s o3crtd.html
US. EPA. (2012a). Total Risk Integrated Methodology (TRIM) Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4). Volume I: User's Guide. EPA-452/B-
12-OOla. Available at:
http://www.epa.gov/ttn/fera/data/apex/APEX45 UsersGuide Voll_Aug2012.pdf.
US EPA. (2012b). Total Risk Integrated Methodology (TRIM) Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4) Volume II: Technical Support Document.
EPA-452/B-12-001b. Available at:
http://www.epa.gov/ttn/fera/data/apex/APEX45 UsersGuide Vol2 Aug2012.pdf.
University of Michigan. (2012). The Panel Study of Income Dynamics (PSID). Data and
documentation available at http://psidonline.isr.umich.edu/
Wiley, J. A., Robinson, J. P., Piazza, T., Garrett, K., Cirksena, K., Cheng, Y.-T., Martin, G.
(1991a). Activity Patterns of California Residents: Final Report. California Air Resources
Board, Sacramento, CA. ARB/R93/487. Available from: NTIS, Springfield, VA., PB94-
108719.
Wiley, J. A., Robinson, J. P., Cheng, Y.-T., Piazza, T., Stork, L., Pladsen, K. (1991b). Study of
Children's Activity Patterns: Final Report. California Air Resources Board, Sacramento,
CA. ARB-R-93/489.
Williams, R., Suggs, J., Creason, J., Rodes, C., Lawless, P., Kwok, R., Zweidinger, R., Sheldon,
L. (2000). The 1998 Baltimore particulate matter epidemiology-exposure study: Part 2.
Personal exposure associated with an elderly population. J Expo Anal Environ Epidemiol.
10(6): 533-543.
Williams, R., Suggs J., Rea, A., Leovic, K., Vette, A., Croghan, C., Sheldon, L., Rodes, C.,
Thornburg, J., Ejire, A., Herbst, M., Sanders Jr., W. (2003a). The Research Triangle
particulate panel study: PM mass concentrations relationships. Atmos Environ. 37:5349-
5363.
Williams, R., Suggs, J., Rea, A., Sheldon, L., Rodes, C., Thornburg, J. (2003b). The Research
Triangle particulate panel study: modeling ambient source contributions to personal and
residential PM mass concentrations. Atmos Environ. 37:5365-5378.
Williams, R., Rea, A., Vette, A., Croghan, C., Whitaker, D., Stevens, C., McDow, S., Fortmann,
R., Sheldon, L., Wilson, H., Thornburg, J., Phillips, M., Lawless, P., Rodes, C., Daughtrey,
H. (2008). The design and field implementation of the Detroit Exposure and Aerosol
Research Study. JExpos Sci Environ Epidem. 19:643-659.
Xue, J., McCurdy, T., Spengler, J., Ozkaynak, H. (2004). Understanding variability in time
spent in selected locations for 7-12-year old children. J Expo Anal Environ Epidemiol.
14(3): 222-33.
5B-38
-------
APPENDIX 5C
Generation of Adult and Child Census-tract Level Asthma
Prevalence using NHIS (2006-2010) and US Census (2000) Data
Table of Contents
5C-1. OVERVIEW 5C-1
5C-2. RAW ASTHMA PREVALENCE DATA SET DESCRIPTION 5C-2
5C-3. LOGISTIC MODELING APPROACH USED TO ESTIMATE ASTHMA
PREVALENCE 5C-4
5C-4. APPLICATION OF LOESS SMOOTHER TO ASTHMA PREVALENCE
ESTIMATION 5C-8
5C-5. CENSUS TRACT LEVEL POVERTY RATIO DATA SET DESCRIPTION AND
PROCESSING 5C-12
5C-6. COMBINED CENSUS TRACT LEVEL POVERTY RATIO AND ASTHMA
PREVALENCE DATA 5C-13
5C-7. REFERENCES 5C-14
5C-i
-------
List of Tables
Table 5C-1. Number of total surveyed persons from NHIS (2006-2010) sample adult and child
files and the number of those responding to asthma survey questions 5C-4
Table 5C-2. Example of alternative logistic models evaluated to estimate child asthma
prevalence using the "EVER" asthma response variable and goodness of fit test
results 5C-7
Table 5C-3. Top 20 model smoothing fits where residual standard error at or a value of 1.0.
5C-10
List of Figures
Figure 5C-1. Normal probability plot of studentized residuals generated using logistic model,
smoothing set to 0.7, and the children 'EVER' asthmatic data set 5C-11
Figure 5C-2. Studentized residuals versus model predicted betas generated using a logistic
model and using the children 'EVER' asthmatic data set, with smoothing set to
0.6 5C-12
5C-ii
-------
5C-1 OVERVIEW
This appendix describes the generation of our census tract level children and adult asthma
prevalence data developed from the 2006-2010 National Health Interview Survey (NHIS) and
census tract level poverty information from the 2000 US Census. The approach is, for the most
part, a reapplication of work performed by Cohen and Rosenbaum (2005), though here we
incorporated a few modifications as described below. Details regarding the earlier asthma
prevalence work are documented in Appendix G of US EPA (2007).
Briefly in the earlier asthma prevalence development work, Cohen and Rosenbaum
(2005) calculated asthma prevalence for children aged 0 to 17 years for each age, gender, and
four US regions using 2003 NHIS survey data. The four regions defined by NHIS were
'Midwest', 'Northeast', 'South', and 'West'. The asthma prevalence was defined as the
probability of a 'Yes' response to the question "EVER been told that [the child] had asthma?"1
among those persons that responded either 'Yes' or 'No' to this question.2 The responses were
weighted to take into account the complex survey design of the NHIS.3 Standard errors and
confidence intervals for the prevalence were calculated using a logistic model (PROC SURVEY
LOGISTIC; SAS, 2012). A scatter-plot technique (LOESS SMOOTHER; SAS, 2012) was
applied to smooth the prevalence curves and compute the standard errors and confidence
intervals for the smoothed prevalence estimates. Logistic analysis of the raw and smoothed
prevalence curves showed statistically significant differences in prevalence by gender and
region, supporting their use as stratification variables in the final data set. These smoothed
prevalence estimates were used as an input to EPA's Air Pollution Exposure Model (APEX) to
estimate air pollutant exposure in asthmatic children (US EPA, 2007; 2008; 2009).
For the current asthma prevalence data set development, several years of recent NHIS
survey data (2006-2010) were combined and used to calculate asthma prevalence. The current
approach estimates asthma prevalence for children (by age in years) as was done previously by
Cohen and Rosenbaum (2005) but now includes an estimate of adult asthma prevalence (by age
groups). In addition, two sets of asthma prevalence for each adults and children were estimated
here. The first data set, as was done previously, was based on responses to the question "EVER
been told that [the child] had asthma". The second data set was developed using the probability
of a 'Yes' response to a question that followed those that answered 'Yes' to the first question
1 The response was recorded as variable "CASHMEV" in the downloaded dataset. Data and documentation are
available at http://www.cdc.gov/nchs/nhis/quest data related 1997 forward.htm.
2 If there were another response to this variable other than "yes" or "no" (i.e., refused, not ascertained, don't know,
and missing), the surveyed individual was excluded from the analysis data set.
3 In the SURVEY LOGISTIC procedure, the variable "WTF_SC" was used for weighting, "PSU" was used for
clustering, and "STRATUM" was used to define the stratum.
5C-1
-------
regarding ever having asthma, specifically, do those persons "STILL have asthma?"4 And
finally, in addition to the nominal variables region and gender (and age and age groups), the
asthma prevalence in this new analysis were further stratified by a family income/poverty ratio
(i.e., whether the family income was considered below or at/above the US Census estimate of
poverty level for the given year).
These new asthma prevalence data sets were linked to the US census tract level poverty
ratios probabilities (US Census, 2007), also stratified by age and age groups. Given 1) the
significant differences in asthma prevalence by age, gender, region, and poverty status, 2) the
variability in the spatial distribution of poverty status across census tracts, stratified by age, and
3) the spatial variability in local scale ambient concentrations of many air pollutants, it is hoped
that the variability in population exposures is now better represented when accounting for and
modeling these newly refined attributes of this susceptible population.
5C-2 RAW ASTHMA PREVALENCE DATA SET DESCRIPTION
In this section we describe the asthma prevalence data sets used and identify the variables
retained for our final data set. First, raw data and associated documentation were downloaded
from the Center for Disease Control (CDC) and Prevention's National Health Interview Survey
(NHIS) website.5 The 'Sample Child' and 'Sample Adult' files were selected because of the
availability of person-level attributes of interest within these files, i.e., age in years ('age_p'),
gender ('sex'), US geographic region ('region'), coupled with the response to questions of
whether or not the surveyed individual ever had and still has asthma. In total, five years of
recent survey data were obtained, comprising over 50,000 children and 120,000 children for
years 2006-2010 (Table 5C-1).
Information regarding personal and family income and poverty ranking are also provided
by the NHIS in separate files. Five files ('INCIMPx.dat') are available for each survey year,
each containing either the actual responses (where recorded or provided by survey participant) or
imputed values for the desired financial variable.6 For this current analysis, the ratio of income
to poverty was used to develop a nominal variable: either the survey participant was below or
at/above a selected poverty threshold. This was done in this manner to be consistent with data
generated as part of a companion data set, i.e., census tract level poverty ratio probabilities
stratified by age (see section 5C-5 below).
4 While we estimated two separate sets of prevalence using the "STILL" and "EVER" variables, only the "STILL"
data were used as input to our exposure model.
5 See http://www.cdc.gov/nchs/nhis.htm (accessed October 4, 2011).
6 Financial information was not collected from all persons; therefore the NHIS provides imputed data. Details into
the available variables and imputation method are provided with each year's data set. For example see "Multiple
Imputation of Family Income and Personal Earnings in the National Health Interview Survey: Methods and
Examples" at http://www.cdc.gov/nchs/data/nhis/tecdoc 2010.pdf.
5C-2
-------
Given the changes in how income data were collected over the five year period of interest
and the presence of imputed data, a data processing methodology was needed to conform each of
the year's data sets to a compatible nominal variable. Briefly, for survey years 2006-2008,
poverty ratios ('RAT_CATF) are provided for each person as a categorical variable, ranging
from <0.5 to 5.0 by increments of either 0.25 (for poverty ratios categories between <0.5 - 2.0)
and 0.50 (for poverty ratios >5.0). For 2009 and 2010 data, the poverty ratio was provided as a
continuous variable ('POVRATI3') rather than a categorical variable.7
When considering the number of stratification variables, the level of asthma prevalence,
and poverty distribution among the survey population, sample size was an important issue. For
the adult data, there were insufficient numbers of persons available to stratify the data by single
ages (for some years of age there were no survey persons). Therefore, the adult survey data were
grouped as follows: ages 18-24, 25-34, 35-44, 45-54, 55-64, 65-74, and, >75.8 To increase the
number of persons within the age, gender, and four region groupings of our characterization of
'below poverty' asthmatics persons, the poverty ratio threshold was selected as <1.5, therefore
including persons that were within 50% above the poverty threshold. As there were five data
sets containing variable imputed poverty ratios (as well as a non-varying values for where
income information was reported) for each year, the method for determining whether a person
was below or above the poverty threshold was as follows. If three or more of the five
imputed/recorded values were <1.5, the person's family income was categorized 'below' the
poverty threshold, if three or more of the 5 values were >1.5, the person's family income was
categorized 'above' the poverty threshold. The person-level income files were then merged with
the sample adult and child files using the 'FFHX' (a household identifier), 'FMX' (a family
identifier), and 'FPX' (an individual identifier) variables. Note, all persons within the sample
adult and child files had corresponding financial survey data.
Two asthma survey response variables were of interest in this analysis and were used to
develop the two separate prevalence data sets for each children and adults. The response to the
first question "Have you EVER been told by a doctor or other health professional that you [or
your child] had asthma?" was recorded as variable name 'CASFFMEV for children and
'AASMEV for adults. Only persons having responses of either 'Yes' or 'No' to this question
were retained to estimate the asthma prevalence. This assumes that the exclusion of those
7 Actually, the 2009 data had continuous values for the poverty ratios ('POVRATI2') but the quality was determined
by us to be questionable: the value varied among family members by orders of magnitude - however, it should be
a constant. The income data ('FAMTNCI2') provided were constant among family members, therefore we
combined these data with poverty thresholds obtained from the US Census (available at:
http://www.census.gov/hhes/www/povertv/data/threshld/thresh08.html') for year 2008 by family size (note, income
is the annual salary from the prior year) and calculated an appropriate poverty ratio for each family member.
8 These same age groupings were used to create the companion file containing the census tract level poverty ratio
probabilities (section 5C-5).
5C-3
-------
responding otherwise, i.e., those that 'refused' to answer, instances where it was "not
ascertained', or the person 'does not know', does not affect the estimated prevalence rate if either
'Yes' or 'No' answers could actually be given by these persons. There were very few persons
(<0.3%) that did provide an unusable response (Table 5C-1), thus the above assumption is
reasonable. A second question was asked as a follow to persons responding "Yes" to the first
question, specifically, "Do you STILL have asthma?" and noted as variables 'CASSTILL' and
'AASSTILL' for children and adults, respectively. Again, while only persons responding 'Yes'
and 'No' were retained for further analysis, the representativeness of the screened data set is
assumed unchanged from the raw survey data given the few persons having unusable data
Table 5C-1. Number of total surveyed persons from NHIS (2006-2010) sample adult and
child files and the number of those responding to asthma survey questions.
Study Group/Respondents
Children
All Persons
Yes/No Asthma
Yes/No to Still Have + No Asthma
Adults
All Persons
Yes/No Asthma
Yes/No to Still Have + No Asthma
Number of Surveyed Persons
2010
11,277
1 1 ,256
11,253
2010
27,157
27,157
27,113
2009
11,156
11,142
11,129
2009
27,731
27,715
27,686
2008
8,815
8,800
8,793
2008
21,781
21,766
21,726
2007
9,417
9,404
9,394
2007
23,393
23,372
23,349
2006
9,837
9,815
9,797
2006
24,275
24,242
24,208
TOTAL
50,502
50,417
50,366
TOTAL
124,337
124,252
124,082
5C-3 LOGISTIC MODELING APPROACH USED TO ESTIMATE ASTHMA
PREVALENCE
As described in the previous section, four person-level analytical data sets were created
from the raw NHIS data files, generally containing similar variables: a 'Yes' or 'No' asthma
response variable (either 'EVER' or 'STILL'), an age (or age group for adults), their gender
('male' or 'female'), US geographic region ('Midwest', 'Northeast', 'South', and 'West'), and
poverty status ('below' or above'). One approach to calculate prevalence rates and their
uncertainties for a given gender, region, poverty status, and age is to calculate the proportion of
'Yes' responses among the 'Yes' and 'No' responses for that demographic group, appropriately
weighting each response by the survey weight. This simplified approach was initially used to
develop 'raw' asthma prevalence rates however this approach may not be completely
appropriate. The two main issues with such a simplified approach are that the distributions of
5C-4
-------
the estimated prevalence rates would not be well approximated by normal distributions and that
the estimated confidence intervals based on a normal approximation would often extend outside
the [0, 1] interval. A better approach for such survey data is to use a logistic transformation and
fit the model:
Prob(asthma) = exp(beta) / (1 + exp(beta) ),
where, beta may depend on the explanatory variables for age, gender, poverty status, or region.
This is equivalent to the model:
Beta = logit {prob(asthma) } = log {prob(asthma) / [1 -prob(asthma)] }
The distribution of the estimated values of beta is more closely approximated by a normal
distribution than the distribution of the corresponding estimates of prob (asthma). By applying a
logit transformation to the confidence intervals for beta, the corresponding confidence intervals
for prob(asthma) will always be inside [0, 1]. Another advantage of the logistic modeling is that
it can be used to compare alternative statistical models, such as models where the prevalence
probability depends upon age, region, poverty status, and gender, or on age, region, poverty
status but not gender.
A variety of logistic models were fit and compared to use in estimating asthma
prevalence, where the transformed probability variable beta is a given function of age, gender,
poverty status, and region. I used the SAS procedure SURVEYLOGISTIC to fit the various
logistic models, taking into account the NHIS survey weights and survey design (using both
stratification and clustering options), as well as considering various combinations of the selected
explanatory variables.
As an example, Table 5C-2 lists the models fit and their log-likelihood goodness-of-fit
measures using the sample child data and for the "EVER" asthma response variable. A total of
32 models were fit, depending on the inclusion of selected explanatory variables and how age
was considered in the model. The 'Strata' column lists the eight possible stratifications: no
stratification, stratified by gender, by region, by poverty status, by region and gender, by region
and poverty status, by gender and poverty status, and by region, gender and poverty status. For
example, "5. region, gender" indicates that separate prevalence estimates were made for each
combination of region and gender. As another example, "2. gender" means that separate
prevalence estimates were made for each gender, so that for each gender, the prevalence is
assumed to be the same for each region. Note the prevalence estimates are independently
5C-5
-------
calculated for each stratum. The 'Description' column of Table 5C-2 indicates how beta
depends upon the age:
Linear in age Beta = a + |3 x age, where a and |3 vary with strata.
Quadratic in age Beta = a + |3 x age + y x age2, where a |3 and y vary with strata.
Cubic in age Beta = a + |3 x age + y x age2 + 5 x age3, where a, |3, y, and 5 vary
with the strata.
f(age) Beta = arbitrary function of age, with different functions for
different strata
The category f(age) is equivalent to making age one of the stratification variables, and is
also equivalent to making beta a polynomial of degree 16 in age (since the maximum age for
children is 17), with coefficients that may vary with the strata.
The fitted models are listed in order of complexity, where the simplest model (i.e., model
1) is an unstratified linear model in age and the most complex model (model 32) has a
prevalence that is an arbitrary function of age, gender, poverty status, and region. Model 32 is
equivalent to calculating independent prevalence estimates for each of the 288 combinations of
age, gender, poverty status, and region.
Table 5C-2 also includes the -2 Log Likelihood statistic, a goodness-of-fit measure, and
the associated degrees of freedom (DF), which is the total number of estimated parameters. Any
two models can be compared using their -2 Log Likelihood values: models having lower values
are preferred. If the first model is a special case of the second model, then the approximate
statistical significance of the first model is estimated by comparing the difference in the -2 Log
Likelihood values with a chi-squared random variable having r degrees of freedom, where r is
the difference in the DF (hence a likelihood ratio test). For all pairs of models from Table 5C-2,
all the differences in the -2 Log Likelihood statistic are at least 600,000 and thus significant at p-
values well below 1 percent. Based on its having the lowest -2 Log Likelihood value, the last
model fit (model 32: retaining all explanatory variables and usingf(age)) was preferred and used
to estimate the asthma prevalence.9
9 Similar results were obtained when estimating prevalence using the 'STILL' have asthma variable as well as when
investigating model fit using the adult data sets. Note that because age was a categorical variable in the adult data
sets it could only be evaluated using}"(age_group). See Attachment B, Tables 5CB-1 to 5CB-4 for all model fit
results.
5C-6
-------
Table 5C-2. Example of alternative logistic models evaluated to estimate child asthma
prevalence using the "EVER" asthma response variable and goodness of fit test results.
Model No.
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
Description
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
Strata
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
- 2 Log Likelihood
288740115.1
287062346.4
288120804.1
287385013.1
286367652.6
286283543.6
285696164.7
284477928.1
286862135.1
285098650.6
286207721.5
285352164
284330346.1
284182547.5
283587631 .7
282241318.6
286227019.6
284470413
285546716.1
284688169.9
283662673.5
283404487.5
282890785.3
281407414.3
285821686.2
283843266.2
284761522.8
284045849.2
282099156.1
281929968.5
281963915.7
278655423.1
DF1
2
4
8
4
16
16
8
32
3
6
12
6
24
24
12
48
4
8
16
8
32
32
16
64
18
36
72
36
144
144
72
288
model degrees of freedom.
5C-7
-------
The SURVEYLOGISTIC procedure produces estimates of the beta values and their 95%
confidence intervals for each combination of age, region, poverty status, and gender. By
applying the inverse logit transformation,
Prob(asthma) = exp(beta) / (1 + exp(beta) ),
one can convert the beta values and associated 95% confidence intervals into predictions
and 95% confidence intervals for the prevalence. The standard error for the prevalence was
estimated as
StdError {Prob(asthma)} = StdError (beta) x exp(- beta) / (1 + exp(beta) )2
which follows from the delta method (i.e., a first order Taylor series approximation).
Estimated asthma prevalence using this approach and termed here as 'unsmoothed' are provided
in Attachment A. Asthma prevalence for children is provided in Attachment A, Tables 5CA-1
('EVER' had Asthma) and 5CA-2 ('STILL' have asthma) while adult asthma prevalence is
provided in Attachment A, Tables 5CA-3 ('EVER' had Asthma) and 5CA-4 ('STILL' have
asthma). Graphical representation of each study group is also provided in a series of plots within
Attachment A, Figures 5CA-1 to 5CA-4. The variables provided in the tabular presentation are:
• Region
• Gender
• Age (in years) or Age_group (age categories)
• Poverty Status
• Prevalence = predicted prevalence
• SE = standard error of predicted prevalence
• LowerCI = lower bound of 95 % confidence interval for predicted prevalence
• UpperCI = upper bound of 95 % confidence interval for predicted prevalence
5C-4 APPLICATION OF LOESS SMOOTHER TO ASTHMA PREVALENCE
ESTIMATION
The estimated prevalence curves shows that the prevalence is not necessarily a smooth
function of age. The linear, quadratic, and cubic functions of age modeled by
SURVEYLOGISTIC were identified as a potential method for smoothing the curves, but they
did not provide the best fit to the data. One reason for this might be due to the attempt to fit a
global regression curve to all the age groups, which means that the predictions for age A are
affected by data for very different ages. A local regression approach that separately fits a
5C-8
-------
regression curve to each age A and its neighboring ages was used, giving a regression weight of
1 to the age^4, and lower weights to the neighboring ages using a tri-weight function:
Weight = {1 - [ \age -A /q] 3}, where age -A <= q
The parameter q defines the number of points in the neighborhood of the age A. Instead
of calling q the smoothing parameter, S AS defines the smoothing parameter as the proportion of
points in each neighborhood. A quadratic function of age to each age neighborhood was fit
separately for each gender and region combination. These local regression curves were fit to the
beta values, the logits of the asthma prevalence estimates, and then converted them back to
estimated prevalence rates by applying the inverse logit function exp(beta) / (1 + exp(beta)). In
addition to the tri-weight variable, each beta value was assigned a weight of
1 / [std error (beta)]2, to account for their uncertainties.
In this application of LOESS, weights of 1 / [std error (beta)]2 were used such that a2 =
1. The LOESS procedure estimates a2 from the weighted sum of squares. Because it is assumed
a2 = 1, the estimated standard errors are multiplied by 1 / estimated a and adjusted the widths of
the confidence intervals by the same factor.
One data issue was an overly influential point that needed to be adjusted to avoid
imposing wild variation in the "smoothed" curves: for the West region, males, age 0, above
poverty threshold, there were 249 children surveyed that all gave 'No' answers to the asthma
question, leading to an estimated value of-14.203 for beta with a standard error of 0.09. In this
case the raw probability of asthma equals zero, so the corresponding estimated beta would be
negative infinity, but SAS's software gives -14.203 instead. To reduce the excessive impact of
this single data point, we replaced the estimated standard error by 4, which is approximately four
times the maximum standard error for all other region, gender, poverty status, and age
combinations.
There are several potential values that can be selected for the smoothing parameter; the
optimum value was determined by evaluating three regression diagnostics: the residual standard
error, normal probability plots, and studentized residuals. To generate these statistics, the
LOESS procedure was applied to estimated smoothed curves for beta, the logit of the prevalence,
as a function of age, separately for each region, gender, and poverty classification. For the
children data sets, curves were fit using the choices of 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 for the
smoothing parameter. This selected range of values was bounded using the following
observations. With only 18 points (i.e., the number of ages), a smoothing parameter of 0.2
cannot be used because the weight function assigns zero weights to all ages except age^4, and a
quadratic model cannot be uniquely fit to a single value. A smoothing parameter of 0.3 also
5C-9
-------
cannot be used because that choice assigns a neighborhood of 5 points only (0.3 x 18 = 5,
rounded down), of which the two outside ages have assigned weight zero, making the local
quadratic model fit exactly at every point except for the end points (ages 0, 1, 16 and 17).
Usually one uses a smoothing parameter below 1 so that not all the data are used for the local
regression at a given x value. Note also that a smoothing parameter of 0 can be used to generate
the unsmoothed prevalence. The selection of the smoothing parameter used for the adult curves
would follow a similar logic, although the lower bound could effectively be extended only to 0.9
given the number of age groups. This limits the selection of smoothing parameter applied to the
two adult data sets to a value of 0.9, though values of 0.8 to 1.0 were nevertheless compared for
good measure.
The first regression diagnostic used was the residual standard error, which is the LOESS
estimate of a. As discussed above, the true value of a equals 1, so the best choice of smoothing
parameter should have residual standard errors as close to 1 as possible. Attachment B, Tables
5CB-5 to 5CB-8 contain the residual standard errors output from the LOESS procedure,
considering region, gender, poverty status and each data set examined. For children 'EVER'
having asthma and when considering the best 20 models (of the 112 possible) using this criterion
(note also within 0.06 RSE units of 1), the best choice varies with gender, region, and poverty
status between smoothing parameters of 0.6, 0.7, and 0.8 (Table 5C-3). Similar results were
observed for the 'STILL' data set, though a value of 0.6 would be slightly preferred. Either adult
data set could be smoothed using a value of 0.8 or 0.9 given the limited selection of smoothing
values, though 0.9 appears a better value for the 'STILL' data set.
Table 5C-3. Top 20 model smoothing fits where residual standard error at or a value of
1.0.
Data
Set
Children
Adults
Asthma
EVER
STILL
EVER
STILL
LOESS Smoothing Parameter
0.4
2
2
n/a
n/a
0.5
2
3
n/a
n/a
0.6
5
4
n/a
n/a
0.7
5
2
n/a
n/a
0.8
4
3
6
5
0.9
1
3
6
7
1.0
1
3
8
8
The second regression diagnostic was developed from an approximate studentized
residual. The residual errors from the LOESS model were divided by standard error (beta) to
make their variances approximately constant. These approximately studentized residuals should
be approximately normally distributed with a mean of zero and a variance of a2 = 1. To test this
assumption, normal probability plots of the residuals were created for each smoothing parameter,
5C-10
-------
combining all the studentized residuals across genders, regions, poverty status, and ages. These
normal probability plots are provided in Attachment B, Figures CB-1 to CB-4. The results for
the children data indicate little distinction or affect by the selection of a particular smoothing
parameter (e.g., see Figure 5C-1 below), although linearity in the plotted curve is best expressed
with smoothing parameters at or above values of 0.6. When considering the adult data sets,
again the appropriate value would be 0.9, as Attachment B, Figures 5CB-3 and 5CB-4 supports
this conclusion.
i -
o-
-i
-3
Ql
1O
I
25
3D
75
9O 95
\
99
99.9
Figure 5C-1. Normal probability plot of studentized residuals generated using logistic
model, smoothing set to 0.7, and the children 'EVER' asthmatic data set.
The third regression diagnostic, presented in Attachment B, Figures 5CB-5 to 5CB-8 are
plots of the studentized residuals against the smoothed beta values. All the studentized residuals
for a given smoothing parameter are plotted together within the same graph. Also plotted is a
LOESS smoothed curve fit to the same set of points, with SAS's optimal smoothing parameter
choice, to indicate the typical pattern. Ideally there should be no obvious pattern and an average
studentized residual close to zero with no regression slope (e.g., see Figure 5C-2). For the
children data sets, these plots generally indicate no unusual patterns, and the results for
smoothing parameters 0.4 through 0.6 indicate a fit LOESS curve closest to the studentized
residual equals zero line. When considering the adult data sets, again the appropriate value
would be 0.9, as Attachment B, Figures 5CB-7 and 5CB-8 supports this conclusion.
5C-11
-------
stu±rt
5-
4-
3-
2-
1-
o-
-1-
-2-
-3-
-4-
-5-
X
O
o
-500000
-300000
-200000
-100000
O
o o o
OOO
XXX
XXX
Figure 5C-2. Studentized residuals versus model predicted betas generated using a logistic
model and using the children 'EVER' asthmatic data set, with smoothing set to 0.6.
When considering both children asthma prevalence responses evaluated, the residual
standard error (estimated values for sigma) suggests the choice of smoothing parameter as 0.6 to
0.8. The normal probability plots of the Studentized residuals suggest preference for smoothing
at or above 0.6. The plots of residuals against smoothed predictions suggest the choices of 0.4
through 0.6. We therefore chose the final value of 0.6 to use for smoothing the children's asthma
prevalence. For the adults, 0.9 was selected for smoothing.
Smoothed asthma prevalence and associated graphical presentation are provided in
Attachment C, following a similar format as the unsmoothed data provided in Attachment A.
5C-5 CENSUS TRACT LEVEL POVERTY RATIO DATA SET DESCRIPTION AND
PROCESSING
This section describes the approach used to generate census tract level poverty ratios for
all US census tracts, stratified by age and age groups where available. The data set generation
involved primarily two types of data downloaded from the 2000 US Census, each are described
below.
First, individual state level SF3 geographic data ("geo") .uf3 files and associated
documentation were downloaded10 and, following import by SAS (SAS, 2012), were screened
10 Geographic data were obtained from http://www2.census.gov/census 2000/datasets/Summary File 3/.
Information regarding variable names is given in Figure 2-5 of US Census (2007).
5C-12
-------
for tract level information using the "sumlev" variable equal to ' 140'. For quality control
purposes and ease of matching with the poverty level data, our geo data set retained the
following variables: stusab, sumlev, logrecno, state, county, tract, name, latitude, and longitude.
Second, the individual state level SF3 files ("30") were downloaded, retaining the
number of persons across the variable "PCT50" for all state "logrecno".11 The data provided by
the PCT50 variable is stratified by age or age groups (ages <5, 5, 6-11, 12-14, 15, 16-17, 18-24,
25-34, 35-44, 45-54, 55-64, 65-74, and >75) and income/poverty ratios, given in increments of
0.25. We calculated two new variables for each state logrecno using the number of persons from
the PCT50 stratifications; the fraction of those persons having poverty ratios < 1.5 and > 1.5 by
summing the appropriate PCT50 variable and dividing by the total number of persons in that
age/age group. Finally the poverty ratio data were combined with the above described census
tract level geographic data using the "stusab" and "logrecno" variables. The final output was a
single file containing relevant tract level poverty probabilities by age groups for all US census
tracts (where available).
5C-6 COMBINED CENSUS TRACT LEVEL POVERTY RATIO AND ASTHMA
PREVALENCE DATA
Because the prevalence data are stratified by standard US Census defined regions,12 we
first mapped the tract level poverty level data to an appropriate region based on the State.
Further, as APEX requires the input data files to be complete, additional processing of the
poverty probability file was needed. For where there was missing tract level poverty
information,13 we substituted an age-specific value using the average for the particular county the
tract was located within. The frequency of missing data substitution comprised 1.7% of the total
poverty probability data set. The two data sets were merged and the final asthma prevalence was
calculated using the following weighting scheme:
Prevalence = round((pov_prob*prev_poor)+((l~pov_prob)*prev_notpoor),9JMM)l);
11 Poverty ratio data were obtained from http://www2.census.gov/census 2000/datasets/Summary File 3/.
Information regarding poverty ratio names variable names is given in chapter 6 of US Census Bureau (2007). We
used the variable "PCT50", an income to poverty ratio variable stratified by various ages and age groups and
described in chapter 7 of US Census Bureau (2007).
12 For example, see http://www.cdc.gov/std/stats 10/census.htm.
13 Whether there were no data collected by the Census or whether there were simply no persons in that age group is
relatively inconsequential to estimating the asthmatic persons exposed, particularly considering latter case as no
persons in that age group would be modeled.
5C-13
-------
whereas each US census tract value now expresses a tract specific poverty-weighted
prevalence, stratified by ages (children 0-17), age groups (adults), and two genders. These final
prevalence data are found within the APEX asthmaprevalence.txt file.
5C-7 REFERENCES
Cohen, J., and Rosenbaum, A. (2005). Analysis of NHIS Asthma Prevalence Data.
Memorandum to John Langstaff by ICF Incorporated. For US EPA Work Assignment 3-
08 under EPA contract 68D01052.
SAS. (2012). SAS/STAT 9.2 User's Guide, Second Edition. Available at:
http://support.sas.com/documentation/cdl/en/statug/63033/PDF/default/statug.pdf
US Census Bureau. (2007). 2000 Census of Population and Housing. Summary File 3 (SF3)
Technical Documentation, available at: http://www.census.gov/prod/cen2000/doc/sf3.pdf
Individual SF3 files '30' (for income/poverty variables pctSO) for each state were
downloaded from: http://www2.census.gov/census 2000/datasets/Summary File 3/.
US EPA. (2007). Ozone Population Exposure Analysis for Selected Urban Areas (July 2007).
Office of Air Quality Planning and Standards, Research Triangle Park, NC. EPA-452/R-
07-010. Available at: http://epa.gOv/ttn/naaqs/standards/ozone/s o3crtd.html.
US EPA. (2008). Risk and Exposure Assessment to Support the Review of the NO2 Primary
National Ambient Air Quality Standard. Report no. EPA-452/R-08-008a. November
2008. Available at:
http://www.epa.gov/ttn/naaqs/standards/nox/data/20081121 NO2REA final.pdf
US EPA. (2009). Risk and Exposure Assessment to Support the Review of the SO2 Primary
National Ambient Air Quality Standard. Report no. EPA-452/R-09-007. August 2009.
Available at:
http://www.epa.gov/ttn/naaqs/standards/so2/data/200908SO2REAFinalReport.pdf
5C-14
-------
Appendix 5C, Attachment A
Unsmoothed Asthma Prevalence Tables and Figures
Appendix 5C, Attachment A, Table CA-1. Unsmoothed prevalence for children "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
Prevalence
0.0018
0.0387
0.0367
0.0395
0.0815
0.0885
0.0438
0.1374
0.0820
0.1027
0.0995
0.1129
0.1752
0.1331
0.1944
0.1383
0.1731
0.1311
0.0564
0.0585
0.1256
0.1127
0.1746
0.1584
0.1229
0.0867
0.1523
0.2070
0.2293
0.1359
0.1501
0.1527
0.1197
0.2103
0.2054
0.1844
0.0061
0.0258
0.0848
0.0996
0.0876
0.1593
0.0977
0.1793
0.1503
0.1418
0.1569
0.1717
0.2054
0.1846
0.1671
0.1454
0.1557
0.1320
0.0293
0.1051
0.1786
SE
0.0018
0.0233
0.0148
0.0186
0.0298
0.0207
0.0200
0.0277
0.0246
0.0220
0.0193
0.0277
0.0391
0.0256
0.0477
0.0302
0.0341
0.0256
0.0353
0.0197
0.0487
0.0419
0.0395
0.0447
0.0417
0.0353
0.0392
0.0486
0.1109
0.0470
0.0484
0.0380
0.0462
0.0760
0.0597
0.1134
0.0044
0.0178
0.0231
0.0261
0.0223
0.0313
0.0229
0.0313
0.0356
0.0265
0.0322
0.0371
0.0338
0.0358
0.0291
0.0356
0.0278
0.0233
0.0176
0.0376
0.0652
LowerCI
0.0002
0.0117
0.0165
0.0155
0.0390
0.0556
0.0176
0.0916
0.0450
0.0669
0.0675
0.0688
0.1112
0.0905
0.1173
0.0890
0.1160
0.0885
0.0160
0.0299
0.0567
0.0529
0.1100
0.0888
0.0616
0.0381
0.0902
0.1275
0.0800
0.0670
0.0774
0.0921
0.0544
0.0980
0.1121
0.0491
0.0015
0.0066
0.0491
0.0588
0.0527
0.1069
0.0611
0.1259
0.0930
0.0973
0.1035
0.1106
0.1470
0.1244
0.1175
0.0885
0.1087
0.0926
0.0089
0.0509
0.0835
UpperCI
0.0129
0.1208
0.0797
0.0972
0.1624
0.1382
0.1046
0.2010
0.1450
0.1545
0.1442
0.1797
0.2652
0.1916
0.3049
0.2086
0.2502
0.1898
0.1799
0.1112
0.2552
0.2240
0.2658
0.2664
0.2301
0.1851
0.2456
0.3182
0.5043
0.2562
0.2710
0.2427
0.2431
0.3949
0.3462
0.4976
0.0247
0.0957
0.1426
0.1636
0.1423
0.2306
0.1527
0.2489
0.2340
0.2021
0.2306
0.2568
0.2795
0.2650
0.2322
0.2297
0.2182
0.1848
0.0922
0.2047
0.3418
5C-15
-------
Appendix 5C, Attachment A, Table CA-1. Unsmoothed prevalence for children "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Age
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
Prevalence
0.2066
0.2770
0.2504
0.2186
0.2192
0.2902
0.1242
0.2897
0.2669
0.2589
0.2429
0.1470
0.1965
0.1855
0.3740
0.0055
0.0296
0.0697
0.0723
0.1142
0.1058
0.0933
0.1084
0.0780
0.1362
0.0979
0.1697
0.0535
0.0910
0.1500
0.1733
0.1884
0.1694
0.0315
0.1230
0.0703
0.1860
0.1666
0.2347
0.0682
0.0972
0.2049
0.1695
0.0988
0.2622
0.1377
0.3506
0.1869
0.1965
0.1986
0.1625
0.0256
0.0542
0.0635
0.0835
0.1378
0.1444
0.2175
0.2019
0.1878
0.1286
0.1879
0.2532
0.1801
SE
0.0513
0.0638
0.0499
0.0447
0.0456
0.0649
0.0437
0.0639
0.0613
0.1050
0.0693
0.0490
0.0509
0.0611
0.1042
0.0054
0.0164
0.0252
0.0250
0.0254
0.0296
0.0254
0.0251
0.0221
0.0374
0.0298
0.0382
0.0229
0.0273
0.0207
0.0355
0.0510
0.0395
0.0251
0.0576
0.0277
0.0555
0.0598
0.0636
0.0250
0.0362
0.0604
0.0698
0.0440
0.0734
0.0525
0.0762
0.0537
0.0534
0.0470
0.0602
0.0130
0.0231
0.0220
0.0232
0.0329
0.0357
0.0482
0.0343
0.0373
0.0342
0.0278
0.0420
0.0233
LowerCI
0.1236
0.1703
0.1656
0.1436
0.1428
0.1806
0.0607
0.1815
0.1646
0.1068
0.1329
0.0742
0.1150
0.0935
0.1998
0.0008
0.0099
0.0337
0.0362
0.0731
0.0602
0.0541
0.0681
0.0442
0.0780
0.0530
0.1073
0.0228
0.0499
0.1138
0.1142
0.1077
0.1052
0.0064
0.0469
0.0319
0.1002
0.0791
0.1329
0.0327
0.0458
0.1107
0.0717
0.0400
0.1445
0.0629
0.2188
0.1031
0.1120
0.1221
0.0754
0.0094
0.0231
0.0318
0.0478
0.0849
0.0875
0.1376
0.1429
0.1252
0.0751
0.1394
0.1799
0.1388
UpperCI
0.3247
0.4170
0.3600
0.3184
0.3211
0.4312
0.2374
0.4285
0.4021
0.5051
0.4017
0.2703
0.3151
0.3345
0.5884
0.0368
0.0854
0.1384
0.1394
0.1741
0.1793
0.1563
0.1682
0.1339
0.2272
0.1738
0.2578
0.1204
0.1604
0.1953
0.2541
0.3085
0.2613
0.1404
0.2852
0.1479
0.3193
0.3175
0.3802
0.1366
0.1944
0.3478
0.3505
0.2240
0.4277
0.2752
0.5100
0.3148
0.3217
0.3065
0.3158
0.0679
0.1218
0.1228
0.1418
0.2158
0.2291
0.3263
0.2774
0.2719
0.2115
0.2485
0.3439
0.2303
5C-16
-------
Appendix 5C, Attachment A, Table CA-1. Unsmoothed prevalence for children "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Poverty Status
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Age
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
Prevalence
0.1581
0.2043
0.1752
0.1798
0.1836
0.0375
0.1649
0.2200
0.1124
0.2651
0.2398
0.3209
0.2651
0.2905
0.3810
0.3382
0.2485
0.2819
0.2961
0.2876
0.2632
0.2407
0.3123
0.0129
0.0191
0.0558
0.0793
0.0834
0.0932
0.1446
0.1439
0.1111
0.1258
0.0626
0.1288
0.1064
0.1387
0.1621
0.1399
0.1362
0.1299
0.0495
0.0734
0.0828
0.0973
0.1578
0.1409
0.1536
0.1658
0.1428
0.2123
0.1408
0.2249
0.1741
0.1463
0.2428
0.1947
0.1285
0.1322
0.0135
0.0782
0.1134
0.1063
0.1679
SE
0.0340
0.0447
0.0287
0.0360
0.0282
0.0275
0.0506
0.0503
0.0445
0.0909
0.0651
0.0432
0.0572
0.0969
0.0773
0.1019
0.0708
0.0705
0.0685
0.0713
0.0661
0.0559
0.0734
0.0080
0.0084
0.0147
0.0200
0.0184
0.0222
0.0226
0.0248
0.0194
0.0222
0.0154
0.0210
0.0182
0.0222
0.0243
0.0169
0.0253
0.0197
0.0216
0.0210
0.0207
0.0271
0.0372
0.0300
0.0381
0.0332
0.0302
0.0413
0.0347
0.0466
0.0519
0.0296
0.0437
0.0399
0.0344
0.0323
0.0065
0.0162
0.0190
0.0211
0.0303
LowerCI
0.1022
0.1303
0.1257
0.1195
0.1346
0.0087
0.0877
0.1371
0.0501
0.1262
0.1355
0.2427
0.1686
0.1401
0.2446
0.1732
0.1359
0.1656
0.1808
0.1695
0.1548
0.1483
0.1885
0.0038
0.0080
0.0330
0.0479
0.0537
0.0579
0.1057
0.1017
0.0784
0.0883
0.0383
0.0928
0.0756
0.1006
0.1198
0.1100
0.0938
0.0959
0.0207
0.0415
0.0503
0.0556
0.0976
0.0917
0.0927
0.1104
0.0931
0.1425
0.0855
0.1467
0.0941
0.0972
0.1675
0.1280
0.0747
0.0807
0.0052
0.0517
0.0811
0.0714
0.1165
UpperCI
0.2366
0.3056
0.2387
0.2614
0.2454
0.1477
0.2887
0.3337
0.2330
0.4738
0.3885
0.4107
0.3908
0.5070
0.5392
0.5551
0.4102
0.4371
0.4448
0.4440
0.4107
0.3660
0.4701
0.0427
0.0447
0.0928
0.1286
0.1273
0.1467
0.1948
0.1996
0.1550
0.1762
0.1005
0.1759
0.1478
0.1881
0.2156
0.1763
0.1938
0.1737
0.1137
0.1268
0.1336
0.1649
0.2450
0.2103
0.2439
0.2414
0.2126
0.3042
0.2233
0.3288
0.2997
0.2142
0.3382
0.2847
0.2122
0.2092
0.0342
0.1165
0.1563
0.1554
0.2360
5C-17
-------
Appendix 5C, Attachment A, Table CA-1. Unsmoothed prevalence for children "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Poverty Status
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Age
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Prevalence
0.1644
0.1328
0.1542
0.1502
0.1522
0.1485
0.1767
0.1915
0.1939
0.1381
0.1579
0.1698
0.1530
0.0610
0.1005
0.1102
0.1699
0.1642
0.2510
0.2064
0.1588
0.2518
0.2246
0.2022
0.1890
0.2322
0.2345
0.2265
0.1801
0.1286
0.1916
0.0049
0.0390
0.0269
0.0439
0.0232
0.0988
0.0829
0.1065
0.0960
0.1124
0.0978
0.1186
0.1655
0.0855
0.1258
0.1482
0.1394
0.2285
0.0064
0.0443
0.0523
0.0403
0.0346
0.0887
0.1351
0.1364
0.1106
0.1254
0.0585
0.0747
0.0720
0.1898
0.1431
SE
0.0226
0.0212
0.0270
0.0224
0.0232
0.0240
0.0255
0.0236
0.0255
0.0196
0.0246
0.0193
0.0240
0.0181
0.0206
0.0225
0.0324
0.0288
0.0485
0.0339
0.0309
0.0503
0.0381
0.0368
0.0344
0.0383
0.0454
0.0489
0.0371
0.0303
0.0297
0.0037
0.0202
0.0097
0.0153
0.0079
0.0294
0.0223
0.0281
0.0280
0.0296
0.0285
0.0188
0.0352
0.0196
0.0278
0.0213
0.0254
0.0375
0.0064
0.0195
0.0220
0.0140
0.0177
0.0372
0.0432
0.0360
0.0244
0.0405
0.0204
0.0264
0.0279
0.0591
0.0431
LowerCI
0.1247
0.0964
0.1083
0.1114
0.1121
0.1073
0.1322
0.1495
0.1487
0.1039
0.1154
0.1352
0.1117
0.0338
0.0667
0.0732
0.1154
0.1152
0.1682
0.1477
0.1072
0.1663
0.1588
0.1394
0.1305
0.1656
0.1573
0.1448
0.1183
0.0799
0.1399
0.0011
0.0139
0.0132
0.0219
0.0118
0.0544
0.0484
0.0627
0.0534
0.0662
0.0545
0.0864
0.1074
0.0542
0.0806
0.1111
0.0967
0.1632
0.0009
0.0185
0.0226
0.0202
0.0126
0.0380
0.0703
0.0798
0.0711
0.0650
0.0292
0.0368
0.0331
0.0993
0.0773
UpperCI
0.2136
0.1802
0.2148
0.1994
0.2033
0.2018
0.2323
0.2419
0.2487
0.1813
0.2122
0.2110
0.2061
0.1076
0.1488
0.1626
0.2431
0.2285
0.3572
0.2808
0.2290
0.3622
0.3078
0.2839
0.2658
0.3153
0.3345
0.3361
0.2645
0.2005
0.2566
0.0216
0.1048
0.0541
0.0858
0.0450
0.1730
0.1384
0.1752
0.1666
0.1846
0.1695
0.1606
0.2463
0.1324
0.1911
0.1949
0.1969
0.3101
0.0441
0.1025
0.1166
0.0788
0.0919
0.1934
0.2439
0.2234
0.1682
0.2283
0.1137
0.1460
0.1496
0.3323
0.2495
5C-18
-------
Appendix 5C, Attachment A, Table CA-1. Unsmoothed prevalence for children "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Age
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Prevalence
0.1168
0.0814
0.0637
0.0000
0.0244
0.0517
0.0601
0.1698
0.1236
0.1376
0.1288
0.1018
0.1884
0.1604
0.2121
0.1833
0.2105
0.1475
0.1641
0.1958
0.2113
0.0135
0.0812
0.0417
0.1182
0.1349
0.1562
0.1853
0.1484
0.1549
0.1275
0.1742
0.1909
0.1678
0.1793
0.1919
0.1410
0.1863
0.2030
SE
0.0304
0.0290
0.0235
0.0000
0.0121
0.0155
0.0172
0.0275
0.0288
0.0264
0.0354
0.0223
0.0315
0.0273
0.0298
0.0349
0.0397
0.0309
0.0263
0.0282
0.0289
0.0128
0.0317
0.0131
0.0351
0.0329
0.0401
0.0444
0.0343
0.0343
0.0418
0.0431
0.0554
0.0599
0.0491
0.0454
0.0577
0.0384
0.0493
LowerCI
0.0692
0.0398
0.0305
0.0000
0.0092
0.0285
0.0339
0.1224
0.0772
0.0934
0.0738
0.0657
0.1342
0.1138
0.1596
0.1244
0.1431
0.0966
0.1188
0.1463
0.1602
0.0020
0.0370
0.0224
0.0647
0.0823
0.0926
0.1133
0.0928
0.0988
0.0654
0.1049
0.1046
0.0800
0.1021
0.1180
0.0606
0.1223
0.1229
UpperCI
0.1906
0.1593
0.1285
0.0000
0.0635
0.0920
0.1041
0.2307
0.1918
0.1980
0.2152
0.1547
0.2579
0.2215
0.2762
0.2618
0.2987
0.2187
0.2224
0.2569
0.2733
0.0832
0.1691
0.0765
0.2061
0.2131
0.2514
0.2883
0.2288
0.2346
0.2338
0.2751
0.3227
0.3185
0.2959
0.2966
0.2946
0.2734
0.3165
5C-19
-------
Appendix 5C, Attachment A, Table CA-2. Unsmoothed prevalence for children "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
Prevalence
0.0018
0.0387
0.0302
0.0395
0.0531
0.0617
0.0386
0.0801
0.0492
0.0789
0.0625
0.0856
0.1269
0.1089
0.1580
0.0863
0.1300
0.0989
0.0564
0.0486
0.0959
0.0697
0.1697
0.0819
0.0809
0.0680
0.1257
0.1394
0.1871
0.0726
0.1101
0.1258
0.0999
0.1648
0.1647
0.1747
0.0061
0.0214
0.0752
0.0692
0.0527
0.1293
0.0710
0.1369
0.1047
0.1096
0.1004
0.1340
0.1093
0.1029
0.1230
0.1007
0.1141
0.0644
0.0274
0.0892
0.1786
0.1620
0.2557
0.1914
0.1432
0.1788
0.2414
0.1114
SE
0.0018
0.0233
0.0135
0.0186
0.0214
0.0173
0.0192
0.0239
0.0151
0.0200
0.0162
0.0232
0.0357
0.0264
0.0478
0.0213
0.0319
0.0236
0.0353
0.0183
0.0434
0.0338
0.0387
0.0265
0.0357
0.0325
0.0346
0.0398
0.1071
0.0266
0.0452
0.0354
0.0435
0.0745
0.0576
0.1141
0.0044
0.0175
0.0222
0.0203
0.0201
0.0303
0.0193
0.0301
0.0299
0.0269
0.0281
0.0348
0.0242
0.0210
0.0236
0.0305
0.0268
0.0193
0.0175
0.0369
0.0652
0.0475
0.0634
0.0400
0.0333
0.0378
0.0604
0.0404
LowerCI
0.0002
0.0117
0.0125
0.0155
0.0238
0.0354
0.0143
0.0442
0.0267
0.0476
0.0373
0.0498
0.0717
0.0669
0.0849
0.0526
0.0792
0.0613
0.0160
0.0229
0.0383
0.0263
0.1065
0.0428
0.0332
0.0261
0.0719
0.0779
0.0548
0.0349
0.0477
0.0711
0.0413
0.0640
0.0799
0.0429
0.0015
0.0042
0.0417
0.0385
0.0247
0.0805
0.0413
0.0878
0.0589
0.0669
0.0571
0.0791
0.0700
0.0684
0.0837
0.0548
0.0711
0.0354
0.0077
0.0386
0.0835
0.0888
0.1517
0.1248
0.0894
0.1162
0.1429
0.0533
UpperCI
0.0129
0.1208
0.0715
0.0972
0.1142
0.1055
0.0999
0.1411
0.0888
0.1280
0.1029
0.1433
0.2145
0.1724
0.2751
0.1382
0.2062
0.1556
0.1799
0.1000
0.2206
0.1723
0.2594
0.1512
0.1840
0.1661
0.2105
0.2369
0.4777
0.1451
0.2340
0.2130
0.2226
0.3629
0.3094
0.4997
0.0247
0.1008
0.1319
0.1213
0.1090
0.2011
0.1193
0.2072
0.1793
0.1745
0.1704
0.2179
0.1665
0.1520
0.1771
0.1780
0.1780
0.1143
0.0925
0.1927
0.3418
0.2772
0.3974
0.2821
0.2215
0.2649
0.3780
0.2180
5C-20
-------
Appendix 5C, Attachment A, Table CA-2. Unsmoothed prevalence for children "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Age
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
Prevalence
0.2022
0.1731
0.2271
0.1627
0.0967
0.1509
0.1167
0.3301
0.0055
0.0296
0.0697
0.0470
0.0717
0.0642
0.0709
0.0697
0.0609
0.0996
0.0740
0.1028
0.0386
0.0187
0.0907
0.1270
0.0974
0.1239
0.0078
0.1230
0.0658
0.1700
0.1139
0.2219
0.0583
0.0495
0.0850
0.0652
0.0988
0.2587
0.0882
0.3162
0.1293
0.1798
0.1429
0.1133
0.0131
0.0505
0.0635
0.0582
0.1007
0.1245
0.1990
0.1240
0.1482
0.0980
0.0999
0.1805
0.1204
0.0855
0.1243
0.1249
0.1198
0.0690
0.0375
0.1649
SE
0.0624
0.0406
0.1064
0.0591
0.0413
0.0506
0.0490
0.1005
0.0054
0.0164
0.0252
0.0158
0.0199
0.0196
0.0254
0.0180
0.0209
0.0334
0.0260
0.0305
0.0187
0.0095
0.0181
0.0344
0.0267
0.0375
0.0078
0.0576
0.0272
0.0576
0.0456
0.0583
0.0290
0.0252
0.0368
0.0294
0.0440
0.0734
0.0426
0.0739
0.0372
0.0479
0.0381
0.0426
0.0101
0.0227
0.0220
0.0216
0.0281
0.0318
0.0511
0.0274
0.0321
0.0321
0.0216
0.0342
0.0211
0.0237
0.0351
0.0247
0.0283
0.0173
0.0275
0.0506
LowerCI
0.1061
0.1072
0.0822
0.0767
0.0406
0.0757
0.0495
0.1683
0.0008
0.0099
0.0337
0.0240
0.0413
0.0349
0.0346
0.0416
0.0307
0.0507
0.0366
0.0565
0.0147
0.0069
0.0609
0.0733
0.0562
0.0671
0.0011
0.0469
0.0287
0.0842
0.0503
0.1282
0.0215
0.0179
0.0354
0.0264
0.0400
0.1416
0.0332
0.1913
0.0722
0.1039
0.0831
0.0527
0.0029
0.0206
0.0318
0.0277
0.0574
0.0742
0.1171
0.0795
0.0956
0.0506
0.0648
0.1229
0.0848
0.0491
0.0702
0.0839
0.0744
0.0418
0.0087
0.0877
UpperCI
0.3511
0.2675
0.4908
0.3125
0.2129
0.2781
0.2512
0.5456
0.0368
0.0854
0.1384
0.0897
0.1218
0.1151
0.1398
0.1143
0.1171
0.1865
0.1439
0.1797
0.0975
0.0500
0.1330
0.2108
0.1636
0.2177
0.0541
0.2852
0.1436
0.3133
0.2376
0.3561
0.1484
0.1294
0.1903
0.1521
0.2240
0.4249
0.2146
0.4746
0.2209
0.2930
0.2348
0.2269
0.0574
0.1185
0.1228
0.1181
0.1705
0.2013
0.3177
0.1885
0.2227
0.1813
0.1509
0.2573
0.1682
0.1449
0.2108
0.1819
0.1872
0.1117
0.1477
0.2887
5C-21
-------
Appendix 5C, Attachment A, Table CA-2. Unsmoothed prevalence for children "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Age
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
Prevalence
0.1621
0.1015
0.2486
0.1479
0.2630
0.1707
0.2056
0.3343
0.2276
0.1643
0.1117
0.1931
0.1714
0.2043
0.1684
0.2140
0.0129
0.0144
0.0452
0.0675
0.0540
0.0572
0.1002
0.0894
0.0762
0.0969
0.0473
0.0847
0.0768
0.0700
0.1059
0.0930
0.0702
0.0867
0.0404
0.0613
0.0704
0.0812
0.1404
0.1276
0.0792
0.1262
0.1185
0.1147
0.1038
0.1461
0.1299
0.1013
0.1699
0.1591
0.0633
0.0975
0.0044
0.0700
0.0911
0.0962
0.1230
0.1321
0.0999
0.1114
0.0946
0.1108
0.1010
0.0946
SE
0.0496
0.0440
0.0909
0.0487
0.0391
0.0507
0.0966
0.0680
0.0786
0.0600
0.0389
0.0430
0.0664
0.0555
0.0501
0.0526
0.0080
0.0076
0.0169
0.0196
0.0150
0.0138
0.0186
0.0191
0.0160
0.0210
0.0135
0.0165
0.0152
0.0158
0.0211
0.0186
0.0156
0.0162
0.0203
0.0183
0.0193
0.0254
0.0367
0.0304
0.0288
0.0305
0.0290
0.0286
0.0301
0.0366
0.0490
0.0262
0.0385
0.0365
0.0273
0.0299
0.0025
0.0162
0.0195
0.0206
0.0259
0.0204
0.0192
0.0214
0.0168
0.0202
0.0186
0.0175
LowerCI
0.0864
0.0420
0.1131
0.0753
0.1939
0.0926
0.0751
0.2162
0.1093
0.0770
0.0552
0.1223
0.0764
0.1162
0.0912
0.1286
0.0038
0.0051
0.0215
0.0379
0.0311
0.0354
0.0692
0.0584
0.0502
0.0627
0.0269
0.0576
0.0518
0.0447
0.0711
0.0624
0.0451
0.0597
0.0149
0.0338
0.0408
0.0434
0.0826
0.0789
0.0381
0.0775
0.0724
0.0694
0.0579
0.0879
0.0600
0.0602
0.1071
0.0998
0.0267
0.0526
0.0014
0.0442
0.0595
0.0627
0.0805
0.0970
0.0681
0.0758
0.0664
0.0770
0.0699
0.0655
UpperCI
0.2835
0.2255
0.4621
0.2701
0.3463
0.2935
0.4521
0.4776
0.4145
0.3164
0.2132
0.2914
0.3410
0.3338
0.2901
0.3345
0.0427
0.0402
0.0926
0.1175
0.0920
0.0911
0.1431
0.1346
0.1141
0.1466
0.0819
0.1231
0.1124
0.1080
0.1550
0.1364
0.1077
0.1242
0.1050
0.1085
0.1189
0.1471
0.2286
0.1997
0.1573
0.1989
0.1881
0.1836
0.1792
0.2331
0.2589
0.1655
0.2590
0.2441
0.1427
0.1737
0.0135
0.1092
0.1373
0.1449
0.1833
0.1774
0.1443
0.1608
0.1330
0.1569
0.1438
0.1348
5C-22
-------
Appendix 5C, Attachment A, Table CA-2. Unsmoothed prevalence for children "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Poverty Status
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Age
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
Prevalence
0.1340
0.1122
0.0713
0.0899
0.0871
0.0700
0.0477
0.0859
0.0820
0.1434
0.1320
0.2314
0.1395
0.1207
0.2064
0.1364
0.1473
0.1390
0.1673
0.1684
0.0936
0.1379
0.0816
0.1057
0.0013
0.0353
0.0159
0.0284
0.0183
0.0689
0.0477
0.0469
0.0756
0.0686
0.0791
0.0763
0.1023
0.0571
0.1012
0.0923
0.0787
0.1303
0.0064
0.0443
0.0249
0.0372
0.0114
0.0491
0.1016
0.0908
0.0874
0.0839
0.0275
0.0339
0.0551
0.1028
0.1312
0.0630
0.0758
0.0328
0.0000
0.0039
0.0305
0.0384
SE
0.0207
0.0226
0.0153
0.0158
0.0147
0.0178
0.0162
0.0197
0.0201
0.0319
0.0265
0.0486
0.0302
0.0269
0.0474
0.0279
0.0315
0.0286
0.0339
0.0449
0.0305
0.0353
0.0275
0.0289
0.0013
0.0202
0.0076
0.0132
0.0071
0.0276
0.0166
0.0144
0.0263
0.0196
0.0250
0.0124
0.0260
0.0163
0.0251
0.0207
0.0214
0.0294
0.0064
0.0195
0.0153
0.0137
0.0102
0.0294
0.0419
0.0302
0.0258
0.0267
0.0137
0.0160
0.0254
0.0393
0.0440
0.0247
0.0287
0.0163
0.0000
0.0040
0.0113
0.0129
LowerCI
0.0983
0.0750
0.0466
0.0635
0.0623
0.0421
0.0242
0.0544
0.0503
0.0914
0.0881
0.1498
0.0902
0.0771
0.1285
0.0903
0.0956
0.0917
0.1109
0.0975
0.0485
0.0820
0.0415
0.0609
0.0002
0.0113
0.0062
0.0113
0.0085
0.0308
0.0239
0.0255
0.0376
0.0388
0.0420
0.0553
0.0614
0.0323
0.0615
0.0590
0.0458
0.0827
0.0009
0.0185
0.0074
0.0179
0.0020
0.0148
0.0440
0.0464
0.0484
0.0443
0.0103
0.0133
0.0219
0.0474
0.0662
0.0288
0.0354
0.0122
0.0000
0.0005
0.0147
0.0197
UpperCI
0.1801
0.1646
0.1077
0.1260
0.1206
0.1141
0.0916
0.1330
0.1309
0.2178
0.1931
0.3397
0.2097
0.1840
0.3145
0.2009
0.2203
0.2051
0.2445
0.2752
0.1729
0.2226
0.1544
0.1772
0.0095
0.1045
0.0401
0.0695
0.0389
0.1468
0.0928
0.0846
0.1459
0.1185
0.1440
0.1043
0.1655
0.0989
0.1622
0.1416
0.1322
0.1993
0.0441
0.1025
0.0805
0.0756
0.0638
0.1506
0.2174
0.1698
0.1529
0.1532
0.0715
0.0839
0.1315
0.2089
0.2435
0.1324
0.1546
0.0850
0.0000
0.0289
0.0623
0.0735
5C-23
-------
Appendix 5C, Attachment A, Table CA-2. Unsmoothed prevalence for children "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Above Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Below Poverty
Age
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Prevalence
0.1363
0.0933
0.0803
0.1014
0.0537
0.1120
0.1202
0.1333
0.1258
0.1039
0.0873
0.0881
0.1066
0.1364
0.0135
0.0812
0.0308
0.0944
0.1056
0.0856
0.1277
0.0943
0.1282
0.0883
0.0697
0.0954
0.0759
0.0600
0.1457
0.1099
0.0957
0.1136
SE
0.0261
0.0268
0.0208
0.0320
0.0182
0.0242
0.0253
0.0271
0.0286
0.0328
0.0217
0.0222
0.0230
0.0284
0.0128
0.0317
0.0080
0.0311
0.0306
0.0256
0.0356
0.0353
0.0343
0.0287
0.0228
0.0365
0.0316
0.0276
0.0391
0.0551
0.0350
0.0421
LowerCI
0.0927
0.0523
0.0478
0.0537
0.0273
0.0726
0.0788
0.0885
0.0796
0.0549
0.0531
0.0532
0.0692
0.0897
0.0020
0.0370
0.0185
0.0486
0.0588
0.0471
0.0726
0.0443
0.0746
0.0459
0.0363
0.0440
0.0329
0.0239
0.0844
0.0394
0.0458
0.0534
UpperCI
0.1960
0.1608
0.1317
0.1834
0.1029
0.1689
0.1791
0.1959
0.1934
0.1879
0.1404
0.1425
0.1607
0.2021
0.0832
0.1691
0.0510
0.1755
0.1822
0.1508
0.2149
0.1897
0.2115
0.1632
0.1298
0.1947
0.1655
0.1427
0.2398
0.2713
0.1894
0.2254
5C-24
-------
Appendix 5C, Attachment A, Table CA-3. Unsmoothed prevalence for adults "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Age_grp
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
Prevalence
0.1633
0.1347
0.1214
0.1157
0.1360
0.1104
0.0990
0.1990
0.1896
0.1789
0.1903
0.2760
0.1459
0.1295
0.1658
0.1254
0.0934
0.0659
0.0856
0.0884
0.0808
0.1672
0.1103
0.0945
0.1445
0.1623
0.1474
0.0830
0.1834
0.1375
0.1297
0.1209
0.1306
0.1244
0.0844
0.1642
0.1726
0.1771
0.2140
0.2174
0.1752
0.0941
0.1658
0.1262
0.0773
0.0976
0.0911
0.0926
0.0689
0.1753
0.1255
0.1317
0.1189
0.1681
0.1383
0.0943
0.1501
0.1290
0.1050
0.1163
0.1279
0.1231
0.0939
0.1511
SE
0.0154
0.0096
0.0084
0.0072
0.0103
0.0107
0.0095
0.0156
0.0177
0.0209
0.0180
0.0255
0.0205
0.0202
0.0158
0.0092
0.0083
0.0057
0.0086
0.0106
0.0110
0.0182
0.0156
0.0191
0.0204
0.0203
0.0307
0.0217
0.0199
0.0107
0.0109
0.0095
0.0106
0.0130
0.0101
0.0194
0.0170
0.0172
0.0204
0.0232
0.0186
0.0132
0.0223
0.0126
0.0094
0.0086
0.0096
0.0128
0.0127
0.0200
0.0178
0.0244
0.0162
0.0490
0.0313
0.0265
0.0121
0.0084
0.0074
0.0060
0.0087
0.0102
0.0092
0.0133
LowerCI
0.1353
0.1169
0.1059
0.1022
0.1171
0.0910
0.0819
0.1701
0.1573
0.1415
0.1576
0.2289
0.1101
0.0948
0.1371
0.1085
0.0784
0.0555
0.0701
0.0697
0.0617
0.1345
0.0832
0.0632
0.1089
0.1263
0.0968
0.0492
0.1476
0.1178
0.1097
0.1034
0.1113
0.1010
0.0666
0.1296
0.1418
0.1459
0.1767
0.1753
0.1417
0.0712
0.1265
0.1034
0.0607
0.0820
0.0740
0.0704
0.0478
0.1395
0.0945
0.0909
0.0906
0.0923
0.0875
0.0536
0.1279
0.1134
0.0914
0.1051
0.1119
0.1044
0.0773
0.1269
UpperCI
0.1958
0.1547
0.1389
0.1306
0.1575
0.1332
0.1193
0.2314
0.2268
0.2237
0.2281
0.3285
0.1908
0.1744
0.1990
0.1446
0.1109
0.0779
0.1040
0.1114
0.1050
0.2060
0.1447
0.1391
0.1893
0.2061
0.2182
0.1367
0.2256
0.1598
0.1527
0.1409
0.1528
0.1523
0.1064
0.2059
0.2084
0.2132
0.2567
0.2664
0.2147
0.1234
0.2142
0.1531
0.0980
0.1158
0.1117
0.1209
0.0982
0.2179
0.1648
0.1872
0.1545
0.2865
0.2118
0.1606
0.1754
0.1464
0.1205
0.1285
0.1459
0.1446
0.1136
0.1790
5C-25
-------
Appendix 5C, Attachment A, Table CA-3. Unsmoothed prevalence for adults "EVER" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age_grp
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
Prevalence
0.1336
0.1452
0.1622
0.2039
0.1616
0.1127
0.1438
0.1095
0.0890
0.0704
0.0782
0.0789
0.0893
0.1473
0.0914
0.0972
0.1062
0.1068
0.0966
0.0702
0.1595
0.1387
0.1368
0.1431
0.1478
0.1541
0.1231
0.1522
0.1191
0.1466
0.1874
0.1747
0.1318
0.1370
0.1499
0.1304
0.0984
0.0944
0.0917
0.1168
0.1208
0.1589
0.0846
0.0760
0.1422
0.0979
0.1349
0.0937
SE
0.0087
0.0125
0.0128
0.0179
0.0163
0.0133
0.0100
0.0078
0.0066
0.0051
0.0071
0.0078
0.0111
0.0152
0.0122
0.0139
0.0138
0.0156
0.0149
0.0130
0.0150
0.0096
0.0109
0.0092
0.0094
0.0130
0.0117
0.0184
0.0118
0.0182
0.0219
0.0181
0.0179
0.0198
0.0188
0.0107
0.0080
0.0081
0.0075
0.0126
0.0160
0.0222
0.0128
0.0135
0.0214
0.0176
0.0323
0.0194
LowerCI
0.1175
0.1224
0.1386
0.1711
0.1321
0.0891
0.1253
0.0952
0.0769
0.0610
0.0654
0.0649
0.0698
0.1199
0.0701
0.0732
0.0821
0.0799
0.0710
0.0486
0.1323
0.1209
0.1168
0.1261
0.1303
0.1302
0.1020
0.1195
0.0978
0.1145
0.1483
0.1419
0.1005
0.1027
0.1167
0.1108
0.0837
0.0796
0.0780
0.0943
0.0928
0.1201
0.0626
0.0535
0.1052
0.0684
0.0831
0.0620
UpperCI
0.1515
0.1714
0.1889
0.2413
0.1962
0.1415
0.1645
0.1258
0.1027
0.0811
0.0932
0.0956
0.1135
0.1797
0.1184
0.1280
0.1363
0.1414
0.1301
0.1004
0.1911
0.1586
0.1595
0.1621
0.1671
0.1813
0.1479
0.1920
0.1441
0.1859
0.2341
0.2131
0.1709
0.1806
0.1905
0.1527
0.1153
0.1116
0.1076
0.1438
0.1558
0.2073
0.1133
0.1069
0.1894
0.1381
0.2116
0.1393
5C-26
-------
Appendix 5C, Attachment A, Table CA-4. Unsmoothed prevalence for adults "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Age_grp
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
Prevalence
0.1062
0.0859
0.0859
0.0858
0.0996
0.0755
0.0643
0.1306
0.1329
0.1354
0.1398
0.2110
0.1190
0.1029
0.0790
0.0599
0.0486
0.0447
0.0555
0.0524
0.0477
0.0938
0.0572
0.0731
0.0969
0.1350
0.1349
0.0643
0.1123
0.0917
0.0944
0.0858
0.0945
0.0898
0.0706
0.1232
0.1180
0.1265
0.1745
0.1744
0.1388
0.0488
0.0888
0.0655
0.0409
0.0564
0.0469
0.0641
0.0527
0.0780
0.0847
0.0795
0.0798
0.1322
0.1055
0.0758
0.0893
0.0731
0.0689
0.0716
0.0865
0.0914
0.0599
0.0996
SE
0.0133
0.0090
0.0081
0.0061
0.0090
0.0083
0.0073
0.0144
0.0143
0.0187
0.0166
0.0221
0.0180
0.0183
0.0125
0.0066
0.0063
0.0049
0.0059
0.0076
0.0088
0.0143
0.0137
0.0162
0.0208
0.0205
0.0294
0.0213
0.0148
0.0102
0.0092
0.0080
0.0086
0.0106
0.0098
0.0182
0.0147
0.0138
0.0185
0.0211
0.0148
0.0088
0.0161
0.0093
0.0061
0.0078
0.0085
0.0105
0.0110
0.0129
0.0171
0.0212
0.0196
0.0492
0.0296
0.0247
0.0090
0.0064
0.0051
0.0049
0.0064
0.0090
0.0072
0.0119
LowerCI
0.0828
0.0699
0.0713
0.0746
0.0832
0.0608
0.0514
0.1049
0.1073
0.1027
0.1102
0.1709
0.0879
0.0722
0.0577
0.0482
0.0377
0.0360
0.0450
0.0394
0.0331
0.0693
0.0355
0.0470
0.0630
0.0997
0.0869
0.0332
0.0864
0.0735
0.0778
0.0714
0.0790
0.0711
0.0537
0.0918
0.0921
0.1018
0.1412
0.1369
0.1123
0.0341
0.0620
0.0495
0.0304
0.0429
0.0328
0.0463
0.0348
0.0562
0.0566
0.0467
0.0489
0.0617
0.0600
0.0395
0.0732
0.0615
0.0595
0.0626
0.0747
0.0753
0.0473
0.0786
UpperCI
0.1354
0.1052
0.1031
0.0986
0.1188
0.0934
0.0802
0.1614
0.1634
0.1764
0.1757
0.2575
0.1590
0.1448
0.1071
0.0743
0.0625
0.0554
0.0683
0.0694
0.0682
0.1258
0.0908
0.1119
0.1461
0.1804
0.2035
0.1208
0.1447
0.1138
0.1141
0.1029
0.1127
0.1128
0.0924
0.1634
0.1499
0.1560
0.2137
0.2196
0.1704
0.0693
0.1257
0.0862
0.0547
0.0738
0.0667
0.0880
0.0789
0.1075
0.1248
0.1322
0.1275
0.2608
0.1789
0.1406
0.1086
0.0866
0.0797
0.0818
0.1000
0.1105
0.0756
0.1254
5C-27
-------
Appendix 5C, Attachment A, Table CA-4. Unsmoothed prevalence for adults "STILL" having asthma.
Smoothed
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age_grp
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
Prevalence
0.0867
0.1152
0.1369
0.1780
0.1303
0.0895
0.0608
0.0471
0.0451
0.0359
0.0413
0.0441
0.0636
0.0617
0.0344
0.0488
0.0800
0.0676
0.0687
0.0331
0.0908
0.0819
0.0994
0.0937
0.1013
0.1103
0.0783
0.0901
0.0861
0.1081
0.1391
0.1293
0.1053
0.1061
0.0620
0.0528
0.0582
0.0499
0.0542
0.0756
0.0711
0.0741
0.0457
0.0344
0.1119
0.0528
0.1159
0.0442
SE
0.0079
0.0113
0.0123
0.0173
0.0152
0.0118
0.0079
0.0053
0.0048
0.0040
0.0055
0.0057
0.0097
0.0086
0.0064
0.0109
0.0131
0.0122
0.0129
0.0083
0.0143
0.0070
0.0090
0.0095
0.0087
0.0114
0.0092
0.0135
0.0111
0.0143
0.0179
0.0164
0.0166
0.0162
0.0104
0.0068
0.0061
0.0065
0.0072
0.0102
0.0133
0.0132
0.0097
0.0089
0.0198
0.0137
0.0336
0.0131
LowerCI
0.0725
0.0948
0.1144
0.1467
0.1033
0.0689
0.0471
0.0377
0.0365
0.0288
0.0317
0.0342
0.0470
0.0468
0.0239
0.0314
0.0579
0.0473
0.0473
0.0202
0.0663
0.0691
0.0830
0.0766
0.0854
0.0898
0.0621
0.0669
0.0667
0.0831
0.1075
0.1005
0.0770
0.0782
0.0445
0.0410
0.0473
0.0386
0.0416
0.0579
0.0491
0.0520
0.0301
0.0207
0.0786
0.0316
0.0644
0.0246
UpperCI
0.1035
0.1393
0.1629
0.2144
0.1631
0.1154
0.0782
0.0587
0.0556
0.0446
0.0535
0.0567
0.0855
0.0810
0.0494
0.0751
0.1097
0.0957
0.0987
0.0539
0.1231
0.0968
0.1186
0.1141
0.1197
0.1347
0.0982
0.1202
0.1105
0.1394
0.1781
0.1648
0.1425
0.1424
0.0858
0.0679
0.0715
0.0642
0.0702
0.0982
0.1019
0.1046
0.0689
0.0568
0.1570
0.0870
0.1996
0.0781
5C-28
-------
I ' ' ' ' I' ' ' ' I ' ' ' ' I '' ' ' I '' ' ' I ' '' ' I ' '' ' I ' ' ' ' I ' ' '' I ' ' ' ' I ' ' ' ' I ' ' ' ' I' ' ' ' I' ' ' ' I ' ' ' ' I '
O 123456789 1O 11 1213M15 16 17
1 I ' ' '' I ' ' '' I '
O 123456789 1O 11 12 B M 15 16 17
Rnrie Iv^e
g^rrfa- ^~^~ Knde ^~^~
l^ix'l. l^ivv^r4liiHTr\lJ< pv\;Juixki7rics:Hilo[iiklLiixkiiriii'\iJs
Q6-
Q3-
Q2-
Ql-
QO
O 123456789 1O 11 1213M15 16 17
Q6-
Q5-
Q4-
Q3-
Q2-
QO
r ' ' • i' ''' i''' ' i'' ' ' i'''' i ' '' ' i''' ' i' ' '' i ' ''' i ''' ' i' ' ' ' i ' ''' i''' ' i'' ' ' i'' '' i ''' ' i''' ' i
O 123456789 1O 11 1213M15 16 17
g^rrkr ^^^^ lenrie
g^rrkr Ifertrie
Appendix 5C, Attachment A, Figure CA-1. Unsmoothed prevalence and confidence intervals for children 'EVER' having asthma.
5C-29
-------
Q4
Q3
Q2-
Ql
oa
O 123456789 1O 11 12 B M 15 16 17
O 1 23 45 6 7891O11121314151617
gprdsr
Rmte ]Vfle
gprctr
Rmfe ]Vfle
ao-
1 r •'' i' •'' i'' '' i'' '' i '' •' i '' •' i •
O 123456789 1O 11 12 B M 15 16 17
gprctr
Rmte ]Vfle
9 10 11 12 13 14 15 16 17
Rmis ]Vfle
Appendix 5C, Attachment A, Figure CA-1, cont. Unsmoothed prevalence and confidence intervals for children 'EVER' having
asthma.
5C-30
-------
O 123456789 1O 11 1213 14 KKI7
Rmte JVfls
O 123456789 1O 11 1213 14 15 16 17
Rnde JVfle
Q5-
O4-
Q3-
Q2-
Ql-
i f ;>-• ::
iH^H^
0123456789 10 11 1213M15 16 17
^prefer Rirrie
0123456789 10 11 1213M15 16 17
gp-rfe- Rirrie
Appendix 5C, Attachment A, Figure CA-2. Unsmoothed prevalence and confidence intervals for children 'STILL' having asthma.
5C-31
-------
Q6-
Q5-
Q4:
Q3
Q2-
Ql
Q6-
Q5-
Q4-
Q3-
Q2-
Ql-
O 123456789 1O 11 1213M15 16 17
O 123456789 1O 11 1213M15 16 17
Rmte Tvfle
o^nrfcr- Knde
QO-
1 r ' •' i •'' • i' •'' i'' •' i •'' • i' ''' i' •'' i '' •' i •'' • i' •'' i'' •' i •'' ' i'''' i'' '' i'
O 123456789 1O 11 1213M15 16 17
Rrate
O 1 23 4 S 6 7 8 9 1O 11 12 ]3 14 ]5 16 17
Hmfe JVfls
Appendix 5C, Attachment A, Figure CA-2, cont. Unsmoothed prevalence and confidence intervals for children 'STILL' having
asthma.
5C-32
-------
QO-
1&-24
Ql-
QO-
1S-24
2534
2534
3544
4554
Rmte
3544
4S54
gprctr
Rrate
Q3-
Q2-
01-
18-24
gprctr
Hmis
Nfle
25-34
35-44
45-54
3&L3P
Hmis
Appendix 5C, Attachment A, Figure CA-3. Unsmoothed prevalence and confidence intervals for adults 'EVER' having asthma.
5C-33
-------
QO-
1&-24
QO-
1S-24
2534
2534
3544
4554
Rmte
3544
4554
gprctr
Rmte
Tvfle
QO-
1S24
2534
3544
4554
gprctr
Rmte
JVfle
gprctr
Rmte
TXfle
Appendix 5C, Attachment A, Figure CA-3, cont. Unsmoothed prevalence and confidence intervals for adults 'EVER' having
asthma.
5C-34
-------
ptev
Q4:
Q3:
gpncfer
Hmfe
JVfle
25-34 3544
3544 45-54
gpncfer
Hmfe
5564 eS-74
JVfle
J3EV
Q4-
3544 45-54 55« 66-71
Rrafe
JVfls
Appendix 5C, Attachment A, Figure CA-4. Unsmoothed prevalence and confidence intervals for adults 'STILL' having asthma.
5C-35
-------
Q4-
Q3-
Q2-
QO-
is-24
QO-
1S-24
2534
2534
3544
4554
Rrate
3544
4554
gprctr
Rrate
Q3-
02-
Ql-
35-44 45-54
gprctr
l\fle
Appendix 5C, Attachment A, Figure CA-4, cont. Unsmoothed prevalence and confidence intervals for adults 'STILL' having
asthma.
5C-36
-------
Appendix 5C, Attachment B
Logistic Model Fit Tables and Figures
Appendix 5C, Attachment B, Table 5CB-1. Alternative logistic models for estimating child asthma prevalence using the
"EVER" asthma response variable and goodness of fit test results.
Description
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
Stratification Variable
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1 . none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1 . none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
-2 log likelihood
288740115.1
287062346.4
288120804.1
287385013.1
286367652.6
286283543.6
2856961 64.7
284477928.1
286862135.1
285098650.6
286207721 .5
2853521 64
284330346.1
284182547.5
283587631 .7
282241318.6
286227019.6
284470413
285546716.1
284688169.9
283662673.5
283404487.5
282890785.3
281407414.3
285821686.2
283843266.2
284761522.8
284045849.2
282099156.1
281929968.5
281963915.7
278655423.1
DF
2
4
8
4
16
16
8
32
3
6
12
6
24
24
12
48
4
8
16
8
32
32
16
64
18
36
72
36
144
144
72
288
Appendix 5C, Attachment B, Table 5CB-2. Alternative logistic models for estimating child asthma prevalence using the
"STILL" asthma response variable and goodness of fit test results.
Description
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
1 . logit(prob) = linear in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
2. logit(prob) = quadratic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
Stratification Variable
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1 . none
2. gender
3. region
-2 log likelihood
181557347.7
180677544.6
180947344.2
180502490.5
179996184.8
179517528
179637601.4
178567573.9
180752073.1
179771977.6
180088080.5
179611530.4
179004935.6
178519078.1
178640744.8
177414967.2
180247874.1
179235170
179583725.1
DF
2
4
8
4
16
16
8
32
3
6
12
6
24
24
12
48
4
8
16
5C-37
-------
Appendix 5C, Attachment B, Table 5CB-2. Alternative logistic models for estimating child asthma prevalence using the
"STILL" asthma response variable and goodness of fit test results.
Description
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
3. logit(prob) = cubic in age
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
4. logit(prob) = f(age)
Stratification Variable
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
-2 log likelihood
179067549.2
178407915.7
177897359.3
1 78029240
176642073.7
179972765.3
178918713.8
178852704.9
178599743.4
177075815.4
176418872.7
177422457.4
173888684.9
DF
8
32
32
16
64
18
36
72
36
144
144
72
288
Appendix 5C, Attachment B, Table 5CB- 3. Alternative logistic models for estimating adult asthma prevalence using the
"EVER" asthma response variable and goodness of fit test results.
Description
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
Stratification Variable
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
-2 log likelihood
825494282
821614711.2
824598583.4
823443004.3
820520390.7
821958349.1
819560679.9
81 772371 0
DF
7
14
28
14
56
56
28
112
Appendix 5C, Attachment B, Table 5CB-4. Alternative logistic models for estimating adult asthma prevalence using the
"STILL" asthma response variable and goodness of fit test results.
Description
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
4. logit(prob) = f(age grp)
Stratification Variable
1. none
2. gender
3. region
4. poverty
5. region, gender
6. region, poverty
7. gender, poverty
8. region, gender, poverty
-2 log likelihood
600538044.1
594277797.3
599561222.3
597511872.6
593112157.6
596008068.6
591 394271 .8
589398969.5
DF
7
14
28
14
56
56
28
112
5C-38
-------
Appendix 5C, Attachment B, Table 5CB-5. Effect on residual standard error by varying LOESS
smoothing parameter while fitting children "EVER" having asthma data set.
Region
South
Northeast
South
Midwest
Midwest
South
South
Midwest
West
South
Midwest
Midwest
Midwest
Midwest
Northeast
South
South
Midwest
West
Northeast
South
Northeast
Midwest
Northeast
Midwest
West
West
South
South
Midwest
Midwest
South
Northeast
Midwest
South
South
Midwest
Midwest
Midwest
Midwest
South
South
Northeast
South
Midwest
West
South
South
South
West
South
West
South
Northeast
Midwest
Midwest
South
Northeast
West
West
Gender
Female
Female
Male
Male
Male
Female
Male
Male
Female
Female
Female
Male
Male
Female
Female
Male
Female
Female
Male
Female
Female
Female
Male
Female
Male
Female
Male
Female
Female
Male
Female
Male
Female
Male
Male
Female
Male
Male
Male
Female
Male
Male
Female
Male
Female
Male
Male
Female
Male
Female
Male
Male
Female
Male
Male
Male
Female
Male
Female
Female
Poverty Ratio
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Smoothing Parameter
0.5
0.7
0.6
0.9
0.8
0.8
0.5
1
0.7
0.7
0.7
0.4
0.8
0.6
0.4
0.7
0.6
0.8
0.6
0.6
0.5
0.8
0.7
0.5
0.9
0.8
0.5
0.9
0.4
0.7
0.9
0.6
0.9
0.5
0.7
0.4
0.6
0.4
1
0.5
0.8
0.5
1
1
1
0.4
0.8
0.6
0.9
0.9
0.4
0.4
0.7
0.6
0.6
0.5
0.8
0.5
0.6
1
Residual Standard Error
0.999919
1 .00088
1 .003839
1 .00548
1 .01 0889
1.012178
0.982885
1 .023284
0.973279
0.97298
1 .028007
0.970948
0.965591
1 .038233
0.961444
1 .040867
0.954946
1 .0451 07
1.052418
0.946315
0.945525
1 .054556
0.940657
0.940383
1 .063971
1.066819
1 .067075
1 .067923
0.930104
0.929292
1 .072631
0.927161
1 .074984
0.917969
0.912266
1 .089646
0.90827
0.906073
1 .094737
1 .096459
1 .099725
0.898228
1.101884
0.896985
1.103976
0.894137
0.893364
0.891551
0.890138
1.111538
0.885511
1.115223
0.86999
0.86934
0.86245
0.857982
0.857778
0.857592
0.852664
1.147894
5C-39
-------
Appendix 5C, Attachment B, Table 5CB-5. Effect on residual standard error by varying LOESS
smoothing parameter while fitting children "EVER" having asthma data set.
Region
South
South
Northeast
West
West
West
South
West
West
South
Northeast
South
West
Northeast
West
Northeast
West
Midwest
Northeast
Northeast
Midwest
West
Midwest
Northeast
South
Northeast
West
Northeast
West
Northeast
West
West
Midwest
West
West
Northeast
West
Northeast
Northeast
Northeast
Northeast
Midwest
Northeast
West
Northeast
West
Northeast
Northeast
Midwest
West
Midwest
Midwest
Gender
Female
Female
Male
Male
Female
Female
Male
Female
Female
Male
Male
Female
Female
Male
Female
Male
Male
Female
Male
Male
Female
Male
Female
Male
Male
Male
Male
Male
Female
Male
Male
Female
Female
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Male
Female
Male
Female
Male
Female
Female
Poverty Ratio
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Smoothing Parameter
1
0.9
0.7
0.7
0.9
0.8
0.4
0.7
1
0.9
0.8
1
0.6
1
0.5
0.9
1
1
1
0.7
0.4
0.5
0.9
0.6
1
0.8
0.9
0.9
0.5
0.4
0.6
0.4
0.8
0.8
0.7
0.5
0.8
0.8
0.7
0.9
0.6
0.7
0.5
0.4
1
0.9
0.4
0.4
0.6
1
0.5
0.4
Residual Standard Error
0.849143
0.847567
0.844668
1.163749
1.163943
1.166005
0.826195
1.174564
1.178045
1.178803
0.820245
1.182254
1.187757
0.811815
0.808706
0.805685
0.804743
0.799988
0.799128
0.798212
1.20612
0.793132
0.788082
0.78547
1 .21 6423
0.78144
0.780843
0.779772
1 .224495
0.769037
0.763027
0.762134
0.758775
0.756848
0.752592
0.729776
1 .2841 53
1 .292845
1 .296274
1 .308752
1 .309671
0.688366
1.314991
1.31595
1 .3271 29
1 .35931
1 .37577
0.618785
0.607758
1 .395061
0.541466
0.522325
5C-40
-------
Appendix 5C, Attachment B, Table 5CB-6. Effect on residual standard error by varying LOESS
smoothing parameter while fitting children "STILL" having asthma data set.
Region
South
Northeast
Northeast
Northeast
Northeast
Midwest
Midwest
Midwest
Midwest
Northeast
South
Midwest
Northeast
South
Northeast
Northeast
Northeast
Midwest
South
Northeast
Northeast
Midwest
Midwest
Midwest
Northeast
Midwest
South
Northeast
South
West
South
South
South
South
West
South
West
West
South
West
Midwest
West
West
Midwest
Northeast
West
Midwest
West
South
Midwest
West
Midwest
Northeast
West
Midwest
Midwest
Midwest
South
Midwest
Gender
Female
Male
Male
Male
Male
Male
Male
Female
Male
Male
Male
Female
Male
Female
Male
Male
Male
Male
Female
Male
Male
Male
Female
Male
Male
Male
Female
Male
Male
Male
Male
Male
Female
Female
Male
Male
Female
Female
Male
Female
Male
Male
Male
Female
Female
Female
Female
Male
Female
Female
Male
Male
Female
Female
Male
Female
Female
Male
Female
Poverty Ratio
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Smoothing
Parameter
1
0.9
0.7
0.9
0.4
0.7
0.8
1
0.6
0.8
0.6
0.5
0.5
0.4
0.6
0.5
0.8
0.9
0.6
1
1
0.5
0.9
1
0.7
0.7
0.7
0.6
1
0.9
0.5
0.7
0.8
0.9
0.8
0.9
0.7
1
0.8
0.9
0.8
0.7
1
0.8
0.4
0.8
0.6
0.7
0.5
0.7
0.6
0.4
0.5
0.7
0.9
0.9
1
0.4
0.8
Residual Standard Error
1.000117
1 .000909
1 .000993
0.997502
0.997275
0.996943
0.996544
1 .003498
0.995815
0.995723
1 .0071 98
0.99235
1 .008536
0.99041
1 .009859
1.01048
1.011028
1.011038
1.013156
1.01445
1 .01 6505
1.01692
0.979917
1 .020707
1.021388
0.977074
0.976479
1 .024042
0.975784
1 .025093
1 .0261 84
0.971057
0.965833
0.965238
1 .03481
0.964953
1 .036384
1 .040924
0.957162
1 .044522
1 .04601
1 .04802
1 .050309
0.946142
0.94543
1.055218
0.938888
1 .063545
1.063816
0.931681
1 .0791 46
1 .080605
1 .083479
1 .084472
1 .084476
0.914962
0.913089
1 .087093
0.912722
5C-41
-------
Appendix 5C, Attachment B, Table 5CB-6. Effect on residual standard error by varying LOESS
smoothing parameter while fitting children "STILL" having asthma data set.
Region
West
Midwest
Midwest
Northeast
South
Midwest
West
Midwest
Northeast
West
West
South
Midwest
Midwest
West
South
Northeast
South
Northeast
Northeast
West
Northeast
West
South
West
South
West
South
West
West
West
Midwest
West
Midwest
South
South
South
South
West
West
Midwest
South
Northeast
Northeast
West
South
Northeast
West
South
Northeast
Northeast
Northeast
Northeast
Gender
Female
Male
Male
Female
Male
Male
Female
Female
Male
Female
Male
Female
Male
Female
Female
Female
Female
Female
Female
Female
Male
Female
Male
Female
Male
Female
Female
Female
Female
Female
Male
Female
Female
Female
Male
Male
Female
Male
Female
Male
Female
Male
Female
Female
Male
Male
Female
Male
Male
Female
Female
Female
Female
Poverty Ratio
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Smoothing
Parameter
0.6
0.6
1
0.6
0.4
0.5
0.6
0.7
0.4
0.5
0.5
0.6
0.4
0.4
0.8
1
0.7
0.5
0.9
1
0.4
0.8
0.4
0.7
0.6
0.9
0.5
0.8
0.4
0.9
0.5
0.6
1
0.4
0.5
0.6
0.4
0.7
0.4
0.8
0.5
0.8
0.7
0.6
0.9
0.9
0.8
1
1
0.5
0.9
1
0.4
Residual Standard Error
0.912605
0.907737
1.103127
1.103286
1.112998
0.878223
1.124127
0.875579
0.874469
0.873529
1.127032
0.87206
0.869726
1.135372
1.136048
0.863066
1.140006
0.858107
1.147352
1.148471
1.152015
1.153553
0.845979
0.842335
0.8413
0.841106
1.166931
0.830955
0.826586
1.183444
0.815615
0.802622
1 .20757
0.78769
1.214019
1.216661
0.781555
1 .242272
1.252141
1 .254244
0.742493
1 .294055
1 .32003
1.355219
1 .356792
1 .365737
1.39015
1 .405599
1 .408469
1.431367
1 .503674
1 .574778
1.605
5C-42
-------
Appendix 5C, Attachment B, Table 5CB-7. Effect on residual standard error by varying LOESS smoothing
parameter while fitting adults "EVER" having asthma data set.
Region
Midwest
South
West
West
South
Midwest
West
West
West
Northeast
West
Midwest
Northeast
South
Midwest
Midwest
South
Northeast
South
South
South
Northeast
West
South
South
Northeast
Northeast
Northeast
West
Northeast
Northeast
South
Midwest
West
South
West
Northeast
Midwest
West
West
South
Northeast
Northeast
Midwest
Midwest
Midwest
Midwest
Midwest
Gender
Female
Female
Female
Male
Female
Female
Male
Female
Male
Female
Male
Female
Male
Male
Male
Male
Female
Male
Female
Female
Female
Male
Female
Male
Male
Female
Male
Male
Female
Female
Male
Male
Male
Female
Male
Female
Female
Male
Male
Male
Male
Female
Female
Male
Male
Female
Female
Female
Poverty Ratio
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Smoothing Parameter
1
1
0.9
0.8
1
0.9
0.8
0.8
0.9
1
1
0.8
0.8
1
0.9
1
0.9
1
0.8
0.9
0.8
1
1
0.9
1
0.9
0.9
0.9
1
1
0.8
0.9
1
0.8
0.8
0.9
0.8
0.9
0.9
1
0.8
0.9
0.8
0.8
0.8
1
0.9
0.8
Residual Standard Error
0.983356
1 .040607
1 .04471 2
0.937658
1 .06598
0.911278
1 .095844
0.893319
0.886119
0.875056
0.858542
0.843191
1.177547
0.813689
1.190978
0.785268
0.77381
1 .241 548
0.751726
0.747912
0.740577
0.732859
1 .275049
0.708509
0.706944
0.699107
1 .301 543
0.677309
0.669638
0.662619
0.646318
0.64328
1 .395026
0.597305
0.58427
0.567466
0.528031
0.49517
1.523816
1 .537805
0.400237
0.394894
0.362058
0.306085
0.169594
1.910643
1 .920542
2.249162
5C-43
-------
Appendix 5C, Attachment B, Table 5CB-8. Effect on residual standard error by varying LOESS smoothing
parameter while fitting adults "STILL" having asthma data set.
Region
South
West
West
West
West
West
West
South
Midwest
Northeast
Midwest
South
Midwest
South
South
West
South
Midwest
Northeast
Northeast
South
South
South
West
South
Northeast
Northeast
Midwest
Northeast
Northeast
Northeast
Midwest
South
Northeast
South
Midwest
Midwest
Northeast
West
Northeast
Northeast
Midwest
Midwest
Midwest
Midwest
West
West
West
Gender
Male
Female
Female
Female
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Female
Male
Female
Male
Male
Female
Male
Female
Female
Male
Female
Male
Male
Male
Male
Male
Female
Male
Male
Female
Female
Male
Male
Female
Female
Male
Female
Male
Male
Male
Poverty Ratio
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Smoothing Parameter
0.8
0.8
0.9
1
1
0.8
0.9
0.9
1
1
0.9
0.8
0.8
1
0.9
1
0.9
1
0.9
1
1
0.9
0.8
0.9
1
0.8
1
1
0.9
0.9
0.8
0.9
1
1
0.8
0.9
0.8
0.8
0.8
0.9
0.8
1
0.9
0.8
0.8
0.8
1
0.9
Residual Standard Error
1.015193
1 .04571 4
1.051807
1 .061 488
0.92928
0.925921
0.915895
1 .097531
0.89825
1.102905
0.876146
1.128781
0.870507
1.130393
0.835583
0.825684
1.192655
0.788217
0.786205
1.21537
1 .23752
0.748499
0.717121
0.670751
0.664236
0.65848
0.653985
0.650735
0.630298
1.370134
1 .375365
0.620174
1 .400273
0.581032
0.568428
0.508247
0.503315
0.478186
0.464598
0.453855
0.396203
1.616706
1 .636938
0.295923
1 .883863
2.16547
2.200364
2.396381
5C-44
-------
2-
\
10
2-
25 S) ~
ISEmai Ffetoatiles
\
10
25 SO "
r~Etnai Fferaatiles
2-
1
-1
Appendix 5C, Attachment B, Figure 5CB-1. Normal probability plots of studentized residuals generated using logistic model and children 'EVER'
asthmatic data set.
5C-45
-------
1-
-1
2-
1 -
o-
-1
-2
o
25 SO ^
T"imd Ktaaliles
01
o
10
25
50
90 95
99
99.9
Appendix 5C, Attachment B, Figure 5CB-1, cont. Normal probability plots of studentized residuals generated using logistic model and children
'EVER' asthmatic data set.
5C-46
-------
NMnai fxatdtifily fiat ofSlutta4ra(lre9dds ty&xnatk
1-
-1-
-3
Ql
\ I
5 10
I I
25 50
I I
90 95
99 999
1-
-1-
Ql
5 10 25
I
50
90 95 99 999
]>arrni fxatdtifily fiat aTslud^rii^xlie^ckds ty&xnatk
-2-
-3
Ql
\ I
5 10
I
25
I
50
I I I
-5 90 95
99 999
3-
2-
1
o-
-1-
-2
-3-
Ql
5 10 25
I
50
90 95
I
99
999
Appendix 5C, Attachment B, Figure 5CB-2. Normal probability plots of studentized residuals generated using logistic model and children
'STILL' asthmatic data set.
5C-47
-------
NMnai fxatdtifily fiat ofSlutta4ra(lre9dds ty&xnatk
Ql
5 10 25
3D
90 95
99
999
]>arrni fxatdtifily fiat aTsluttariraclre^dds ty&xnatk
-4-
Ql
\ I
5 10
I
25
I
50
I
-5
I I
90 95
99
999
-2
-4'
o
Ql
5 10
25
50
90 95
99
999
Appendix 5C, Attachment B, Figure 5CB-2, cont. Normal probability plots of studentized residuals generated using logistic model and children
'STILL' asthmatic data set.
5C-48
-------
INiiill |i< Ji Jilily |1< 4 < I 'sUi Li 4L*x I ii>ii k» Is b^nmaOirgpcremfci-
15-
10
05
O
-05
-10
-15
-20
00
n
i i i
5 1O 25
i
SO
9O 95 99 999
IN ii ill |i< Ji Jilily |1< 4 < I 'sin Li *i>»x I ii>it k* Is ly snmaOirgpcremfci-
1-
O
O
\ I I I
Ql 1 5 1O
I I
25 SO
I I
9O 95
99 999
3-
ti
I
-1
O
O
01
5 10 25 50 -& 90 96
99.9
Appendix 5C, Attachment B, Figure 5CB-3. Normal probability plots of studentized residuals generated using logistic model and adult 'EVER'
asthmatic data set.
5C-49
-------
INiiill |i< Ji Jilily |1< 4 < I 'sUi Li 4L*x I ii>ii k» Is b^nmaOirgpcremfci-
20-
15
10-
O5-
o-
-05
-10
-15
-20
o
o
n
i i i
51O25
i
5O
IN ii ill |i< Ji Jilily |1< 4 < I 'sUi Li iL*x I ii>ii k* Is ty smnftingfamfter
-3
O
\
Ql
\ I
5 1O
25 SO
I I
9O 95
99 999
I.-I
-2
Ql
o
o oo
5 10 25 50 T5, 9O 96
999
Appendix 5C, Attachment B, Figure 5CB-4. Normal probability plots of studentized residuals generated using logistic model and adult 'STILL'
asthmatic data set.
5C-50
-------
studsrt
5-
4-
3-
2-
1-
a
-i-
-2-
-3-
-4-
-500000
400000 -300000
000
000
000
+ + +
XXX
XXX
XXX
-200000
000
000
+
+
XXX
XXX
-100000
stLitrt
5-
4-
3-
2-
1-
O:
-1-
-2-
-3-
-4:
-5-
-500000
400000
-300000
000
000
000
XXX
XXX
XXX
-200000
o o o
o o o
+ + 4
+ + 4
XXX
XXX
-LOOOOO
stu±rt
5-
4-
3-
2-
1-
0-
-1
-2-
-3-
-4:
-5'
StLltrt
5-
4-
3-
2-
1-
O:
-1-
-2-
-3-
-4:
-5-
000
000
000
+ + +
XXX
XXX
XXX
O O O
o o o
o o o
+ + +
+ + 4
XXX
XXX
000
000
+ + -1
XXX
XXX
XXX
000
000
•+
•+
XXX
XXX
Appendix 5C, Attachment B, Figure 5CB-5. Studentized residuals generated using logistic model versus model predicted betas and the child
'EVER' asthmatic data set.
5C-51
-------
stu±rt
5-
4-
3-
2-
1-
o-
-1-
-2-
-3:
-5-
-5(
X
X -r"
MOOD ^00000 -300000 -2OOOOO -1OOOOO O
stu±rt
5-
4-
3-
2-
1-
O-
-1-
-2-
-3-
-4-
-5-
000
000
000
000
+ + •+
XXX
XXX
XXX
O O O
O O O
O O O
+ + +
XXX
XXX
stu±rt
5-
4-
3-
2-
1-
o-
-1-
-2-
-3-
-4-i
-5-
-500000
KOODOO -300000
000
+ + •+
XXX
XXX
XXX
-200000
-100000
000
000
+
+
XXX
XXX
000
000
000
000
+ + •+
XXX
XXX
XXX
O O O
O O O
O O O
+ + +
+ + 4-
XXX
XXX
Appendix 5C, Attachment B, Figure 5CB-5, cont. Studentized residuals generated using logistic model versus model predicted betas and the child
'EVER' asthmatic data set.
5C-52
-------
studsrt
5-
4-
3-
2-
1-
o
-1-
-2-
-3-
-4-
-500000
-400000
-300000
000
000
000
+ + +
XXX
XXX
XXX
-200000
000
000
+
+
XXX
XXX
-100000
stLitrt
5-
4-
3-
2-
1-
O:
-1-
-2-
-3-
^t-
-5-
-500000
o
^00000
-300000
000
000
000
XXX
XXX
XXX
-200000
o o o
o o o
+ + 4
+ + 4
XXX
XXX
-LOOOOO
studsrt
5-
4-
3-
2-
1-
0-
-1
-2-
-3-
-4:
-5'
O
o,
StLltrt
5-
4-
3-
2-
1-
oi
-i-
-2-
-3-
^t-
-5-
o
,°nO
000
000
000
+ + +
XXX
XXX
XXX
O O O
o o o
o o o
+ + +
+ + 4
XXX
XXX
000
000
+ + -1
XXX
XXX
XXX
000
000
•+
•+
XXX
XXX
Appendix 5C, Attachment B, Figure 5CB-6. Studentized residuals generated using logistic model versus model predicted betas and the child
'STILL' asthmatic data set.
5C-53
-------
studsrt
5-
4-
3-
2-
1-
o-
-1-
-2-
-3-
-4-
-5'
-500000
400000
000
000
000
+ + +
XXX
XXX
XXX
-300000 -200000
000
000
+
+
XXX
XXX
stLitrt
5-
4-
3-
2:
1-
O:
-1-
-2-
-3-
-100000
-500000
400000
-300000
000
000
000
O O O
+ + +
XXX
-200000
-LOOOOO
O O Q
O O Q
+ + 4
XXX
XXX
Studsrt
5-
4-
3-
2-
1-
0-
-1
-2-
-3-
-4:
-5'
O
000
000
000
+ + +
XXX
XXX
XXX
O O O
O O O
O O O
+ + +
+ + 4
XXX
XXX
Appendix 5C, Attachment B, Figure 5CB-6, cont. Studentized residuals generated using logistic model versus model predicted betas using child
'STILL' asthmatic data set.
5C-54
-------
aTixid^tllo^lscf^Ut]^^^
stu±rt
3-
2J
1-
O:
-1-
-2J
-3-
xxc^fe:
Q^^3feJI|M)€^|%$E§^^iQ fflO O
-------
aTixid^tllo^lscffKUtstffl
stu±rt
3-
2J
1-
-1-
-2J
-3-
X
X-^^^J^Si8fe^A^^o
0
-3.00000
-20H)0
-100000
000
000
000
000
+ + •+
XXX
XXX
XXX
O O O
O O O
O O O
4- + +
XXX
XXX
stu±rt
3-
2J
1-
o-
-1-
-2J
-3-
xX 0
O x + dx
^^A^'^O^v^'^^S^^1'^^'e:^'^X/^/r-\
xx'^^'^^^i G/x ^^T^^v^^ X ^^
X0x0 00
X
000
O O O
O O O
O O O
+ + +
XXX
XXX
XXX
O O O
O O O
O O O
4- + +
+ + +
XXX
XXX
stu±rt
3-
2-
O:
-1:
-2-.
-3-
x 0
xOO-^^^^^^^^1
XX X
000
000
O O O
+ + •+
XXX
XXX
XXX
-3.CD003
-20TDO
-1CD003
000
000
•+ + +
XXX
XXX
Appendix 5C, Attachment B, Figure 5CB-8. Studentized residuals generated using logistic model versus model predicted betas using adult
'STILL' asthmatic data set.
5C-56
-------
Appendix 5C, Attachment C
Smoothed Asthma Prevalence Tables and Figures
Appendix 5C, Attachment C, Table 5CC-1. Smoothed prevalence for children "EVER" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
Prevalence
0.0083
0.0179
0.0327
0.0509
0.0671
0.0854
0.0995
0.1041
0.1024
0.1020
0.1055
0.1192
0.1390
0.1529
0.1603
0.1597
0.1517
0.1374
0.0413
0.0706
0.1047
0.1356
0.1553
0.1488
0.1327
0.1341
0.1535
0.1729
0.1861
0.1691
0.1470
0.1439
0.1541
0.1707
0.1962
0.2323
0.0133
0.0313
0.0585
0.0898
0.1111
0.1256
0.1411
0.1496
0.1502
0.1542
0.1627
0.1760
0.1876
0.1847
0.1764
0.1641
0.1487
0.1318
0.0429
0.0908
0.1530
SE
0.0050
0.0066
0.0076
0.0096
0.0122
0.0134
0.0141
0.0145
0.0132
0.0121
0.0127
0.0137
0.0163
0.0176
0.0176
0.0160
0.0161
0.0229
0.0168
0.0168
0.0173
0.0208
0.0237
0.0229
0.0228
0.0224
0.0239
0.0270
0.0311
0.0300
0.0247
0.0239
0.0244
0.0275
0.0427
0.0813
0.0066
0.0091
0.0102
0.0121
0.0145
0.0149
0.0158
0.0164
0.0161
0.0166
0.0173
0.0181
0.0186
0.0181
0.0170
0.0149
0.0144
0.0201
0.0176
0.0214
0.0235
LowerCI
0.0022
0.0079
0.0195
0.0336
0.0448
0.0602
0.0725
0.0765
0.0769
0.0784
0.0806
0.0922
0.1070
0.1182
0.1254
0.1277
0.1197
0.0945
0.0167
0.0416
0.0724
0.0962
0.1100
0.1053
0.0902
0.0920
0.1080
0.1215
0.1272
0.1131
0.1006
0.0990
0.1078
0.1186
0.1187
0.1002
0.0045
0.0164
0.0398
0.0666
0.0831
0.0964
0.1100
0.1171
0.1182
0.1211
0.1283
0.1397
0.1501
0.1483
0.1422
0.1341
0.1198
0.0937
0.0173
0.0536
0.1084
UpperCI
0.0310
0.0397
0.0541
0.0766
0.0993
0.1198
0.1351
0.1403
0.1352
0.1317
0.1369
0.1527
0.1787
0.1956
0.2026
0.1979
0.1903
0.1956
0.0985
0.1174
0.1491
0.1879
0.2146
0.2062
0.1910
0.1912
0.2136
0.2401
0.2640
0.2451
0.2097
0.2045
0.2156
0.2395
0.3065
0.4512
0.0391
0.0588
0.0851
0.1200
0.1471
0.1621
0.1793
0.1892
0.1891
0.1942
0.2041
0.2193
0.2319
0.2277
0.2167
0.1994
0.1833
0.1823
0.1026
0.1498
0.2118
5C-57
-------
Appendix 5C, Attachment C, Table 5CC-1. Smoothed prevalence for children "EVER" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Age
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
Prevalence
0.2110
0.2428
0.2458
0.2393
0.2261
0.2225
0.2354
0.2499
0.2553
0.2512
0.2149
0.1941
0.2027
0.2364
0.3045
0.0115
0.0278
0.0533
0.0823
0.1027
0.1066
0.1023
0.0979
0.1010
0.1146
0.1179
0.1170
0.1154
0.1246
0.1405
0.1551
0.1714
0.1883
0.0394
0.0754
0.1188
0.1539
0.1684
0.1503
0.1355
0.1263
0.1322
0.1583
0.1818
0.2030
0.2293
0.2437
0.2368
0.2188
0.1906
0.1572
0.0279
0.0444
0.0668
0.0948
0.1269
0.1665
0.1891
0.1901
0.1858
0.1873
0.1908
0.1926
0.1934
SE
0.0277
0.0303
0.0285
0.0270
0.0268
0.0290
0.0311
0.0339
0.0357
0.0377
0.0355
0.0308
0.0292
0.0390
0.0768
0.0066
0.0095
0.0108
0.0127
0.0152
0.0150
0.0143
0.0137
0.0144
0.0166
0.0171
0.0175
0.0164
0.0148
0.0148
0.0152
0.0209
0.0376
0.0211
0.0229
0.0229
0.0265
0.0295
0.0269
0.0245
0.0231
0.0257
0.0301
0.0342
0.0358
0.0359
0.0366
0.0335
0.0286
0.0298
0.0443
0.0107
0.0103
0.0106
0.0134
0.0174
0.0209
0.0207
0.0204
0.0189
0.0189
0.0180
0.0163
0.0168
LowerCI
0.1566
0.1828
0.1888
0.1853
0.1729
0.1655
0.1741
0.1831
0.1852
0.1779
0.1473
0.1353
0.1462
0.1617
0.1652
0.0032
0.0131
0.0340
0.0584
0.0737
0.0777
0.0749
0.0715
0.0734
0.0828
0.0852
0.0836
0.0838
0.0955
0.1109
0.1245
0.1302
0.1189
0.0119
0.0383
0.0770
0.1043
0.1131
0.1003
0.0902
0.0836
0.0853
0.1029
0.1183
0.1355
0.1600
0.1726
0.1713
0.1625
0.1335
0.0822
0.0119
0.0265
0.0470
0.0692
0.0933
0.1257
0.1478
0.1494
0.1479
0.1494
0.1545
0.1595
0.1592
UpperCI
0.2780
0.3150
0.3133
0.3033
0.2900
0.2924
0.3101
0.3311
0.3409
0.3423
0.3025
0.2703
0.2741
0.3320
0.4921
0.0402
0.0583
0.0827
0.1150
0.1413
0.1445
0.1383
0.1325
0.1375
0.1566
0.1611
0.1615
0.1568
0.1611
0.1765
0.1916
0.2223
0.2851
0.1222
0.1433
0.1789
0.2214
0.2432
0.2193
0.1987
0.1862
0.1993
0.2358
0.2689
0.2926
0.3172
0.3323
0.3179
0.2879
0.2645
0.2796
0.0639
0.0733
0.0940
0.1284
0.1702
0.2173
0.2387
0.2389
0.2307
0.2322
0.2333
0.2307
0.2329
5C-58
-------
Appendix 5C, Attachment C, Table 5CC-1. Smoothed prevalence for children "EVER" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Age
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
Prevalence
0.1847
0.1797
0.1781
0.1795
0.1838
0.0946
0.1345
0.1759
0.2132
0.2353
0.2638
0.2909
0.3169
0.3272
0.3238
0.3163
0.3022
0.2846
0.2779
0.2702
0.2698
0.2745
0.2843
0.0137
0.0266
0.0453
0.0687
0.0928
0.1142
0.1298
0.1333
0.1231
0.1095
0.1033
0.1086
0.1212
0.1368
0.1437
0.1448
0.1395
0.1283
0.0496
0.0682
0.0893
0.1111
0.1319
0.1473
0.1553
0.1592
0.1650
0.1766
0.1825
0.1805
0.1837
0.1932
0.1891
0.1760
0.1560
0.1298
0.0335
0.0629
0.0985
0.1306
0.1472
SE
0.0172
0.0168
0.0156
0.0162
0.0251
0.0396
0.0296
0.0264
0.0326
0.0361
0.0316
0.0305
0.0339
0.0405
0.0439
0.0429
0.0412
0.0388
0.0367
0.0343
0.0316
0.0349
0.0575
0.0056
0.0064
0.0068
0.0086
0.0112
0.0123
0.0128
0.0123
0.0117
0.0109
0.0102
0.0103
0.0110
0.0113
0.0111
0.0104
0.0113
0.0172
0.0153
0.0123
0.0116
0.0141
0.0171
0.0181
0.0183
0.0183
0.0188
0.0198
0.0216
0.0219
0.0221
0.0218
0.0202
0.0181
0.0195
0.0271
0.0089
0.0093
0.0094
0.0116
0.0133
LowerCI
0.1499
0.1458
0.1465
0.1467
0.1350
0.0365
0.0817
0.1251
0.1503
0.1653
0.2004
0.2287
0.2475
0.2451
0.2356
0.2304
0.2199
0.2074
0.2048
0.2016
0.2062
0.2048
0.1760
0.0056
0.0156
0.0325
0.0522
0.0710
0.0900
0.1042
0.1085
0.0996
0.0877
0.0830
0.0881
0.0991
0.1138
0.1210
0.1235
0.1166
0.0952
0.0250
0.0458
0.0670
0.0838
0.0987
0.1120
0.1193
0.1231
0.1277
0.1374
0.1398
0.1373
0.1401
0.1499
0.1487
0.1398
0.1178
0.0810
0.0186
0.0453
0.0797
0.1073
0.1204
UpperCI
0.2253
0.2195
0.2149
0.2178
0.2452
0.2240
0.2134
0.2416
0.2932
0.3236
0.3388
0.3621
0.3954
0.4214
0.4265
0.4169
0.3995
0.3769
0.3651
0.3518
0.3445
0.3573
0.4250
0.0334
0.0450
0.0629
0.0901
0.1203
0.1439
0.1605
0.1627
0.1512
0.1359
0.1279
0.1332
0.1475
0.1635
0.1699
0.1690
0.1661
0.1709
0.0962
0.1004
0.1181
0.1459
0.1740
0.1914
0.1997
0.2035
0.2104
0.2241
0.2347
0.2336
0.2371
0.2453
0.2374
0.2192
0.2037
0.2015
0.0596
0.0867
0.1212
0.1581
0.1787
5C-59
-------
Appendix 5C, Attachment C, Table 5CC-1. Smoothed prevalence for children "EVER" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Prevalence
0.1523
0.1539
0.1485
0.1461
0.1517
0.1639
0.1772
0.1794
0.1752
0.1705
0.1652
0.1600
0.1562
0.0629
0.0922
0.1253
0.1578
0.1852
0.1975
0.2038
0.2087
0.2078
0.2080
0.2122
0.2137
0.2192
0.2199
0.2059
0.1946
0.1827
0.1709
0.0131
0.0188
0.0264
0.0361
0.0469
0.0647
0.0857
0.1008
0.1032
0.1063
0.1166
0.1181
0.1196
0.1202
0.1241
0.1389
0.1665
0.2118
0.0250
0.0309
0.0387
0.0488
0.0602
0.0843
0.1143
0.1295
0.1195
0.0950
0.0786
0.0812
0.0979
0.1278
0.1324
SE
0.0130
0.0128
0.0125
0.0123
0.0124
0.0129
0.0134
0.0128
0.0127
0.0120
0.0108
0.0118
0.0190
0.0140
0.0118
0.0123
0.0156
0.0186
0.0190
0.0198
0.0204
0.0203
0.0206
0.0203
0.0202
0.0214
0.0220
0.0209
0.0186
0.0177
0.0246
0.0067
0.0057
0.0053
0.0064
0.0083
0.0105
0.0130
0.0144
0.0151
0.0144
0.0140
0.0129
0.0131
0.0130
0.0127
0.0125
0.0152
0.0305
0.0138
0.0099
0.0082
0.0099
0.0129
0.0169
0.0197
0.0191
0.0175
0.0151
0.0139
0.0150
0.0179
0.0221
0.0211
LowerCI
0.1259
0.1278
0.1231
0.1212
0.1265
0.1375
0.1496
0.1530
0.1491
0.1458
0.1428
0.1358
0.1189
0.0383
0.0694
0.1008
0.1265
0.1479
0.1592
0.1639
0.1675
0.1669
0.1664
0.1711
0.1727
0.1759
0.1755
0.1639
0.1571
0.1471
0.1235
0.0042
0.0096
0.0171
0.0245
0.0317
0.0451
0.0611
0.0733
0.0746
0.0786
0.0893
0.0927
0.0938
0.0945
0.0987
0.1136
0.1358
0.1525
0.0073
0.0152
0.0243
0.0312
0.0374
0.0538
0.0776
0.0930
0.0861
0.0666
0.0530
0.0537
0.0651
0.0866
0.0925
UpperCI
0.1831
0.1842
0.1782
0.1752
0.1810
0.1943
0.2085
0.2093
0.2049
0.1984
0.1902
0.1876
0.2026
0.1016
0.1215
0.1547
0.1951
0.2294
0.2424
0.2506
0.2570
0.2558
0.2567
0.2601
0.2612
0.2698
0.2718
0.2554
0.2385
0.2246
0.2317
0.0400
0.0365
0.0407
0.0531
0.0689
0.0919
0.1189
0.1372
0.1412
0.1424
0.1509
0.1494
0.1513
0.1519
0.1548
0.1687
0.2025
0.2864
0.0819
0.0618
0.0612
0.0757
0.0955
0.1296
0.1652
0.1775
0.1636
0.1338
0.1150
0.1209
0.1447
0.1848
0.1859
5C-60
-------
Appendix 5C, Attachment C, Table 5CC-1. Smoothed prevalence for children "EVER" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Prevalence
0.1188
0.0917
0.0600
0.0057
0.0191
0.0479
0.0903
0.1300
0.1437
0.1374
0.1290
0.1365
0.1560
0.1794
0.1980
0.1948
0.1818
0.1771
0.1801
0.1897
0.2081
0.0258
0.0442
0.0700
0.1005
0.1323
0.1609
0.1663
0.1582
0.1536
0.1543
0.1630
0.1746
0.1828
0.1809
0.1800
0.1828
0.1881
0.1964
SE
0.0176
0.0164
0.0186
0.0035
0.0067
0.0092
0.0114
0.0149
0.0158
0.0157
0.0148
0.0148
0.0154
0.0160
0.0175
0.0180
0.0175
0.0164
0.0148
0.0149
0.0248
0.0126
0.0124
0.0119
0.0144
0.0190
0.0218
0.0213
0.0205
0.0204
0.0214
0.0240
0.0270
0.0270
0.0276
0.0259
0.0233
0.0242
0.0396
LowerCI
0.0853
0.0615
0.0300
0.0014
0.0084
0.0306
0.0673
0.0993
0.1110
0.1050
0.0985
0.1058
0.1236
0.1454
0.1608
0.1566
0.1449
0.1423
0.1484
0.1577
0.1567
0.0087
0.0237
0.0479
0.0729
0.0959
0.1186
0.1247
0.1182
0.1140
0.1128
0.1168
0.1230
0.1306
0.1280
0.1298
0.1371
0.1405
0.1234
UpperCI
0.1631
0.1347
0.1163
0.0229
0.0428
0.0743
0.1201
0.1685
0.1842
0.1779
0.1671
0.1743
0.1950
0.2193
0.2413
0.2396
0.2256
0.2183
0.2167
0.2264
0.2709
0.0738
0.0812
0.1013
0.1370
0.1799
0.2147
0.2184
0.2086
0.2040
0.2075
0.2228
0.2420
0.2498
0.2495
0.2440
0.2396
0.2471
0.2978
5C-61
-------
Appendix 5C, Attachment C, Table 5CC-2. Smoothed prevalence for children "STILL" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
Prevalence
0.0082
0.0168
0.0289
0.0420
0.0509
0.0573
0.0611
0.0624
0.0629
0.0663
0.0737
0.0889
0.1056
0.1157
0.1191
0.1177
0.1107
0.0999
0.0381
0.0620
0.0875
0.1079
0.1187
0.1117
0.0940
0.0974
0.1144
0.1237
0.1196
0.1074
0.1025
0.1096
0.1236
0.1412
0.1633
0.1906
0.0122
0.0268
0.0480
0.0710
0.0842
0.0934
0.1056
0.1117
0.1111
0.1138
0.1126
0.1108
0.1129
0.1139
0.1128
0.1054
0.0935
0.0782
0.0402
0.0824
0.1338
0.1774
0.1949
0.1867
0.1807
0.1734
0.1739
0.1814
SE
0.0051
0.0064
0.0070
0.0086
0.0103
0.0108
0.0109
0.0107
0.0100
0.0096
0.0108
0.0126
0.0151
0.0163
0.0160
0.0144
0.0143
0.0205
0.0164
0.0160
0.0160
0.0183
0.0202
0.0194
0.0188
0.0187
0.0205
0.0220
0.0237
0.0225
0.0199
0.0211
0.0229
0.0266
0.0413
0.0779
0.0064
0.0083
0.0091
0.0113
0.0134
0.0138
0.0144
0.0149
0.0152
0.0155
0.0153
0.0146
0.0137
0.0132
0.0127
0.0118
0.0133
0.0184
0.0177
0.0213
0.0225
0.0255
0.0267
0.0237
0.0222
0.0221
0.0248
0.0269
LowerCI
0.0021
0.0073
0.0169
0.0267
0.0326
0.0378
0.0412
0.0427
0.0443
0.0481
0.0533
0.0649
0.0768
0.0845
0.0882
0.0896
0.0831
0.0632
0.0146
0.0349
0.0581
0.0738
0.0811
0.0758
0.0602
0.0634
0.0765
0.0830
0.0766
0.0672
0.0664
0.0712
0.0815
0.0924
0.0914
0.0722
0.0038
0.0135
0.0315
0.0500
0.0591
0.0673
0.0779
0.0829
0.0820
0.0840
0.0831
0.0826
0.0861
0.0880
0.0878
0.0822
0.0682
0.0462
0.0151
0.0463
0.0917
0.1282
0.1429
0.1402
0.1371
0.1301
0.1260
0.1297
UpperCI
0.0319
0.0382
0.0490
0.0655
0.0788
0.0859
0.0897
0.0902
0.0886
0.0907
0.1012
0.1206
0.1435
0.1565
0.1588
0.1530
0.1461
0.1544
0.0956
0.1076
0.1295
0.1550
0.1704
0.1616
0.1439
0.1469
0.1676
0.1805
0.1821
0.1673
0.1551
0.1649
0.1830
0.2099
0.2746
0.4158
0.0384
0.0525
0.0725
0.1001
0.1187
0.1282
0.1416
0.1489
0.1489
0.1525
0.1507
0.1472
0.1466
0.1462
0.1438
0.1343
0.1269
0.1292
0.1028
0.1425
0.1911
0.2401
0.2601
0.2443
0.2344
0.2273
0.2350
0.2478
5C-62
-------
Appendix 5C, Attachment C, Table 5CC-2. Smoothed prevalence for children "STILL" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Age
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
Prevalence
0.1813
0.1749
0.1702
0.1499
0.1366
0.1484
0.1846
0.2590
0.0153
0.0281
0.0437
0.0584
0.0657
0.0668
0.0678
0.0696
0.0737
0.0840
0.0807
0.0710
0.0629
0.0680
0.0786
0.0913
0.1095
0.1328
0.0234
0.0564
0.1040
0.1466
0.1618
0.1441
0.1124
0.0751
0.0633
0.0838
0.1288
0.1778
0.2073
0.2063
0.1929
0.1703
0.1414
0.1108
0.0225
0.0368
0.0562
0.0797
0.1035
0.1289
0.1472
0.1423
0.1290
0.1251
0.1288
0.1262
0.1246
0.1230
0.1207
0.1114
0.0983
0.0823
0.0930
0.1202
SE
0.0282
0.0282
0.0298
0.0296
0.0269
0.0268
0.0359
0.0740
0.0089
0.0096
0.0090
0.0098
0.0112
0.0111
0.0111
0.0114
0.0124
0.0147
0.0144
0.0134
0.0116
0.0113
0.0117
0.0120
0.0165
0.0330
0.0142
0.0190
0.0219
0.0272
0.0304
0.0280
0.0238
0.0174
0.0157
0.0188
0.0270
0.0336
0.0349
0.0328
0.0287
0.0235
0.0234
0.0327
0.0108
0.0105
0.0104
0.0127
0.0162
0.0187
0.0190
0.0181
0.0163
0.0159
0.0155
0.0139
0.0139
0.0149
0.0144
0.0126
0.0124
0.0171
0.0402
0.0280
LowerCI
0.1275
0.1214
0.1143
0.0959
0.0876
0.0987
0.1185
0.1306
0.0042
0.0132
0.0276
0.0402
0.0449
0.0461
0.0471
0.0482
0.0506
0.0569
0.0541
0.0466
0.0416
0.0469
0.0564
0.0681
0.0781
0.0753
0.0061
0.0266
0.0648
0.0964
0.1056
0.0928
0.0698
0.0447
0.0364
0.0507
0.0802
0.1154
0.1410
0.1435
0.1375
0.1248
0.0974
0.0567
0.0078
0.0195
0.0373
0.0559
0.0730
0.0931
0.1102
0.1070
0.0973
0.0943
0.0985
0.0989
0.0971
0.0939
0.0925
0.0868
0.0743
0.0518
0.0347
0.0710
UpperCI
0.2514
0.2454
0.2457
0.2268
0.2066
0.2169
0.2761
0.4484
0.0537
0.0589
0.0683
0.0840
0.0950
0.0958
0.0967
0.0993
0.1062
0.1224
0.1187
0.1068
0.0938
0.0976
0.1085
0.1214
0.1513
0.2234
0.0856
0.1157
0.1627
0.2167
0.2400
0.2168
0.1761
0.1234
0.1078
0.1355
0.2004
0.2638
0.2941
0.2873
0.2637
0.2281
0.2009
0.2051
0.0633
0.0682
0.0838
0.1123
0.1449
0.1757
0.1938
0.1868
0.1690
0.1641
0.1668
0.1598
0.1584
0.1594
0.1560
0.1420
0.1291
0.1285
0.2262
0.1964
5C-63
-------
Appendix 5C, Attachment C, Table 5CC-2. Smoothed prevalence for children "STILL" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Age
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
Prevalence
0.1475
0.1714
0.1860
0.2060
0.2256
0.2496
0.2727
0.2579
0.2318
0.1902
0.1624
0.1641
0.1699
0.1797
0.1933
0.2097
0.0131
0.0228
0.0352
0.0495
0.0633
0.0740
0.0826
0.0888
0.0860
0.0791
0.0747
0.0736
0.0776
0.0851
0.0871
0.0876
0.0859
0.0819
0.0396
0.0573
0.0772
0.0963
0.1120
0.1206
0.1219
0.1152
0.1131
0.1190
0.1208
0.1195
0.1275
0.1405
0.1394
0.1296
0.1136
0.0923
0.0228
0.0476
0.0793
0.1076
0.1193
0.1194
0.1145
0.1071
0.1011
0.1000
0.1059
0.1122
SE
0.0256
0.0311
0.0335
0.0276
0.0276
0.0317
0.0387
0.0395
0.0366
0.0310
0.0268
0.0254
0.0251
0.0244
0.0276
0.0451
0.0059
0.0063
0.0064
0.0074
0.0089
0.0092
0.0096
0.0099
0.0100
0.0095
0.0088
0.0085
0.0087
0.0093
0.0093
0.0087
0.0091
0.0136
0.0135
0.0113
0.0109
0.0136
0.0165
0.0174
0.0173
0.0162
0.0157
0.0161
0.0175
0.0178
0.0192
0.0197
0.0184
0.0166
0.0184
0.0249
0.0070
0.0082
0.0089
0.0109
0.0123
0.0117
0.0111
0.0105
0.0099
0.0098
0.0102
0.0106
LowerCI
0.0997
0.1134
0.1232
0.1519
0.1708
0.1866
0.1964
0.1810
0.1611
0.1311
0.1116
0.1155
0.1216
0.1321
0.1397
0.1274
0.0048
0.0124
0.0236
0.0355
0.0464
0.0561
0.0638
0.0695
0.0666
0.0606
0.0576
0.0570
0.0606
0.0669
0.0688
0.0702
0.0681
0.0567
0.0186
0.0371
0.0564
0.0704
0.0805
0.0874
0.0888
0.0842
0.0829
0.0880
0.0874
0.0857
0.0910
0.1026
0.1037
0.0973
0.0791
0.0503
0.0116
0.0325
0.0619
0.0859
0.0949
0.0960
0.0924
0.0861
0.0813
0.0806
0.0855
0.0910
UpperCI
0.2130
0.2508
0.2708
0.2732
0.2919
0.3255
0.3653
0.3535
0.3216
0.2678
0.2302
0.2278
0.2323
0.2396
0.2612
0.3253
0.0349
0.0415
0.0522
0.0685
0.0857
0.0969
0.1063
0.1129
0.1105
0.1025
0.0963
0.0944
0.0989
0.1078
0.1099
0.1087
0.1080
0.1169
0.0823
0.0876
0.1048
0.1306
0.1536
0.1641
0.1652
0.1556
0.1524
0.1591
0.1646
0.1642
0.1757
0.1893
0.1848
0.1706
0.1605
0.1634
0.0443
0.0693
0.1011
0.1341
0.1490
0.1475
0.1411
0.1323
0.1251
0.1236
0.1305
0.1376
5C-64
-------
Appendix 5C, Attachment C, Table 5CC-2. Smoothed prevalence for children "STILL" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Age
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
Prevalence
0.1103
0.1052
0.0983
0.0899
0.0811
0.0727
0.0499
0.0749
0.1033
0.1305
0.1519
0.1595
0.1598
0.1540
0.1466
0.1457
0.1504
0.1508
0.1506
0.1470
0.1345
0.1215
0.1080
0.0948
0.0077
0.0122
0.0181
0.0248
0.0305
0.0382
0.0482
0.0573
0.0628
0.0697
0.0768
0.0786
0.0808
0.0829
0.0845
0.0908
0.1016
0.1180
0.0244
0.0270
0.0306
0.0354
0.0407
0.0577
0.0807
0.0954
0.0876
0.0648
0.0495
0.0473
0.0606
0.0845
0.0931
0.0846
0.0629
0.0376
0.0007
0.0052
0.0225
0.0596
SE
0.0105
0.0105
0.0094
0.0081
0.0089
0.0136
0.0126
0.0110
0.0116
0.0149
0.0177
0.0180
0.0185
0.0180
0.0170
0.0170
0.0171
0.0171
0.0184
0.0192
0.0179
0.0159
0.0164
0.0227
0.0049
0.0046
0.0045
0.0055
0.0068
0.0077
0.0091
0.0098
0.0106
0.0106
0.0099
0.0094
0.0100
0.0108
0.0111
0.0110
0.0129
0.0236
0.0144
0.0091
0.0074
0.0090
0.0112
0.0146
0.0185
0.0181
0.0159
0.0127
0.0107
0.0110
0.0137
0.0179
0.0180
0.0154
0.0143
0.0146
0.0007
0.0027
0.0063
0.0095
LowerCI
0.0893
0.0843
0.0795
0.0737
0.0636
0.0479
0.0285
0.0542
0.0805
0.1012
0.1171
0.1240
0.1234
0.1186
0.1130
0.1122
0.1167
0.1171
0.1146
0.1097
0.0999
0.0907
0.0770
0.0555
0.0019
0.0053
0.0105
0.0153
0.0186
0.0245
0.0318
0.0393
0.0432
0.0497
0.0577
0.0603
0.0615
0.0621
0.0632
0.0694
0.0766
0.0753
0.0066
0.0128
0.0179
0.0201
0.0221
0.0328
0.0483
0.0624
0.0583
0.0419
0.0306
0.0282
0.0366
0.0526
0.0603
0.0562
0.0379
0.0158
0.0001
0.0014
0.0112
0.0398
UpperCI
0.1356
0.1305
0.1210
0.1093
0.1028
0.1089
0.0860
0.1027
0.1316
0.1666
0.1948
0.2029
0.2045
0.1977
0.1879
0.1870
0.1917
0.1921
0.1955
0.1943
0.1788
0.1607
0.1494
0.1573
0.0306
0.0278
0.0310
0.0401
0.0494
0.0590
0.0724
0.0829
0.0904
0.0970
0.1016
0.1018
0.1056
0.1100
0.1121
0.1179
0.1337
0.1803
0.0862
0.0561
0.0518
0.0615
0.0738
0.0996
0.1319
0.1434
0.1296
0.0989
0.0792
0.0781
0.0988
0.1329
0.1411
0.1253
0.1026
0.0868
0.0067
0.0192
0.0447
0.0884
5C-65
-------
Appendix 5C, Attachment C, Table 5CC-2. Smoothed prevalence for children "STILL" having asthma.
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Prevalence
0.0989
0.1070
0.0959
0.0830
0.0877
0.1029
0.1189
0.1292
0.1214
0.1050
0.0981
0.0997
0.1091
0.1290
0.0263
0.0374
0.0518
0.0681
0.0871
0.1074
0.1167
0.1138
0.1073
0.0964
0.0830
0.0745
0.0825
0.1000
0.1074
0.1120
0.1127
0.1084
SE
0.0140
0.0147
0.0141
0.0126
0.0124
0.0135
0.0140
0.0153
0.0154
0.0139
0.0127
0.0116
0.0128
0.0231
0.0130
0.0101
0.0086
0.0105
0.0143
0.0173
0.0183
0.0186
0.0177
0.0164
0.0149
0.0151
0.0165
0.0197
0.0200
0.0193
0.0222
0.0340
LowerCI
0.0691
0.0754
0.0660
0.0565
0.0613
0.0737
0.0883
0.0955
0.0879
0.0749
0.0707
0.0742
0.0810
0.0814
0.0088
0.0204
0.0358
0.0483
0.0604
0.0749
0.0820
0.0789
0.0741
0.0659
0.0557
0.0474
0.0527
0.0643
0.0707
0.0760
0.0724
0.0531
UpperCI
0.1397
0.1496
0.1372
0.1203
0.1239
0.1419
0.1584
0.1724
0.1653
0.1452
0.1346
0.1327
0.1454
0.1984
0.0761
0.0673
0.0742
0.0952
0.1240
0.1517
0.1635
0.1615
0.1529
0.1389
0.1221
0.1152
0.1268
0.1524
0.1600
0.1620
0.1714
0.2088
5C-66
-------
Appendix 5C, Attachment C, Table 5CC-3. Smoothed prevalence for adults "EVER" having asthma
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Age_group
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
Prevalence
0.1642
0.1341
0.1193
0.1204
0.1246
0.1165
0.0980
0.2014
0.1812
0.1782
0.2104
0.2295
0.1892
0.1176
0.1705
0.1209
0.0886
0.0727
0.0770
0.0828
0.0847
0.1654
0.1143
0.1066
0.1376
0.1643
0.1396
0.0853
0.1791
0.1423
0.1256
0.1246
0.1281
0.1151
0.0879
0.1646
0.1705
0.1842
0.2084
0.2180
0.1695
0.0960
0.1728
0.1163
0.0932
0.0901
0.0963
0.0874
0.0708
0.1734
0.1323
0.1182
0.1254
0.1361
0.1305
0.0988
0.1533
0.1235
0.1114
0.1149
0.1261
0.1188
0.0959
0.1491
SE
0.0141
0.0063
0.0058
0.0057
0.0066
0.0062
0.0089
0.0153
0.0114
0.0130
0.0146
0.0164
0.0145
0.0173
0.0149
0.0063
0.0053
0.0046
0.0054
0.0058
0.0106
0.0175
0.0109
0.0122
0.0146
0.0164
0.0160
0.0205
0.0176
0.0076
0.0072
0.0071
0.0076
0.0070
0.0098
0.0182
0.0110
0.0126
0.0143
0.0156
0.0118
0.0125
0.0210
0.0081
0.0070
0.0063
0.0072
0.0073
0.0118
0.0193
0.0138
0.0135
0.0144
0.0198
0.0195
0.0255
0.0114
0.0054
0.0050
0.0047
0.0058
0.0058
0.0087
0.0122
LowerCI
0.1219
0.1142
0.1012
0.1025
0.1040
0.0971
0.0719
0.1531
0.1445
0.1370
0.1638
0.1770
0.1435
0.0690
0.1249
0.1008
0.0719
0.0583
0.0602
0.0647
0.0545
0.1122
0.0808
0.0703
0.0936
0.1141
0.0918
0.0353
0.1265
0.1183
0.1029
0.1024
0.1043
0.0934
0.0598
0.1104
0.1356
0.1442
0.1629
0.1684
0.1321
0.0603
0.1126
0.0914
0.0721
0.0710
0.0744
0.0656
0.0398
0.1138
0.0896
0.0772
0.0816
0.0786
0.0743
0.0373
0.1185
0.1065
0.0956
0.0998
0.1077
0.1004
0.0701
0.1107
UpperCI
0.2176
0.1568
0.1402
0.1409
0.1486
0.1392
0.1322
0.2603
0.2248
0.2284
0.2662
0.2920
0.2453
0.1933
0.2284
0.1444
0.1087
0.0904
0.0980
0.1053
0.1292
0.2370
0.1593
0.1585
0.1979
0.2309
0.2068
0.1920
0.2474
0.1701
0.1525
0.1509
0.1565
0.1412
0.1273
0.2383
0.2123
0.2323
0.2627
0.2773
0.2149
0.1495
0.2560
0.1469
0.1197
0.1139
0.1237
0.1155
0.1229
0.2552
0.1911
0.1768
0.1879
0.2253
0.2191
0.2366
0.1959
0.1429
0.1295
0.1320
0.1472
0.1400
0.1297
0.1978
5C-67
-------
Appendix 5C, Attachment C, Table 5CC-3. Smoothed prevalence for adults "EVER" having asthma
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age_group
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
18-24
25-34
35-44
45-54
55-64
65-74
75+
Prevalence
0.1365
0.1414
0.1686
0.1881
0.1651
0.1125
0.1445
0.1086
0.0860
0.0742
0.0733
0.0790
0.0900
0.1433
0.1031
0.0934
0.1055
0.1072
0.0942
0.0712
0.1571
0.1415
0.1373
0.1423
0.1497
0.1445
0.1266
0.1434
0.1318
0.1440
0.1806
0.1713
0.1511
0.1292
0.1566
0.1233
0.1025
0.0908
0.0955
0.1067
0.1265
0.1521
0.0942
0.0885
0.1133
0.1237
0.1134
0.0961
SE
0.0066
0.0078
0.0097
0.0115
0.0101
0.0124
0.0095
0.0050
0.0044
0.0040
0.0045
0.0048
0.0102
0.0144
0.0087
0.0090
0.0101
0.0108
0.0092
0.0123
0.0135
0.0067
0.0070
0.0067
0.0071
0.0070
0.0112
0.0164
0.0092
0.0117
0.0144
0.0136
0.0117
0.0177
0.0173
0.0069
0.0060
0.0054
0.0059
0.0068
0.0152
0.0204
0.0095
0.0102
0.0130
0.0156
0.0142
0.0190
LowerCI
0.1149
0.1159
0.1369
0.1505
0.1325
0.0755
0.1147
0.0926
0.0720
0.0616
0.0594
0.0639
0.0606
0.1000
0.0766
0.0664
0.0751
0.0750
0.0666
0.0385
0.1163
0.1201
0.1150
0.1207
0.1268
0.1220
0.0929
0.0945
0.1026
0.1074
0.1350
0.1284
0.1141
0.0785
0.1067
0.1019
0.0839
0.0741
0.0774
0.0860
0.0834
0.0938
0.0660
0.0590
0.0753
0.0789
0.0726
0.0474
UpperCI
0.1614
0.1714
0.2059
0.2324
0.2039
0.1644
0.1805
0.1269
0.1025
0.0891
0.0902
0.0974
0.1316
0.2013
0.1376
0.1300
0.1462
0.1510
0.1314
0.1279
0.2089
0.1660
0.1631
0.1670
0.1758
0.1704
0.1702
0.2117
0.1678
0.1903
0.2374
0.2248
0.1974
0.2054
0.2240
0.1485
0.1247
0.1107
0.1174
0.1318
0.1871
0.2373
0.1327
0.1308
0.1670
0.1888
0.1727
0.1849
5C-68
-------
Appendix 5C, Attachment C, Table 5CC-4. Smoothed prevalence for adults "STILL" having asthma
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Midwest
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
Northeast
South
South
South
South
South
South
South
South
Gender
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Poverty Status
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Age_group
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
Prevalence
0.1046
0.0888
0.0835
0.0893
0.0909
0.0811
0.0630
0.1327
0.1280
0.1315
0.1600
0.1777
0.1488
0.0940
0.0807
0.0584
0.0479
0.0472
0.0522
0.0528
0.0481
0.0912
0.0683
0.0694
0.1015
0.1338
0.1202
0.0709
0.1098
0.0965
0.0899
0.0901
0.0917
0.0862
0.0726
0.1212
0.1199
0.1338
0.1655
0.1824
0.1273
0.0529
0.0922
0.0600
0.0488
0.0483
0.0563
0.0576
0.0554
0.0791
0.0800
0.0805
0.0857
0.1064
0.1040
0.0771
0.0891
0.0735
0.0684
0.0732
0.0846
0.0817
0.0641
0.0948
SE
0.0121
0.0057
0.0052
0.0050
0.0057
0.0051
0.0067
0.0139
0.0095
0.0114
0.0134
0.0146
0.0128
0.0157
0.0115
0.0045
0.0040
0.0038
0.0042
0.0045
0.0081
0.0136
0.0091
0.0109
0.0141
0.0165
0.0161
0.0210
0.0134
0.0065
0.0063
0.0060
0.0062
0.0059
0.0093
0.0166
0.0093
0.0106
0.0127
0.0143
0.0098
0.0086
0.0154
0.0058
0.0050
0.0051
0.0065
0.0063
0.0106
0.0128
0.0119
0.0135
0.0162
0.0224
0.0200
0.0236
0.0083
0.0039
0.0036
0.0037
0.0046
0.0047
0.0070
0.0105
LowerCI
0.0703
0.0714
0.0675
0.0738
0.0736
0.0654
0.0438
0.0907
0.0980
0.0961
0.1181
0.1318
0.1091
0.0513
0.0491
0.0448
0.0359
0.0358
0.0395
0.0393
0.0268
0.0542
0.0430
0.0402
0.0624
0.0866
0.0751
0.0250
0.0721
0.0765
0.0708
0.0718
0.0727
0.0681
0.0467
0.0744
0.0914
0.1013
0.1260
0.1381
0.0972
0.0300
0.0509
0.0428
0.0340
0.0334
0.0376
0.0393
0.0281
0.0430
0.0459
0.0427
0.0419
0.0475
0.0501
0.0241
0.0649
0.0615
0.0571
0.0617
0.0705
0.0674
0.0443
0.0641
UpperCI
0.1528
0.1100
0.1030
0.1077
0.1118
0.1002
0.0898
0.1899
0.1656
0.1772
0.2132
0.2352
0.1998
0.1659
0.1299
0.0758
0.0637
0.0620
0.0687
0.0706
0.0847
0.1496
0.1067
0.1173
0.1610
0.2010
0.1869
0.1850
0.1638
0.1210
0.1136
0.1124
0.1151
0.1085
0.1110
0.1915
0.1559
0.1747
0.2143
0.2370
0.1650
0.0917
0.1616
0.0836
0.0696
0.0693
0.0834
0.0837
0.1062
0.1409
0.1360
0.1465
0.1672
0.2211
0.2035
0.2203
0.1212
0.0876
0.0817
0.0866
0.1012
0.0987
0.0920
0.1380
5C-69
-------
Appendix 5C, Attachment C, Table 5CC-4. Smoothed prevalence for adults "STILL" having asthma
Smoothed
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Region
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
South
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
West
Gender
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Male
Poverty Status
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Above Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Below Poverty Level
Age_group
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
18-24
25-34
35-44
45-44
55-64
65-74
75+
Prevalence
0.0942
0.1086
0.1446
0.1618
0.1379
0.0881
0.0600
0.0490
0.0421
0.0386
0.0384
0.0457
0.0627
0.0583
0.0443
0.0492
0.0720
0.0771
0.0608
0.0353
0.0842
0.0876
0.0931
0.0981
0.1028
0.0984
0.0825
0.0863
0.0934
0.1091
0.1332
0.1292
0.1169
0.1021
0.0597
0.0569
0.0549
0.0525
0.0562
0.0660
0.0783
0.0720
0.0484
0.0539
0.0784
0.0936
0.0758
0.0489
SE
0.0059
0.0073
0.0095
0.0112
0.0095
0.0109
0.0073
0.0035
0.0033
0.0031
0.0034
0.0038
0.0089
0.0080
0.0053
0.0067
0.0090
0.0096
0.0075
0.0082
0.0115
0.0054
0.0062
0.0065
0.0067
0.0061
0.0090
0.0121
0.0078
0.0100
0.0120
0.0120
0.0104
0.0148
0.0092
0.0046
0.0045
0.0046
0.0053
0.0058
0.0131
0.0125
0.0068
0.0084
0.0115
0.0155
0.0129
0.0136
LowerCI
0.0758
0.0859
0.1149
0.1267
0.1082
0.0570
0.0392
0.0381
0.0322
0.0292
0.0282
0.0343
0.0382
0.0358
0.0290
0.0303
0.0460
0.0492
0.0390
0.0154
0.0522
0.0708
0.0742
0.0781
0.0820
0.0795
0.0565
0.0524
0.0695
0.0789
0.0967
0.0929
0.0854
0.0609
0.0351
0.0432
0.0414
0.0389
0.0407
0.0487
0.0437
0.0389
0.0294
0.0311
0.0465
0.0517
0.0413
0.0182
UpperCI
0.1166
0.1365
0.1806
0.2043
0.1742
0.1337
0.0907
0.0629
0.0550
0.0510
0.0520
0.0607
0.1013
0.0937
0.0672
0.0790
0.1112
0.1188
0.0937
0.0787
0.1328
0.1080
0.1163
0.1226
0.1281
0.1213
0.1189
0.1387
0.1243
0.1489
0.1806
0.1770
0.1580
0.1662
0.0998
0.0745
0.0723
0.0704
0.0770
0.0889
0.1364
0.1295
0.0787
0.0919
0.1293
0.1635
0.1350
0.1250
5C-70
-------
rjT-n-r-p-T-n-r-n-r-r-p
O 123456789 1O 11 1213M15 16 17 O 123456789 1O 11 1213M15 16 17
Rmfe Tvfle
g^rcfer ^~^~ lende
0123456789 1O 11 12133435 16 17 0123456789 1O 11 12131415 16 17
Rrate
Appendix 5C, Attachment C, Figure 5CC-1. Smoothed prevalence and confidence intervals for children 'EVER' having asthma.
5C-71
-------
O 1 23 4 5 6 7891O1112B14K1617
gprcfer
Hmfe
JVfls
O 123456789 1O 11 1213M15 16 17
Rrate
]Vfle
O 1 23 45 6 7891O11121314151617
gprctr
Hmis
Nfle
O 1 23 45 6 789 10 11 1213 14 15 16 17
Hmis
Nfle
Appendix 5C, Attachment C, Figure 5CC-1, cont. Smoothed prevalence and confidence intervals for children 'EVER' having
asthma.
5C-72
-------
0123456789 10 11 1213M15 16 17
0123456789 10 11 1233M15 16 17
^prefer Rirrie
gp-rfe- Rirrie
0123456789 10 11 1233M15 16 17
gp-rfe- Rirrie
pev
Q6-
Q5-
Q4
Q3-
Q2-
Ql-
QO-
O 1 23 45 6 7891O1112B14151S17
Hmfe JVfle
Appendix 5C, Attachment C, Figure 5CC-2. Smoothed prevalence and confidence intervals for children 'STILL' having asthma.
5C-73
-------
Q6:
05-
04-
03-
02-
Q1:
QO-
O 1 23 45 6 7891O11121314151617
gprda-
Rmte ]Vfle
Q6-
Q5-
Q4-
Q3:
O2-
Ql
QO-
O 1 23 4 5 6 7891O1112BM151617
agP
Hmfe JVfls
H3»e2. SriJotheclasttiTn'SllLL!,
Q6:
Q5-
Q4^
Q3:
O2:
Qi:
QO:
4^
O 123456789 1O 11 1213M15 16 17
O 123456789 1O 11 12 13 14 15 16 17
o^nrfcr- Knde
]\fle
Kirrie - Mie
Appendix 5C, Attachment C, Figure 5CC-2, cont. Smoothed prevalence and confidence intervals for children 'STILL' having
asthma.
5C-74
-------
Q3-
Q2-
01-
Q3-
Ql-
18-24
gprctr
25-34-
3544
1 ' ' I ' ' '
45-54
3&L3P
Nfle
Nfle
Q3-
02-
ai-
03:
Q2:
QO
K-24
gprctr
Rmfe
]Vfle
2534
3544
45-54
Rrafe
Appendix 5C, Attachment C, Figure 5CC-3. Smoothed prevalence and confidence intervals for Adults 'EVER' having asthma.
5C-75
-------
Rrafe
Rrafe
ptev
O4:
ptev
Q4-
Hmfe
3544 45-54
gpncfer
Hmfe
JVfle
JVfle
Appendix 5C, Attachment C, Figure 5CC-3, cont. Smoothed prevalence and confidence intervals for Adults 'EVER' having
asthma.
5C-76
-------
03-
Q2-
Ql-
Ql-
18-24
25-34-
3544
~
45-54
3&L3P
Nfle
Nfle
03-
02-
Ql-
03:
Q2:
QO
K-24
3544 45-54
gprctr
Hmis
55St 65-74
Nfle
2534
3544
45-54
Rrafe
Appendix 5C, Attachment C, Figure 5CC-4. Smoothed prevalence and confidence intervals for Adults 'STILL' having asthma.
5C-77
-------
ps^
Q4:
02-
"f
Ql
18-24
gprctr
25-34-
3544
45-54
3&L3P
55St 65-74
Nfle
Nfle
Q4-
03-
02-
01-
oo-
1
f
F=
£-24
1
1
2534
T J J T 1
— i 1 1 ^ ^
3544 45-54 5B-& ffi-^4 ^5
Q4:
gprctr
Hmis
Nfle
gpncfer
Hmfe
JVfls
Appendix 5C, Attachment C, Figure 5CC-4, cont. Smoothed prevalence and confidence intervals for Adults 'STILL' having
asthma.
5C-78
-------
APPENDIX 5D
Variability Analysis and Uncertainty Characterization
Table of Contents
5D-1. OVERVIEW 5D-1
5D-2. TREATMENT OF VARIABILITY AND CO-VARIABILITY 5D-1
5D-3. CHARACTERIZATION OF UNCERTAINTY 5D-7
5D-4. REFERENCES 5D-10
5D-i
-------
List of Tables
Table 5D-1. Components of exposure variability modeled by APEX 5D-4
Table 5D-2. Important components of co-variability in exposure modeling 5D-6
5D-ii
-------
5D-1. OVERVIEW
An important issue associated with any population exposure or risk assessment is the
characterization of variability and uncertainty. Variability refers to the inherent heterogeneity in
a population or variable of interest (e.g., residential air exchange rates). The degree of variability
cannot be reduced through further research, only better characterized with additional
measurement. Uncertainty refers to the lack of knowledge regarding the values of model input
variables (i.e., parameter uncertainty)., the physical systems or relationships used (i.e., use of
input variables to estimate exposure or risk or model uncertainty), and in specifying the scenario
that is consistent with purpose of the assessment (i.e., scenario uncertainty). Uncertainty is,
ideally, reduced to the maximum extent possible through improved measurement of key
parameters and iterative model refinement. The approaches used to assess variability and to
characterize uncertainty in this HREA are discussed in the following two sections. The primary
purpose of this characterization is to provide a summary of variability and uncertainty
evaluations conducted to date regarding our Os exposure assessments and APEX exposure
modeling and to identify the most important elements of uncertainty in need of further
characterization. Each section contains a concise tabular summary of the identified components
and how, for elements of uncertainty, each source may affect the estimated exposures.
5D-2. TREATMENT OF VARIABILITY AND CO-VARIABILITY
The purpose for addressing variability in this HREA is to ensure that the estimates of
exposure and risk reflect the variability of ambient Os concentrations, population characteristics,
associated Os exposure and intake dose, and potential health risk across the study area and for
the simulated at-risk populations. In this HREA, there are several algorithms that account for
variability of input data when generating the number of estimated benchmark exceedances or
health risk outputs. For example, variability may arise from differences in the population
residing within census tracts (e.g., age distribution) and the activities that may affect population
exposure to Cb and the resulting intake dose estimate (e.g., time spent outdoors, performing
moderate or greater exertion level activities outdoors). A complete range of potential exposure
levels and associated risk estimates can be generated when appropriately addressing variability in
exposure and risk assessments; note however that the range of values obtained would be within
the constraints of the input parameters, algorithms, or modeling system used, not necessarily the
complete range of the true exposure or risk values.
Where possible, staff identified and incorporated the observed variability in input data
sets rather than employing standard default assumptions and/or using point estimates to describe
model inputs. The details regarding variability distributions used in data inputs are described in
5D-1
-------
Appendix 5B, while details regarding the variability addressed within its algorithms and
processes are found in the APEX TSD (US EPA, 2012).
Briefly, APEX has been designed to account for variability in most of the input data,
including the physiological variables that are important inputs to determining exertion levels and
associated ventilation rates. APEX simulates individuals and then calculates Os exposures for
each of these simulated individuals. The individuals are selected to represent a random sample
from a defined population. The collection of individuals represents the variability of the target
population, and accounts for several types of variability, including demographic, physiological,
and human behavior. In this assessment, we simulated 200,000 individuals to reasonably capture
the variability expected in the population exposure distribution for each study area. APEX
incorporates stochastic processes representing the natural variability of personal profile
characteristics, activity patterns, and microenvironment parameters. In this way, APEX is able
to represent much of the variability in the exposure estimates resulting from the variability of the
factors effecting human exposure.
We note also that correlations and non-linear relationships between variables input to the
model can result in the model producing incorrect results if the inherent relationships between
these variables are not preserved. That is why APEX is also designed to account for co-
variability, or linear and nonlinear correlation among the model inputs, provided that enough is
known about these relationships to specify them. This is accomplished by providing inputs that
enable the correlation to be modeled explicitly within APEX. For example, there is a non-linear
relationship between the outdoor temperature and air exchange rate in homes. One factor that
contributes to this non-linear relationship is that windows tend to be closed more often when
temperatures are at either low or high extremes than when temperatures are moderate. This
relationship is explicitly modeled in APEX by specifying different probability distributions of air
exchange rates for different ambient temperatures. In any event, APEX models variability and
co-variability in two ways:
• Stochastically. The user provides APEX with probability distributions
characterizing the variability of many input parameters. These are treated
stochastically in the model and the estimated exposure distributions reflect this
variability. For example, the rate of Os removal in houses can depend on a
number of factors which we are not able to explicitly model at this time, due to a
lack of data. However, we can specify a distribution of removal rates which
reflects observed variations in Os decay. APEX randomly samples from this
distribution to obtain values which are used in the mass balance model. Further,
co-variability can be modeled stochastically through the use of conditional
distributions. If two or more parameters are related, conditional distributions that
5D-2
-------
depend on the values of the related parameters are input to APEX. For example,
the distribution of air exchange rates (AERs) in a house depends on the outdoor
temperature and whether or not air conditioning (A/C) is in use. In this case, a set
of AER distributions is provided to APEX for different ranges of temperatures
and A/C use, and the selection of the distribution in APEX is driven by the
temperature and A/C status at that time. The spatial variability of A/C prevalence
is modeled by supplying APEX with A/C prevalence for each Census tract in the
modeled area.
• Explicitly. For some variables used in modeling exposure, APEX models
variability and co-variability explicitly and not stochastically. For example,
hourly-average ambient Os concentrations and temperatures are used in model
calculations. These are input to the model for every hour in the time period
modeled at different spatial locations, and in this way the variability and co-
variability of hourly concentrations and temperatures are modeled explicitly.
Important sources of the variability and co-variability accounted for by APEX and used
for this exposure analysis are summarized in Table 5D-1 and Table 5D-2 below, respectively.
5D-3
-------
Table 5D-1. Components of exposure variability modeled by APEX.
Component
Variability Source
Comment
Population data
Individuals are randomly sampled from US census tracts
used in each model study area, stratified by age (single
years), gender, and employment status probability
distributions (US Census Bureau, 2007a).
Commuting data
Employed individuals are probabilistically assigned ambient
concentrations originating from either their home or work tract
based on US Census derived commuter data (US Census
Bureau, 2007a).
Simulated
Individuals
Activity patterns
Data diaries are randomly selected from CHAD master
(>38,000 diaries) using six diary pools stratified by two day-
types (weekday, weekend) and three temperature ranges (<
55.0 °F, between 55.0 and 83.9°F, and >84.0 °F). The CHAD
diaries capture real locations that people visit and the
activities they perform, ranging from 1 minute to 1 hour in
duration (US EPA, 2002).
Longitudinal profiles
A sequence of diaries is linked together for each individual
that preserves both the inter- and intra-personal variability in
human activities (Glen et al., 2008).
Asthma prevalence
Asthma prevalence is stratified by two genders, single age
years (0-17), seven age groups, (18-24, 25-34, 35-44, 45-54,
55-64, 65-74, and, >75), four regions (Midwest, Northeast,
South, and West), and US census tract level poverty ratios
(CDC, 2011; US Census Bureau, 2007b).
Measured ambient Os
concentrations
Ambient Input
Temporal: 1-hour concentrations for an entire Os season or
year predicted using ambient monitoring data.
Spatial: Several monitors are used to represent ambient
conditions within each study area; each monitor was assigned
a 30 km zone of influence, though value from closest monitor
is used for each tract. Four US study areas assess regional
differences in ambient conditions.
Meteorological data
Spatial: Values from closest available local surface National
Weather Service (NWS) station were used.
Temporal: 1-hour temperature data input for each year; daily
values calculated by APEX.
Microenvironmental
Approach
Microenvironments:
General
Twenty-eight total microenvironments are represented,
including those expected to be associated with high exposure
concentrations (i.e., outdoors and outdoor near-road). Where
this type of variability is incorporated within particular
microenvironmental algorithm inputs, this results in
differential exposure estimates for each individual (and event)
as persons spend varying time frequency within each
microenvironment and ambient concentrations vary spatially
within and between study areas.
Microenvironments:
Spatial Variability
Ambient concentrations used in microenvironmental
algorithms vary spatially within (where more than one site
available) and among study areas. Concentrations near
roadways are adjusted to account fortitration by NO.
5D-4
-------
Component
Physiological
Factors and
Algorithms
Variability Source
Microenvironments:
Temporal Variability
Air exchange rates
Proximity factors for
on- and near roads
Resting metabolic
rate (RMR)
Maximum normalized
oxygen consumption
rate (NVO2)
Maximum oxygen
debt (MOXD)
Recovery time
METS by activity
Oxygen uptake per
unit of energy
expended (UCF)
Body mass
Height
Body surface area
Comment
All exposure calculations are performed at the event-level
when using either factors or mass balance approach
(durations can be as short as one minute). In addition, for the
indoor microenvironments, using a mass balance model
accounts for Os concentrations occurring during a previous
hour (and of ambient origin) to calculate a current event's
indoor Os concentrations.
Several lognormal distributions are sampled based on five
daily mean temperature ranges, study area, and study-area
specific A/C prevalence rates.
Three distributions are used, stratified by road-type (urban,
interstate, and rural), selected based on VMTto address
expected ozone titration by NO near roads.
Regression equations for three age-group (18-29, 30-59, and
60+) and two genders were used with body mass as the
independent variable (see Johnson et al. (2000) and section
5.3 of APEX TSD).
Single year age- and gender-specific normal distributions are
randomly sampled for each person (Isaacs and Smith, 2005
and section 7.2 of APEX TSD). This variable is used to
calculate maximum metabolic equivalents (METS).
Normal distributions for maximum obtainable oxygen,
stratified by 3 age groups (ages 0-11, 12-18, 19-100) and two
genders (Isaacs and Smith, 2007 and section 7.2 of APEX
TSD). Used when adjusting METS to address fatigue and
EPOC.
One uniform distribution randomly sampled to estimate the
time required to recover a maximum oxygen deficit (Isaacs
and Smith, 2007 and section 7.2 of APEX TSD).
Values randomly sampled from distributions developed for
specific activities (a few are age-group specific) (McCurdy,
2000; US EPA, 2002).
Values randomly sampled from a uniform distribution to
convert energy expenditure to oxygen consumption (Johnson
et al., 2000 and section 5.3 of APEX TSD).
Randomly selected from population-weighted lognormal
distributions with age- and gender-specific geometric mean
(GM) and geometric standard deviation (GSD) derived from
the National Health and Nutrition Examination Survey
(NHANES) for the years 1999-2004 (Isaacs and Smith (2005)
and section 5.3 of APEX TSD).
Values randomly sampled from distributions used are based
on equations developed for each gender by Johnson (1 998)
using height and weight data from Brainard and Burmaster
(1 992) (also see Appendix B of 201 0 CO REA).
Point estimates of exponential parameters used for
calculating body surface area as a function of body mass
(Burmaster, 1998)
5D-5
-------
Component
Variability Source
Comment
Ventilation rate
Event-level activity-specific regression equations stratified by
four age groups, using age, gender, body mass normalized
oxygen consumption rate as independent variables, and
accounting for intra and interpersonal variability (Graham and
McCurdy, 2005).
Fatigue and EPOC
APEX approximates the onset of fatigue, controlling for
unrealistic or excessive exercise events in each persons
activity time-series while also estimating excess post-
exercise oxygen consumption (EPOC) that may occur
following vigorous exertion activities (Isaacs et al., 2007 and
section 7.2 of APEX TSD).
Table 5D-2. Important components of co-variability in exposure modeling.
Type of Co-variability
Within-person correlations 1
Between-person correlations
Correlations between profile variables and
microenvironment parameters
Correlations between demographic
variables (e.g., age, gender) and activities
Correlations between activities and
microenvironment parameters
Correlations among microenvironment
parameters in the same microenvironment
Correlations between demographic
variables and air quality
Correlations between meteorological
variables and activities
Correlations between meteorological
variables and microenvironment
parameters
Correlations between drive times in CHAD
and commute distances traveled
Consistency of occupation/school
microenvironmental time and time spent
commuting/busing for individuals from one
working/school day to the next.
Modeled
by
APEX?
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Treatment in APEX / Comments
Sequence of activities performed,
microenvironments visited, and general
physiological parameters (body mass, height,
ventilation rates).
Judged as not important.
Profiles are assigned microenvironment
parameters.
Age and gender are used in activity diary selection.
Perhaps important, but do not have data. For
example, frequency of opening windows when
cooking or smoking tobacco products.
Modeled with joint conditional variables.
Modeled with the spatially varying demographic
variables and air quality input to APEX.
Temperature is used in activity diary selection.
The distributions of microenvironment parameters
can be functions of temperature.
CHAD diary selection is weighted by commute
times for employed persons during weekdays.
Simulated individuals are assigned activity diaries
longitudinally without regard to occupation or
school schedule (note though, longitudinal variable
used to develop annual profile is time spent
outdoors).
1 The term correlation is used to represent linear and nonlinear relationships.
5D-6
-------
5D-3. CHARACTERIZATION OF UNCERTAINTY
While it may be possible to capture a range of exposure or risk values by accounting for
variability inherent to influential factors, the true exposure or risk for any given individual within
a study area is unknown, though can be estimated. To characterize health risks, exposure and
risk assessors commonly use an iterative process of gathering data, developing models, and
estimating exposures and risks, given the goals of the assessment, scale of the assessment
performed, and limitations of the input data available. However, significant uncertainty often
remains and emphasis is then placed on characterizing the nature of that uncertainty and its
impact on exposure and risk estimates.
In the final 2008 Os NAAQS rule,1 EPA staff performed such a characterization and at
that time, identified the most important uncertainties affecting the exposure estimates. The key
elements of uncertainty were 1) the modeling of human activity patterns over an Os season, 2)
the modeling of variations in ambient Os concentrations near roadways, 3) the modeling of air
exchange rates that affect the amount of Os that penetrates indoors, and 4) the characterization of
energy expenditure (and related ventilation rate estimates) for children engaged in various
activities. Further, the primary findings of a quantitative Monte Carlo analysis also performed at
that time indicated that the overall uncertainty of the APEX estimated exposure distributions was
relatively small: the percent of children or asthmatic children with exposures above 0.06, 0.07, or
0.08 ppm-8hr under moderate exertion have 95% were estimated by APEX to have uncertainty
intervals of at most ±6 percentage points. Details for these previously identified uncertainties are
discussed in the 2007 Os Staff Paper (section 4.6) and in a technical memorandum describing the
2007 Os exposure modeling uncertainty analysis (Langstaff, 2007).
The REA's conducted for the most recent NO2 (US EPA, 2008), SO2 (US EPA, 2009),
and CO (US EPA, 2010) NAAQS reviews also presented characterizations of the uncertainties
associated with APEX exposure modeling (among other pollutant specific issues), albeit mainly
qualitative evaluations. Conclusions drawn from all of these assessments regarding exposure
modeling uncertainty have been integrated here, following the standard approach used by EPA
staff since 2008 and outlined by WHO (2008) to identify, evaluate, and prioritize the most
important uncertainties relevant to the estimated potential health effect endpoints used in this Os
HREA. Staff selected a mainly qualitative approach here supplemented by various model
sensitivity analyses and input data evaluations, all complimentary to quantitative uncertainty
characterizations conducted for the 2007 Os REA by Langstaff (2007).
1 Federal Register Vol. 73, No. 60. Available at: http://www.epa.gov/ttn/naaqs/standards/ozone/fr/20080327.pdf
5D-7
-------
The qualitative approach used here varies from that described by WHO (2008) in that a
greater focus was placed on evaluating the direction and the magnitude2 of the uncertainty; that
is, qualitatively rating how the source of uncertainty, in the presence of alternative information,
may affect the estimated exposures and health risk results. In addition and consistent with the
WHO (2008) guidance, staff discuss the uncertainty in the knowledge base (e.g., the accuracy of
the data used, acknowledgement of data gaps) and decisions made where possible (e.g., selection
of particular model forms), although qualitative ratings were assigned only to uncertainty
regarding the knowledge base.
First, staff identified the key aspects of the assessment approach that may contribute to
uncertainty in the exposure and risk estimates and provided the rationale for their inclusion.
Then, staff characterized the magnitude and direction of the influence on the assessment results
for each of these identified sources of uncertainty. Consistent with the WHO (2008) guidance,
staff subjectively scaled the overall impact of the uncertainty by considering the degree of
uncertainty as implied by the relationship between the source of uncertainty and the exposure
concentrations.
Where the magnitude of uncertainty was rated low, it was judged that changes within the
source of uncertainty would have only a small effect on the exposure results. For example, we
have commonly employed statistical procedure to substitute missing concentration values to
complete the meteorological data sets. Staff has consistently compared the air quality
distributions and found negligible differences between the substituted data set and the one with
missing values (e.g., Tables 5-13 through 5-16 of US EPA, 2010), primarily because of the
infrequency of missing value substitutions needed to complete a data set. There is still
uncertainty in the approach used, and there may be alternative, and possibly better, methods
available to perform such a task. However, in this instance, staff judged that the quantitative
comparison of the ambient concentration data sets indicates that there would likely be little
influence on exposure estimates by the data substitution procedure used.
A magnitude designation of moderate implies that a change within the source of
uncertainty would likely have a moderate (or proportional) effect on the results. For example,
the magnitude of uncertainty associated with using the quadratic approach to represent a
hypothetical future air quality scenario was rated as low-moderate. While we do not have
information regarding how the ambient Os concentration distribution might look in the future, we
do know however what the distribution might look like based on historical trends and the
emission sources. These historical data and trends serve to generate algorithms used to adjust air
quality. If these trends in observed concentrations and emissions were to remain constant in the
future, then the magnitude of the impact to estimated exposures in this assessment would be
2 This is synonymous with the "level of uncertainty" discussed in WHO (2008), section 5.1.2.2.
5D-8
-------
judged as likely low or having negligible impact on the estimated exposures. However, if there
are entirely new emission sources in the future or if the approach developed is not equally
appropriate across the range of assessed study areas, the magnitude of influence might be judged
as greater. For example, when comparing exposure estimates for one year that used three
different 3-year periods to adjust that year's air quality levels to just meet the current standard,
staff observed mainly proportional differences (e.g., a factor of two or three) in the estimated
number of persons exposed in more than half of the twelve study areas (Langstaff, 2007).
Assuming that these types of ambient concentration adjustments could reflect the addition of a
new or unaccounted for emission source in a particular study area, staff also judged the
magnitude of influence in using the quadratic approach to adjust air quality data to represent a
hypothetical future scenario as moderate. A characterization of high implies that a small change
in the source would have a large effect on results, potentially an order of magnitude or more.
This rating would be used where the model estimates were extremely sensitive to the identified
source of uncertainty.
In addition to characterizing the magnitude of uncertainty, staff also included the
direction of influence, indicating how the source of uncertainty was judged to affect estimated
exposures or risk estimates; either the estimated values were possibly over- or under-estimated.
In the instance where the component of uncertainty can affect the assessment endpoint in either
direction, the influence was judged as both. Staff characterized the direction of influence as
unknown when there was no evidence available to judge the directional nature of uncertainty
associated with the particular source. Staff also subjectively scaled the knowledge-base
uncertainty associated with each identified source using a three-level scale: low indicated
significant confidence in the data used and its applicability to the assessment endpoints,
moderate implied that there were some limitations regarding consistency and completeness of
the data used or scientific evidence presented, and high indicated the extent of the knowledge-
base was extremely limited.
The output of the uncertainty characterization is a summary describing, for each
identified source of uncertainty, the magnitude of the impact and the direction of influence the
uncertainty may have on the exposure and risk characterization results.
5D-9
-------
5D-4. REFERENCES
Brainard J and Burmaster D. (1992). Bivariate distributions for height and weight of men and
women in the United States. Risk Analysis. 12(2):267-275.
Burmaster DE. (1998). Lognormal distributions for skin area as a function of body weight. Risk
Analysis. 18(l):27-32.
CDC. (2011). Summary Health Statistics for U.S. Adults: National Health Interview Survey,
years 2006-10. U.S. Department of Health and Human Services, Hyattsville, MD. Data
and documentation available at: http://www.cdc.gov/nchs/nhis.htm (accessed October 4,
2011).
Glen G, Smith L, Isaacs K, McCurdy T, Langstaff J. (2008). A new method of longitudinal
diary assembly for human exposure modeling. J Expos Sci Environ Epidem. 18:299-311.
Graham SE and T McCurdy. (2005). Revised ventilation rate (VE) equations for use in
inhalation-oriented exposure models. Report no. EPA/600/X-05/008. Report is found
within Appendix A of US EPA (2009). Metabolically Derived Human Ventilation Rates:
A Revised Approach Based Upon Oxygen Consumption Rates (Final Report). Report no.
EPA/600/R-06/129F. Appendix D contains "Response to peer-review comments on
Appendix A", prepared by S. Graham (US EPA). Available at:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=202543
Issacs K and Smith L. (2005). New Values for Physiological Parameters for the Exposure Model
Input File Physiology.txt. Technical memorandum to Tom McCurdy, NERL WA10.
December 20, 2005. Provided in Appendix A of the CO REA (US EPA, 2010).
Isaacs K, Glen G, McCurdy T., and Smith L. (2007). Modeling energy expenditure and oxygen
consumption in human exposure models: Accounting for fatigue and EPOC. J Expos Sci
Environ Epidemiol. 18(3):289-98.
Johnson T. (1998). Memo No. 5: Equations for Converting Weight to Height Proposed for the
1998 Version of pNEM/CO. Memorandum Submitted to U.S. Environmental Protection
Agency. TRJ Environmental, Inc., 713 Shadylawn Road, Chapel Hill, North Carolina
27514.
Johnson T, Mihlan G, LaPointe J, Fletcher K, Capel J, Rosenbaum A, Cohen J, Stiefer P. (2000).
Estimation of carbon monoxide exposures and associated carboxyhemoglobin levels for
residents of Denver and Los Angeles using pNEM/CO. Appendices. EPA contract 68-D6-
0064.
Langstaff JE. (2007). OAQPS Staff Memorandum to Ozone NAAQS Review Docket (OAR-
2005-0172). Subject: Analysis of Uncertainty in Ozone Population Exposure Modeling.
[January 31,2007]. Available at:
http://www.epa.gov/ttn/naaqs/standards/ozone/s_o3_cr_td.html
McCurdy T. (2000). Conceptual basis for multi-route intake dose modeling using an energy
expenditure approach. J Expos Anal Environ Epidemiol. 10:1-12.
Schofield WN. (1985). Predicting basal metabolic rate, new standards, and review of previous
work. HumNutrClinNutr. 39C(S1):5-41.
US Census Bureau. (2007a). Employment Status: 2000- Supplemental Tables. Available at:
http://www.census.gov/population/www/cen2000/phc-t28.html.
5D-10
-------
US Census Bureau. (2007b). 2000 Census of Population and Housing. Summary File 3 (SF3)
Technical Documentation, available at: http://www.census.gov/prod/cen2000/doc/sf3.pdf.
Individual SF3 files '30' (for income/poverty variables pct49-pct51) for each state were
downloaded from: http://www2.census.gov/census 2000/datasets/Summary File 3/.
US EPA. (2002). EPA's Consolidated Human Activities Database. Available at:
http://www.epa.gov/chad/.
US EPA. (2007). Ozone Population Exposure Analysis for Selected Urban Areas. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, NC. Available at: http://www.epa.gov/ttn/naaqs/standards/ozone/s_o3_cr_td.html
US EPA. (2008). Risk and Exposure Assessment to Support the Review of the NO2 Primary
National Ambient Air Quality Standard. Report no. EPA-452/R-08-008a. November
2008. Available at:
http://www.epa.gov/ttn/naaqs/standards/nox/data/20081121 NO2 REA final.pdf.
US EPA. (2009). Risk and Exposure Assessment to Support the Review of the SO2 Primary
National Ambient Air Quality Standard. Report no. EPA-452/R-09-007. August 2009.
Available
athttp://www. epa.gov/ttn/naaqs/standards/so2/data/200908SO2RE AFinalReport.pdf.
US EPA. (2010). Quantitative Risk and Exposure Assessment for Carbon Monoxide -
Amended. EPA Office of Air Quality Planning and Standards. EPA-452/R-10-009. July
2010. Available at: http://www.epa.gov/ttn/naaqs/standards/co/data/CO-REA-Amended-
Julv2010.pdf
US EPA. (2012). Total Risk Integrated Methodology (TRIM) - Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4.4) Volume I: User's Guide. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park,NC. EPA-452/B-12-00la. Available at:
http://www.epa.gov/ttn/fera/human apex.html
WHO. (2008). Harmonization Project Document No. 6. Part 1: Guidance document on
characterizing and communicating uncertainty in exposure assessment. Available at:
http://www.who.int/ipcs/methods/harmonization/areas/exposure/en/.
5D-11
-------
This page left intentionally blank
5D-12
-------
APPENDIX 5E
Updated Analysis of Air Exchange Rate Data: Memorandum from
ICF International
5E-i
-------
^i* §**-*-"-
ICF
INTERNATIONAL
MEMORANDUM
To: John Langstaff
From: Jonathan Cohen, Hemant Mallya, Arlene Rosenbaum
Date: 28 December, 2012
Re: Updated Analysis of Air Exchange Rate Data
EPA is planning to use the APEX exposure model to estimate ozone exposure in 16
cities / metropolitan areas: Atlanta, GA; Baltimore, MD; Boston, MA; Chicago, IL;
Cleveland, OH; Dallas, TX: Detroit, Ml; Denver, CO: Houston, TX; Los Angeles, CA;
New York, NY; Philadelphia, PA; Sacramento, CA; Seattle, WA; St. Louis, MO-IL;
Washington, DC. As part of this effort, ICF International has developed distributions of
residential and non-residential air exchange rates (AER) for use as APEX inputs for the
cities to be modeled. This memorandum describes the analysis of the AER data and
the proposed APEX input distributions. Also included in this memorandum are
proposed APEX inputs for penetration and proximity factors for selected
microenvironments.
Residential Air Exchange Rates
Studies. Residential air exchange rate (AER) data were obtained from the following
seven studies and summarized in Table 1:
Avol: Avol et al., 1998. In this study, ozone concentrations and AERs were
measured at 126 residences in the greater Los Angeles metropolitan area
between February and December, 1994. Measurements were taken in four
communities: Lancaster, Lake Gregory, Riverside, and San Dimas. Data
included the daily average outdoor temperature, the presence or absence of an
air conditioner (either central or room), and the presence or absence of a swamp
(evaporative) cooler. Air exchange rates were computed based on the total
house volume and based on the total house volume corrected for the furniture.
These data analyses used the study corrected AERs.
RTP Panel: Williams et al., 2003a, 2003b. In this study particulate matter
concentrations and daily average AERs were measured at 37 residences in
central North Carolina during 2000 and 2001 (averaging about 23 AER
measurements per residence). The residences belong to two specific cohorts: a
mostly Caucasian, non-smoking group aged at least 50 years having cardiac
defibrillators living in Chapel Hill; a group of non-smoking, African Americans
aged at least 50 years with controlled hypertension living in a low-to-moderate
SES neighborhood in Raleigh. Data included the daily average outdoor
temperature, and the number of air conditioner units (either central or room).
Every residence had at least one air conditioner unit.
60 Broadway Street—— San Francisco, CA 94111-1429 «—— 415-677-7100—— 415-677-7177 fax—— www.idmnsultjng.com
5E-1
-------
RIOPA: Meng et al., 2004, Weisel et al., 2004. The Relationship of Indoor,
Outdoor, and Personal Air (RIOPA) study was undertaken to estimate the impact
of outdoor sources of air toxics to indoor concentrations and personal exposures.
Volatile organic compounds, carbonyls, fine particles and AERs were measured
once or twice at 310 non-smoking residences from summer 1999 to spring 2001.
Measurements were made at residences in Elizabeth, NJ, Houst on TX, and Los
Angeles CA. Residences in California were randomly selected. Residences in
New Jersey and Texas were preferentially selected to be close (< 0.5 km) to
sources of air toxics. The AER measurements (generally over 24 hours) used a
PMCH tracer. Data included the daily average outdoor temperature, and the
presence or absence of central air conditioning, room air conditioning, or a
swamp (evaporative) cooler.
TEACH: Chillrud at al., 2004, Kinney et al., 2002, Sax et al., 2004. The Toxic
Exposure Assessment, a Columbia/Harvard (TEACH) study was designed to
characterize levels of and factors influencing exposures to air toxics among high
school students living in inner-city neighborhoods of New York City and Los
Angeles, CA. Volatile organic compounds, aldehydes, fine particles, selected
trace elements, and AER were measured at 87 high school student's residences
in New York City and Los Angeles in 1999 and 2000. Data included the
presence or absence of an air conditioner (central or room) and hourly outdoor
temperatures (which were converted to daily averages for these analyses).
Wilson 1984: Wilson et al., 1986, 1996. In this 1984 study, AER and other data
were collected at about 600 southern California homes with three seven-day
tests (in March and July 1984, and January 1985) for each home. We obtained
the data directly from Mr. Wilson. The available data consisted of the three
seven-day averages, the month, the residence zip code, the presence or
absence of a central air conditioner, and the presence or absence of a room air
conditioner. We matched these data by month and zip code to the
corresponding monthly average temperatures obtained from EPA's SCRAM
website as well as from the archives in www.wunderground.com (personal and
airport meteorological stations). Residences more than 25 miles away from the
nearest available meteorological station were excluded from the analysis. For
our analyses, the city/location was defined by the meteorological station, since
grouping the data by zip code would not have produced sufficient data for most
of the zip codes.
Wilson 1991: Wilson etal., 1996, Colomeetal., 1993, 1994. In this 1991 study,
AER and other data were collected at about 300 California homes with one two-
day test in the winter for each home. We obtained the data directly from Mr.
Wilson. The available data consisted of the two-day averages, the date, city
name, the residence zip code, the presence or absence of a central air
conditioner, the presence or absence of a swamp (evaporative) cooler, and the
presence or absence of a room air conditioner. We matched these data by date,
city, and zip code to the corresponding daily average temperatures obtained from
EPA's SCRAM website as well as from the archives in www.wunderground.com
(personal and airport meteorological stations). Residences more than 25 miles
5E-2
-------
away from the nearest available meteorological station were excluded from the
analysis. For our analyses, the city/location was defined by the meteorological
station, since grouping the data by zip code would not have produced sufficient
data for most of the zip codes.
Murray and Burmaster: Murray and Burmaster (1995). For this article, Murray
and Burmaster corrected and compiled nationwide residential AER data from
several studies conducted between 1982 and 1987. These data were originally
compiled by the Lawrence Berkeley National Laboratory. We acknowledge Mr.
Murray's assistance in obtaining these data for us. The available data consisted
of AER measurements, dates, cities, and degree-days. Information on air
conditioner presence or absence was not available.
DEARS: Sheldon (2007). The Detroit Exposure and Aerosol Research Study
(DEARS) collected air exchange rate data as well as PM2.5 and other air pollutant
data at about 120 homes in Detroit, Michigan, for 3 years starting in the summer
of 2004. Each home was sampled for 5 days in the winter and/or 5 days in the
summer. The available data included AER measurements, dates, average,
minimum, and maximum temperatures, the use or non-use of air/conditioners
during each measurement day, the use or non-use of window fans during each
measurement day, and the number of minutes per day that the windows were
open.
For each of the studies, air conditioner usage, window status (open or closed), and fan
status (on or off) was not part of the experimental design, although some of these
studies included information on whether air conditioners or fans were used (and for how
long) and whether windows were closed during the AER measurements (and for how
long).
As indicated in the above summaries, random selection was not used to identify homes
to include in some of the studies: The RTP Panel study selected two specific cohorts of
older subjects with specific diseases. The RIOPA study was biased towards residences
near air toxics sources. The TEACH study focused on inner-city neighborhoods.
Nevertheless, we included all these studies because we determined that any potential
selection bias would be likely to be small and we preferred to keep as much data as
possible. The DEARS study selected homes from certain neighborhoods in Detroit.
The proportion of the DEARS study homes that used A/C on one or more survey days
was 57%, and the proportion of DEARS study homes with some daily temperatures
above 25 °C that used A/C on one or more of those hot days was 73%. The American
Housing Survey shows that for the Detroit metropolitan area, the proportion of homes
with A/C was 90% (see Table 2 below), but 71 % for the city of Detroit. This suggests
that the DEARS study sample may be representative of an older housing stock than the
overall Detroit metropolitan area.
All data and statistical results are compiled into the attached Excel spreadsheet
Summary_Statistic.Dec 2012.xls.
5E-3
-------
Table 1. Summary of Studies of Residential Air Exchange Rates.
Study/
Attribute
Locations
Years
Months/
Seasons
Homes with
AER
Measurements
Total AER
Measurements
Average AER
Measurements
per Home
AER
Measurement
Duration
AER
Measurement
Technique
Min AER (lr1)
Max AER (tr1)
Mean AER (Ir1)
Min Temp. (°C)
Max Temp. (°C)
Air Conditioner
Categories
Avol
Lancaster, Lake
Gregory,
Riverside, San
Dimas. All in
Southern CA
1994
Feb; Mar; Apr;
May; Jun; Jul;
Aug; Sep; Oct;
Nov
86
161
1.87
Not Available
Not Available
0.01
2.70
0.80
-0.04
36.25
No A/C; Central
or Room A/C;
Swamp Cooler
only; Swamp +
[Central or
Room]
RTP Panel
Research
Triangle Park, NC
2000; 2001
2000 (Jun; Jul;
Aug; Sep; Oct;
Nov), 2001 (Jan;
Feb; Apr; May)
37
854
23.08
24 hour
Perflourocarbon
tracer
0.02
21.44
0.72
-2.18
30.81
Central or Room
A/C (Y/N)
RIOPA
CA; NJ; TX
1999; 2000;
2001
1999 (July to
Dec); 2000 (all
months); 2001
(Jan and Feb)
284
524
1.85
24 to 96 hours
PMCH tracer
0.08
87.50
1.41
-6.82
32.50
Window A/C
(Y/N); Evap
Coolers (Y/N)
TEACH
Los Angeles, CA;
New York City, NY
1999; 2000
1999 (Feb; Mar;
Apr; Jul; Aug);
2000 (Jan; Feb;
Mar; Sep; Oct)
85
151
1.78
Sample time (hours)
reported. Ranges
from about 1 to 7
days
Perflourocarbon
tracer
0.12
8.87
1.71
-1.36
32.00
Central or Room
A/C (Y/N)
Wilson 1984
Southern CA
1984, 1985
Mar 1984, Jul 1984, Jan
1985
581
1,362
2.34
7 days
Perflourocarbon tracer
0.03
11.77
1.05
11.00
28.00
Central A/C (Y/N); Room
A/C (Y/N);
Wilson 1991
Southern CA
1984
Jan, Mar, Jul
288
316
1.10
7 days
Perflourocarbon tracer
0.01
2.91
0.57
3.00
25.00
Central A/C (Y/N); Room
A/C (Y/N); Swamp
Cooler(Y/N)
Murray and
Burmaster
AZ, CA, CO,
CT, FL, ID,
MD, MN.MT,
NJ
1982-1987
Various
1,884
2,844
1.51
Not available
Not available
0.01
11.77
0.76
Not available
Not available
Not available
DEARS
Detroit, Ml
2004- 2006
Jan. Feb, Mar,
Jul. Aug
127
868
6.83
24 hour
Perflourocarbon
tracer
0.08
13.56
0.80
-12.92
30.79
Central or
Room A/C
(Y/N)
5E-4
-------
Study/
Attribute
Air Conditioner
Measurements
Made
Fan
Categories
Fan
Measurements
Made
Window Open/
Closed Data
Comments
Avol
A/C use in
minutes
Not available
Time on or off for
various fan types
during sampling
was recorded,
but not included
in database
provided.
Duration open
between times
6am-12 pm;
12pm -6 pm;
and 6pm -6am
RTP Panel
Not Available
Fan (Y/N)
Not Available
Windows (open /
closed along with
duration open in
inch-hours units
RIOPA
Duration
measurements in
Mrs and Mins
Fan (Y/N)
Duration
measurements in
Hours and
Minutes
Windows (Open /
Closed) along
with window
open duration
measurements
CA sample was
a random sample
of homes. NJ
and TX homes
were deliberately
chosen to be
near to ambient
sources.
TEACH
Not Available
Not Available
Not Available
Not Available
Restricted to inner-
city homes with high
school students.
Wilson 1984
Not Available
Not Available
Not Available
Not Available
Contemporaneous
temperature data
obtained for these
analyses from SCRAM
and
www.wunderground.com
meteorological data.
Wilson 1991
Not Available
Not Available
Not Available
Not Available
Contemporaneous
temperature data
obtained for these
analyses from SCRAM
and
www.wunderground.com
meteorological data.
Murray and
Burmaster
Not available
Not available
Not available
Not available
DEARS
Not available
Fan (Y/N)
Daily duration
in minutes
Windows (Open
/ Closed) along
with window
open duration
measurements
5E-5
-------
We compiled the data from these eight studies to create the following variables, of
which some had missing values:
• Study
• Date
• Time - Time of the day that the AER measurement was made
• HouseJD - Residence identifier
• MeasurementJD - Uniquely identifies each AER measurement for a given study
• AER - Air Exchange Rate (per hour)
• AER_Duration - Length of AER measurement period (hours)
• Have_AC - Indicates if the residence has any type of air conditioner (A/C), either
a room A/C or central A/C or swamp cooler or any of them in combination. "Y" =
"Yes." "N" = "No"
• Type_of_AC1 - Indicates the types of A/C or swamp cooler available in each
house measured. Possible values: "Central A/C" "Central and Room A/C"
"Central or Room A/C" "No A/C" "Swamp + (Central or Room)" "Swamp Cooler
only" "Window A/C" "Window and Evap""" (missing)
• Type_of_AC2 - Indicates if a house measured has either no A/C or some A/C.
Possible values: "No A/C" "Central or Room A/C""" (missing)
• Type_of_AC3 - Indicates if a house measured has either no A/C, central A/C, or
room A/C. If separate A/C information is available for central A/C and for room
A/C, then homes with both central and room A/C are coded as having "Central
A/C." Possible values: "No A/C" "Central A/C" "Room A/C""" (missing)
• Have_Fan - Indicates if the house studied has any fans
• Mean_Temp - Daily average outside temperature (°C)
• Min_Temp - Minimum hourly outside temperature (°C)
• Max_Temp - Maximum hourly outside temperature (°C)
• State
• City
• Location - Two character abbreviation
• Flag - Data status. Murray and Burmaster study: "Used" or "Not Used." Other
studies: "Used"; "Missing" (missing values for AER, Type_of_AC2, and/or
Mean_Temp); "Outlier".
Note that in the Wilson 1991 study, one of the Los Angeles values was recorded as
having an "Unknown" air conditioning type. In this analysis, this measurement is treated
as having missing A/C information and is not included in the main data analyses.
Note that for the DEARS study, the available data did not include the presence or
absence of an air conditioner or window fan. Instead, the data set included a flag
indicating whether or not an air conditioner or window fan was used on that day. For the
data analysis, we assumed that if the air conditioner or window fan was used on at least
one day, then the air conditioner or window fan was present; otherwise the air
conditioner or window fan was absent. Since the maximum observed temperature
during the study days was 31 °C and some of the homes did not have measurements on
hot summer days, some of the homes designated as having no A/C may have been
5E-6
-------
mistakenly designated. For the DEARS study the type of air conditioner (central or
room) was not provided.
The compiled data base is the attached PC SAS dataset
complete_database2.sas7bdat.
The main data analysis was based on the seven studies other than the Murray and
Burmaster study. The Murray and Burmaster data were excluded because of the
absence of information on air conditioner presence. (However, a subset of these data
was used for a supplementary analysis described below.)
Based on our review of the AER data we excluded nine values with unrealistically high
AER values - above 10 per hour (see the worksheet "OUTLIERS"). The main data
analysis used all the remaining data that had non-missing values for AER,
Type_of_AC2, and Mean_Temp. For the main data analysis, we decided to base the
A/C type variable on the broad characterization "No A/C" versus "Central or Room A/C"
since this variable could be calculated from all of the studies (excluding Murray and
Burmaster). Information on the presence or absence of swamp coolers was not
available from all the studies, and, also importantly, the corresponding information on
swamp cooler prevalence for the subsequent ozone modeling cities was not available
from the American Housing Survey. It is plausible that AER distributions depend upon
the presence or absence of a swamp cooler. It is also plausible that AER distributions
also depend upon whether the residence specifically has a central A/C, room or window
A/C, or both. However we determined to use the broader A/C type definition, which in
effect assumes that the exact A/C type and the presence of a swamp cooler are
approximately proportionately represented in the surveyed residences. The detailed
A/C type variable Type_of_AC3 was used for the Los Angeles AER distributions.
Most of the studies had more than one AER measurement for the same house. It is
reasonable to assume that the AER varies with the house as well as other factors such
as the temperature. (The A/C type can be assumed to be the same for each
measurement of the same house). We expected the temperature to be an important
factor since the AER will be affected by the use of the available ventilation (air
conditioners, windows, fans), which in turn will depend upon the outside meteorology.
Therefore it is not appropriate to average data for the same house under different
conditions, which might have been one way to account for dependence between
multiple measurements on the same house. To simplify the data analysis, we chose to
ignore possible dependence between measurements on the same house on different
days and treat all the AER values as if they were statistically independent.
Summary Statistics. We computed summary statistics for AER and its natural
logarithm LOG_AER on selected strata defined from the study, city, A/C type, and mean
temperature. Cities were defined as in the original databases, except that for Los
Angeles we combined all the data in the Los Angeles ozone modeling region, i.e., the
counties of Los Angeles, Orange, Ventura, Riverside, and San Bernardino. A/C type
was defined from the Type_of_AC2 variable, which we abbreviated as "NA" = "No A/C"
and "AC" = "Central or Room A/C". The mean temperature was grouped into the
following temperature bins: -15 to 0 °C, 0 to 10 °C, 10 to 20 °C, 20 to 25 °C, 25 to 30 °C,
30 to 40 °C (Values equal to the lower bounds are excluded from each interval). Also
5E-7
-------
included were strata defined by study = "AH" and/or city = "All," and/or A/C type = "AH"
and/or temperature bin = "All". The following summary statistics for AER and LOG_AER
respectively are shown in the worksheets "SUMMARY_STATISTICS_AER" and
"SUMMARY_STATISTICS_LOG_AER":
• Number of values
• Arithmetic Mean
• Arithmetic Standard Deviation
• Arithmetic Variance
• Deciles (Min, 10th, 20th ... 90th percentiles, Max)
These calculations exclude all nine outliers and results are not shown for strata with 10
or fewer values, since those summary statistics are extremely unreliable.
Examination of these summary tables clearly demonstrates that the AER distributions
vary greatly across cities and A/C types and temperatures, so that the selected AER
distributions for the modeled cities should also depend upon the city, A/C type and
temperature. For example, the mean AER for residences with A/C ranges from 0.39 for
Los Angeles between 30 and 40 °C to 1.73 for New York between 20 and 25 °C. The
mean AER for residences without A/C ranges from 0.46 for San Francisco between 10
and 20 °C upwards to 2.54 for Detroit between 25 and 30 °C. The need to account for
the city as well as the A/C type and temperature is illustrated by the result that for
residences with A/C and between 20 and 25 °C, the mean AER ranges from 0.52 for
Research Triangle Park to 1.73 for New York. Statistical comparisons are described
below.
Statistical Comparisons. Various statistical comparisons between the different strata
are shown in the worksheets COMPARISON_STATISTICS_AER and
COMPARISON_STATISTICS_LOG_AER, for the AER and its logarithm, respectively.
The various strata are defined as in the Summary Statistics section, excluding the "AN"
cases. For each analysis, we fixed one or two of the variables Study, City, A/C type,
temperature, and tested for statistically significant differences among other variables.
The comparisons are listed in Table 2.
Table 2. Summary of Com
Comparison
Analysis
Number
1
2
3
4
5
6
Comparison
Variable(s)
"Groups
Compared"
City
Temp. Range
Type of A/C
City
City
Type of A/C AND
Temp. Range
parisons of Means.
Stratification
Variable(s)
(not missing in
worksheet)
Type of A/C AND
Temp. Range
Study AND City
Study AND City
Type of A/C
Temp. Range
Study AND City
Total
Comparisons
12
13
16
2
6
18
Cases with significantly
different means (5 %
level)
AER
8
6
6
2
6
7
Log AER
9
6
6
2
6
7
5E-8
-------
For example, the first set of comparisons fix the Type of A/C and the temperature range;
there are twelve such combinations. For each of these twelve combinations, we
compare the AER distributions across different cities. This analysis determines whether
the AER distribution is appropriately defined by the A/C type and temperature range,
without specifying the city. Similarly, for the sixth set of comparisons, the study and city
are held fixed (18 combinations) and in each case we compare AER distributions across
groups defined by the combination of the A/C type and the temperature range.
The F-Statistic comparisons compare the mean values between groups using a one
way analysis of variance (ANOVA). This test assumes that the AER or log(AER) values
are normally distributed with a mean that may vary with the comparison variable(s) and
a constant variance. Shown in the worksheets are the F-Statistic and its p-value. P-
values above 0.05 indicate cases where all the group means are not statistically
significantly different at the 5 percent level. Those results are summarized in the last
two columns of the above table "Summary of Comparisons of Means" which gives the
number of cases where the means are significantly different. Comparison analyses 2,
3, and 6 show that for a given study and city, slightly less than half of the comparisons
show significant differences in the means across temperature ranges, A/C types, or
both. Comparison analyses 1, 4, and 5 show that for the majority of cases, means vary
significantly across cities, whether you first stratify by temperature range, A/C type, or
both.
The Kruskal-Wallis Statistic comparisons are non-parametric tests that are extensions
of the more familiar Wilcoxon tests to two or more groups. The analysis is valid if the
AER minus the group median has the same distribution for each group, and tests
whether the group medians are equal. (The test is also consistent under weaker
assumptions against more general alternatives). The P-values show similar patterns to
the parametric F-test comparisons of the means. Since the logarithm is a strictly
increasing function and the test is non-parametric, the Kruskal-Wallis tests give identical
results for AER and Log (AER).
The Mood Statistic comparisons are non-parametric tests that compare the scale
statistics for two or more groups. The scale statistic measures variation about the
central value, which is a non-parametric generalization of the standard deviation.
Specifically, suppose there is a total of N AER or log(AER) values, summing across all
the groups. These N values are ranked from 1 to N, and the j'th highest value is given a
score of {j - (N+1 )/2}2. The Mood statistic uses a one way ANOVA statistic to compare
the total scores for each group. Generally, the Mood statistics show that in most cases
the scale statistics are not statistically significantly different. Since the logarithm is a
strictly increasing function and the test is non-parametric, the Mood tests give identical
results for AER and Log (AER).
Fitting Distributions. Based on the summary statistics and the statistical comparisons,
the need to fit different AER distributions to each combination of A/C type, city, and
temperature is apparent. For each combination with a minimum of 11 AER values, we
fitted and compared exponential, log-normal, normal, and Weibull distributions to the
AER values.
5E-9
-------
The first analysis used the same stratifications as in the above "Summary Statistics" and
"Statistical Comparisons" sections. Results are not reported for all strata because of the
minimum data requirement of 11 values. Results for each combination of A/C type, city,
and temperature (i.e., A, C, and T) are given in the worksheet FIT_STATISTICS_ACT.
Each combination has four rows, one for each fitted distribution. For each distribution
we report the fitted parameters (mean, standard deviation, scale, shape) and the p-
value for three standard goodness-of-fit tests: Kolmogorov-Smirnov (K-S), Cramer-Von-
Mises (C-M), Anderson-Darling (A-D). Each goodness-of-fit test compares the empirical
distribution of the AER values to the fitted distribution. The K-S and C-M tests are
different tests examining the overall fit, while the Anderson-Darling test gives more
weight to the fit in the tails of the distribution. For each combination, the best-fitting of
the four distributions has the highest p-value and is marked by an x in the final three
columns. The mean and standard deviation (Std_Dev) are the values for the fitted
distribution. The scale and shape parameters are defined by:
• Exponential: density = a'1 exp(-x/a), where shape = mean = a
• Log-normal: density = {axV(2:r)}'1 exp{ -(log x - Q2 / (2a2)}, where shape = a and
scale = C,. Thus the geometric mean and geometric standard deviation are given
by exp(Q and exp(a), respectively.
• Normal: density = {aV(2:r)}'1 exp{ -(x - jj,)2 / (2a2)}, where mean = jj, and standard
deviation = a
• Weibull: density = (c/a) (x/a)c~1 exp{-(x/a)c}, where shape = c and scale = a
Generally, the log-normal distribution was the best-fitting of the four distributions, and
so, for consistency, we recommend using the fitted log-normal distributions for all the
cases. For the log-normal distributions, the worksheet includes the geometric mean
and geometric standard deviation.
One limitation of the initial analysis was that distributions were available only for
selected cities, and yet the summary statistics and comparisons demonstrate that the
AER distributions depend upon the city as well as the temperature range and A/C type.
As one option to address this issue, we considered modeling cities not listed in the
FIT_STATISTICS_ACT worksheet by using the AER distributions across all cities and
dates for a given temperature range and A/C type. Those fitted distributions are
presented in the FIT_STATISTICS_AT worksheet.
Another important limitation of the initial analysis was that distributions were not fitted to
all of the temperature ranges due to inadequate data. There are missing values
between temperature ranges, and the temperature ranges are all bounded. To address
this issue, the temperature ranges were regrouped to cover the entire range of
temperatures from minus to plus infinity, although obviously the available data to fit
these ranges have finite temperatures. Stratifying by A/C type, city, and the new
temperature ranges produces the results in the worksheet FIT_STATISTICS_ACT2.
Results are reported for five cities: Detroit, AC and NA; Houston, AC and NA; Los
Angeles, AC and NA; New York, AC and NA; Research Triangle Park, AC. As noted
above the DEARS study sample is likely to be representative of an older housing stock
than for the Detroit metropolitan area as a whole. For this reason, the worksheet
FIT_STATISTICS_ACT2 also reports results for the aggregation of the Detroit and New
5E-10
-------
York data ("Detroit or NYC," AC and NA); these results can be used for modeling
Chicago, Cleveland and Detroit.
Corresponding to each of the fitted distributions in FIT_STATISTICS _ACT2, we created
histograms to compare each of the fitted distributions with the empirical distributions.
For these graphs, the A/C type, city, and temperature range combinations were
assigned a letter code, as shown in the first column of the worksheet. The first three
digits are a numerical code used for sorting the graphs, and the remaining characters
code the A/C type, temperature range, and city in a fairly obvious manner. These
graphs are in the attached file Graphs_ACT2.Dec 2012.rtf, which can be read directly
by Word.
AER Distributions for Eight Cities. Based upon the FIT_STATISTICS_ACT2 results
for the five cities and the corresponding graphs, we propose using those fitted
distributions for the two cities Houston and New York. For Los Angeles, more data
were available for the type of air conditioning system, and we decided to use
distributions for the more detailed air conditioning type, Type_of_AC3, as described
below.
For Atlanta, Boston, and Philadelphia, we propose using the distribution for one of the
cities thought to have similar characteristics to the city to be modeled with respect to
factors that might influence AERs. These factors include the age composition of
housing stock, construction methods, and other meteorological variables not explicitly
treated in the analysis, such as humidity and wind speed patterns.
As noted above the DEARS study sample is likely to be representative of an older
housing stock than for the Detroit metropolitan area as a whole. Therefore, we
combined the DEARS study data with the data from New York to create an aggregate
distribution for application to Detroit and other cities with similar characteristics, i.e.,
Chicago and Cleveland.
The distributions proposed for the eight cities are as follows:
• Atlanta, GA: Use log-normal distributions for Research Triangle Park.
Residences with A/C only.
• Boston, MA: Use log-normal distributions for New York.
• Chicago, IL: Use log-normal distributions for "Detroit or NYC."
• Cleveland, OH: Use log-normal distributions for "Detroit or NYC."
• Detroit, Ml: Use log-normal distributions for "Detroit or NYC."
• Houston, TX: Use log-normal distributions for Houston.
• New York, NY: Use log-normal distributions for New York.
• Philadelphia, PA: Use log-normal distributions for New York.
All the above distributions are to be found in the FIT_STATISTICS_ACT2 worksheet.
Since the AER data for Research Triangle Park was only available for residences with
air conditioning, AER distributions for Atlanta residences without air conditioning are
discussed below.
5E-11
-------
To avoid unusually extreme simulated AER values, we propose to set a minimum AER
value of 0.01 and a maximum AER value of 10.
Obviously, we would be prefer to model each city using data from the same city, but this
approach was chosen as a reasonable alternative, given the available AER data.
AER Distributions for Sacramento and St. Louis. For these two cities, a direct
mapping to one of the five cities Detroit, Houston, Los Angeles, New York, and
Research Triangle Park is not recommended because the cities are likely to be too
dissimilar. Instead, we decided to use the distribution for the inland parts of Los
Angeles to represent Sacramento and to use the aggregate distributions for all cities
outside of California to represent St. Louis. The results are presented in the worksheet
FIT_STATISTICS_ACT3 and corresponding histogram graphs Graphs_ACT3.Dec
2012.rtf. The results for the city denoted by "Sacramento" were obtained by combining
all the available AER data for Sacramento, Riverside, and San Bernardino counties.
The results for the city denoted by "St. Louis" were obtained by combining all non-
California AER data. Thus our proposal is:
• Sacramento, CA: Use log-normal distributions for "Sacramento" in
FIT_STATISTICS_ACT3
• St. Louis, MO-IL: Use log-normal distributions for "St. Louis" in
FIT_STATISTICS_ACT3
To avoid unusually extreme simulated AER values, we propose to set a minimum AER
value of 0.01 and a maximum AER value of 10.
AER Distributions for Baltimore and Washington DC. Baltimore and Washington
DC were judged likely to have similar characteristics to Detroit, Research Triangle Park
and New York City. To choose between these three cities, we compared the Murray
and Burmaster AER data for Maryland with AER data from each of those cities. The
Murray and Burmaster study included AER data for Baltimore and for Gaithersburg and
Rockville, primarily collected in March. April, and May 1987, although there is no
information on daily mean temperatures or A/C type. We collected all the March, April,
and May AER data for Detroit, Research Triangle Park and New York City, and
compared those three distributions with the Murray and Burmaster Maryland data for
the same three months. The summary statistics for AER and Log (AER) are given in
the worksheets SUM_STAT_MURRAY_AER and SUM_STAT_MURRAY_LOG_AER,
using the same formats as the other summary statistics worksheets. The corresponding
statistical comparisons between Detroit and Maryland, Research Triangle Park and
Maryland, and between New York and Maryland, are shown in the worksheets
COMP_STAT_MURRAY_AER and COMP_STAT_MURRAY_LOG_AER.
The results for the means and central values show significant differences at the 5
percent level between the New York and Maryland distributions and between the Detroit
and Maryland distributions. Between Research Triangle Park and Maryland, the central
values and the mean AER values are not statistically significantly different, and the
differences in the mean log (AER) values are much less statistically significant than
between New York and Maryland. The scale statistic comparisons are not statistically
significantly different between New York and Maryland or between Detroit and
5E-12
-------
Maryland, but were statistically significantly different between Research Triangle Park
and Maryland. Since matching central and mean values is generally more important
than matching the scales, we propose to model both Baltimore and Washington DC
residences with air conditioning using the Research Triangle Park distributions, stratified
by temperature:
• Washington DC: Use log-normal distributions for Research Triangle Park in
FIT_STATISTICS_ACT2. Residences with A/C only.
• Baltimore, MD: Use log-normal distributions for Research Triangle Park in
FIT_STATISTICS_ACT2. Residences with A/C only.
Since the AER data for Research Triangle Park was only available for residences with
air conditioning, the estimated AER distributions for Baltimore and Washington DC
residences without air conditioning are discussed below.
To avoid unusually extreme simulated AER values, we propose to set a minimum AER
value of 0.01 and a maximum AER value of 10.
AER Distributions for Washington DC, Baltimore MD, and Atlanta GA Residences
With No A/C. For Atlanta, Baltimore, and Washington DC we have proposed to use the
AER distributions for Research Triangle Park. However, all the Research Triangle Park
data (from the RTP Panel study) were from houses with air conditioning, so there are no
available distributions for the "No A/C" cases. For these three cities, one option is to
use AER distributions fitted to all the study data for residences without A/C, stratified by
temperature. These fitted distributions are given in the worksheet
FIT_STATISTICS_AT2. The distributions for "No A/C" and "Central or Room A/C" are
both presented for completeness, although we only propose applying the "No A/C"
distributions for modeling these two cities for residences without A/C. However, since
Atlanta, Baltimore, and Washington DC residences are expected to be better
represented by residences outside of California, we instead propose to use the "No A/C"
AER distributions aggregated across cities outside of California, which is the same as
the recommended choice for the St. Louis "No A/C" AER distributions.
• Washington DC, No A/C: Use log-normal distributions for "St. Louis" in
FIT_STATISTICS_ACT3. Residences without A/C only.
• Atlanta, GA, No A/C: Use log-normal distributions for "St. Louis" in
FIT_STATISTICS_ACT3. Residences without A/C only.
• Baltimore, MD. No A/C: Use log-normal distributions for "St. Louis" in
FIT_STATISTICS_ACT3. Residences without A/C only.
To avoid unusually extreme simulated AER values, we propose to set a minimum AER
value of 0.01 and a maximum AER value of 10.
AER Distributions for Los Angeles. Los Angeles data were collected in the Avol,
RIOPA, TEACH, Wilson 1984, and Wilson 1991 studies, and in most cases the data
included whether or not the residence had central A/C, and whether or not the
residence had room A/C. We decided to evaluate whether to stratify the Los Angeles
air exchange rate distribution by this more detailed A/C type, defined by the variable
Type_of_AC3, as well as by the temperature range. The Avol study included
5E-13
-------
information on whether or not the residence had an A/C but not on whether or not the
residence has Central A/C, Room A/C, or both. To avoid potential bias due to location
we decided to exclude all the Avol data from these analyses, although we could have
decided to include the 62 Avol study residences without A/C.
Summary statistics of the AER and its logarithm for Los Angeles, stratified by
Type_of_AC3 and temperature range are shown in the worksheets
SUM_STAT_LOS_ANGELES_AER and SUM_STAT_LOS_ANGELES_LOG_AER,
respectively. The corresponding statistical comparisons between residences with
Central A/C only and Room A/C only are shown in the worksheets COMPARE_LA_AER
and COMPARE_LA_LOG_AER. These comparisons did not include the residences
with no A/C, since the goal was to evaluate the need to further stratify the AER
distribution by the specific A/C types, rather than only by the presence or absence of
any A/C. For each of the temperature ranges 10-20 and 20-25, the statistical
comparisons show statistically significant differences between Central A/C and Room
A/C residences in the means and central values, except for the mean comparison of
AER in the range 10-20 (p-value 0.06), and no statistically significant differences in the
scale statistics. For the temperature range 25-30, the statistical comparisons show no
statistically significant differences between Central A/C and Room A/C residences in the
means and central values, and statistically significant differences in the scale statistics.
For the temperature range 0-10, the statistical comparisons show no statistically
significant differences between Central A/C and Room A/C residences. On this basis,
we decided to stratify the Los Angeles AER distributions by the Type_of_AC3 air
conditioner type variable as well as by the temperature range.
The worksheet FIT_STATISTICS_ACT4 compares exponential, log-normal, normal, and
Weibull distributions fitted to each stratum of the Los Angeles data, defined by
Type_of_AC3 and temperature range, using the temperature ranges "<= 20" "20-25"
and ">25". Histograms of the fitted distributions are in the attached Graphs_ACT4.Dec
2012.rtf file, using the letter codes given in the first column of FIT_STATISTICS_ACT4.
Since the tabulated goodness-of-fit statistics and histograms generally support the
lognormal distribution, we propose to model Los Angeles using the Los Angeles data
stratified by the detailed A/C type, Type_of_AC3, and by the temperature range:
• Los Angeles, CA: Use log-normal distributions for Los Angeles in
FIT_STATISTICS_ACT4
A/C Type and Temperature Distributions. Since the proposed AER distribution is
conditional on the A/C type and temperature range, these values also need to be
simulated using APEX in order to select the appropriate AER distribution. Mean daily
temperatures are one of the available APEX inputs for each modeled city, so that the
temperature range can be determined for each modeled day according to the mean
daily temperature. To simulate the A/C type, we obtained estimates of A/C prevalence
from the American Housing Survey. Thus for each city/metropolitan area other than Los
Angeles, we obtained the estimated fraction of residences with Central or Room A/C
(see Table 3), which gives the probability p for selecting the A/C type "Central or Room
A/C". Obviously, 1-p is the probability for "No A/C".
5E-14
-------
For comparison with Atlanta, Baltimore, and Washington DC, we have included the A/C
type percentage for Charlotte, NC (representing Research Triangle Park, NC). As
discussed above, we propose modeling the 96-98% of Atlanta, Baltimore and
Washington DC residences with A/C using the Research Triangle Park AER
distributions, and modeling the 2-4 % of Atlanta, Baltimore, and Washington DC
residences without A/C using the combined study No A/C AER distributions (denoted as
"St Louis" in FIT_STATISTICS_ACT3).
Table 3. Fraction of occupied residences with central or room A/C (from American
Housing Survey).
CITY
Atlanta
Baltimore
Boston
Chicago
Cleveland
Denver
Detroit
Houston
New York
Philadelphia
Sacramento
St. Louis
Washington DC
Research Triangle Park
SURVEY AREA & YEAR
Atlanta, 2004
Baltimore, 2007
Boston, 2007
Chicago, 2009
Cleveland, 2004
Denver, 2004
Detroit, 2009
Houston, 2007
New York, 2009
Philadelphia, 2009
Sacramento, 2004
St. Louis, 2004
Washington DC, 2007
Charlotte NC, 2002
PERCENTAGE
98.03, replace by 100*
96.70, replace by 100*
85.67
94.74
79.80
84.56
89.61
98.77
86.93
95.00
92.81
98.73
98.15, replace by 100*
97.89
* See text
For Los Angeles, we have proposed to use different AER distributions depending upon
whether or not the residence had central A/C, and whether or not the residence had
room A/C. To simulate the A/C type, we obtained estimates of A/C prevalence from the
American Housing Survey as shown in Table 4:
Table 4. Fraction of occupied Los Angeles CMSA residences with central or room
A/C1
TYPE OF A/C
No A/C
Room A/C
Central A/C
PERCENTAGE
33.89
16.44
49.68
1 American Housing Survey, Los Angeles-Long Beach, 2003, Anaheim-Santa Ana 2002, Riverside-San
Bernardino-Ontario 2002
Other AER Studies
We have not used information from some additional residential and non-residential AER
studies that might provide additional information or data. Indoor / outdoor ozone and
PAN distributions were studied by Jakobi and Fabian (1997). Liu et al (1995) studied
5E-15
-------
residential ozone and AER distributions in Toronto, Canada. Weschler and Shields
(2000) describes a modeling study of ventilation and air exchange rates. Weschler
(2000) includes a useful overview of residential and non-residential AER studies.
AER Distributions for Other Indoor Environments
To estimate AER distributions for non-residential, indoor environments (e.g., offices and
schools), we obtained three AER data sets:
The early "Turk" data set (Turk et al, 1989) includes 40 AER measurements from offices
(25 values), schools (7 values), libraries (3 values), and multi-purpose (5 values), each
measured using an SFe tracer over two- or four-hours in different seasons of the year.
The more recent "Persily" data (Persily and Gorfain 2004; Persily et al. 2005) were
derived from the US EPA Building Assessment Survey and Evaluation (BASE) study,
which was conducted to assess indoor air quality, including ventilation, in a large
number of randomly selected office buildings throughout the U.S. The data base
consists of a total of 390 AER measurements in 96 large, mechanically ventilated
offices; each office was measured up to four times over two days, Wednesday and
Thursday AM and PM. The office spaces were relatively large, with at least 25
occupants, and preferably 50 to 60 occupants. AERs were measured both by a
volumetric method and by a C02 ratio method, and included their uncertainty estimates.
For these analyses, we used the recommended "Best Estimates" defined by the values
with the lower estimated uncertainty; in the vast majority of cases the best estimate was
from the volumetric method.
The most recent "Bennett" data (Bennett et al 2011) was a field study of 37 small and
medium commercial buildings throughout California conducted in 2008 to 2010. The
data base includes information on ventilation rate, temperature, and heating, ventilating,
and air conditioning (HVAC) system characteristics. The study included: seven retail
establishments; five restaurants; eight offices; two each of gas stations, hair salons,
healthcare facilities, grocery stores, dental offices, and fitness gyms; and five other
buildings. The selection of buildings was semi-randomized, providing a minimum
coverage of the vast variety of SMCBs across space, age, size, and building use
category. Buildings were almost evenly distributed across each of five regions of the
state; north-coastal, north-inland, south-coastal, south-inland, and central-inland.
Measurements were made in summer and winter, and three of the buildings were
sampled twice, once in each season. Thus there were a total of 40 measured AER
values. Whole building ventilation rates were determined with a tracer decay method
using SFe, a well established method for commercial buildings.
The office AER SAS databases used for these analyses are attached:
turkdata.sas7bdat for the Turk study, basedata.sasfbdat for the Persily study, and
bennettdata.sas7bdat for the Bennett study.
Due to the small sample size of the Turk data, the data were analyzed without
stratification by building type and/or season. For the Persily data, the AER values for
each office space were averaged, rather using the individual measurements, to account
for the strong dependence of the AER measurements for the same office space over a
5E-16
-------
relatively short period. For the Bennett data, analyses of variance of the AER values
and of their logarithms confirmed the finding of the study authors (Bennett et al 2011)
that the AER varied statistically significantly between restaurants and other buildings.
Analyses of variance using a categorized outdoor temperature variable (15-25, 25-35,
and 35 or more) also showed that the AER did not vary statistically significantly with the
outdoor temperature. Although the study authors also reported significant differences
depending upon whether or not doors were open, the stratification by whether or not
doors are open would not be feasible for APEX modeling since that would require a
model for whether or not doors are kept open (perhaps based on the outdoor
temperature and some building characteristics). The study authors did not find a
significant effect due to the presence or absence of HVAC systems. Therefore the
Bennett data were stratified by whether or not the building is a restaurant.
Summary statistics of AER and log (AER) for the three studies are presented in Table 5.
The overall mean values are similar for the three studies, but the mean value for the
restaurants in the Bennett study is almost four times the mean value for the non-
restaurants. Compared to the Turk study, the standard deviations are about twice as
high for the Persily data and for the Bennett study restaurants, but are much lower for
the Bennett study non-restaurants. The proposed AER distributions were derived only
from the most recent Bennett data study stratified into restaurants and non-restaurants.
Similarly to the analyses of the residential AER distributions, we fitted exponential, log-
normal, normal, and Weibull distributions to the AER values. The results are shown in
Tables 6 and 7.
Table 5. AER summary statistics for offices and other non-residential buildings.
Study
Bennett
Bennett
Bennett
Persily
Turk
Bennett
Bennett
Bennett
Persily
Turk
Variable
AER
AER
AER
AER
AER
Log AER
Log AER
Log AER
Log AER
Log AER
Subgroup
All
Not
restaurant
Restaurant
All
All
All
Not
restaurant
Restaurant
All
All
N
40
34
6
96
40
40
34
6
96
40
Mean
1.622
1.144
4.328
1.962
1.540
0.152
-0.052
1.312
0.104
0.254
Std Dev
1.649
0.768
2.642
2.325
0.881
0.785
0.619
0.618
1.104
0.639
Min
0.300
0.300
1.460
0.071
0.300
-1.204
-1.204
0.378
-2.642
-1.204
25th
%ile
0.705
0.620
2.640
0.501
0.850
-0.350
-0.478
0.971
-0.694
-0.164
Median
1.035
0.995
3.855
1.080
1.500
0.034
-0.005
1.344
0.077
0.405
75th
%ile
1.890
1.460
5.090
2.756
2.050
0.637
0.378
1.627
1.012
0.715
Max
9.070
4.020
9.070
13.824
4.100
2.205
1.391
2.205
2.626
1.411
5E-17
-------
Table 6. Best fitting restaurant AER distributions from the Bennett et al. (2011).
Scale
4.328
1.312
4.907
Shape
0.618
1.922
Mean
4.328
4.492
4.328
4.353
Std Dev
4.328
3.062
2.642
2.358
Distribution
Exponential
Lognormal
Normal
Weibull
P-Value
Kolmogorov-
Smirnov
0.25
0.15
0.15
P-Value
Cramer-
von
Mises
0.23
0.50
0.25
0.25
P-Value
Anderson-
Darling
0.25
0.50
0.25
0.25
Table 7. Best fitting n on-restaurant AER distributions from the Bennett et al.
(2011).
Scale
1.144
-0.052
1.290
Shape
0.619
1.654
Mean
1.144
1.149
1.144
1.153
Std Dev
1.144
0.785
0.768
0.716
Distribution
Exponential
Lognormal
Normal
Weibull
P-Value
Kolmogorov-
Smirnov
0.00
0.15
0.01
P-Value
Cramer-
von
Mises
0.00
0.50
0.01
0.21
P-Value
Anderson-
Darling
0.00
0.50
0.01
0.18
(For an explanation of the Kolmogorov-Smirnov, Cramer-von Mises, and Anderson-
Darling P-values see the discussion of residential AER distributions above.) For
restaurants, according to two of the three goodness-of-fit measures the best-fitting
distribution is the log-normal; the best-fitting distribution is exponential for the
Kolmogorov-Smirnov test. For non-restaurants, according to all three goodness-of-fit
measures the best-fitting distribution is the log-normal. Reasonable choices for the
lower and upper bounds are the observed minimum and maximum AER values.
We therefore propose the following indoor, non-residential AER distributions.
• AER distribution for indoor, non-residential, restaurant microenvironments:
Lognormal, with scale and shape parameters 1.312 and 0.618, i.e., geometric
mean = 3.712, geometric standard deviation = 1.855. Lower Bound = 1.46. Upper
bound = 9.07.
• AER distribution for indoor, non-residential, non-restaurant microenvironments:
Lognormal, with scale and shape parameters -0.052 and 0.618, i.e., geometric
mean = 0.949, geometric standard deviation = 1.857. Lower Bound = 0.30. Upper
bound = 4.02.
Application of the proposed distributions in APEX would require estimates of the
proportions of restaurants or similar facilities among the non-residential buildings in
each city.
Proximity and Penetration Factors For Outdoors, In-vehicle, and Mass Transit
For the APEX modeling of the outdoor, in-vehicle, and mass transit micro-environments,
an approach using proximity and penetration factors is proposed, as follows.
5E-18
-------
Outdoors Near Road
Penetration factor = 1.
For the Proximity factor, we propose using ratio distributions developed from the
Cincinnati Ozone Study (American Petroleum Institute, 1997, Appendix B; Johnson et
al. 1995). The field study was conducted in the greater Cincinnati metropolitan area in
August and September, 1994. Vehicle tests were conducted according to an
experimental design specifying the vehicle type, road type, vehicle speed, and
ventilation mode. Vehicle types were defined by the three study vehicles: a minivan, a
full-size car, and a compact car. Road types were interstate highways (interstate),
principal urban arterial roads (urban), and local roads (local). Nominal vehicle speeds
(typically met over one minute intervals within 5 mph) were at 35 mph, 45 mph, or 55
mph. Ventilation modes were as follows:
• Vent Open: Air conditioner off. Ventilation fan at medium. Driver's window half
open. Other windows closed.
• Normal A/C. Air conditioner at normal. All windows closed.
• Max A/C: Air conditioner at maximum. All windows closed.
Ozone concentrations were measured inside the vehicle, outside the vehicle, and at six
fixed site monitors in the Cincinnati area.
The proximity factor can be estimated from the distributions of the ratios of the outside-
vehicle ozone concentrations to the fixed-site ozone concentrations, reported in Table 8
of Johnson et al. (1995). Ratio distributions were computed by road type (local, urban,
interstate, all) and by the fixed-site monitor (each of the six sites, as well as the nearest
monitor to the test location). For this analysis we propose to use the ratios of outside-
vehicle concentrations to the concentrations at the nearest fixed site monitor, as shown
in Table 8.
Table 8. Ratio of outside-vehicle ozone to ozone at nearest fixed site1
Road
Type1
Local
Urban
Interstate
All
Number
of
cases1
191
299
241
731
Mean1
0.755
0.754
0.364
0.626
Standard
Deviation1
0.203
0.243
0.165
0.278
25th
Percentile1
0.645
0.585
0.232
0.417
50th
Percentile1
0.742
0.722
0.369
0.623
75th
Percentile1
0.911
0.896
0.484
0.808
Estimated
5th
Percentile2
0.422
0.355
0.093
0.170
1 From Table 8 of Johnson et al. (1995). Data excluded if fixed-site concentration < 40 ppb.
2 Estimated using a normal approximation as Mean - 1.64 x Standard Deviation.
For the outdoors-near- road microenvironment, we recommend using the distribution for
local roads, since most of the outdoors-near-road ozone exposure will occur on local
roads. The summary data from the Cincinnati Ozone Study are too limited to allow fitting
of distributions, but the 25th and 75th percentiles appear to be approximately equidistant
from the median (50th percentile). Therefore we propose using a normal distribution with
the observed mean and standard deviation. A plausible upper bound for the proximity
factor equals 1. Although the normal distribution allows small positive values and can
5E-19
-------
even produce impossible, negative values (with a very low probability), the titration of
ozone concentrations near a road is limited. Therefore, as an empirical approach, we
recommend a lower bound of the estimated 5th percentile, as shown in the final column
of the above table. Therefore in summary we propose:
• Penetration factor for outdoors, near road: 1.
• Proximity factor for outdoors, near road: Normal distribution. Mean = 0.755.
Standard Deviation = 0.203. Lower Bound = 0.422. Upper Bound = 1.
Outdoors, Public Garage / Parking Lot
This micro-environment is similar to the outdoors-near-road microenvironment. We
therefore recommend the same distributions as for outdoors-near-road:
• Penetration factor for outdoors, public garage / parking lot: 1.
• Proximity factor for outdoors, public garage / parking lot: Normal distribution.
Mean = 0.755. Standard Deviation = 0.203. Lower Bound = 0.422. Upper Bound
= 1.
Outdoors, Other
The outdoors, other ozone concentrations should be well represented by the ambient
monitors. Therefore we propose:
• Penetration factor for outdoors, other: 1.
• Proximity factor for outdoors, other: 1.
In-Vehicle
For the proximity factor for in-vehicle, we also recommend using the results of the
Cincinnati Ozone Study presented in Table 5. For this microenvironment, the ratios
depend upon the road type, and the relative prevalences of the road types can be
estimated by the proportions of vehicle miles traveled in each modeled city. The
proximity factors are assumed, as before, to be normally distributed, the upper bound to
be1, and the lower bound to be the estimated 5th percentile.
• Proximity factor for in-vehicle, local roads: Normal distribution. Mean = 0.755.
Standard Deviation = 0.203. Lower Bound = 0.422. Upper Bound = 1.
• Proximity factor for in-vehicle, urban roads: Normal distribution. Mean = 0.754.
Standard Deviation = 0.243. Lower Bound = 0.355. Upper Bound = 1.
• Proximity factor for in-vehicle, interstates: Normal distribution. Mean = 0.364.
Standard Deviation = 0.165. Lower Bound = 0.093. Upper Bound = 1.
To complete the specification, the distribution of road type needs to be estimated for
each city to be modeled. Vehicle miles traveled (VMT) by city (defined by the Federal-
AID urbanized area) and road type were obtained from the Federal Highway
Administration. For local and interstate road types, the VMT for the same DOT
categories were used. For urban roads, the VMT for all other road types was summed
5E-20
-------
(Other freeways/expressways, Other principal arterial, Minor arterial, Collector). The
computed VMT ratios for each city are shown in Table 9
Table 9. Vehicle Miles Traveled by City and Road Type (2008)1.
FEDERAL-AID URBANIZED
AREA
Atlanta
Baltimore
Boston
Chicago
Cleveland
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
FRACTION VMT BY ROAD
TYPE
INTERSTATE
0.32
0.34
0.32
0.31
0.38
0.25
0.24
0.24
0.29
0.18
0.23
0.24
0.37
0.30
URBAN
0.45
0.59
0.54
0.57
0.45
0.65
0.66
0.73
0.66
0.67
0.65
0.69
0.45
0.62
LOCAL
0.23
0.07
0.14
0.12
0.17
0.10
0.10
0.03
0.05
0.15
0.12
0.07
0.18
0.08
1 http://www.fhwa.dot.gov/policvinformation/statistics/2008/pdf/hm71.pdf
Thus to simulate the proximity factor in APEX, we propose to first select the road type
according to the above probability table of road types, then select the AER distribution
(normal) for that road type as defined in the last set of bullets.
For the penetration factor for in-vehicle, we recommend using the inside-vehicle to
outside-vehicle ratios from the Cincinnati Ozone Study. The ratio distributions were
summarized for all the data and for stratifications by vehicle type, vehicle speed, road
type, traffic (light, moderate, or heavy), and ventilation. The overall results and results
by ventilation type are shown in Table 10.
Table 10. Ratio of inside-vehicle ozone to outside-vehicle ozone1.
Ventilation1
Vent Open
Normal A/C
Maximum
A/C
All
Number
of
cases1
226
332
254
812
Mean1
0.361
0.417
0.093
0.300
Standard
Deviation1
0.217
0.211
0.088
0.232
25th
Percent! le1
0.199
0.236
0.016
0.117
50th
Percent! le1
0.307
0.408
0.071
0.251
75th
Percentile1
0.519
0.585
0.149
0.463
Estimated
5th
Percentile2
0.005
0.071
O.OOO3
O.OOO3
1. From Table 7 of Johnson et al.(1995). Data excluded if outside-vehicle concentration <
20 ppb.
2. Estimated using a normal approximation as Mean - 1.64 x Standard Deviation
3. Negative estimate (impossible value) replaced by zero.
Although the data in Table 7 indicate that the inside-to-outside ozone ratios strongly
depend upon the ventilation type, it would be very difficult to find suitable data to
5E-21
-------
estimate the ventilation type distributions for each modeled city. Furthermore, since the
Cincinnati Ozone Study was scripted, the ventilation conditions may not represent real-
world vehicle ventilation scenarios. Therefore, we propose to use the overall average
distributions.
• Penetration factor for in-vehicle: Normal distribution. Mean = 0.300. Standard
Deviation = 0.232. Lower Bound = 0.000. Upper Bound = 1.
Mass Transit
The mass transit microenvironment is expected to be similar to the in-vehicle
microenvironment. Therefore we recommend using the same APEX modeling
approach:
• Proximity factor for mass transit, local roads: Normal distribution. Mean = 0.755.
Standard Deviation = 0.203. Lower Bound = 0.422. Upper Bound = 1.
• Proximity factor for mass transit, urban roads: Normal distribution. Mean = 0.754.
Standard Deviation = 0.243. Lower Bound = 0.355. Upper Bound = 1.
• Proximity factor for mass transit, interstates: Normal distribution. Mean = 0.364.
Standard Deviation = 0.165. Lower Bound = 0.093. Upper Bound = 1.
• Road type distributions for mass transit: See Table 8.
• Penetration factor for mass transit: Normal distribution. Mean = 0.300. Standard
Deviation = 0.232. Lower Bound = 0.000. Upper Bound = 1.
5E-22
-------
References
American Petroleum Institute (1997). Sensitivity testing ofpNEM/03 exposure to
changes in the model algorithms. Health and Environmental Sciences Department.
Avol, E. L, W. C. Navioli, and S. D. Colome (1998) Modeling ozone levels in and
around southern California homes. Environ. Sci. Technol. 32, 463-468.
Bennett, D. H, W. Fisk, M. G. Apte, X. Wu, A. Trout, D. Faulkner, D. Sulivan (2011).
Ventilation, Temperature, and HVAC Characteristics in Small and Medium
Commercial Buildings (SMCBs) in California. Journal submission.
Chilrud, S. N., D. Epstein, J. M. Ross, S. N. Sax, D. Pederson, J. D. Spengler, P. L.
Kinney (2004). Elevated airborne exposures of teenagers to manganese, chromium,
and iron from steel dust and New York City's subway system. Environ. Sci. Technol.
38, 732-737.
Colome, S.D., A. L. Wilson, Y. Tian (1993). California Residential Indoor Air Quality
Study, Volume 1, Methodology and Descriptive Statistics. Report prepared for the
Gas Research Institute, Pacific Gas & Electric Co., San Diego Gas & Electric Co.,
Southern California Gas Co.
Jakobi, G and Fabian, P. (1997). Indoor/outdoor concentrations of ozone and
peroxyacetyl nitrate (PAN). Int. J. Biometeorol. 40: 162-165..
Johnson, T., A. Pakrasi, A. Wisbeth, G. Meiners, W. M. Ollison (1995). Ozone
exposures within motor vehicles - results of a field study in Cincinnati, Ohio.
Proceedings 88th annual meeting and exposition of the Air & Waste Management
Association, June 18-23, 1995. San Antonio, TX. Preprint paper 95-WA84A.02.
Kinney, P. L., S. N. Chillrud, S. Ramstrom, J. Ross, J. D. Spengler (2002). Exposures to
multiple air toxics in New York City. Environ Health Perspect 110, 539-546.
Liu, L.-J. S, P. Koutrakis, J. Leech, I. Broder, (1995) Assessment of ozone exposures in
the greater metropolitan Toronto area. J. Air Waste Manage. Assoc. 45: 223-234
Meng, Q. Y., B. J. Turpin, L. Korn, C. P. Weisel, M. Morandi, S. Colome, J. J. Zhang, T.
Stock, D. Spektor, A. Winer, L. Zhang, J. H. Lee, R. Giovanetti, W. Cui, J. Kwon, S.
Alimokhtari, D. Shendell, J. Jones, C. Farrar, S. Maberti (2004). Influence of ambient
(outdoor) sources on residential indoor and personal PlVb.s concentrations: Analyses
of RIOPA data. Journal of Exposure Analysis and Environ Epidemiology. Preprint.
Murray, D. M. and D. E. Burmaster (1995). Residential Air Exchange Rates in the
United States: Empirical and Estimated Parametric Distributions by Season and
Climatic Region. Risk Analysis, Vol. 15, No. 4, 459-465.
Persily, A. and J. Gorfain.(2004). Analysis of ventilation data from the U.S.
Environmental Protection Agency Building Assessment Survey and Evaluation
(BASE) Study. National Institute of Standards and Technology, NISTIR 7145,
December 2004.
Persily, A., J. Gorfain, G. Brunner.(2005). Ventilation design and performance in U.S.
office buildings. ASHRAE Journal. April 2005, 30-35.
Sax, S. N., D. H. Bennett, S. N. Chillrud, P. L. Kinney, J. D. Spengler (2004).
Differences in source emission rates of volatile organic compounds in inner-city
5E-23
-------
residences of New York City and Los Angeles. Journal of Exposure Analysis and
Environ Epidemiology. Preprint.
Sheldon, L. (2007). The Detroit Exposure and Aerosol Research Study (DEARS).
Presentation.
Turk, B. H., D. T. Grimsrud, J. T. Brown, K. L. Geisling-Sobotka, J. Harrison, R. J. Prill
(1989). Commercial building ventilation rates and particle concentrations. ASHRAE,
No. 3248.
Weschler, C. J. (2000) Ozone in indoor environments: concentration and chemistry.
Indoor Air 10: 269-288.
Weschler, C. J. and Shields, H. C. (2000) The influence of ventilation on reactions
among indoor pollutants: modeling and experimental observations. Indoor Air. 10:
92-100.
Weisel, C. P., J. J. Zhang, B. J. Turpin, M. T. Morandi, S. Colome, T. H. Stock, D. M.
Spektor, L. Korn, A. Winer, S. Alimokhtari, J. Kwon, K. Mohan, R. Harrington, R.
Giovanetti, W. Cui, M. Afshar, S. Maberti, D. Shendell (2004). Relationship of Indoor,
Outdoor and Personal Air (RIOPA) study; study design, methods and quality
assurance / control results. Journal of Exposure Analysis and Environ Epidemiology.
Preprint.
Williams, R., J. Suggs, A. Rea, K. Leovic, A. Vette, C. Croghan, L. Sheldon, C. Rodes,
J. Thornburg, A. Ejire, M. Herbst, W. Sanders Jr. (2003a). The Research Triangle
Park particulate matter panel study: PM mass concentration relationships. Atmos
Env 37, 5349-5363.
Williams, R., J. Suggs, A. Rea, L. Sheldon, C. Rodes, J. Thornburg (2003b). The
Research Triangle Park particulate patter panel study: modeling ambient source
contribution to personal and residential PM mass concentrations. Atmos Env 37,
5365-5378.
Wilson, A. L., S. D. Colome, P. E. Baker, E. W. Becker (1986). Residential Indoor Air
Quality Characterization Study of Nitrogen Dioxide, Phase I, Final Report. Prepared
for Southern California Gas Company, Los Angeles.
Wilson, A. L., S. D. Colome, Y. Tian (1994). California Residential Indoor Air Quality
Study, Volume 2, Carbon Monoxide and Air Exchange Rate: A Univariate &
Multivariate Analysis. Report prepared for the Gas Research Institute, Pacific Gas &
Electric Co., San Diego Gas & Electric Co., Southern California Gas Co.
Wilson, A. L., S. D. Colome, Y. Tian, P. E. Baker, E. W. Becker, D. W. Behrens, I. H.
Billick, C. A. Garrison (1996). California residential air exchange rates and residence
volumes. Journal of Exposure Analysis and Environ Epidemiology. Vol. 6, No. 3.
5E-24
-------
This page left intentionally blank
5E-25
-------
APPENDIX 5F
Detailed Exposure Results
Table of Contents
5F-1 Exposure Modeling Results for Base Air Quality 5F-1
5F-2 Exposure Modeling Results for Adjusted Air Quality 5F-14
5F-i
-------
List of Tables
Table 5F-1. Percent of all school-age children with Os exposures at or above 60, 70, and 80
ppb-8hr while at moderate or greater exertion, years 2006-2010, base air quality.
5F-10
Table 5F-2. Percent of all school-age children with Os exposures at or above 60, 70, and 80
ppb-8hr while at moderate or greater exertion, years 2006-2010, adjusted air
quality 5F-22
Table 5F-3. Mean and maximum number of all school-age children (and associated days per
Os season) with at least one daily maximum 8-hr average Os exposure at or above
60 ppb-8hr while at moderate or greater exertion 5F-51
Table 5F-4. Total number of school-age children and asthmatic school-age children
experiencing at least one or two daily maximum 8-hr average Os exposures in all
study areas by year, base air quality and air quality adjusted to just meeting the
existing 75 ppb standard 5F-53
Table 5F-5. Mean and maximum number of school-age children and asthmatic school-age
children with at least one or two daily maximum 8-hr average O3 exposures at or
above benchmark levels, all 15 urban study areas combined 5F-55
List of Figures
Figure 5F-1. Percent of all school-age children with at least one daily maximum 8-hr average
Os exposure at or above 60, 70, and 80 ppb while at moderate or greater exertion,
years 2006-2010, base air quality 5F-4
Figure 5F-2. Percent of asthmatic school-age children with at least one daily maximum 8-hr
average Os exposure at or above 60, 70, and 80 ppb while at moderate or greater
exertion, years 2006-2010, base air quality 5F-5
Figure 5F-3. Percent of asthmatic adults with at least one daily maximum 8-hr average Os
exposure at or above 60, 70, and s 80 ppb while at moderate or greater exertion,
years 2006-2010, base air quality 5F-6
Figure 5F-4. Percent of older adults with at least one daily maximum 8-hr average Os exposure
at or above 60, 70, and 80 ppb while at moderate or greater exertion, years 2006-
2010, base air quality 5F-7
Figure 5F-5. Percent of school-age children with multiple daily maximum 8-hr average Os
exposures at or above 60 ppb per study area Os season, while at moderate or
greater exertion, years 2006-2010, base air quality 5F-8
Figure 5F-6. Percent of school-age children with multiple daily maximum 8-hr average Os
exposures at or above 70 ppb per study area Os season, while at moderate or
greater exertion, years 2006-2010, base air quality 5F-9
Figure 5F-7. Incremental increases in percent of all school-age children with at least one daily
maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb
5F-ii
-------
(middle panel), or 80 ppb (bottom panel) using the maximum percent exposed for
each study area, year 2006-2010 adjusted air quality 5F-15
Figure 5F-8. Incremental increases in percent of asthmatic school-age children with at least one
daily maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the maximum percent exposed for
each study area, year 2006-2010 adjusted air quality 5F-16
Figure 5F-9. Incremental increases in percent of asthmatic adults with at least one daily
maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the maximum percent exposed for
each study area, year 2006-2010 adjusted air quality 5F-17
Figure 5F-10. Incremental increases in percent of all older adults with at least one daily
maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the maximum percent exposed for
each study area, year 2006-2010 adjusted air quality 5F-18
Figure 5F-11. Incremental increases in percent of all school-age children with at least one daily
maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the mean percent exposed for each
study area, year 2006-2010 adjusted air quality 5F-19
Figure 5F-12. Incremental increases in percent of all school-age children with at least two daily
maximum 8-hr average Os exposures at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the maximum percent exposed for
each study area, year 2006-2010 adjusted air quality 5F-20
Figure 5F-13. Incremental increases in percent of all school-age children with at least two daily
maximum 8-hr average Os exposures at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the mean percent exposed for each
study area, year 2006-2010 adjusted air quality 5F-21
-------
This appendix contains the detailed results for the primary APEX simulations performed
to estimate exposures associated with base air quality (section 5F-1) and for air quality just
meeting the existing and alternative standard levels (section 5F-2).
5F-1 EXPOSURE MODELING RESULTS FOR BASE AIR QUALITY
As described in the main body of the HREA, comprehensive multi-panel displays of
exposure results are presented for each of the study groups of interest, i.e., all school-age
children (ages 5 to 18), asthmatic school-age children, asthmatic adults (ages 19 to 95), and older
adults (ages 65 to 95) (Figure 5F-1 to Figure 5F-4, respectively). Included in each display are
the three benchmark levels (60, 70, and 80 ppb-8hr), the five years of air quality (2006-2010), for
the 15 study areas. Modeled exposures in the 15 study areas and considering each benchmark
level are presented on the same scale to allow for direct comparisons across the multi-panel
display. The most notable patterns in the exposure results are described here using one study
group (i.e., school-age children), as there is a general consistency in the year-to-year variability
within each study area across all four study groups. Any deviation from the observed pattern
will be discussed for the subsequent study group. Table 5F-1 is also provided and contains the
complete exposure output for all study areas and years for school-age children.
Figure 5F-1 presents the percent of school-age children experiencing at least one Os
exposure at or above the selected benchmark levels while at moderate or greater exertion.
Consistent with the previously discussed observations regarding year-to-year variability in
ambient concentrations (Chapter 4), most study areas have the greatest percent of school-age
children experiencing concentrations at or above the three benchmark levels during 2006 or 2007
along with having the lowest percent of school-age children exposed during 2009. Three
Western U.S. study areas, Dallas, Los Angeles, and Sacramento, differ slightly from this pattern
in that they exhibit a minimum percent of school-age children exposed during 2010, while in
Houston and Chicago the minimum exposures occur during year 2008. In general, between 20 to
40% of school-age children experience at least one daily maximum 8-hr average Os exposure at
or above 60 ppb, 10 to 20% experience at least one Os exposure at or above 70 ppb, and 0 to
10% experience at least one Cb exposure at or above 80 ppb, all while at moderate or greater
exertion and considering the base air quality (2006-2010).
The percent of asthmatic school-age children experiencing at least one daily maximum 8-
hr average Os exposure at or above the selected benchmark levels while at moderate or greater
exertion (Figure 5F-2) is virtually indistinguishable from that of all school-age children (Figure
5F-1) regarding both the year-to-year pattern and percent of persons exposed. This is the result
of having both simulated study groups use an identical time-location-activity diary pool to
construct each simulated individual's time series of activities performed and locations visited.
Different however would be the relative number of asthmatic school-age children exposed in
5F-1
-------
each study area if compared with non-asthmatic school-age children, as the asthma prevalence
rates vary by U.S. location (HREA, Table 5-2) though on average are about 10% of all school-
age children.
As mentioned above, the overall year-to-year pattern of exposure for asthmatic adults is
similar to that observed for school-age children, though the percent of asthmatic adults
experiencing exposures at or above the health effect benchmark levels is lower by a factor of
about three or more (Figure 5F-3). Having a lower percent of asthmatic adults exposed is
expected given that outdoor time expenditure is an important determinant of Os exposure (FIREA
section 5.3.2) and that adults spend less time outdoors than children (REA section 5.3.1). In
general, between 5 to 10% of asthmatic adults experience at least one daily maximum 8-hr
average Os exposure at or above 60 ppb, 0 to 5% experience at least one daily maximum 8-hr
average Os exposure at or above 70 ppb, and 0 to 2% experience at least one daily maximum 8-
hr average Os exposure at or above 80 ppb, all while at moderate or greater exertion.
While the percent of asthmatic adults exposed is much lower, the number of asthmatic
adults at or above the exposure benchmarks is generally just below that estimated number of
asthmatic school-age children. As an example, for year 2006 in Atlanta, approximately 44% of
asthmatic school-age children (or about 37,000) were estimated to experience at least one daily
maximum 8-hr average exposure at or above 60 ppb. Though a much smaller percent of
asthmatic adults were estimated to experience a similar exposure for the same year (i.e., about
16%), this is equivalent to nearly 31,000 asthmatic adults exposed, at least one time, to an 8-hr
average Os concentration at or above 60 ppb.
The percent of older adults (ages 65 to 95) experiencing exposures at or above the
selected benchmark levels (Figure 5F-4) is lower by a fewer percentage points when compared
with the results for asthmatic adults. Again, older adults, on average, would tend to spend less
time outdoors when compared with both adults and children (REA section 5.3.1), in addition to
fewer older adults performing activities at moderate or greater exertion for extended periods of
time, thus leading to fewer older adults exposed to concentrations of concern. In general, less
than 10% of older adults experience at least one Os exposure at or above 60 ppb-8hr, less than
5% experience at least one Os exposure at or above 70 ppb-8hr, and about 2% or less experience
at least one Cb exposure at or above 80 ppb-8hr, all while at moderate or greater exertion
considering base air quality.
Given the similar year-to-year patterns of the single and multiple exposure occurrences
and when considering any of the four study groups, we present the graphic multi-day exposure
results here considering school-age children only. All multi-day exposure results are provided in
Table 5F-1. Figure 5F-5 illustrates the percent of school-age children having multiple exposures
at or above 60 ppb-8hr for each of the 15 study areas, considering base air quality (2006-2010).
5F-2
-------
Depending on the year and study area, about 10 to 25% of school-age children could experience
at least two exposures above the 60 ppb benchmark during the ozone season, while about 5 to
10% could experience at least four. Most study areas and years are estimated to have fewer than
5% of school-age children experience six or more exposures above 60 ppb considering the base
air quality. When considering the multi-day exposures for school-age children at or above the 70
ppb benchmark (Figure 5F-6), about 2 to 10% of school-age children could experience at least
two exposures during the ozone season, while four or more exposures were generally limited to
fewer than 4% of school-age children.
5F-3
-------
< re
o<
x
LJJ
(0
re
•*? o
co£
0) 0)
c x
O uj
+•» *-
(0 0)
re
o>
re
o>
oo 2
in o
II
0) —
O 0)
<: >
T 0)
ij_ 0)
O
d>
Q.
I ! I I
CT •? <$-<§>?
OOQO QQOOO OQOQO
«y fy *y fy Cy *y *y (y fy (y *y Cy fy *V
Figure 5F-1. Percent of all school-age children with at least one daily maximum 8-hr
average Os exposure at or above 60, 70, and 80 ppb while at moderate or greater exertion,
years 2006-2010, base air quality.
5F-4
-------
Chicago
J5 £$ <5 <£> -O SS' SS' JS' JS* *i JS' ,f+ jw jjr ^*
Figure 5F-2. Percent of asthmatic school-age children with at least one daily maximum 8-
hr average Os exposure at or above 60, 70, and 80 ppb while at moderate or greater
exertion, years 2006-2010, base air quality.
5F-5
-------
to A^ ^ 0) d to A, C& O
'V *V 'V *V *V ^V 'V ^ 'V 'V 'V 'V *V ^V *V
Figure 5F-3. Percent of asthmatic adults with at least one daily maximum 8-hr average
exposure at or above 60, 70, and s 80 ppb while at moderate or greater exertion, years
2006-2010, base air quality.
5F-6
-------
* * / / /
-------
/ / / / / / / / / / / / / / /
Figure 5F-5. Percent of all school-age children with multiple daily maximum 8-hr average
Os exposures at or above 60 ppb per study area Os season, while at moderate or greater
exertion, years 2006-2010, base air quality.
5F-8
-------
/ /
«^ <£<§** <^
-------
Table 5F-1. Percent of all school-age children with Os exposures at or above 60, 70, and 80
ppb-8hr while at moderate or greater exertion, years 2006-2010, base air quality.
Study Area
Atlanta
Baltimore
Boston
Chicago
Exposure
Benchmark
(ppb-8hr)
60
70
80
60
70
80
60
70
80
60
70
80
Year
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
% of school-age children experiencing multiple exposures per O3 season
at or above benchmarks, base air quality
>1
42.4
41.8
27.3
16.7
20.3
25.2
23.8
9.4
3.8
4.4
10.6
9.3
1.8
0.5
0.4
36.8
29
22.6
10.1
31
19
11.4
8.4
1.1
11.8
5.8
2.7
2
0.1
2.5
22.7
31.8
18.3
12.3
12.9
7.1
14.9
4.6
2.6
2.3
1
4.8
0.6
0.1
0.1
14.7
26.5
3.6
7.8
15.6
2.7
9.2
0
0.8
1.5
0.1
1.2
0
0.1
>2
28.9
28.5
15
6.7
10.1
13
11.9
2.5
0.5
0.8
2.7
2.4
0.1
0
0
23.5
16.5
11.1
2.7
18.6
8.1
3.9
2.1
0.1
3.6
0.9
0.4
0.1
0
0.2
9.6
17.5
7.1
3.2
4
1.2
4.8
0.6
0.2
0.1
0
0.6
0
0
0
4.5
13.1
0.4
1.4
5
0.2
2
0
0
0.1
0
0
0
0
>3
21.9
21.5
9.2
2.9
5.6
7.3
6.4
0.7
0.1
0.2
0.8
0.7
0
0
0
16.6
10.9
6
0.8
12.5
3.8
1.7
0.6
0
1.3
0.1
0
0
0
0
4.5
10.8
2.8
0.8
1.3
0.2
1.8
0.1
0
0
0
0.1
0
0
0
1.4
7.1
0.1
0.3
1.7
0
0.4
0
0
0
0
0
0
0
>4
17.2
16.8
5.7
1.3
3.2
4.3
3.6
0.2
0
0
0.2
0.2
0
0
0
12.2
7.2
3.6
0.3
8.6
1.8
0.7
0.2
0
0.5
0
0
0
0
0
2.1
6.8
1.2
0.1
0.4
0
0.7
0
0
0
0
0
0
0
0
0.5
3.9
0
0
0.5
0
0.1
0
0
0
0
0
0
0
>5
13.8
13.3
3.6
0.5
1.8
2.4
2
0.1
0
0
0
0.1
0
0
0
9
5.1
2.2
0.1
5.8
0.8
0.3
0.1
0
0.2
0
0
0
0
0
0.9
4.2
0.5
0
0.1
0
0.2
0
0
0
0
0
0
0
0
0.1
2.1
0
0
0.1
0
0
0
0
0
0
0
0
0
>6
11
10.5
2.3
0.2
1.1
1.3
1.1
0
0
0
0
0
0
0
0
6.7
3.7
1.3
0
4
0.4
0.2
0
0
0.1
0
0
0
0
0
0.4
2.6
0.2
0
0
0
0.1
0
0
0
0
0
0
0
0
0
1.2
0
0
0
0
0
0
0
0
0
0
0
0
5F-10
-------
Study Area
Cleveland
Dallas
Denver
Detroit
Exposure
Benchmark
(ppb-8hr)
60
70
80
60
70
80
60
70
80
60
70
80
Year
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
% of school-age children experiencing multiple exposures per O3 season
at or above benchmarks, base air quality
>1
0
21.4
29.7
21.5
10.2
20.6
4.2
12
6
0.8
4.1
0.2
2.5
0.6
0
0.3
41.3
27.1
21.9
30
21.1
22.1
10.4
5.6
12.2
5.3
6.3
2.9
0.5
2.5
0.6
31.9
24.1
27.3
14
18.4
11.7
5.2
5.3
1.6
1.6
1.6
0.3
0.6
0
0
22.3
34.4
15.6
12.8
16.8
6.1
16.3
3
2.2
2.6
0.8
4
0
0.1
0.1
>2
0
9.1
16.4
9.7
2.6
8.9
0.4
3.6
0.9
0
0.6
0
0.2
0
0
0
28.4
14.1
10.3
17
9.6
11.1
2.3
0.9
3.6
0.9
1.3
0.1
0
0.2
0
19.6
13
15.7
5.1
8.6
4
1
0.7
0.1
0.1
0.1
0
0
0
0
9.1
20.1
5.3
3.6
6.3
0.8
6.1
0.2
0.2
0.3
0
0.6
0
0
0
>3
0
4.2
10.2
4.7
0.8
4.3
0.1
1.2
0.1
0
0.1
0
0
0
0
0
21.7
8
5.4
10.6
4.6
6.1
0.5
0.2
1.1
0.2
0.4
0
0
0
0
13.3
8.1
10.2
2
4.4
1.5
0.3
0.1
0
0
0
0
0
0
0
4.1
13.1
1.8
1
2.6
0.1
2.5
0
0
0
0
0.1
0
0
0
>4
0
1.8
6.4
2.2
0.2
2.1
0
0.4
0
0
0
0
0
0
0
0
17.2
4.5
2.8
6.7
2.2
3.4
0.1
0
0.3
0
0.1
0
0
0
0
9.5
5.3
6.8
0.8
2.4
0.6
0.1
0
0
0
0
0
0
0
0
1.7
8.5
0.6
0.3
1
0
1
0
0
0
0
0
0
0
0
>5
0
0.8
3.9
1.1
0.1
1
0
0.2
0
0
0
0
0
0
0
0
13.8
2.5
1.5
4.2
1
1.9
0
0
0.1
0
0
0
0
0
0
7
3.4
4.6
0.3
1.2
0.2
0
0
0
0
0
0
0
0
0
0.7
5.6
0.2
0.1
0.4
0
0.4
0
0
0
0
0
0
0
0
>6
0
0.3
2.5
0.5
0
0.4
0
0.1
0
0
0
0
0
0
0
0
11
1.3
0.7
2.6
0.4
1
0
0
0
0
0
0
0
0
0
5.2
2.2
3.1
0.1
0.6
0.1
0
0
0
0
0
0
0
0
0
0.3
3.6
0
0
0.2
0
0.1
0
0
0
0
0
0
0
0
5F-11
-------
Study Area
Houston
Los Angeles
New York
Philadelphia
Exposure
Benchmark
(ppb-8hr)
60
70
80
60
70
80
60
70
80
60
70
80
Year
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
% of school-age children experiencing multiple exposures per O3 season
at or above benchmarks, base air quality
>1
37.2
25.9
20.6
24.7
23.1
20.5
9.6
5.5
11.4
7.8
9.3
1.5
0.7
3.8
1.1
39.3
35
38.8
35.4
23.4
23.2
19.1
21.3
18.2
9.6
11.8
8.2
9.8
7.4
2.2
34.8
35.6
31.9
13.5
33.6
18
16.1
13
1.8
12.8
6.6
4.2
2.5
0
2.4
36.6
40.9
34.1
12.7
33.5
16.1
20.9
14.5
1
13.5
3.4
7
3
0
2.7
>2
23
13.4
8.9
11
11.1
8.6
2.2
0.8
2.2
1.5
2.3
0.1
0
0.2
0
24.1
20.1
23.6
20.4
12.2
11.3
7.8
10
7.4
3.1
3.9
1.8
3
1.5
0.3
19.1
21.1
17.2
4
19.4
5.9
5.6
3.4
0.1
3.8
0.8
0.5
0.2
0
0.2
22.4
26
19.8
3.7
19.8
5.9
8.6
4.6
0
4.3
0.4
1.1
0.3
0
0.3
>3
15.7
7.5
4.3
5.1
6
4
0.6
0.1
0.4
0.3
0.6
0
0
0
0
16.9
13.1
16.5
13.8
7.6
6.4
3.8
5.6
3.8
1.3
1.5
0.4
1.1
0.4
0.1
11.4
13.6
9.8
1.2
12.5
2
2
0.9
0
1.2
0.1
0.1
0
0
0
15.4
18.1
12.7
1.1
13.1
2.3
3.9
1.6
0
1.5
0
0.2
0
0
0
>4
11
4.2
2
2.3
3.2
1.9
0.1
0
0.1
0
0.1
0
0
0
0
12.6
9.4
12.2
10
5.1
3.9
1.9
3.3
2
0.6
0.5
0.1
0.5
0.1
0
7
9
5.7
0.3
8.3
0.6
0.7
0.2
0
0.3
0
0
0
0
0
10.8
13.2
8.4
0.3
8.9
0.9
1.8
0.6
0
0.5
0
0
0
0
0
>5
7.8
2.4
0.9
1
1.6
0.9
0
0
0
0
0
0
0
0
0
9.6
6.9
9.2
7.5
3.7
2.4
1
2.1
1.1
0.3
0.2
0
0.2
0
0
4.1
6.1
3.3
0.1
5.5
0.2
0.2
0
0
0.1
0
0
0
0
0
7.6
9.8
5.5
0.1
6.2
0.4
0.7
0.2
0
0.2
0
0
0
0
0
>6
5.5
1.2
0.4
0.4
0.8
0.4
0
0
0
0
0
0
0
0
0
7.4
5.2
7.2
5.7
2.7
1.5
0.5
1.3
0.6
0.1
0.1
0
0.1
0
0
2.5
4
1.8
0
3.6
0.1
0.1
0
0
0
0
0
0
0
0
5.4
7.1
3.7
0
4.3
0.1
0.3
0.1
0
0.1
0
0
0
0
0
5F-12
-------
Study Area
Sacramento
St. Louis
Washington DC
Exposure
Benchmark
(ppb-8hr)
60
70
80
60
70
80
60
70
80
Year
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
% of school-age children experiencing multiple exposures per O3 season
at or above benchmarks, base air quality
>1
40.9
24.7
36.3
35
20.8
22.8
9.2
19.1
16.7
7.4
8.8
2.7
7.6
5.2
2
32.4
38.4
11
11.3
22.9
12.7
20.9
1.4
1.9
4.9
2.2
6.8
0
0.2
0.1
35.4
35.7
26.2
8.9
32.1
18.4
16.6
10.3
0.7
12.5
5.7
3.9
2.5
0.1
2.4
>2
26.9
12.4
22.8
21.4
9.9
11.7
2.3
8.8
6.8
2
2.7
0.2
2.3
1
0.3
19.7
24.8
3
2.8
11
3.9
9.9
0.1
0
0.7
0.1
1.4
0
0
0
22.1
22.8
14.1
2
19.8
7.7
6.4
2.7
0
4.2
1
0.5
0.2
0
0.3
>3
20
7.2
16.1
14.5
5.6
6.8
0.7
4.7
3.2
0.7
0.9
0
0.9
0.3
0
13.2
17.8
0.9
0.8
5.8
1.3
5
0
0
0.1
0
0.3
0
0
0
15.3
16.1
8.4
0.5
13.5
3.5
2.7
0.7
0
1.6
0.1
0.1
0
0
0
>4
15.3
4.3
11.7
10.4
3.5
4.2
0.2
2.7
1.7
0.3
0.3
0
0.4
0.1
0
9
13
0.2
0.2
3.1
0.4
2.6
0
0
0
0
0.1
0
0
0
11
11.7
5
0.1
9.6
1.6
1.2
0.2
0
0.7
0
0
0
0
0
>5
12
2.7
8.9
7.6
2.3
2.7
0.1
1.7
0.8
0.1
0.2
0
0.1
0
0
6
9.9
0.1
0
1.6
0.1
1.3
0
0
0
0
0
0
0
0
8
8.8
2.9
0
6.9
0.7
0.6
0
0
0.3
0
0
0
0
0
>6
9.5
1.7
6.8
5.7
1.4
1.7
0
1.1
0.5
0
0
0
0.1
0
0
4
7.4
0
0
0.9
0
0.6
0
0
0
0
0
0
0
0
5.8
6.5
1.7
0
4.9
0.3
0.2
0
0
0.1
0
0
0
0
0
5F-13
-------
5F-2 EXPOSURE MODELING RESULTS FOR ADJUSTED AIR QUALITY
In this section, we present the exposures estimated when considering the air quality
adjusted to just meeting the existing Cb NAAQS standard, as well as when considering potential
alternative standard levels (55, 60, 65, 70 ppb 8-hr) of the existing standard. We note that one
study area (Chicago) Os ambient monitor design values were below that of the existing standard
during the 2008-2010, therefore APEX simulations could not be performed for that 3-year
period. We could not simulate just meeting a standard level of 60 ppb-8hr or below in the New
York study area, thus APEX simulations for these air quality scenarios could not be performed
for the New York study area.
First are presented three-paneled figures for each of the four exposure study groups of
interest (i.e. school-age children, asthmatic school-age children, asthmatic adults, older adults),
one panel of which was briefly summarized at the end of Chapter 5 in the key observation
section (all school-age children, 60 ppb-8hr benchmark). Presented for each of the three
exposure benchmarks (60 ppb-8hr, 70 ppb-8hr, 80 ppb-8hr) are the highest estimated percent
exposed while at moderate or greater exertion in each study area, considering just meeting the
existing and alternative standards (Figure 5F-7 to Figure 5F-10).
Exposures for the all school-age children study group were additionally characterized by
calculating the mean percent (averaged over the study years) experiencing at least one exposure
at or above each of the three benchmarks (60 ppb-8hr, 70 ppb-8hr, 80 ppb-8hr) while at
moderate or greater exertion (Figure 5F-11). Further, the maximum (Figure 5F-12) and mean
(Figure 5F-13) percent of all school-age children experiencing at least two exposures at or above
the three health effect benchmark levels are presented. Following these figures, Table 5F-2
provides the complete exposure output for all study areas, years, benchmark levels, and adjusted
air quality scenarios for all school-age children, the study group containing the greatest percent
and number of persons exposed in the FtREA.
And finally, the mean and maximum number of all school-age children and associated
person days with at least one exposure at or above each of the benchmark levels is provided in
Table 5F-3, by study area and air quality scenario. Table 5F-4 contains the total number of
persons experiencing at least one or two 8-hour exposures in all study areas by year, base air
quality and air quality adjusted to just meeting the existing 75 ppb standard. And finally, Table
5F-5 contains the number of school-age children and asthmatic school-age children experiencing
at least one or two exposures above the selected benchmark levels, summed for all 15 study
areas, and representing all simulated air quality scenarios.
5F-14
-------
Washington
0% 2% 4% 6% 8% 10°/c 12% 14% : 6% 1 8% 20% 22% 24% 26%
Percent ot'AJl School-Age Children with at Least One 8-hr Daily Max Exposure >= 5C ppb
standard level (ppb;; ^H 63 ^H 65 ^H 70 ^H 75
AtlanU
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroi;
Houston
Los Angeles
New Yuri
Philadelphia
Sacramento
3t Louis
I I
Washington |
] :
0% 1% 2% 3% 4% 5% 6% 7% 8°/t 9%
Percent ot All School-Age Children with a;Lesst One 8-hr Daily Max Exposure >=70 ppb
standard level = 30 ppb
standard level (ppb) •• 63 •• 65 ^M70 ^^75
Figure 5F-7. Incremental increases in percent of all school-age children with at least one
daily maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb (middle
panel), or 80 ppb (bottom panel) using the maximum percent exposed for each study area,
year 2006-2010 adjusted air quality.
5F-15
-------
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroi;
Houston
Los Angeles
New York
Philadelphia
Sacramento
3t Louis
Washington
1 |
1 1
1 I
1 1
1 1
1
1 1
1 1
1
1 1
1 I
0% 2% 4% 6% 8% 10°/c 12% 14% :&/, 18% 20% 22% 24% 2
Percent of Asthmatic Children (5-18) will at Least One 8-hr Daly Max Exposure > 60 pnb
standard level (ppb; i^B c'O ^H 65 i^B70 i^B75
Atlanta
Baltimore
Boston
C!hirago
Clew land
1
I
1
I i
1 1
Dallas JJ |
Denver |
Detroi;
Houston
Los Angeles
New Yori
Philadelphia
Sacramento
St Louis
Washington
] |
I
i
0% 1% 2% 3°/o 4% 5% 6% 7% S% 9%
Percent ?.f Asthmatic Children (5-1$) with at Least Ore 5-hr Daily Max Zxpcsure >70 ppb
standard level (jpb; i^B <:~O ••165 BB170 ••§ 75
St Louis
Washington
3.0M 0.:% 0.2M 0.3% 0/W O.fK 0.6% 0.7% 0.8% 0.9M, l.OM, 1.1% 1.214 1.314 1.1%
Percent of Asthmatic Children (5-1S) with at Least On3 8-hr Doily Max Exposure > 3C ppb
standard level (ppb; •• 63 I 1 65 I »70 •• 75
Figure 5F-8. Incremental increases in percent of asthmatic school-age children with at
least one daily maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb
(middle panel), or 80 ppb (bottom panel) using the maximum percent exposed for each
study area, year 2006-2010 adjusted air quality.
5F-16
-------
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
1
1
I
i
0.0% 0.2% 0.4% 0.6% 0.8% 1.0% 1.2% 1.4% 1.6% 1.8% 2.0% 2.2%
Percent of Asthmatic Adults (19-95) with at Least One 8-hr Daily Max Exposure > 70 ppb
standard level (ppb) I 1 60 [=] 65 •—I 70 I 1 75
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
•
1
ZD
_l 1
p
1
1
1
0.00% 0.02% 0.04% 0.06% 0.08% 0.10% 0.12% 0.14% 0.16% 0.18% 0.20% 0.2;
Percent of Asthmatic Adults (19-95) with at Least One 8-hr Daily Max Exposure > 80 ppb
standard level (ppb) i 60 i 65 i 70 i 75
Figure 5F-9. Incremental increases in percent of asthmatic adults with at least one daily
maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb (middle panel),
or 80 ppb (bottom panel) using the maximum percent exposed for each study area, year
2006-2010 adjusted air quality.
5F-17
-------
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
Percent of Older Adults (65-95) with at Least One 8-hr Daily Max Exposure > 60 ppb
standard level (ppb) I I 60 I 1 65 ^M 70 I I 75
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
0.0% 0.2% 0.4% 0.6% 0.8% 1.0% 1.2% 1.4% 1.6% l.i
Percent of Older Adults (65-95) with at Least One 8-hr Daily Max Exposure > 70 ppb
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
0.00% 0.05% 0.10% 0.15% 0.20% 0.25% 0.
Percent of Older Adults (65-95) with at Least One 8-hr Daily Max Exposure > 80 ppb
standard level (ppb) i i 60 i i 65 i i 70 i i 75
Figure 5F-10. Incremental increases in percent of all older adults with at least one daily
maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb (middle panel),
or 80 ppb (bottom panel) using the maximum percent exposed for each study area, year
2006-2010 adjusted air quality.
5F-18
-------
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
I
I
1
i
I
i
1
i
Percent of All School-Age Children with at Least One 8-hr Daily Max Exposure >= 60 ppb
standard level (ppb) I I 60 I 1 65 ^B 70 I I 75
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
i
i
I
i
1
i
i
1
1
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5%
Percent of All School-Age Children with at Least One 8-hr Daily Max Exposure >= 70 ppb
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
0.0
H
1
1
0% 0.05% 0.10% 0.15% 0.20% 0.25% 0.30% 0.35
Percent of All School- Age Children with at Least One 8-hr Daily Max Exposure >= 80 ppb
standard level (ppb) i i 60 i i 65 i i 70 i i 75
Figure 5F-11. Incremental increases in percent of all school-age children with at least one
daily maximum 8-hr average Os exposure at or above 60 ppb (top panel), 70 ppb (middle
panel), or 80 ppb (bottom panel) using the mean percent exposed for each study area, year
2006-2010 adjusted air quality.
5F-19
-------
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroi;
Houston
Los Angeles
New York
Philadelphia
Sacramento
3t Louis
Washington
0°
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroi;
Houston
Los Angeles
New Yore
Philadelphia
Sacramento
St Louis
Washington
| | |
1
1 1
1
1
J | |
| |
1 1
•b 1% 2% 3% 4% 5"b 6% 7% 8% 9% 10% 11% 12% 13% 14% -.5°
Percent of All School-Age Children with at Least Two 8-hr Daily Max Exposure >= 60 pnb
standard level (ppb; ^H 6j ^H 65 ^•70 ^B 7:1
1
ID
1
_l 1
1 1
5.0°/o 0.2% C.4% 0.6% 0.8W 1.0% 1.2% 1.4% 1.6% 1.8% 2.C% 2.2%
Percent of All School-Age Children with at L?a;t Two 8-hr Daily Max Exposure >=7Cppb
standard level (ppb) ^^ 65 ^^ 65 ^H 70 ^^ 75
0.00^0 C.01W 0.029/0 0.039/0 O.CVW 0.35% 0.06% D.07%
Percent of All School-Age Children with at Least Two S-hrDail/McxZxpcsu-c -
standard level (jspb) ^^M 63 i i 65 i i 70 i i 75
Figure 5F-12. Incremental increases in percent of all school-age children with at least two
daily maximum 8-hr average Os exposures at or above 60 ppb (top panel), 70 ppb (middle
panel), or 80 ppb (bottom panel) using the maximum percent exposed for each study area,
year 2006-2010 adjusted air quality.
5F-20
-------
Percent of All School-Age Children with at Least Two 8-hr Daily Max Exposure >= 60 ppb
standard level (ppb) I I 60 I 1 65 ^M 70 I I 75
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
i
1
1
1
Philadelphia |
Sacramento
St. Louis
I
0.00% 0.10% 0.20% 0.30% 0.40% 0.50% 0.60
Percent of All School-Age Children with at Least Two 8-hr Daily Max Exposure >= 70 ppb
Atlanta
Baltimore
Boston
Chicago
Cleveland
Dallas
Denver
Detroit
Houston
Los Angeles
New York
Philadelphia
Sacramento
St. Louis
Washington
1
ID
1
1
1
0.000% 0.002% 0.004% 0.006% 0.008% 0.010% 0.012% 0.014%
Percent of All School-Age Children with at Least Two 8-hr Daily Max Exposure >= 80 ppb
standard level (ppb) i i 60 i i 65 i i 70 i i 75
Figure 5F-13. Incremental increases in percent of all school-age children with at least two
daily maximum 8-hr average Os exposures at or above 60 ppb (top panel), 70 ppb (middle
panel), or 80 ppb (bottom panel) using the mean percent exposed for each study area, year
2006-2010 adjusted air quality.
5F-21
-------
Table 5F-2. Percent of all school-age children with Os exposures at or above 60, 70, and 80
ppb-8hr while at moderate or greater exertion, years 2006-2010, adjusted air quality.
Study Area
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Exposure
Benchmark
(ppb-8hr)
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
Air Quality
Scenario1
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.3
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1.2
0.8
0
0.8
0.2
0.3
0.1
0
0
0
0
0
0
0
0
0
0
0
4.8
4.2
0.5
3.3
1.4
2.2
0.5
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0.7
0
0.6
0.1
0.3
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.2
0
0.1
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-22
-------
Study Area
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Exposure
Benchmark
(ppb-8hr)
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
Air Quality
Scenario1
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
Year
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.2
0
0.2
0
0
0
0
0
0
0
0
10.8
9.4
1.8
10
4.4
6.9
1.4
1
0.1
1.2
0.4
0.3
0.2
0.1
0
0.1
0
0
19.3
17.7
5.4
19.3
10.6
14
4.4
3.7
0.4
4.2
1.6
1.9
0.7
0.4
0
0.5
0.1
>2
0
0
0
0
0
0
0
0
0
0
0
3.3
2.8
0.2
3.1
0.8
1.8
0.1
0.1
0
0.1
0
0
0
0
0
0
0
0
8.9
7.9
1.2
8.7
3
5.5
0.7
0.6
0
0.7
0.1
0.2
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
1.2
0.9
0
1.1
0.2
0.5
0
0
0
0
0
0
0
0
0
0
0
0
4.5
3.9
0.3
4.5
1
2.5
0.1
0.1
0
0.1
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0.4
0.3
0
0.4
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
2.4
2.1
0.1
2.4
0.4
1.2
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
0
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
1.4
1.1
0
1.3
0.1
0.6
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.7
0.5
0
0.7
0
0.3
0
0
0
0
0
0
0
0
0
0
0
5F-23
-------
Study Area
Atlanta
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Exposure
Benchmark
(ppb-8hr)
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
Air Quality
Scenario1
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
Year
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.1
0.1
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
1
0.4
0.3
0.6
0
1.2
0
0
0
0
0
0
0
0
0
0
0
0
5.4
3.3
2.1
2.5
0.1
3.8
0.3
0.2
0.1
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.9
0.4
0.1
0.2
0
0.5
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0.1
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-24
-------
Study Area
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Boston
Exposure
Benchmark
(ppb-8hr)
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
Air Quality
Scenario1
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
Year
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.1
0
0.3
0
0
0
0
0
0
11.8
7.7
5.3
6.3
0.6
9.5
1.2
0.6
0.4
0.6
0
1
0.1
0.1
0
0
0
0.1
19
13.6
9.7
10.7
1.9
16.2
4
2
1.4
1.7
0.1
2.6
0.3
0.2
0.1
0.1
0
0.4
0
>2
0
0
0
0
0
0
0
0
0
3.7
1.8
0.8
1.2
0
2.5
0.1
0
0
0
0
0
0
0
0
0
0
0
8.4
5.1
2.5
3
0.2
6.4
0.5
0.2
0.1
0.1
0
0.2
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
1.2
0.5
0.2
0.3
0
0.7
0
0
0
0
0
0
0
0
0
0
0
0
4.1
2.1
0.8
1
0
2.8
0.1
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0.5
0.2
0
0.1
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
2.1
0.9
0.3
0.3
0
1.3
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0.2
0.1
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
1.1
0.4
0.1
0.1
0
0.6
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0.2
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-25
-------
Study Area
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Exposure
Benchmark
(ppb-8hr)
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
Air Quality
Scenario1
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
Year
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.2
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
1.7
0
0.3
1.3
0.4
0
0
0
0
0
0
0
0
0
0
0
0
1.2
6.7
0.8
2.9
4.6
2.5
0
0.5
0
0.1
0.2
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
1.1
0
0.3
0.5
0.2
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-26
-------
Study Area
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Boston
Chicago
Chicago
Chicago
Exposure
Benchmark
(ppb-8hr)
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
Air Quality
Scenario1
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
Year
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0
0
0
6.7
15.7
4.7
9.1
9
6.7
0.3
3.2
0.2
0.8
1.3
0.7
0
0.2
0
0.1
0
0
11.9
21.9
9.1
15.9
11.4
11.2
1.4
6.6
0.8
3.2
2.1
1.7
0
1
0.1
0.4
0.1
0.1
0.1
0.3
0
>2
0
0
0
0
0
0
0
1.2
5.5
0.7
2.1
1.8
1.1
0
0.4
0
0
0
0
0
0
0
0
0
0
3.1
9.7
2.1
5.5
2.8
3
0.1
1.1
0
0.3
0.1
0.1
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0.2
2
0.1
0.5
0.2
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0.9
4.6
0.5
2.1
0.6
0.9
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0.8
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
2.3
0.1
0.8
0.1
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1.1
0
0.3
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-27
-------
Study Area
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Exposure
Benchmark
(ppb-8hr)
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
Air Quality
Scenario1
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
Year
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0.3
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0.7
2.2
0
0.1
1.1
2
0
0.1
0
0
0.1
0
0
0
0
0
0
0
2.9
8.1
0.2
0.8
2.9
6.5
0.1
0.4
0
0
0.3
0.2
0
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0
0
0.1
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0.3
1.8
0
0
0.4
1.5
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0
0
0.1
0.4
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0.1
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-28
-------
Study Area
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Exposure
Benchmark
(ppb-8hr)
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
70
70
70
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
Air Quality
Scenario1
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(06-08)
75(06-08)
75(06-08)
75(06-08)
75(06-08)
75(06-08)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
Year
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2006
2007
2008
2006
2007
2008
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0
7.7
16
1.1
2.9
6.6
13.6
0.5
2.7
0
0
0.6
1
0
0.1
0
0
0.1
0
13.5
24.7
3
2.1
7.5
0
0
0.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>2
0
0
0
0
0
1.6
5.7
0.1
0.3
1.2
4.1
0
0.2
0
0
0
0.1
0
0
0
0
0
0
3.9
11.6
0.3
0.1
1.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0.4
2.1
0
0
0.2
1.3
0
0
0
0
0
0
0
0
0
0
0
0
1.2
6
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0.1
0.8
0
0
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0.4
3.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0.3
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.1
1.6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-29
-------
Study Area
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Exposure
Benchmark
(ppb-8hr)
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
Air Quality
Scenario1
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0.1
0.4
0
0
0.2
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.5
3
1.2
0.6
0.7
0.5
0
0.2
0
0
0
0
0
0
0
0
0
0
3.1
9.3
5.3
4.1
1.9
2.2
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
2.6
0.9
0.6
0.2
0.3
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.7
0.2
0.1
0
0.1
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-30
-------
Study Area
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Exposure
Benchmark
(ppb-8hr)
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
Air Quality
Scenario1
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.1
0.9
0.4
0.2
0.2
0.1
0
0
0
0
0
0
9.3
18
11.7
10.6
4.5
8.2
0.6
3.7
1.8
1.5
0.5
0.5
0
0.2
0.1
0
0
0
0.1
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>2
0
0
0
0
0
0
0
0
0
0
0
0
2.2
7.5
3.4
2.9
0.9
2
0
0.5
0.2
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0.6
3.4
1.2
1
0.2
0.6
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0.2
1.6
0.4
0.4
0.1
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0.7
0.1
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-31
-------
Study Area
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Exposure
Benchmark
(ppb-8hr)
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
Air Quality
Scenario1
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
1.9
1.1
0.1
0.1
0.7
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
7.6
2.6
0.7
0.7
3.1
1
0.2
0.3
0
0
0.1
0
0
0
0
0
0
0
16
6.4
3
3
8.4
3.7
1.5
1
>2
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0.1
0
0
0.4
0.1
0
0
0
0
0
0
0
0
0
0
0
0
7.1
1
0.4
0.4
1.9
0.5
0.1
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3.4
0.1
0.1
0.1
0.5
0.1
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1.7
0
0
0
0.1
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0
0
0
0
0
0
0
5F-32
-------
Study Area
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Exposure
Benchmark
(ppb-8hr)
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
Air Quality
Scenario1
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.1
0.1
0.6
0.1
0
0.1
0
0
0
0
22.9
10.9
6.7
7.9
14.9
8.3
4.5
1.9
0.3
0.5
2.2
0.6
0.2
0.3
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>2
0
0
0
0
0
0
0
0
0
0
12.2
2.8
1.4
1.9
5.5
2
0.8
0.1
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
7.3
0.8
0.4
0.6
2.2
0.5
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
4.4
0.2
0.1
0.2
0.8
0.2
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
2.7
0
0
0
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
1.6
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-33
-------
Study Area
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Exposure
Benchmark
(ppb-8hr)
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
Air Quality
Scenario1
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.5
0.1
0.4
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4.4
2
2.8
9.5
3.1
3.3
0
0
0
0.4
0
0
0
0
0
0
0
0
12.9
7
8.9
18.9
7.8
9.5
0.7
0.2
0.4
1.7
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0.3
0.3
2.8
0.4
0.7
0
0
0
0
0
0
0
0
0
0
0
0
4.4
2
2.5
9.2
2
3.1
0
0
0
0.1
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.1
0.1
1
0.1
0.2
0
0
0
0
0
0
0
0
0
0
0
0
1.8
0.7
0.9
5
0.6
1.2
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0.3
0.3
2.8
0.2
0.4
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0.1
0.1
1.5
0.1
0.2
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0.1
0.9
0
0.1
0
0
0
0
5F-34
-------
Study Area
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Denver
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Exposure
Benchmark
(ppb-8hr)
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
Air Quality
Scenario1
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.4
0.2
0
0
0
0.1
0
0
21.3
13.8
16.7
25.6
12.5
16.3
2.9
0.9
1.2
4.1
1.2
1
0.1
0
0
0.5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.7
>2
0
0
0
0
0
0
0
0
10.4
5.4
7.6
14.4
4.2
7.1
0.3
0.1
0
0.4
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
5.7
2.5
3.8
9.1
1.5
3.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
3.1
1.3
1.9
5.9
0.6
1.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
1.8
0.7
1
3.8
0.2
0.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
1
0.4
0.6
2.5
0.1
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-35
-------
Study Area
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Exposure
Benchmark
(ppb-8hr)
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
Air Quality
Scenario1
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-1 0)
65(08-10)
65(08-1 0)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-1 0)
70(08-1 0)
70(08-1 0)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0.4
0.3
0.2
0
0
0
0
0
0
0
0
0
0
0
0
2
4.6
0.7
3.6
2.8
3.1
0
0.2
0
0
0.1
0
0
0
0
0
0
0
4.9
10.3
2.5
9
6.9
8.7
0.2
0.9
0
0.9
0.5
0.5
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
1.1
0
0.5
0.3
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0.8
3.6
0.2
2.3
1.4
2.5
0
0.1
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.1
1.4
0
0.6
0.3
0.8
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0
0.1
0.1
0.2
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0.1
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
5F-36
-------
Study Area
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Exposure
Benchmark
(ppb-8hr)
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
Air Quality
Scenario1
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-1 0)
75(08-1 0)
75(08-1 0)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0
0
10.6
19.1
6.7
15.6
12.8
16.8
1.4
4.2
0.3
3
2.2
2.6
0
0.2
0
0
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0
0.1
0.2
>2
0
0
0
0
0
0
3
8.6
1.4
5.3
3.6
6.3
0
0.8
0
0.2
0.2
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0.9
4.3
0.3
1.8
1
2.6
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0.3
2.3
0.1
0.6
0.3
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0.1
1.3
0
0.2
0.1
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0.7
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-37
-------
Study Area
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Exposure
Benchmark
(ppb-8hr)
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
Air Quality
Scenario1
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.7
0.2
0
0
0
0
0
0
0
0
0
0
0
0
3
0.2
0.3
2.1
5.7
3.3
0.2
0
0
0
0.4
0
0
0
0
0
0
0
7
1.4
0.8
6.9
11.9
9
1
0
0
0.5
2.1
0.6
0.1
0
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.6
0
0
0.2
0.7
0.3
0
0
0
0
0
0
0
0
0
0
0
0
1.9
0.1
0
1.4
2.9
2.2
0.1
0
0
0
0.1
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0.6
0
0
0.3
0.8
0.6
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0.2
0.2
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-38
-------
Study Area
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Houston
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Exposure
Benchmark
(ppb-8hr)
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
Air Quality
Scenario1
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0.1
0
12
4.1
2.5
13.2
17.8
15.3
2.5
0.1
0.1
1.9
5.5
2.4
0.3
0
0
0.2
0.7
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
>2
0
0
0
0
4.1
0.7
0.2
4.4
6.3
5.6
0.4
0
0
0.1
0.6
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
1.7
0.2
0
1.5
2.4
2.2
0.1
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0.7
0
0
0.5
0.9
0.9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0.3
0
0
0.2
0.3
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0.1
0
0
0.1
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-39
-------
Study Area
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Exposure
Benchmark
(ppb-8hr)
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
Air Quality
Scenario1
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0
0
0
0
0
0
0
0
0.9
1.5
0.9
1.2
1.2
0.8
0
0
0
0
0
0
0
0
0
0
0
0
4.4
4.9
4.2
4.9
5
3.2
0
0.2
0
0
0
0
0
0
0
0
>2
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.3
0.3
0.3
0.3
0.2
0
0
0
0
0
0
0
0
0
0
0
0
1.6
1.6
1.5
1.8
1.8
1.3
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
0.1
0.1
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0.8
0.8
0.9
0.9
0.7
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0.5
0.4
0.6
0.5
0.4
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0.3
0.3
0.3
0.3
0.2
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.2
0.2
0.2
0.2
0.1
0
0
0
0
0
0
0
0
0
0
5F-40
-------
Study Area
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
Exposure
Benchmark
(ppb-8hr)
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
Air Quality
Scenario1
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-1 0)
70(08-1 0)
70(08-1 0)
70(06-08)
70(06-08)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
10.2
10.2
9.9
10.2
10
6.9
0.5
1
0.5
0.6
0.5
0.4
0
0.1
0
0
0
0
0
0
0
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
2.6
2
2.3
5.7
1.2
6.6
0.2
0
>2
0
0
4.5
4.3
4.3
4.5
4.4
2.9
0.1
0.1
0.1
0.1
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.3
0.3
1.1
0.1
1.4
0
0
>3
0
0
2.5
2.4
2.4
2.5
2.5
1.7
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
0.3
0
0.3
0
0
>4
0
0
1.6
1.4
1.5
1.6
1.5
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0.1
0
0
>5
0
0
1.1
0.9
1
1.1
1
0.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0.7
0.7
0.7
0.7
0.7
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-41
-------
Study Area
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Exposure
Benchmark
(ppb-8hr)
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
Air Quality
Scenario1
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-1 0)
75(08-1 0)
75(08-1 0)
75(06-08)
75(06-08)
75(06-08)
75(08-1 0)
75(08-10)
75(08-1 0)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.1
0.4
0
0.5
0
0
0
0
0
0
9.1
8.5
7.8
17.2
5.4
19
1.5
0.4
0.7
3.4
0.3
3.7
0.1
0
0
0.3
0
0.3
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
>2
0
0
0
0
0
0
0
0
0
0
1.9
2.2
1.7
6.3
0.9
8
0.1
0
0
0.4
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0.4
0.7
0.5
2.4
0.2
3.6
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0.1
0.2
0.2
1
0
1.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
0.4
0
0.9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-42
-------
Study Area
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Exposure
Benchmark
(ppb-8hr)
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
Air Quality
Scenario1
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.1
0.7
0.1
1
0
1.7
0
0
0
0
0
0.1
0
0
0
0
0
0
1.3
3.3
1
4.3
0
4.6
0
0.1
0
0.1
0
0.3
0
0
0
0
0
0
5.3
9.9
4.3
11.8
0.8
11.6
0.1
0.9
0.1
1.1
>2
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.4
0
0.6
0
0.6
0
0
0
0
0
0
0
0
0
0
0
0
0.9
2.4
0.6
3.2
0
3.3
0
0.1
0
0.1
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.7
0.1
1.1
0
1.1
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0.3
0
0.4
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0.1
0
0.1
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
5F-43
-------
Study Area
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Exposure
Benchmark
(ppb-8hr)
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
Air Quality
Scenario1
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
1.5
0
0
0
0
0
0.1
12.1
17.5
10
20.5
4
20.2
0.8
3
0.8
4.1
0
4.2
0
0.3
0
0.3
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0.4
>2
0
0.1
0
0
0
0
0
0
3.7
6.6
2.4
8.6
0.4
8.7
0
0.3
0
0.5
0
0.5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
1.3
2.7
0.7
3.9
0
4.3
0
0.1
0
0.1
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0.4
1.2
0.2
1.9
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0.2
0.6
0.1
0.8
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0.1
0.2
0
0.4
0
0.5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-44
-------
Study Area
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Exposure
Benchmark
(ppb-8hr)
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
Air Quality
Scenario1
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.8
0.8
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
4.7
1.6
4.3
3.8
2.2
0.9
0
0.2
0.2
0.2
0
0
0
0
0
0
0
0
10
3
8.3
7.6
5.9
2.2
0.4
0.5
0.9
0.8
0.1
0.1
>2
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.9
0
0.9
0.7
0.2
0.1
0
0
0
0
0
0
0
0
0
0
0
0
3.4
0.3
2.4
2.1
1.3
0.3
0
0
0.1
0.1
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0.2
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1.3
0
0.9
0.8
0.3
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0
0.4
0.3
0.1
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0.1
0.1
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0.1
0
0
0
0
0
0
0
0
0
5F-45
-------
Study Area
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
Exposure
Benchmark
(ppb-8hr)
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
Air Quality
Scenario1
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-10)
55(08-10)
55(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0
0
16.5
6
13.5
12.8
11.2
4.5
2.6
1.1
2.7
2.3
1.1
0.5
0
0.2
0.2
0.2
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
1.5
0
0.4
>2
0
0
0
0
0
0
7.4
1.2
5.4
4.9
3.9
0.9
0.4
0
0.5
0.4
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
>3
0
0
0
0
0
0
3.7
0.3
2.5
2.2
1.5
0.2
0
0
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
2
0.1
1.2
1.1
0.6
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
1.1
0
0.7
0.6
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0.6
0
0.4
0.3
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-46
-------
Study Area
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
Exposure
Benchmark
(ppb-8hr)
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
Air Quality
Scenario1
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
60(06-08)
60(06-08)
60(06-08)
60(08-10)
60(08-10)
60(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
65(06-08)
65(06-08)
65(06-08)
65(08-10)
65(08-10)
65(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
70(06-08)
70(08-10)
70(08-10)
70(08-10)
70(06-08)
70(06-08)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0.4
0.5
0
0
0
0
0
0
0
0
0
0
0
0
2.5
7.3
0.2
2.4
2.4
6.3
0
0.4
0
0.1
0.2
0.1
0
0
0
0
0
0
9.3
16.9
1.2
6
6.2
14.9
0.5
2.7
0
0.5
0.7
0.9
0
0.2
>2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
2
0
0.2
0.1
1.1
0
0
0
0
0
0
0
0
0
0
0
0
2.3
7
0.1
1.1
0.9
5.2
0
0.3
0
0
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.6
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0.7
3.4
0
0.2
0.1
1.9
0
0.1
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
1.7
0
0.1
0
0.7
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.8
0
0
0
0.2
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0
0
0
0.1
0
0
0
0
0
0
0
0
5F-47
-------
Study Area
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Exposure
Benchmark
(ppb-8hr)
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
Air Quality
Scenario1
70(06-08)
70(08-10)
70(08-10)
70(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
55(06-08)
55(06-08)
55(06-08)
55(08-1 0)
55(08-1 0)
55(08-1 0)
55(06-08)
55(06-08)
55(06-08)
55(08-1 0)
55(08-1 0)
55(08-1 0)
55(06-08)
55(06-08)
55(06-08)
55(08-1 0)
55(08-1 0)
55(08-1 0)
60(06-08)
60(06-08)
60(06-08)
60(08-1 0)
60(08-1 0)
60(08-1 0)
Year
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0.1
0
18.1
25.8
3.5
9.5
10
21.1
2.9
8.1
0.2
1.1
1.4
3.6
0.1
1.1
0
0
0.2
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.6
0.1
0.1
0.3
0
0.6
>2
0
0
0
0
7.8
13.8
0.5
2.4
2.2
9.6
0.2
2.2
0
0
0
0.4
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>3
0
0
0
0
3.7
8.2
0.1
0.6
0.5
4.8
0
0.7
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
1.6
5
0
0.2
0.1
2.3
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0.8
3
0
0.1
0
1.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0.4
1.8
0
0
0
0.6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5F-48
-------
Study Area
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Exposure
Benchmark
(ppb-8hr)
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
Air Quality
Scenario1
60(06-08)
60(06-08)
60(06-08)
60(08-1 0)
60(08-1 0)
60(08-1 0)
60(06-08)
60(06-08)
60(06-08)
60(08-1 0)
60(08-1 0)
60(08-1 0)
65(06-08)
65(06-08)
65(06-08)
65(08-1 0)
65(08-1 0)
65(08-1 0)
65(06-08)
65(06-08)
65(06-08)
65(08-1 0)
65(08-1 0)
65(08-1 0)
65(06-08)
65(06-08)
65(06-08)
65(08-1 0)
65(08-1 0)
65(08-1 0)
70(06-08)
70(06-08)
70(06-08)
70(08-1 0)
70(08-1 0)
70(08-1 0)
70(06-08)
70(06-08)
70(06-08)
70(08-1 0)
70(08-1 0)
70(08-1 0)
70(06-08)
70(06-08)
70(06-08)
70(08-1 0)
Year
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0
0
0
0
0
0
0
0
0
0
0
2.4
1.8
0.9
3.4
0.1
5
0.1
0
0
0.2
0
0.2
0
0
0
0
0
0
6.7
6.5
3.4
9.2
1
12.5
0.7
0.1
0.2
1.3
0
1.4
0
0
0
0.1
>2
0
0
0
0
0
0
0
0
0
0
0
0
0.3
0.2
0.1
0.6
0
1.2
0
0
0
0
0
0
0
0
0
0
0
0
1.8
1.7
0.6
2.7
0.1
5
0
0
0
0.1
0
0.1
0
0
0
0
>3
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0.1
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0.6
0.1
0.8
0
2.4
0
0
0
0
0
0
0
0
0
0
>4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0.2
0
0.2
0
1.3
0
0
0
0
0
0
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
0
0.1
0
0.7
0
0
0
0
0
0
0
0
0
0
>6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0
0
0
0
0
0
0
0
0
0
5F-49
-------
Study Area
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Exposure
Benchmark
(ppb-8hr)
80
80
60
60
60
60
60
60
70
70
70
70
70
70
80
80
80
80
80
80
Air Quality
Scenario1
70(08-1 0)
70(08-1 0)
75(06-08)
75(06-08)
75(06-08)
75(08-1 0)
75(08-1 0)
75(08-1 0)
75(06-08)
75(06-08)
75(06-08)
75(08-1 0)
75(08-1 0)
75(08-1 0)
75(06-08)
75(06-08)
75(06-08)
75(08-10)
75(08-10)
75(08-10)
Year
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
2006
2007
2008
2008
2009
2010
% of all school-age children experiencing multiple exposures per
O3 season at or above benchmarks, adjusted air quality
>1
0
0.1
12.5
12.9
7.3
18.5
4.1
23.4
2
1.1
0.8
4.8
0.2
6
0.1
0
0
0.8
0
0.6
>2
0
0
4.6
4.9
1.8
8.2
0.6
12.5
0.2
0.1
0
0.8
0
1.4
0
0
0
0
0
0
>3
0
0
2.1
2.2
0.5
4.1
0.1
7.7
0
0
0
0.1
0
0.4
0
0
0
0
0
0
>4
0
0
1
1.1
0.1
2
0
4.9
0
0
0
0
0
0.1
0
0
0
0
0
0
>5
0
0
0.4
0.5
0
0.9
0
3.3
0
0
0
0
0
0
0
0
0
0
0
0
>6
0
0
0.2
0.3
0
0.4
0
2.1
0
0
0
0
0
0
0
0
0
0
0
0
1 Abbreviation indicates 8 hour standard level and three year averaging period. For example, 75(08-10) represents
simulated ambient concentrations just meeting the existing standard (75 ppb-8hr) using air quality years 2008-2010.
Values indicated as zero represent values of < 0.05% or actual zero (no one experienced an exposure.
5F-50
-------
Table 5F-3. Mean and maximum number of all school-age children (and associated days
per Os season) with at least one daily maximum 8-hr average Os exposure at or above 60
ppb-8hr while at moderate or greater exertion.
Study area
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Atlanta
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Boston
Boston
Boston
Boston
Boston
Boston
Chicago
Chicago
Chicago
Chicago
Chicago
Chicago
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Dallas
Dallas
Dallas
Dallas
Dallas
Dallas
Denver
Denver
Denver
Denver
Denver
Denver
Air quality
scenario/
standard
level (ppb)
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
Mean
Number of
children
255487
1 27378
64409
24933
5064
818
1 30826
61511
35864
15050
3109
346
1 77358
1 24529
81220
30411
6710
538
258923
260946
174401
79122
22578
3831
119740
59189
24462
6525
838
81
310037
1 41 1 1 9
82161
32893
8561
846
1 2961 5
95296
57266
21348
1177
21
Number of
days per year
827740
229661
90926
29770
5359
841
357038
1 03377
49008
17414
3218
346
331 700
1 90220
1 08757
34767
7064
538
429030
449687
246130
99309
26014
4223
234478
88930
31229
7396
864
83
853320
247643
1 22988
41095
9159
860
328720
1 87450
91933
27215
1192
21
Maximum
Number of
children
36491 6
166169
93000
41687
10128
2522
1 86072
95781
59793
27416
5888
817
28771 3
198171
141821
60377
15506
1713
503935
47001 1
303354
1 53346
41894
5842
172100
1 04388
53736
17557
2195
188
452737
251 505
175109
83162
20508
2913
1 78689
1 43603
1 06034
53234
2863
39
Number of
days per year
1496000
325800
1 36700
50720
10840
2599
609600
1 79700
88270
33050
5998
817
699100
365900
220300
72810
16620
1713
1046000
922700
477300
1 98700
48110
6885
423200
1 83800
75140
20980
2289
189
1 894000
588800
327400
117000
23050
2913
563800
367300
221 600
78130
2928
39
5F-51
-------
Study area
Detroit
Detroit
Detroit
Detroit
Detroit
Detroit
Houston
Houston
Houston
Houston
Houston
Houston
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
New York
New York
New York
New York
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Philadelphia
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
Sacramento
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
St. Louis
Washington
Washington
Washington
Washington
Washington
Washington
Air quality
scenario/
standard
level (ppb)
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
Mean
Number of
children
207174
1 43352
74557
29881
3329
82
254991
110832
64399
25986
2565
41
1244571
342236
1 59498
39327
1548
15
1 1 48294
41 8702
125784
1602
388598
170184
87649
29333
7291
699
1 47074
47859
27070
12503
1744
0
1 22408
86067
53629
20760
2915
140
267667
1 27234
63711
22251
2645
0
Number of
days per year
402320
23001 5
1 02732
35337
3457
82
574580
1 70062
84835
29950
2610
41
3978000
741990
290320
57289
1547
15
2740240
635750
1 53098
1880
1 094200
279328
1 1 7594
33504
7693
699
469140
80022
37802
14970
1830
0
315640
1 62499
82013
25857
3100
143
771500
246782
97776
27721
2776
0
Maximum
Number of
children
349520
1 94330
1 04733
46936
6699
320
360732
173115
115161
55481
6888
236
1 4231 98
368974
179329
54045
5831
75
1368877
729630
253458
3241
503583
252907
145466
56832
20486
2891
1 90752
76891
46556
22069
3892
0
202543
136172
89003
38381
7840
503
345115
226043
1 21 074
48425
5578
0
Number of
days per year
922500
374700
1 63400
61570
7225
320
1082000
270900
1 52500
63640
6947
236
4981 000
814100
334200
72580
5831
75
3664000
1311000
324700
3704
1 630000
459900
203900
65860
21890
2891
764900
1 48400
71980
27510
4296
0
689100
316800
1 60900
54050
8718
515
1141000
556700
219700
65920
5803
0
5F-52
-------
Table 5F-4. Total number of school-age children and asthmatic school-age children
experiencing at least one or two daily maximum 8-hr average Os exposures in all study
areas by year, base air quality and air quality adjusted to just meeting the existing 75 ppb
standard.
Exposure Study Group
Asthmatic School-age
Children
All School-age Children
Air Quality
Scenario (3-
year averaging
period)
75(2006-2008)
base
75(2006-2008)
Year
2006
2007
2008
2009
2010
2006
2007
2008
2009
2010
2006
2007
8-hr Average
Exposure
Benchmark
Level (ppb)
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
Number of Persons
Exposed Per Os Season1
At least
once
254454
39422
2399
281062
51351
4236
155918
12603
428
160535
19889
1159
265139
45688
3981
664077
322187
116538
677867
311929
96028
523387
209734
59813
364614
115454
34482
504479
165488
30418
2434809
350957
20489
2624485
481637
At least
twice
94876
4113
120
113215
8788
133
42881
713
13
49447
1532
39
110871
6737
143
381016
126882
26233
397000
113719
16731
280744
67291
11174
157977
33741
5779
263154
49314
3270
900706
38834
708
1045539
74715
lumbers of children are summed across urban case study areas in each year. Because Chicago does not have a
simulation of the existing standard for the 2008-2010 three-year standard averaging period, year 2008 study area
sums were based only the year 2008 simulations that used the 2006-2008 three-year standard averaging period.
5F-53
-------
Exposure Study Group
Air Quality
Scenario 13-
year averaging
period)
base
Year
2008
2009
2010
2006
2007
2008
2009
2010
8-hr Average
Exposure
Benchmark
Level (ppb)
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
60
70
80
Number of Persons
Exposed Per Os Season1
At least
once
45085
1492203
120656
3615
1529789
188071
12385
2439477
395217
27981
6294372
3104290
1140328
6307074
2926601
892814
4987568
2021415
597040
3553528
1191875
370166
4671276
1528455
273651
At least
twice
992
438479
10568
52
469352
16908
141
997079
50200
670
3653902
1265681
264098
3658057
1067241
149031
2672573
684051
132098
1574327
372643
63075
2417136
421992
27126
1 Numbers of children are summed across urban case study areas in each year. Because Chicago does not have a
simulation of the existing standard for the 2008-2010 three-year standard averaging period, year 2008 study area
sums were based only the year 2008 simulations that used the 2006-2008 thee-year standard averaging period.
5F-54
-------
Table 5F-5. Mean and maximum number of school-age children and asthmatic school-age
children with at least one or two daily maximum 8-hr average O3 exposures at or above
benchmark levels, all 15 urban study areas combined.
Study Group
(Total Simulated
Population)
All School-age
Children
(19,049,557)
Asthmatic
School-age
Children
(1,992,762)
Benchmark
Level (ppb)
60
70
80
60
70
80
Air Quality
Scenario
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
base
75
70
65
60
55
Number of People with
at least One Exposure
Mean Year
5,163,000
2,316,000
1,176,000
392,000
70,000
7,500
2,155,000
362,000
94,000
14,000
1,400
200
655,000
27,000
3,700
300
100
0
547,000
246,000
126,000
42,000
7,700
900
225,000
40,000
10,000
1,700
200
0
67,000
2,900
300
0
0
0
Maximum
year
6,890,000
3,588,000
1,988,000
762,000
156,000
18,000
3,548,000
792,000
236,000
41,000
4,500
700
1,311,000
79,000
12,000
1,300
300
0
738,000
385,000
214,000
81,000
17,000
2,200
370,000
88,000
27,000
4,500
500
0
135,000
8,900
1,100
0
0
0
Number of People with
at least Two Exposures
Mean Year
2,795,000
865,000
320,000
67,000
5,100
400
762,000
46,000
5,400
300
0
0
127,000
600
0
0
0
0
296,000
93,000
35,000
7,500
700
100
78,000
5,300
600
100
0
0
13,000
100
0
0
0
0
Maximum
year
4,197,000
1,643,000
674,000
161,000
14,000
1,100
1,470,000
127,000
18,000
1,100
0
0
293,000
2,300
200
0
0
0
448,000
179,000
74,000
18,000
2,100
300
151,000
15,000
2,300
300
0
0
29,000
400
0
0
0
0
For adjusted air quality, the average of the two values for year 2008 (i.e., the 2006-08 and 2008-10 three-year averaging periods)
was first calculated within each study area; then the study area average calculated using all 5 simulated years (2006-2010).
Values > 10,000 were rounded to the nearest thousand, values <10,000 were rounded to the nearest 100.
5F-55
-------
APPENDIX 5G
Targeted Evaluation of Exposure Model Input and Output Data
Table of Contents
5G-1 ANALYSIS OF TIME-LOGATON-ACTIVITY DATA 5G-1
5G-1.1 Personal Attributes of Survey Participants in CHAD and Used by APEX 5G-1
5G-1.2 Afternoon Time Spent Outdoors for CHAD Survey Participants 5G-4
5G-1.3 Afternoon Time Spent Outdoors For ATUS Survey Participants 5G-7
5G-1.4 Outdoor Time and Exertion Level of Asthmatics and Non-Asthmatics In
CHAD 5G-9
5G-2 CHARACTERIZATION OF FACTORS INFLUENCING HIGH EXPOSURES . 5G-14
5G-3 ANALYSIS OF APEX SIMULATED LONGITUDINAL ACTIVITY PATTERNS
5G-23
5G-4 EXPOSURE RESULTS FOR ADDITIONAL AT-RISK POPULATIONS AND
LIFESTAGES, EXPOSURE SCENARIOS, AND AIR QUALITY INPUT DATA
USED 5G-29
5G-4.1 Exposure Estimated For All School-Age Children During Summer Months,
Neither Attending School nor Performing Paid Work 5G-29
5G-4.2 Exposures Estimated For Adult Outdoor Workers During Summer Months. 5G-31
5G-4.3 Averting Behavior and Potential Impact to Exposure Estimates 5G-39
5G-5 COMPARISON OF PERSONAL EXPOSURE MEASUREMENT AND APEX
MODELED EXPOSURES 5G-44
5G-6 REFERENCES 5G-47
5G-i
-------
List of Tables
Table 5G-1.
Table 5G-2.
Table 5G-3.
Table 5G-4.
Table 5G-5.
Table 5G-6.
Table 5G-7.
Table 5G-8.
Table 5G-9.
Table 5G-10.
Personal attributes of survey participants within CHAD and used by APEX.. 5G-3
Comparison of outdoor time expenditure and exertion level among asthmatic and
non-asthmatic diary days for CHAD diaries used by APEX 5G-11
Percent of waking hours spent outdoors at an elevated activity level. A
comparison of CHAD with Shamoo et al. (1994) study asthmatics 5G-12
Percent of waking hours spent outdoors at an elevated activity level: a comparison
of CHAD with EPRI (1992) study asthmatics 5G-13
Percent of waking hours spent outdoors at an elevated activity level: a comparison
of CHAD with EPRI (1988) study asthmatics 5G-13
Range of R2 fit statistics for ANOVA models used to evaluate daily maximum 8-
hour Os exposure concentrations stratified by study area, air quality scenario, and
exposure level 5G-16
Range of R2 fit statistics for ANOVA models used to evaluate daily maximum 8-
hour Os exposure concentrations in Los Angeles stratified by age group, air
quality scenario, and exposure level 5G-16
Distribution of days per week spent performing outdoor work considering the
BLS/O*NET data set and stratified by APEX/CHAD occupation groups 5G-33
Personal attributes and mean time spent working outdoors for CHAD diaries
reporting at least two hours of outdoor work 5G-34
Distribution of days per week spent performing outdoor work considering the
APEX simulated population and stratified by APEX/CHAD occupation and age
groups 5G-38
List of Figures
Figure 5G-1. Participation rate in outdoor activities (top) and mean time spent outdoors
(bottom) for CHAD diaries having at least one minute outdoors (left) and CHAD
diaries having at least two hours outdoors (right) during the afternoon 5G-5
Figure 5G-2. Participation rate in outdoor activities (top) and mean time spent outdoors
(bottom) for ATUS diaries having at least one minute outdoors (left) and ATUS
diaries having at least two hours outdoors (right) 5G-8
Figure 5G-3. Contribution of influential factors to daily maximum 8-hour ozone exposures
using base air quality (left), air quality adjusted to just meet the existing standard
(right), considering all person days (top) and those days where daily maximum 8-
hour exposure exceeded 50 ppb (bottom) in Boston 5G-18
Figure 5G-4. Contribution of influential factors to daily maximum 8-hour ozone exposures
using base air quality (left), air quality adjusted to just meet the existing standard
(right), considering all person days (top) and those days where daily maximum 8-
hour exposure exceeded 50 ppb (bottom) in Atlanta 5G-19
Figure 5G-5. Distributions of afternoon outdoor time expenditure and daily maximum 8-hour
ambient Os concentrations for simulated Boston school-age children (ages 5 to
5G-ii
-------
18) (left) and adults (ages 19 to 35) (right) using base air quality (top) and
concentrations adjusted to just meet the existing standard (bottom) for person
days having daily maximum 8-hour exposures either below or above 50 ppb....
5G-21
Figure 5G-6. Afternoon microenvironmental time (top) and activities performed during
afternoon time outdoors (bottom) for school-age children (left) and adults (right)
experiencing 8-hour daily maximum Os exposures > 50 ppb, Boston base air
quality, 2006 5G-23
Figure 5G-7. Cumulative distribution of median time spent outdoors (top row), afternoon
outdoor participation > 1 minute/day (2nd row), and afternoon outdoor
participation > 2 hours/day (3nd row) for male and female school-age children in
Atlanta (left column), Boston (middle column) and Denver (right column) study
areas. Percent of school-age children with > 2 hours outdoors during afternoon
hours (4th row) and the number of particular CHAD study diary days used (bottom
row) for each exposure simulation day 5G-27
Figure 5G-8. Cumulative distribution of median time spent outdoors (top row), afternoon
outdoor participation > 1 minute/day (2nd row), and afternoon outdoor
participation > 2 hours/day (3nd row) for male and female school-age children in
Houston (left column), Philadelphia (middle column) and Sacramento (right
column) study areas. Percent of school-age children with > 2 hours outdoors
during afternoon hours (4th row) and the number of particular CHAD study diary
days used (bottom row) for each exposure simulation day 5G-28
Figure 5G-9. Comparison of the percent of all school-age children having daily maximum Os
concentration at or above 60 ppb-8hr (top), 70 ppb-8hr (middle), or 80 ppb-8hr
(bottom) during June, July, and August in Detroit 2007: using any available
CHAD diary ("All CHAD Diaries") or using CHAD diaries having no time spent
in school or performing paid work ("No School/Work Diaries") 5G-30
Figure 5G-10. Percent of persons between age 19-35 (left) and 36-55 (right) experiencing
exposures at or above selected benchmark levels while at moderate or greater
exertion using an outdoor worker scenario-based approach (top) and a general
population-based approach (bottom) considering air quality adjusted to just meet
the existing standard in Atlanta, GA, Jun-Aug, 2006 5G-39
Figure 5G-11. Distribution of daily personal Os exposures (top row), outdoor time (2nd row from
top), ambient Os concentrations (3rd row from top), and air exchange rate (bottom
row) for DEARS study participants (left column) and APEX simulated
individuals (right column) in Wayne County, MI, July-August 2006 5G-46
-------
This Appendix presents the complete results of several targeted evaluations and exposure
simulations designed to provide additional insights to APEX input data or approaches used to
estimate exposures, algorithm and model performance evaluations, and estimated exposures for
additional exposure study groups and lifestages of interest.
5G-1 ANALYSIS OF TIME-LOCATON-ACTIVITY DATA
We first present an overview of the data currently available in the CHAD database used
by APEX, including comparison with the version of CHAD used to estimate exposures in the 1st
draft Os HREA. This is followed by an analysis of time spent outdoors - one of the most
important attributes influencing exposures at or above benchmark levels - using CHAD and
recent time-location-activity pattern data from the American Time Use Survey (ATUS). And
finally, CHAD diaries identified as coming from asthmatics are compared with that of non-
asthmatics for afternoon outdoor time and activity level as well as compared with available
independent studies of asthmatic activity patterns.
5G-1.1 Personal Attributes of Survey Participants in CHAD and Used by APEX
The survey participants whose diary data are within CHAD were asked a number of
questions regarding their personal attributes. The number and type of attributes present for
diaries in CHAD is driven largely by the original intent of the individual study. In our exposure
assessment, we have strict requirements to simulate individuals using several personal attributes,
namely age, sex, temperature (as a surrogate for seasonal variation in activity patterns), and day-
of-week. These attributes are considered as important drivers influencing daily activity patterns
(Graham & McCurdy, 2004) and when diaries do not have these particular attributes for a
particular day, they will not be used by APEX.
This APEX modeling requirement serves as an initial screen to the number of available
diaries in the complete CHAD master database (i.e., 54,373) and considering the age range of the
simulated exposure study groups (persons between the ages of 5 and 95), the actual number of
diary days having complete information and used by APEX in the final Os HREA is 41,474.1
This represents an increase of about 8,700 diaries currently used by APEX compared with what
was used by APEX in the 1st draft Os HREA. Additionally, there have been eight new study data
sets incorporated into CHAD and used in our current exposure assessment since the previous Os
NAAQS review conducted in 2007, most of which were from recently conducted activity pattern
studies (Appendix 5B, Section 5B-4). The diary data included from these new studies have more
1 Diaries from persons age 4 are included in this evaluation because they may be used in a simulation to represent a
person aged 5 due to the probabilistic nature of APEX. Typically, a diary matching the attributes of the simulated
individual has a greater probability of selection. Accommodations are allowed to increase the diary pool size (e.g.,
expand the age window of diaries available by value (one year) or percent (15%) of the simulated person's age.
5G-1
-------
than doubled the total activity pattern data used for 2007 Os exposure modeling and has
increased the number of children's diaries by about a factor of five.
Table 5G-1 presents a summary of the important personal attributes used by APEX in
creating activity patterns for simulated persons, along with other attributes of potential interest
(e.g., race/ethnicity). First, we compared the representation of several attributes in the current
CHAD used by APEX versus that used in the 1st draft Ch HREA. Outside of increases in the
number of persons, the general distribution of diaries within the APEX diary selection attributes
(e.g., age, sex, temperature, day-of-week) is similar in both databases. Worth noting is the
number and percent of diaries from each of the three decades analyzed. Currently, the majority
of diaries (54%) from CHAD are taken from surveys conducted in the past decade, while the pre-
1990s represent less than 15% of the total diaries available by APEX.
While there may be other personal or situational attributes that affect daily time
expenditure, these are typically not included in our assessment to generate simulated individuals
simply because the response to the attribute is missing for most persons. For example, income
level is missing for just over 66% of the study participants and only about 30% of employed
workers (persons ages 19 to 64) reported their occupation (Table 5G-1). Missing response data
in CHAD results from either the study not having an income/occupation related survey question
or perhaps the participant refused to answer the question. Note also, when any attribute is added
to the development of a person's profile, the pool of diaries available for selection in simulating
an individual is reduced. This could lead to an increased repetition of diaries used for simulated
individuals, potentially artificially reducing variability in time expenditure. In addition, the
desired study group to be simulated may have too few diaries within a diary pool if most diaries
are missing the needed attribute, leading to a simulation failure. This is why personal attributes
are carefully selected and prioritized according to both their prevalence in CHAD and whether
attribute has a known significant influence on activity patterns.
5G-2
-------
Table 5G-1. Personal attributes of survey participants within CHAD and used by APEX.
Personal
Attribute
Age (years)
Sex
Daily
Maximum
Temperature
(°F)
Day of Week
Decade
Employed
(ages 19-64)
Occupation
(employed)
Income Group
Calendar
Month
Race/Ethnicity
Group Within
Attribute
4-18
19-34
35-50
51 -64
65+
Female
Male
84+
55-83
<55
Weekday
Weekend
1980s
1990s
2000s
No
Yes
Missing/Unknown
Known
Missing/Unknown
> 1 .5 x Poverty
< 1 .5 x Poverty
Missing/Unknown
Calendar Months 5 -
9
Calendar Months 1 -
4, 10-12
Asian
Black
Hispanic
Native American
White
Other
Missing/Unknown
Current APEX CHAD
(41 ,474 total days)
person days
(n)
17680
4490
7238
6181
5885
21466
20008
13817
17827
9830
29031
12443
5999
12831
22644
4651
12755
503
3867
8888
10347
3713
27414
19151
22323
670
6993
2476
28
25009
770
5528
percent of
attribute
42.6
10.8
17.5
14.9
14.2
51.8
48.2
33.3
43
23.7
70.0
30.0
14.5
30.9
54.6
26.0
71.2
2.8
30.3
69.7
24.9
9.0
66.1
46.2
53.8
1.6
16.9
6.0
0.1
60.3
1.9
13.3
1st Draft APEX CHAD
(32,788 total days)
person days
(n)
14111
4001
5957
5016
3703
16840
15948
12113
13078
7597
23794
8994
6167
12390
14231
3747
11227
n/a
3012
8215
10416
3730
18642
16812
15976
349
4040
1339
7
19569
609
6875
percent of
attribute
43
12.2
18.2
15.3
11.3
51.4
48.6
36.9
39.9
23.2
72.6
27.4
18.8
37.8
43.4
25.0
75.0
n/a
26.8
73.2
31.8
11.4
56.9
51.3
48.7
1.1
12.3
4.1
0
59.7
1.9
21
5G-3
-------
5G-1.2 Afternoon Time Spent Outdoors for CHAD Survey Participants
There have been questions raised regarding the representativeness of the diaries from
studies conducted in the 1980s and whether there are any recognizable patterns in time
expenditure in the CHAD diaries across the time period when data were collected. Because time
spent outdoors is a significant factor influencing daily maximum 8-hour Os exposures, we
evaluated the current collection of CHAD diaries used by APEX for two metrics: outdoor
participation rate and mean time spent outdoors. The participation rate is the percent of the
person-days having at least one minute outdoors, and because high Os concentrations commonly
occur during the afternoon hours of summer months, we restricted the analysis to those times of
day (12 PM to 8 PM) and year (May through September). The same data set was used to
calculate a mean outdoor time, though the calculation was further restricted to person days
meeting an additional criterion: person-days having at least one minute outdoors and person-days
having at least 2-hours outdoors. Separating the data into these sub-groups give us insight to the
diaries most likely to be used in simulating a person that exceeds a selected benchmark level and
protects (to a limited degree) from study sample design bias (15-minute time block diaries versus
minute-by-minute event level diaries). Data were further stratified by five age groups (4-18, 19-
34, 35-50, 51-64, 65+) and three decades (1980s, 1990s, and 2000s) using the year the particular
activity pattern study was conducted. As a reminder, CHAD is composed of primarily cross-
sectional data, thus the trend evaluated over the three decades is changes (if any) in participation
rate and the time spent outdoors by the study group, not individuals.
Figure 5G-1 illustrates the trends in afternoon outdoor activity participation and mean
time expended outdoors, considering three decades, five age groups, and whether the total
afternoon time spent outdoors was at least one minute or two hours. Regardless of decade and
duration of time spent outdoors, participation in outdoor activities follows an expected pattern
considering age groups, that is, children tend to have the highest participation rate when
compared with the other age groups, while the oldest persons (aged 65 or greater) tend to have
the lowest participation rate (Figure 5G-1, top left panel). When considering decade and CHAD
diaries having at least one minute spent outdoors, the participation rate appears to have a non-
linear concave trend, whereas CHAD diaries collected during the 1990s exhibit the lowest
outdoor participation rate (ranging from about 40-70%) while much greater participation is found
with the CHAD diaries collected during the 1980s (80-90%) and 2000s (70-80%).
5G-4
-------
CHAD, > 1-minsoutdoortime, months May-Sep, hours 12-8PM
_100
CHAD, > 1-min outdoortime, months May-Sep, hours 12-8PM
CHAD, > 2-hrs outdoortime, months May-Sep, hours 12-8PM
CHAD, > 2-hrs outdoortime, months May-Sep, hours 12-8PM
ages 4 to 18
•ages 19 to 34
•ages 35 to 50
•ages 51 to 64 *ages65+
Figure 5G-1. Participation rate in outdoor activities (top) and mean time spent outdoors
(bottom) for CHAD diaries having at least one minute outdoors (left) and CHAD diaries
having at least two hours outdoors (right) during the afternoon.
It is possible that this observed pattern may be the result of the original study survey
design. All of the CHAD diaries collected during the 1980s used an 'event' level approach, that
is locations visited and activities performed were reported on a minute or longer basis and many
of those same diaries also used a 'contemporaneous diary' approach, that is, real-time data
reporting. The CHAD diaries collected during the 1990s used a mixture of event level and 15-
minute time block data, though mostly using a 'recall' approach, that is, participants were asked
about their activities performed the day before. It is likely that these diaries exhibit the lowest
outdoor participation rate due to the participant missing or ignoring short duration events (< 15
minutes) that may have occurred outdoors (e.g., outdoors in a parking lot and walking to their
vehicle). The CHAD diaries collected during the 2000s, while also a combination of studies that
used the event level and 15-minute time block approach did have a few studies using the real-
5G-5
-------
time diary approach, possibly responsible for the observed increase in outdoor participation rate
during this decade.
When restricting the data set to person days having at least two hours of afternoon time
spent outdoors, the above mentioned temporal pattern nearly disappears (Figure 5G-1, top right
panel) as most age groups exhibit little to no variability in outdoor time participation rate across
the three decades (-20-30% of person-days). However, the diaries from the 1980s for children
ages 4-18 indicate the highest outdoor participation rate (i.e., 50%) compared to all other age
groups and decade of collection. Most of these diaries (i.e., 96%) are from the California
Children's Study and Cincinnati Activity Patterns Study. To date it remains unexplained why
these two studies would have this unusual outdoor participation rate compared to the other
studies and decades and could simply be a function of the two study selecting particularly active
children by chance.2 When considering the entire pool of all diaries available for this age group
and used by APEX, these two studies contribute to about 19% of diaries having two or more
hours of time spent outdoors during the afternoon. This translates to a small difference in the
overall outdoor participation rate for diary pools that would include these earlier studies (39%)
compared to the participation rate excluding these studies (36%), with both values similar to the
findings reported recently by Marino et al. (2012) of 37.5% using a similar metric, though for
pre-school age children. Thus, when considering participation in outdoor activities and the
representativeness of the CHAD study data from the 1980s, it is unlikely that use of these older
diaries would adversely influence exposure model estimates.
When considering the mean time spent outdoors during the afternoon hours for
diaries having at least one minute recorded outdoors, the observed pattern is generally the
inverse of participation. This pattern would be consistent with the above proposed reasoning:
relatively more persons reporting shorter duration events leads to higher overall outdoor
participation rate coinciding with a decrease in the mean time spent outdoors (Figure 5G-1,
bottom left panel). In general, when restricting the data set to person-days that recorded at least
two hours outdoors during the afternoon (Figure 5G-1, bottom right panel), there is variability in
the amount of outdoor time over the three decades, with diaries from the 2000s exhibiting
perhaps the lowest range of mean outdoor time (190-220 min/day) compared with the 1980s
(210-240 min/day) and 1990s (212-258 min/day), a trend perhaps most notable trend when
considering the children's diaries (a decrease in time spent outdoor of about 30 minutes over the
'• Restricting the data by any criterion will result in fewer diary days available that subset, thus any remaining or
newly observed trends (e.g., an apparent linear decline in participation by the diaries from persons aged 19-34)
should include an assessment of potential confounding factors. Given our current and other researcher's past
evaluations of the CHAD data (Graham and McCurdy, 2003; Isaacs et al., 2012; McCurdy and Graham, 2004)
often times, as is the case here, additional influential factors have not been measured and the observed
phenomenon can only be identified as study related.
5G-6
-------
period). However, the coefficient of variation (COV) for each of the age groups and across all
decades for the cross-sectional data was consistently about 40% (data not shown), supporting a
general conclusion of no large differences in the mean time spent outdoors for this set of diaries
over the three decades of data collection. Thus, when considering all diaries having at least two
hours of afternoon outdoors time and the representativeness of the CHAD study data from the
1980s, inclusion of these earlier diaries is also unlikely to have a significant adverse influence on
exposure modeling outcomes.
5G-1.3 Afternoon Time Spent Outdoors For ATUS Survey Participants
We evaluated recent year (2002-2011) time expenditure data from the American Time
Use Survey (ATUS) (US BLS, 2012a). As was done with the CHAD data set, the purpose of the
evaluation was to evaluate trends in outdoor time over the period of time data were collected. A
few strengths of the ATUS data are (1) its recent and ongoing data collection efforts, (2) large
sample size (>120,000 diary days), (3) national representativeness, and (4) that varying diary
approaches would not be an influential or confounding factor in evaluating trends over time.
ATUS does however have a few noteworthy limitations when compared with the CHAD
data: (1) there are no survey participants under 15 years of age, (2) time spent at home locations
is neither distinguished as indoors or outdoors, and (3) missing or unknown location data can
comprise a significant portion of a persons' day (on average, about 40% (George and McCurdy,
2009)). To overcome the limitation afforded by the ambiguous home location, we identified
particular activity codes most likely to occur outdoors (e.g., participation in a sport) to better
approximate each ATUS individual's outdoor time expenditure. Missing several hours of
location and activity information can be problematic when modeling exposures, an issue that
renders the ATUS diaries generally unusable by APEX. However, the particular time of day the
missing data occurs is more accommodating to the purpose of this analysis. While most diaries
are missing location information for 6 or more hours per day, on average about 85% of the
missing time information occurs outside of the hours of interest here (i.e., 12 PM-8 PM), with
most missing time occurring between early morning (4 AM-9 AM) or late evening hours (10
PM-12 AM). Still though, we restricted the ATUS outdoor time analysis to diaries having no
more than 1-hour of missing afternoon time while also only retaining diaries from ATUS
identified metropolitan areas. Data were then stratified by the same five age groups as was done
for the CHAD data, though here the time trends were assessed over individual survey years.
Figure 5G-2 illustrates the results of the ATUS diary outdoor participation rates (top row) and
the mean time spent outdoors during afternoon hours (bottom row) for persons having at least
one minute of afternoon time spent outdoors (left column) or two hours or more outdoors (right
column).
5G-7
-------
ATUS, > 1-min outdoor time, months May-Sep, hours 12-8PM
" T^n
.1
81"
1 &
o^
ean Afternoo
(minut
3 I fe
£
^t^S*^^^.
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
ATUS,> 1-min outdoor time, months May-Sep, hours 12-8PM
1
u
* 70
s. 70
S 60
E ^- 60
p
O 4->
O ro
-a o: 40
° 30
O
o 20
Q)
S 10
X^x_
* — *="7*^ i 1 ,• T l
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
ATUS, >2-hrsoutdoortime, months May-Sep, hours 12-8PM
o 90
(0
•c 80
'€ 70
S 70
11 fin
IS
0 *
o ro
o
Afternoon
3 S g t
. • ~^^' — ' — =^~
' — ^•—- -+f- \ — T~:r^ j — \ =^*
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
ATUS, > 2-hrs outdoortime, months May-Sep, hours 12-8PM
" 250
P
0
~2 %
Sz.
o 5150
2 =
oi E
<
c
(0
r
^^^
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
-ages 15 to 18 -»-ages 19 to 34 -«-ages 35 to 50 -*-ages 51 to 64 -"-ages 65+
Figure 5G-2. Participation rate in outdoor activities (top) and mean time spent outdoors
(bottom) for ATUS diaries having at least one minute outdoors (left) and ATUS diaries
having at least two hours outdoors (right).
5G-8
-------
Not surprisingly given the lack of distinction regarding time indoors and outdoors while
at home for ATUS participants as well as the diary approach used,3 the outdoor activity
participation rate for ATUS study subjects is lower than that of CHAD study subjects; about 30-
40% of ATUS person-days have at least one minute of outdoor time (Figure 5G-2, top left
panel). As was observed with the CHAD data, children (ATUS ages 15 to 18) are more likely to
participate in outdoor activities. The mean time outdoors for persons that reported any amount
of outdoor time is similar to the range indicated by CHAD diaries, generally between 100-150
minutes per day (Figure 5G-2, bottom left panel). When considering person-days having at least
2 hours of time spent outdoors, the range in ATUS diary outdoor participation rate (10-20%,
Figure 5G-2, top right panel)) is lower than that observed for the CHAD data (generally between
20-40%), while the range in mean time spent outdoors (190-240 minutes per day, Figure 5G-2,
top right panel) was similar to that of the CHAD data. Consistent also across the two studies is
the participation rate of children being greater than that of the other age groups. There are no
consistent trends over the nine year ATUS study period regarding either the participation rate or
the mean time spent outdoors for any of the age groups.
5G-1.4 Outdoor Time and Exertion Level of Asthmatics and Non-Asthmatics in CHAD
Due to limited number of CHAD diaries with survey requested health information, all
CHAD diaries are assumed appropriate for any simulated individual (i.e., whether asthmatic,
non-asthmatic, or not indicated), provided they concur with age, sex, temperature, and day-of-
week selection criteria. In general, the assumption of modeling asthmatics similarly to healthy
individuals (i.e., using the same time-location-activity profiles) is supported by the findings of
van Gent et al. (2007), at least when considering children 7 to 10 years in age. These researchers
used three different activity-level measurement techniques; an accelerometer recording 1-minute
time intervals, a written diary considering 15-minute time blocks, and a categorical scale of
activity level. Based on analysis of 5-days of monitoring, van Gent et al. (2007) showed no
difference in the activity data collection methods used as well as no difference between asthmatic
children and healthy children when comparing their respective activity levels. Contrary to this,
an analysis of 2000 BRFSS data by Ford et al. (2003) indicated a statistically significant
difference between the percent of current asthmatics (30.9%) and non-asthmatics (27.8%)
characterized as inactive. In addition, these researchers found small but statistically significant
differences in the percent of asthmatic (26.6%) and non-asthmatic (28.1%) adults achieving
recommended levels of physical activity (i.e., either moderate or greater activity levels).
3 The ATUS time-use information was collected by subject recall of the prior day's activities and "conversational
interviewing" rather than asking scripted questions.
5G-9
-------
Note though, the salient issue is not just outdoor time and activity levels, but the
intersection of the two as well as recognizing the performance capabilities of persons with
asthma. A person's overall physical activity level is strongly linked with their time spent
outdoors and is considered an important correlate in encouraging increased physical activity
among children and adults alike (e.g., Sallis et al., 1998). In addition, introducing regular
exercise has been shown to improve physical fitness in asthmatic children, with statistically
significant increases in ventilation measures such as maximum minute ventilation rate (VEmax)
maximum oxygen uptake (VChmax) (e.g., van Vledhoven et al., 2001). Further, in other related
research, Santuz et al. (1997) indicated no statistically significant difference between asthmatic
and non-asthmatic children when comparing maximum exercise performance levels, provided
the individuals were conditioned through habitual exercise. Thus it appears that asthmatics
perform activities at elevated levels and do so in outdoor microenvironments in similar fashion to
non-asthmatics.
To provide further support to the assumption that any CHAD diary day can be used to
represent the asthmatic study groups regardless of the study participants' characterization of
having asthma or not, we first compared the amount of afternoon outdoor time and participation
in elevated exertion levels among asthmatics and non-asthmatics. Because six of the 19 studies
incorporated in CHAD reported whether the individual was asthmatic or non-asthmatic, we
categorized the data and results using three categories (i.e., asthmatic, non-asthmatic, not
classifiable). Afternoon hours were characterized as was done for above CHAD analyses, that is,
the time between 12 PM and 8 PM and only those persons that did spend some time outdoors
were retained. As is done by APEX in simulating individuals, level of exertion was estimated by
sampling from the specific METS distributions assigned for each person's activity performed.
Then, we selected for activities having a METS value of greater than 3 as times where a person
was at moderate or greater exertion levels (US DHHS, 1999). Afternoon outdoor time was then
stratified by exertion level, summed for two study groups of interest (children and adults), and
presented in percent form within Table 5G-2.
When considering CHAD diaries used by APEX in our simulations, about 18% of the
diaries are from either an asthmatic child or an asthmatic adult. Far fewer children's diaries are
from persons whose asthmatic status is unknown (12%) when compared to adults (30%) though
still, persons having unknown health status are a smaller proportion of the total available person-
days. On average, about 43% of all children spent some afternoon time outdoors while asthmatic
children have a higher participation rate (48.5%) when compared to non-asthmatic children
(41.2%). About half of the adults whose asthmatic condition was known did spend afternoon
time spent outdoors with participation rate generally similar for both asthmatic and non-
asthmatic adults. Outdoor participation rate for persons having unknown asthma status varied
5G-10
-------
from that of known persons; about 60% of the children's diaries and 31% of the adult diaries
indicate some afternoon time was spent outdoors.
Table 5G-2. Comparison of outdoor time expenditure and exertion level among asthmatic
and non-asthmatic diary days for CHAD diaries used by APEX.
Asthmatic?
Persons (n)
Outdoor Habitue (n)
Outdoor Habitue (%)
Percent of Afternoon
Hours Spent
Outdoors (%)
Percent of Afternoon
Outdoor Time at
Moderate or Greater
Exertion (%)
CHAD: Children (4 to 18)1
Yes
3,206
1,564
48.8
28.5
80.3
No
12,346
5,092
41.2
27.5
78.2
Unknown
2,128
1,267
59.5
28.9
79.2
CHAD: Adults (19 to 95)2
Yes
1,254
602
48.0
26.2
62.7
No
15,465
7,949
51.4
27.2
63.8
Unknown
7,075
2,176
30.8
22.2
60.3
1 CHAD studies for where a survey questionnaire response of whether or not child was asthmatic include CIN, ISR,
NHA, NHW, OAB, and SEA (see HREA Table 5-3 for study names).
2 CHAD studies for where survey a questionnaire response of whether or not adult was asthmatic include CIN, EPA,
ISR, NHA, NHW, NSA, and SEA.
The amount of time spent outdoors by the persons that did so varied little across the two
study groups and three asthma categories. On average, diaries from children indicate
approximately 2Vi hours of afternoon time is spent outdoors, 80% of which is at a moderate or
greater exertion level, regardless of their asthma status. Slightly less afternoon time is spent
outdoors by adults (about 125-130 minutes) when compared with children whose asthma status
is known, though more notable is the lowered percent of afternoon time adults perform moderate
or greater exertion level activities (about 63%). As noted above regarding the reduced
participation rate for adults whose asthma status is unknown, diaries for these adults also have
about 20 fewer minutes of afternoon time spent outdoors compared with those persons whose
asthma status is known.
Outdoor time and activity levels of respective cohorts from three independent asthma
activity pattern studies were compared to CHAD diary days using similar metrics. To make the
CHAD data compatible with the independent asthma study data, the entire diary day was
evaluated, not just the afternoon hours, and all persons (not just outdoor habitue) were
considered. In addition and where possible, the demographics of the independent study
participants was used to select for the most representative CHAD diaries (e.g., person's age, sex,
month- or day-of-year, etc.). Table 5G-3, Table 5G-4, and Table 5G-5 summarize the data
reported from the three asthma activity pattern studies and the compatible results generated using
CHAD and the indicated asthma status of the study persons.
5G-11
-------
Table 5G-3. Percent of waking hours spent outdoors at an elevated activity level. A
comparison of CHAD with Shamoo et al. (1994) study asthmatics.
Study
Location
Time of Year
Asthmatic?
Person days
Mean age
(min-max)
Exertion Level
Low
Moderate
Strenuous
Shamoo (1994)1
Los Angeles
Summer
Months
(5-9)
Yes
336
Winter
Months
(11-3)
Yes
314
33
(18-50)
CHAD2
Any Diary
Summer Months (May-Sep)
Yes
375
37.3
(1 8 - 50)
No
4,812
37.9
(18-50)
Unknown
1,049
35.2
(1 8 - 50)
Winter Months (Nov-Mar)
Yes
211
30.4
(18-50)
No
2,512
32.3
(18-50)
Unknown
1,998
34.3
(18-50)
Percent of Asthmatic Waking Hours Spent Outdoors at Given Exertion Level
8.5 6.0
1.9 1.7
0.2 0.2
4.6
6.2
1.1
5.6
6.9
1.2
5.4
4.6
1.0
2.0
2.0
0.6
2.1
2.7
0.7
1.4
2.0
0.5
1 Based on number of minutes performing three self-rated activity levels for three locations per hour (indoor,
outdoor, in-vehicles) over seven days. Non-random sample of 49 subjects selected from voluntary clinical studies.
2 Combination of random and non-random selection studies, national and city-specific, as well as varying diary
protocol (see HREA Appendix 5B). The APEX CHAD file (n=41,474) was additionally screened for persons
having no sleep reported (n=141). Randomly sampled METS values from each activity-specific distribution were
assigned to each person's activities. Moderate and vigorous activity levels were selected based on activities having
a METS value of 3 to <6 and >6, respectively.
When considering the three independent asthma studies, the amount of time spent
outdoors at moderate activity level ranges from a low of approximately 2% to a high of about
11% of waking hours. The estimates of outdoor time associated with moderate activity level
using a similarly constructed cohort of CHAD diaries fall within that range (i.e., between 2.0 and
7.2%). A small but consistent trend of fewer minutes spent outdoors at moderate exertion was
observed for the CHAD asthmatic cohort when compared with the CHAD non-asthmatic group.
The general range in percent of outdoor time associated with strenuous activities using the
CHAD person days (0.4% to 7.6%) was greater when compared to asthmatic persons from the
three independent studies (0.2% to 3.3%).
We recognize that there are a number of differences that exist among the three asthmatic
studies used for comparison along with the use of CHAD diary data from either asthmatics, non-
asthmatics, or unclassified asthmatics that could contribute to variation in the time spent
outdoors at elevated activity levels. This would include: the diary/survey collection methods
used, the classification of activities performed and associated activity levels, the number of study
subjects, and sample selection methods. The particulars regarding how each of these were
addressed across the various studies is wide ranging and could potentially influence the results
generated here. However, based on the mostly comparable results observed in time spent
outdoors at moderate or greater exertion activity levels, we judge the use of a CHAD diary
5G-12
-------
regardless of a persons' asthma condition is reasonably justified based on the available data
analyzed.
Table 5G-4. Percent of waking hours spent outdoors at an elevated activity level: a
comparison of CHAD with EPRI (1992) study asthmatics.
Study
Location
Time of Year
Asthmatic?
Person days
Mean age
(min-max)
Exertion Level
Moderate
Strenuous
EPRI(1992)1
Cincinnati
August
(Wed-Sun)
Yes
408
26
(1 - 78)
CHAD
Any Diary
July and August
(Thu-Sat)
Yes
711
17.9
(1 - 77)
No
3,085
34.9
(1 - 78)
Unknown
1,209
32.3
(1-78)
Percent of Asthmatic Waking Hours Spent
Outdoors at Given Exertion Level
11
3.3
5.0
7.6
7.2
3.7
7.2
2.6
1 Hour diary questionnaire form used up to three activities per hour. Random digit dialing and multiplicity sampling
were used. Three consecutive diary days were collected from 136 asthmatics, mostly Thu-Sat though some
Wednesday and Sunday data were included.
Table 5G-5. Percent of waking hours spent outdoors at an elevated activity level: a
comparison of CHAD with EPRI (1988) study asthmatics.
Study
Location
Time of Year
Asthmatic?
Person days
Mean age
(min-max)
Exertion Level
Moderate
Strenuous
EPRI(1988)1
Los Angeles
April (Fri-Mon)
Yes
156
(18-37)2
CHAD
Any
April and May (Fri-Mon)
Yes
26.2
(18-39)
No
29.7
(18-40)
Unknown
30.9
(1 8 - 40)
Percent of Asthmatic Waking Hours Spent
Outdoors at Given Exertion Level
7
2.4
2.8
0.4
4.5
0.6
4.4
1.2
1 Hour diary questionnaire form used up to three activities per hour. Non-random sample of 26 mild/moderate, 26
moderate/severe asthmatics selected from voluntary clinical studies. Three consecutive diary days were collected
per person (either Fri-Sun or Sat-Mon).
2 General age range approximated from twenty-nine subjects noted by EPRI (1988) as from Linn et al. (1987).
5G-13
-------
5G-2 CHARACTERIZATION OF FACTORS INFLUENCING HIGH EXPOSURES
We investigated the factors that influence estimated exposures, with a focus on persons
experiencing the highest daily maximum 8-hour exposures six selected study areas - Atlanta,
Boston, Denver, Houston, Philadelphia, and Sacramento.4 This analysis required the generation
of detailed APEX output files having varying time intervals, that is, the daily, hourly, and
minute-by-minute (or events) files. Given that the size of these time-series files is dependent on
the number of persons simulated, we simulated 5,000 persons and restricted the analysis to a
single year (2006) to make this evaluation tractable.5 Both the base case (unadjusted or 'as is'
recent air quality conditions) and ambient Os adjusted to just meet the existing standard (75 ppb-
8hr) air quality scenarios were evaluated in each of six study areas. All APEX conditions (e.g.,
ME descriptions, AERs, MET data) were consistent with the 200,000 person APEX simulations
that generated all of summary output discussed in the main body of this chapter.
We were interested in identifying the specific microenvironments and activities most
important to Os exposure and evaluating their duration and particular times of the day people
were engaged in them. Because ambient Os concentrations peak mainly during the afternoon
hours, we focused our microenvironmental time expenditure analysis on the hours between 12
PM and 8 PM. For every person and day of the exposure simulation, we aggregated the time
spent outdoors, indoors, near-roadways, and inside vehicles during these afternoon hours (i.e.,
the time of interest summed to 480 minutes per person day). Data from several APEX output
files were then combined to generate a single daily file for each person containing a variety of
personal attributes (e.g., age, sex), their daily maximum 8-hour ambient and exposure
concentrations, and the aforementioned time expenditure metrics.
We performed an analysis of variance (ANOVA) using SAS PROC GLM (SAS, 2012)
to determine the factors contributing the greatest to the observed variability in the dependent
variable, i.e., each person's daily maximum 8-hour Cb exposure concentrations. This analysis
was distinct for four age-groups of interest (i.e., 5-18, 19-35, 36-64, >65 years of age). The final
statistical models6 included a total of seven explanatory variables: the main effects of (1) daily
4 For the 1st draft Os HREA, this analysis was performed for four study areas: Atlanta, Denver, Los Angeles, and
Philadelphia. One important difference between the exposure simulations at that time compared with this final Os
HREA was the air quality data input to APEX: ambient monitoring data were used for the 1st draft Os HREA along
with a quadratic approach for adjusting air quality to just meet the existing standard.
5 We recognize that there is year-to-year variability in ambient Os concentrations and it is possible that fewer
persons simulated could result in differences in exposures compared to large-scale multi-year model simulations.
Based on a similar detailed evaluation performed for the Carbon Monoxide REA (US EPA, 2010), it is expected
any differences that exist between exposures estimated in a large simulation versus that using a smaller subset of
persons would be small and of limited importance to this particular evaluation.
6 In this investigation, we also evaluated the influence of sex, work and home districts, meteorological zones, each
with varying statistical significance, though overall adding little to explaining variability beyond the variables
selected for the final ANOVA model.
5G-14
-------
maximum 8-hour ambient Os, (2 to 4) afternoon time spent outdoors, near-roads, and inside
vehicles,7 and (5) physical activity index (PAI), while also including interaction effects from (6)
afternoon time outdoors by daily maximum 8-hour ambient concentration, and (7) PAI by
afternoon time outdoors. Two conditions were considered: all person days of the simulation, and
only those days where a person's 8-hour maximum exposure concentration was >50 ppb.8
Selected output from this ANOVA included parameter estimates for each variable, model R-
square statistic (R2), and Type III model sums of squares (SS3).9
Model fits, as indicated by an R2 value, were reasonable across each of the study areas
(Table 5G-6). The selected factors explain about 40-80% of the total variability in 8-hour daily
maximum exposures. Model fits were best when using all person days of the simulation though
results were similar for both air quality scenarios. When considering only those days where
persons had 8-hour daily maximum Cb exposures >50 ppb, consistently less variability in
maximum exposure concentrations was explained by the factors included in each model, though
overall model fits were acceptable. Furthermore, the most robust models tended to be those
developed using either school-age children aged 5 to 18 or adults 19 to 35 years old (e.g., see
Table 5G-7 for Atlanta model R2 results stratified by age groups).
We evaluated the relative contribution each variable had on the total explained variability
using the SS3 in each respective model.10 As with the R2 statistics generated above, the percent
contribution results were separated into four exposure scenarios for each study area, with
estimates for Boston illustrated in Figure 5G-3. When considering all person days of the
simulation (top row), the daily maximum 8-hour ambient Os concentration variable contributes
the greatest to the explained model variance, consistently estimated to be about 85% across all
age groups and for either the base or existing standard air quality scenarios. The interaction of
this variable with afternoon outdoor time contributes an additional 7-10% to the explained
variance, indicating that both ambient concentration and time spent outdoors collectively
contribute to 90% or more of the explained model variance when evaluating all (e.g., high, mid-
range, and low level) daily maximum 8-hour Os exposure concentrations. The main effect of
outdoor time contributed < 1% to the explained variance under these conditions as did
contributions from the other included variables, except for time spent near-roads (about a 5%
7 Including indoor afternoon time creates a strict linear dependence among these four variables and generates biased
estimates, thus it was neither included nor needed in this analysis.
8 This exposure concentration was selected due to the reduced sample size needed for these simulations (i.e., 5,000
total persons), an issue of increasing importance when selecting for persons with the highest exposures.
9 In each of the ANOVA models constructed, type II = type III = type IV sums of squares.
10 Type III sums of squares (SS3) for a given effect are adjusted for all other effects evaluated in the model,
regardless of whether they contain the given effect or not. Thus, the SS3 for each variable represents the
individual effect sums of squares that sum to the total effect sums of squares (or the total model explained
variance).
5G-15
-------
contribution). These results suggest that when considering the Boston exposure study groups
broadly, the daily maximum 8-hour ambient Os concentration is the most important driver in
estimating population-based Os exposures, nearly regardless of specific microenvironmental
locations where exposure might occur.
Table 5G-6. Range of R2 fit statistics for ANOVA models used to evaluate daily maximum
8-hour Os exposure concentrations stratified by study area, air quality scenario, and
exposure level.
Study Area
(Os season)
Atlanta
(245 days)
Boston
(183 days)
Denver
(21 4 days)
Houston
(365 days)
Philadelphia
(21 4 days)
Sacramento
(365 days)
Base Case Model R2 (sample sizes)
All Person
Days
0.71 -0.78
(1,225,000)
0.64-0.70
(915,000)
0.61 -0.68
(1,070,000)
0.75-0.80
(1,825,000)
0.66-0.73
(1,070,000)
0.75-0.81
(1,825,000)
Person days with 8-hr daily
max exposure > 50 ppb
0.62-0.70
(43,646)
0.44-0.62
(23,496)
0.49-0.62
(24,850)
0.57-0.67
(17,779)
0.51 -0.67
(39,561)
0.50-0.71
(19,734)
Existing Standard Model R2 (sample sizes)
All Person
Days
0.67-0.74
(1,225,000)
0.58-0.65
(915,000)
0.55-0.62
(1,070,000)
0.65-0.71
(1,825,000)
0.57-0.64
(1,070,000)
0.68-0.74
(1,825,000)
Person days with 8-hr daily
max exposure > 50 ppb
0.61 -0.70
(18,758)
0.41 -0.59
(16,184)
0.47-0.61
(18,487)
0.38-0.65
(7,322)
0.48-0.66
(16,841)
0.48-0.73
(6,503)
Table 5G-7. Range of R2 fit statistics for ANOVA models used to evaluate daily maximum
8-hour Os exposure concentrations in Los Angeles stratified by age group, air quality
scenario, and exposure level.
Study Area
Atlanta
Age Group
(years)
5-18
19-35
36-64
>65
Base Case Model R2
All Person
Days
0.78
0.71
0.73
0.74
Person days with
8-hr daily max
exposure > 50 ppb
0.64
0.70
0.64
0.62
Existing Standard Model R2
All Person
Days
0.74
0.67
0.68
0.70
Person days with
8-hr daily max
exposure > 50 ppb
0.62
0.70
0.64
0.61
5G-16
-------
When considering only person days having daily maximum 8-hour Os exposures > 50
ppb and for either air quality scenario in Boston, collectively the main effects of ambient
concentrations and outdoor time combined with their interaction similarly contribute to
approximately 85% of the total explained variance (Figure 5G-3, bottom row). However, the
main effect of the 8-hour daily maximum ambient Os concentration variable has a sharply lower
contribution (generally about 5-15%) along with greater contribution from the main effects
variable outdoor time (15-25% contribution) and its interaction with the ambient concentration
variable (40-60%). These results suggest that for highly exposed persons in Boston, the most
important influential factors are time spent outdoors corresponding with high daily maximum 8-
hour ambient Os concentrations.
Results for Atlanta (Figure 5G-4), were generally similar to Boston with notable
differences discussed here.11 The contribution of the maximum 8-hour ambient Os concentration
variable to the total explained variance (about 40-50%) was less than that observed in Boston
when considering all person days (Figure 5G-3 and Figure 5G-4, top rows), while the
contribution from the outdoor time/ambient Os interaction variable was greater in Atlanta (about
20-40% versus 10% in Boston).
This observed dissimilarity in the contribution by ambient concentrations and afternoon
outdoor time may be driven by the A/C prevalence rates and AER distributions used for each
study area.12 Boston has lower A/C prevalence though overall higher AERs (even when
considering mechanical ventilation), thus a greater contribution to exposure is expected from
ambient concentrations by infiltrating to indoor microenvironments and hence, reflected in the
strong main effects for the 8-hour daily maximum ambient Os concentration variable in Boston.
Afternoon time spent near Atlanta roads was estimated to contribute to about 20-30% of the total
explained variance when considering all person days and exposures, a value greater than that
estimated for Boston (generally about 5%) again possibly reflecting an increased importance of
this outdoor microenvironment in Atlanta (and Houston, Sacramento, not shown) relative to that
in Boston (and Philadelphia, not shown).
11 The discussion regarding the relative contribution of the variables to the total explained model variance also
extends to the other four study areas, whereas results for Philadelphia were generally similar to Boston,
Sacramento and Houston were similar to Atlanta, and Denver generally fell somewhere in between these extremes
presented.
12 A/C prevalence is highest in Houston (99%), Atlanta (98%), Philadelphia (95%), and Sacramento (93%)
compared to Boston (86%) and Denver (67%). Boston and Philadelphia used the same (and highest) AER
distributions; Sacramento, Houston, and Atlanta used separate but similar (and lower) AER distributions, while
Denver AER distributions fall somewhere in between these two extremes (see Appendix B, Tables 5B-3 to 5B-5).
5G-17
-------
jjf 19 to 35
jf 19 to 35
Boston, Base Air Quality,
All Person Days
Model Explained Variance
Boston, Base Air Quality,
Person Daysw/8-hrExposures>50 ppb
Model Explained Variance
jf 19 to 35
Boston, Exist!ng75 ppb Standard,
Person Days w/8-hr Exposures > 50 ppb
Model Explained Variance
Boston, Existing 75 ppb Standard,
All Person Days
Model Explained Variance
I Max8-hr Ambient O3
Max 8-hr Ambient O3*Time Outdoors 12-8PM
Time Outdoors 12-8PM*PAI
Time In-Vehicles 12-8PM
I Time Outdoors 12-8PM
I Physical Activity Index (PAI)
Time Near Roads 12-8PM
Figure 5G-3. Contribution of influential factors to daily maximum 8-hour ozone exposures
using base air quality (left), air quality adjusted to just meet the existing standard (right),
considering all person days (top) and those days where daily maximum 8-hour exposure
exceeded 50 ppb (bottom) in Boston.
5G-18
-------
Atlanta, Base Air Quality,
All Person Days
jjf 19 to 35
Model Explained Variance
Atlanta, Base Air Quality,
Person Daysw/8-hrExposures>50 ppb
jf 19 to 35
Model Explained Variance
Atlanta, Existing 75 ppb Standard,
All Person Days
Model Explained Variance
Atlanta, Existing75 ppb Standard,
Person Days w/8-hr Exposures > 50 ppb
jf 19 to 35
Model Explained Variance
I Max8-hr Ambient O3
Max 8-hr Ambient O3*Time Outdoors 12-8PM
Time Outdoors 12-8PM*PAI
Time In-Vehicles 12-8PM
I Time Outdoors 12-8PM
I Physical Activity Index (PAI)
Time Near Roads 12-8PM
Figure 5G-4. Contribution of influential factors to daily maximum 8-hour ozone exposures
using base air quality (left), air quality adjusted to just meet the existing standard (right),
considering all person days (top) and those days where daily maximum 8-hour exposure
exceeded 50 ppb (bottom) in Atlanta.
Because afternoon outdoor time expenditure and daily maximum 8-hour ambient Os
concentrations are an important determinant for high Os exposures regardless of air quality
scenario considered, we compared the distributions of these two variables using person days
where daily maximum 8-hour Os exposures were either below or above 50 ppb. Figure 5G-5
presents this comparison for Boston13 school-age children (ages 5 to 18) and adults (ages 19 to
35) and considering 2006 base air quality and air quality adjusted to just meet the existing Os 8-
hour standard. For school-age children that did not experience a daily maximum 8-hour
13 The overall features of the outdoor time and ambient concentration distributions illustrated by simulated persons
in Boston are similar to each of the other study areas (data not shown).
5G-19
-------
exposure at or above 50 ppb (e.g., top left panel, base air quality), over half of them did not
spend afternoon time spent outdoors, while just under 20% of them spent at least two hours of
their afternoon time spent outdoors, with fewer than 5% spending more than four hours of their
afternoon time outdoors. In addition, nearly 70% would have their daily maximum 8-hour
ambient concentrations below 50 ppb (please note, ambient is not exposure).
Not surprisingly, the distributions for both the outdoor time and ambient concentration
variables are shifted to the right of the figure for school-age children's person days where daily
maximum 8-hour exposures > 50 ppb (e.g., Figure 5G-1, top left panel, base air quality), as for
more than half of the days, highly exposed simulated individuals spend about 250 minutes
outdoors during the afternoon hours along with experiencing daily maximum 8-hour ambient Os
concentrations > 75 ppb.
By design, when air quality is simulated to just meet the existing standard (e.g., Figure
5G-5, bottom left panel), upper percentile ambient concentrations are reduced compared to those
comprising the base air quality such that the majority of ambient concentrations fall well below
the existing standard level of 75 ppb. Given so few occurrences of very high 8-hour ambient Os
concentrations for this air quality scenario, only those school-age children having a majority of
their time spent outdoors experienced the highest daily maximum 8-hour Os exposure
concentrations (Figure 5G-5, bottom left panel, right-most solid line). For additional
completeness, we note the time and concentration distributions for adult person days (Figure
5G-5, right column) were similar with that estimated for school-age children.
5G-20
-------
Boston, Children 5-18 Years Old, Base Air Quality
Daily Maximum 8-Hour Ambient Ozone (ppb)
0 25 50 75 100 125 150
OJ
^
s*
s
^*
^..
J^~
S*
i
i
I
if
IT*'
i
7
^=^2*
/
^Q
S t
l
1
l
S
*ZZ=^T~'
;
• i
1
1 (*»•
^
S
s^
s*^~
0 80 160 240 320 400 480
Total Afternoon Time Spent Outdoors (minutes)
Boston, Children 5-18 Years Old, Existing75 ppb Standard
Daily Maximum 8-Hour Ambient Ozone (ppb)
0 25 50 75 100 125 15
/
i
i
_jf
t
1
1
y
i_ -"""' j-"—
1
1
1
1
t
t
f
~^
^^
f
J^
0
• 100
0 80 160 240 320 400 480
Total Afternoon Time Spent Outdoors (minutes)
Boston, Adults 19-35 Years Old, Existing75 ppb Standard
Daily Maximum 8-Hour Ambient Ozone (ppb)
0 25 50 75 100 125 15
^
/^
/
*^~
^~~
I
I
/
/
1
!^-
~~!
/
. ^
z
L i
t
1
1
1
1
1
t
1
t
1
t
t
£
s
"J"
—
s
s*
r^
0
• 100
0 80 160 240 320 400 480
Total Afternoon Time Spent Outdoors (minutes)
time: Max 8-hr Exposure < 50 ppb ^— Out time: Max 8-hr Exposure > 50 ppb
8-hr Amb: Max 8-hr Exposure < 50 ppb --'Max 8-hr Amb: Max 8-hr Exposure >50 ppb
Figure 5G-5. Distributions of afternoon outdoor time expenditure and daily maximum 8-
hour ambient Os concentrations for simulated Boston school-age children (ages 5 to 18)
(left) and adults (ages 19 to 35) (right) using base air quality (top) and concentrations
adjusted to just meet the existing standard (bottom) for person days having daily
maximum 8-hour exposures either below or above 50 ppb.
By definition, any 8-hour average exposure is time-averaged across all
microenvironmental concentrations; thus several different microenvironments may contribute to
each person's daily maximum level. Understandably based on the above analyses, the outdoor
microenvironment is most important for persons experiencing the highest Os exposures, but we
are also interested in the percentage of time expenditure spent among detailed indoor, outdoor,
and vehicular locations people may inhabit during the afternoon. We summed the afternoon time
expended for highly exposed persons, considering a total of 12 microenvironments (i.e., 3
indoor, 5 outdoor, 2 near road, and 2 vehicular). As an example, Figure 5G-6 presents this
microenvironmental information for Boston school-age children (Figure 5G-6, top left panel) and
adults (Figure 5G-6, top right panel) for persons experiencing daily maximum 8-hour average Cb
exposures > 50 ppb and considering base air quality conditions. On average, approximately 50%
5G-21
-------
of school-age children's total afternoon time is spent outdoors, of which half of this portion is
spent outdoors at home, with parks and other non-residential outdoor locations comprising the
remaining portion. Approximately 45% of the school-age children's afternoon time on high
exposure days is spent indoors, while just less than 10% of afternoon time is spent near-roads or
inside motor vehicles. Afternoon microenvironmental time expenditure for highly exposed
adults (ages 19-35) in Boston was generally similar with that estimated for school-age children
(Figure 5G-6, top right panel).
A person's activity level plays an important role in estimating the risk of adverse health
responses to inhaled ozone. As such, we evaluated the activities performed by highly exposed
individuals while they spent time outdoors during the afternoon hours. Note there are over 100
specific activity codes used in CHAD/APEX, though not all of these will be used in an exposure
modeling simulation depending on the particular diaries that are selected to represent the
simulated study group. We summed the time spent in each specific activity across all highly
exposed persons when spending afternoon time outdoors, ranked the activity sums, and identified
the top eleven activities performed. An aggregate of any remaining less often performed
activities was generated to complete this analysis of activity time expenditure.
Figure 5G-6 shows results for Boston school-age children (bottom left panel), indicating
that greater than half of the time highly exposed children spent outdoors specifically involves
performing a moderate or greater exertion level activity, such as a sporting activity. The same
type of analysis was done for highly exposed adults in Boston (Figure 5G-6, bottom right panel),
whereas about 30% of the outdoor time expenditure was spent engaged in a paid work related
activity (though not necessarily a high exertion level activity), about 15% of the time was spent
playing sports or other moderate or greater exertion level activity, with much of the remaining
specific activities associated with low exertion level (e.g., eating, sitting, visiting) or other less
frequently performed activities of variable exertion level. These results support our
identification of school-age children as an important exposure group, largely a result of the
combined outdoor time expenditure along with concomitantly performing moderate or high
exertion level activities. It is worth noting, one important group not directly assessed in the
general population-based exposure modeling and remaining as a limitation to the main body
FIREA results is outdoor workers. This exposure study group was explicitly modeled using a
scenario based approach and summarized in section 5G-4.2.
5G-22
-------
Outdoor (Restaurant.
or Cafe)
Afternoon Time Expenditure: Children 5-18, 8-hr Exposures> 50 ppb,
Vehicle (Cars and Boston, Base Air Quality
.Vehicle Other (4 MEs)
Outdoor (School,
grounds)
Outdoor (Park or Golf.
loor(Otherindoor-
9MEs)
Afternoon Outdoor Activities: Children 5-18, 8-hr Exposures> 50 ppb,
Boston, Base Air Quality
Othersportsand_
active leisure "^^(
5% Play/OutdoorLeisui
Afternoon Time Expenditure: Adults 19-35, 8-hr Exposures >50 ppb,
ve hi cie(cars and Boston, Base Air Quality
Outdoor (Restaurant.
or Cafe)
Outdoor(ParkorGolf
course]
Light Duty Trucks).
Indoor (Office
luilding. Bank, Post
office)
VNear-roadOther(4
MEs)
lear-road (Within 10
yards of street)
Afternoon Outdoor Activities: Adults 19-35, 8-hr Exposures>50 ppb,
Boston, Base Air Quality
Repair/maintain
Otherentertainmi
/social events
Figure 5G-6. Afternoon microenvironmental time (top) and activities performed during
afternoon time outdoors (bottom) for school-age children (left) and adults (right)
experiencing 8-hour daily maximum Os exposures > 50 ppb, Boston base air quality, 2006.
5G-3 ANALYSIS OF APEX SIMULATED LONGITUDINAL ACTIVITY PATTERNS
IN SCHOOL-AGE CHILDREN
We evaluated the APEX approach used for linking together cross-sectional activity
pattern diaries to generate longitudinal profiles for our simulated individuals. Of particular
interest were how well variability in outdoor participation rate and the amount of time expended
were represented in our population-based exposure simulations. Our goal in developing the most
reasonable longitudinal profiles is to capture expected, important features of population activity
patterns, i.e., there is correlation within an individual's day-to-day activity patterns (though not
exactly repeated nor entirely random) and variability across the modeled study group in day-to-
day activity patterns (not every simulated individual in the study group does the same thing on
the same day). As a reminder, the longitudinal approach is probabilistic, though guided by key
variables influencing activity patterns (i.e., age, sex, day-of-week, commute time [employed
person only], daily maximum temperature, and in our application, considers within and between
variability in outdoor time expenditure). See HREA Appendix 5B, section 5B-4.2.
5G-23
-------
We used the same event-level output data that was generated for the high exposure
analysis above (section 5G-2), which includes the same six study areas - Atlanta, Boston,
Denver, Houston, Philadelphia, and Sacramento - and focused the analysis on school-age
children (ages 5-18). Total time spent outdoors during the afternoon hours (12 PM-8 PM) was
calculated for each person-day of the simulation in each study area. Results of this analysis are
presented in five individual plots for each study area, though combined in a multi-panel display,
one per three study areas, designed to fit on a single page. The five individual plots generated
for each study area are described as follows.
1) Cumulative distribution summarizing each child's median time spent outdoors
across an Os season:14 We first selected simulated individuals within the age group
of interest (5-18) and then stratified these persons by sex. The median value (50th
percentile) of afternoon time spent outdoors was determined for each simulated
individual using all days in their study area's Os season. This data set, comprised of
individual median values (in minutes) was ranked and plotted, stratified by sex.
2) Cumulative distribution summarizing each child's afternoon outdoor
participation (at least one minute/day) across an Os season: We subset school-age
children from the data set and stratified by sex. A categorical variable was developed
by assigning a numeric value of 1 when an individual spent at least one minute during
the afternoon hours outdoors on that given day. Then for each simulated individual,
outdoor participation was determined by summing this variable across the simulation
period (i.e., the number of days per Os season the persons spent at least one minute
outdoor during the afternoon hours) and dividing by the total number of days in that
study area's Os season. This data set, comprised of individual participation values
(provided as a percent) was ranked and plotted, stratified by sex.
3) Cumulative distribution summarizing each child's afternoon outdoor
participation (at least two hours/day) across an Os season: Calculated and
presented in the same manner as #2 above, only that the categorical variable was
assigned a numeric value of 1 if the simulated individual spent at least two hours
outdoor during the afternoon hours on that given day.
4) Daily time series of afternoon outdoor participation (at least two hours/day) by
study group across Os season: Using the categorical variable determined in #3, we
calculated the study group's outdoor participation for every day of each study area's
Os season. Data similar in fashion to our earlier analyses of outdoor time expenditure
(e.g., section 5G-2), only differing in that presented are day-to-day variability in
14 The number of days in an Os season varies across the six study areas. See Table 5G-6.
5G-24
-------
outdoor participation for the simulated study group across each study area's Os
season.
5) Daily time series of the number of study-specific CHAD diaries used across Os
season: The APEX daily file output can include the identity of the specific CHAD
diary used to simulate every individual's daily activity. For every day simulated, we
summed the number of CHAD diaries used to model the school age children's
activity patterns for each day, though stratified by CHAD study identifier (e.g., see
HREA Appendix 5B, Table 5B-1). Plotted is the day-to-day variability in particular
CHAD study diaries used across each study area's Os season.
The results of this longitudinal activity pattern analysis in given in Figure 5G-7 and
Figure 5G-8. To begin, a few generalities regarding the features of each plot and where
consistency is exhibited across study areas. In general, simulated female school-age children
tend to spend less afternoon time outdoors than their male counterparts, consistent of course with
expectations and the data used to develop these simulated profiles (Graham and McCurdy,
2004). About half of the simulated study group spends about half of their days with no afternoon
time spent outdoors across their study area's ozone season, while about 10-20% spent just over 2
hours afternoon time spent outdoors for half of their days (top row of Figure 5G-7 and Figure
5G-8). Nearly every simulated individual participates in at least one afternoon outdoor activity
across the Os season and exhibits a mostly monotonic relationship (2nd row of Figure 5G-7 and
Figure 5G-8), though when considering durations of 2 hours or more, longitudinal outdoor
participation drops dramatically (an non-linearly) for most persons comprising the study group
(3rd row of Figure 5G-7 and Figure 5G-8). For more than half of simulated school age children,
only approximately 1 out of every 5 days was spent outdoors during the afternoon hours for at
least two hours, while a maximum value (around 3 of 5 days) was simulated for only about 10%
or fewer children comprising the study group. Study group participation in at least two hours of
afternoon time outdoors day-to-day ranges from about 5-40% across each study area's Os season
(4th row of Figure 5G-7 and Figure 5G-8), though not surprisingly highest during typical summer
months (June through September). And finally, the majority of CHAD diaries that are used
come from the recently conducted ISR and OAB studies (bottom row of Figure 5G-7 and Figure
5G-8). This is also expected given that these two studies were designed to collect children's
activity patterns and contribute to the bulk of the children's diaries in CHAD. Worthy of note is
the shift in the source of diaries used across the calendar year; the contribution from the OAB
study increases during the summer months while that of the ISR wanes. This is because of the
days/seasons of the year the original study data were collected; most of the ISR data were
collected during non-summer months while the OAB study was conducted during peak Os
5G-25
-------
concentration days. While the use of these different studies in varying numbers over the
simulation period likely drives some of the observed variability in the outdoor participation, at
this time (and previously) staff treat the CHAD study data equally without bias, following our
initial screen of the CHAD master data base that selected for the most complete data available.
Variability in the five longitudinal display plots across the six study areas is evident,
though to a much smaller degree than that observed when considering the magnitude of the
within study area variability. While different lengths of each study areas' Os season may negate
direct comparability of the distributions presented, it is reasonable to conclude that simulated
school-age children in the Atlanta, Houston, and Sacramento study areas had slightly overall
greater participation in outdoor activities and spent more time outdoors than counterparts in
Boston, Denver, and Philadelphia. That said, when considering the daily time series of
participation rates, on many summer days Philadelphia and Sacramento school-age participation
rates for males are as great or greater than participation rates observed most other study areas,
including study areas likely having considerably warmer summer temperatures (Figure 5G-7 and
Figure 5G-8, 4th row). It is possible that for study areas such as Atlanta, summertime maximum
daily temperatures exceed the range affording outdoor comfort, yielding slightly lower rates of
participation in outdoor activities.
Overall, the simulated longitudinal profiles indicate the method for linking together
cross-sectional diaries generates a diverse mixture of persons having variable, though expected,
activity patterns: a small fraction of the simulated population spend a limited amount of
afternoon time outdoors and occurring at a low frequency across an Os season, a small fraction
consistently spends a greater amount (> 2 hours) of time outdoors and occurring at greater
frequency (e.g., 4/5 days per week), while the remaining simulated individuals fall somewhere in
between these two lower and upper bounds regarding participation and total time. While we are
not aware of a population database available to compare with these simulated results, we are
comfortable with the method performance in representing the intended variability in longitudinal
activity patterns
5G-26
-------
Atlanta, 2007
Boston, 2007
80
^gf*
•^***.^
&^
\j$K^
^^"*"tnCC^
_^fiSSJ
*£e3*>~ajm
-•-School-Age Children (Females)
^School-Age Children (Males)
Denver, 2007
90
80
40 •
^^
-rf^"*!^
*PjP*lH^
*$&i^^
-^**^8^s!ff
n^**^08^
^ndT
pJdT^
kiuu
-•-School-Age Children (Females)
^>School-Age Children (Males)
60 120 180
Median Afternoon Time Outdoors
(Minutes/Day)
60 120 180
Median Afternoon Time Outdoors
(Minutes/Day)
60 120 180 240
Median Afternoon Time Outdoors
(Minutes/Day)
Afternoon Outdoor Participation
(Percent of Days > 1 Minute/Day)
Afternoon Outdoor Participation
(Percent of Days > 1 Minute/Day)
Afternoon Outdoor Participation
(Percent of Days > 1 Minute/Day)
yu
to
20
10-
*
J&
M?
//
I
?. . .
J^
-rf^"*1
1^^^
^x^^
^^^
*-School-Age Children (Females)
^-School-Age Children (Males)
S bu
a) 5U
1 20 -
= 10 -
_^
^^" ^
/y
//
//
?
.s*
^^^s&
,4^
jr
^*^&6
r^^B*B^'^
^
--School-Age Children
o-School-Age Children
Females)
(Males)
o) yu
c 7n
z bu
a) 5U
1 20 -
= 10-
jt
jS.
Sj/
fjf
{/
w
L .
-X
-S^J
'^
~^~
f*^^^*~
r
j^JSJH
•-School-Age Children
^School-Age Children
Females)
(Males)
Afternoon Outdoor Participation
(Percent of Days > 2 Hours/Day)
Afternoon Outdoor Participation
(Percent of Days > 2 Hours/Day)
Afternoon Outdoor Participation
(Percent of Days > 2 Hours/Day)
•* IT) (D N- CO O)
Exposure Simulation Date
^ in 5 i^ co o5 o
Exposure Simulation Date ^~
§ §
IT) (D N- CO O) O
Exposure Simulation Date ^~
Exposure Simulation Date
Exposure Simulation Date
Exposure Simulation Date
Figure 5G-7. Cumulative distribution of median time spent outdoors (top row), afternoon
outdoor participation > 1 minute/day (2nd row), and afternoon outdoor participation > 2
hours/day (3nd row) for male and female school-age children in Atlanta (left column),
Boston (middle column) and Denver (right column) study areas. Percent of school-age
children with > 2 hours outdoors during afternoon hours (4th row) and the number of
particular CHAD study diary days used (bottom row) for each exposure simulation day.
5G-27
-------
Houston, 2007
80
40 •
-*"-",
•^O^
^_
•*f
^ rfG^
«X^
— es*8
55^
^uu~ ~
-*- School-Age Children (Females)
-o School-Age Children (Males)
Philadelphia, 2007
Sacramento, 2007
80
50 •
^p^***"^
r^ ^&*£
f**"^
^ff
^
jp/"^
*r^^
dcrfi"^
J-JJU
-•-School-Age Children (Females
^School-Age Children (Males)
0)
"c
| 30 •
^^J
^
,»?*
_^i-f~£Z3
^J**" iftS^
^* (jlffJ*^
)School-Age Children (Males)
60 120 180
Median Afternoon Time Outdoors
(Minutes/Day)
60 120 180
Median Afternoon Time Outdoors
(Minutes/Day)
60 120 180 240
Median Afternoon Time Outdoors
(Minutes/Day)
Afternoon Outdoor Participation
(Percent of Days > 1 Minute/Day)
Afternoon Outdoor Participation
(Percent of Days > 1 Minute/Day)
Afternoon Outdoor Participation
(Percent of Days > 1 Minute/Day)
yu
to
20
/
/ jf
if
If
r
_^r
S^jS
fS
r-*
^^
*-School-Age Children (Females)
^-School-Age Children (Males)
10 •
X
A^
//
//
,*
.X"^
^" ^
-rf^^
*^^~^ff
^^^
--School-Age Children
o-School-Age Children
Females)
(Males)
c 7n
s.
« 30 -
1 20-
= 10-
n .
J
/x
//^
jy
r
jt
^^^^
*&^*
^r
^+*
<[.,>
^y^"
"y^
--School-Age Children
«- School-Age Children
Females)
(Males)
Afternoon Outdoor Participation
(Percent of Days > 2 Hours/Day)
Afternoon Outdoor Participation
(Percent of Days > 2 Hours/Day)
Afternoon Outdoor Participation
(Percent of Days > 2 Hours/Day)
Exposure Simulation Date
Exposure Simulation Date
Exposure Simulation Date
Figure 5G-8. Cumulative distribution of median time spent outdoors (top row), afternoon
outdoor participation > 1 minute/day (2nd row), and afternoon outdoor participation > 2
hours/day (3nd row) for male and female school-age children in Houston (left column),
Philadelphia (middle column) and Sacramento (right column) study areas. Percent of
school-age children with > 2 hours outdoors during afternoon hours (4th row) and the
number of particular CHAD study diary days used (bottom row) for each exposure
simulation day.
5G-28
-------
5G-4 EXPOSURE RESULTS FOR ADDITIONAL AT-RISK POPULATIONS AND
LIFESTAGES, EXPOSURE SCENARIOS, AND AIR QUALITY INPUT DATA
USED
This section includes results for three additional simulations designed to complement
exposures estimated using our general population-based modeling approach presented in the
main body of the HREA. These simulations include (1) exposures estimated for school-aged
children during summer months only (section 5G-4.1), (2) adult outdoor worker exposures
(section 5G-4.2), and (3) exposures to school-age children and asthmatic school-age children
assuming a portion of these study groups exhibit averting behavior in response to high Cb
concentration days (section 5G-4.3).
5G-4.1 Exposure Estimated For All School-Age Children during Summer Months,
neither Attending School nor Performing Paid Work
A targeted simulation was performed for the Detroit study area during the months of June
through August 2007 to simulate summertime exposures by assuming all children were on a
traditional calendar year summer vacation. To do this, a subset of the CHAD diaries used by
APEX was created by including only those persons that did not have any time spent while at
school or time performing paid work. Even though the school children age range in our
exposure simulation is 5-18 years old, to maximize the number of diaries available for use by
APEX we expanded the CHAD diary selection to include children from 4-19 years old. In
considering these diary selection criteria, the resulting time location activity pattern data set input
to APEX had a total of 10,226 diaries having 379,524 event entries. All simulation conditions
were set identically to those set for the main HREA exposure simulations in Detroit, though
75,000 children were explicitly simulated here for this targeted analysis. Four air quality
scenarios were considered: just meeting the existing 8-hour standard of 75 ppb, and at alternative
levels of 70, 65, and 60 ppb. The exposure results from these targeted simulations were
compared with identical APEX simulations run using all available CHAD diaries during the
same summer months (i.e., diary days that include locations visited and activities performed
from persons reporting either school time or paid work).
Figure 5G-9 contains the exposure results for this simulation ( "No School/Work
Diaries") and for a nearly identical simulation that differed only in that is used all CHAD diaries
( "all CHAD Diaries "). When restricting the CHAD diary pool to include only those diaries
having no time spent at school or performing paid work activities, there is about 1/3 or 33%
increase in the number of children at or above each of the selected benchmark levels, a
relationship also consistent when considering multiple exposures over the simulation period.
5G-29
-------
>= 1 Exposure-All CHAD Diaries
>= 1 Expo sure-NoSchool/Work Diaries
>= Z Exposures-All CHAD Diaries
>= Z Expo sure s-NoSchool/Work Diaries
>= 3 Exposures-All CHAD Diaries
>= 3 Exposures-No School/Work Diaries
Percent of Children with 8-hr Daily Max Exposure > 60 ppb
standard level (ppb)
>= 1 Exposure-All CHAD Diaries
>= 1 Expo sure-NoSchool/Work Diaries
>— Z Exposures-All CHAD Diaries
>= Z Expo sure s-NoSchool/Work Diaries
>= 3 Exposures-All CHAD Diaries
]70
11 j
>= 3 Exposures-No School/Work Diaries
0.0% 0.5% 1.0% 1.5% Z.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0%
Percent of Children with 8-hr Daily Max Exposure > 70 ppb
standard level (ppb) ^^H CO
= 1 Exposure-All CHAD Diaries
= 1 Expo sure-NoSchool/Work Diaries
= 2 Exposures-All CHAD Diaries
= Z Expo sure s-NoSchool/Work Diaries
= 3 Exposures-All CHAD Diaries
= 3 Exposures-No School/Work Diaries
0.00% 0.05% 0.10% 0.15H 0.20% 0.25% 0.30% 0.35%
Percent of Children with 8-hr Daily Max Exposure > 80 ppb
standard level (ppb) I 1 60 I 1 65 I 1 70 I 1 75
Figure 5G-9. Comparison of the percent of all school-age children having daily maximum
O3 concentration at or above 60 ppb-8hr (top), 70 ppb-8hr (middle), or 80 ppb-8hr
(bottom) during June, July, and August in Detroit 2007: using any available CHAD diary
("All CHAD Diaries") or using CHAD diaries having no time spent in school or performing
paid work ("No School/Work Diaries").
5G-30
-------
5G-4.2 Exposures Estimated For Adult Outdoor Workers during Summer Months
A targeted APEX simulation was performed for the Atlanta study area to simulate
summertime exposures for two hypothetical adult outdoor worker study groups (ages 19-35 and
ages 35-54), using 2006 air quality just meeting the existing Os standard. To do this, both the
daily and longitudinal activity patterns used by APEX needed to best reflect patterns expected
for adult outdoor workers (e.g., a standardized work schedule during weekdays) while also
capturing variability in those patterns across various occupation types and the overall simulated
adult outdoor worker study group. The development of reasonable time location activity pattern
data to be input to APEX was a complex undertaking, attempting to account for a number of
influential factors such as the distribution of adult outdoor workers, their varying occupation
types, the probabilities associated with performing outdoor work, the linking of this information
with the existing CHAD diaries and APEX METS distributions, all to be done within the existing
APEX model framework and capabilities.
First, the complete distribution of all employed persons' occupations was estimated using
data provided by the U.S. Bureau of Labor and Statistics (US BLS, 2012b).15 The information of
interest was obtained from the 2010 National Employment Matrix, data originally developed
from the Occupational Employment Statistics (OES) survey and based on the 2010/2000
Standard Occupational Classification (SOC) system. The three variables retained for our
purposes here included the SOC occupation titles and codes and the 2010 estimated number of
persons employed, covering over 750 occupation titles.
Second, the identification of occupations where workers spend time outdoors for at least
one or more days per week was determined using data from the Occupational Information
Network (O*NET).16 The O*NET was developed by a partnership of public and private
organizations17 via sponsorship by the US Department of Labor/Employment and Training
Administration (USDOL/ETA). A wealth of information is provided by the O*NET regarding
specific occupations including human interaction processes (e.g., amount of public speaking in a
particular job, the likelihood of encountering angry people), physical work conditions (e.g.,
approximate time spent standing, whether exposed to radiation), and structural job characteristics
(e.g., the degree of job automation, freedom to make decisions).
An advanced search of O*NET was performed using the web database. We first isolated
the data of interest here by work context and selected physical work conditions. Data tables for
two survey question responses were downloaded: the first consisted of persons responding to the
15 US employment data by SOC codes were obtained from: http://www.bis.gov/emp/#tables: Table 1.2 Employment
by occupation, 2010 and projected 2020.
16 Additional information is available at http://www.onetonline.org.
17 Current O*NET partners include Research Triangle Institute (RTI), the Human Resources Research Organization
(HumRRO), North Carolina State University (NCSU), MCNC, and Maher & Maher.
5G-31
-------
question, "how often does this job require working outdoors, exposed to all weather conditions?"
and the second "how often does this job require working outdoors, under cover (e.g., structure
with roof but no walls)?". The tables contain the responses to these two questions, stratified by
the occupation names/codes, and rated using a context score ranging from 0 to 100. According
to the context scale provided by O*NET, occupations with a score of 75 were characterized as
having at least one day per week outdoors, while a score of 100 indicated that every day work
was performed outdoors by workers in that particular occupation, thus the greater the context
score, the greater the likelihood of outdoor work participation.
To start, there were 862 unique occupation codes with context scores for question #1 (i.e.,
exposed entirely to weather), 30 of which also contained context scores for question #2 (there
were no occupations with context scores for question #2 alone). Assuming ozone exposure
would be similar for outdoor workers whether under cover or totally exposed to weather, we
merged the data responses from the two questions by occupation code and assigned the highest
context score of the two responses to each occupation. Given the context scaling information
provided by O*NET, we then assumed the context scores 76-80, 81-85, 86-90, 91-95, and 96-
100 characterized occupations as having 1, 2, 3, 4, or 5 days per week outdoors. Then, mapping
of the O*NET occupations to the above described BLS SOC occupation data set was performed,
and was generally agreeable, with a few exceptions.18 Following additional processing, there
were 144 unique occupation titles having one to five days per week where work was performed
outdoors and the number of persons constituting each. See Attachment 1 for the final
O*NET/BLS mapping and additional data processing assumptions.
These 144 specific occupations then needed to be mapped to the occupation-related
activity codes used by APEX to generate METs in estimating energy expenditure. When CHAD
was developed in the late 1990s, occupation codes from the 1990 US Census were mapped to
twelve broad occupation categories19 and were assigned METs distributions based on the most
commonly performed activities associated with work tasks. In order to use the APEX/CHAD
METs database in its current format and integrate the newly developed 2010 BLS/O*NET data
set, the 1990 Census occupation codes were translated using two additional mapping files: a
1990 Census to 2000 Census code map file20 and a 2000 Census to SOC code map file.21 The
18 There were seven O*NET codes that did not directly correlate with the BLS SOC codes and several O*NET
occupations that were subcategories to broader BLS SOC codes.
19 The list of 1990 Census codes used to map the occupation titles to the APEX METs distributions are found at
https://usa.ipums.org/usa/volii/99occup.shtml. The twelve occupation groups are Executive Administrative
Managerial (ADMIN), Administrative support (ADMSUP), Professional (PROF), Technical (TECH), Sales
(SALE), Protective Services (PROTECT), Service (SERV), Farming Forestry Fishing (FARM), Precision
Production (PREC), Machine Operators (MACH), Transportation (TRANS), and Laborers (LABOR). Household
workers (HSHLD), while an APEX/CHAD occupation group, do not have any work days where time is spent
outdoors.
20 http://www.census.gov/people/io/files/techtab02.pdf.
5G-32
-------
mapping was generally agreeable among the various occupations and across the source files over
the years however, in characterizing several of the O*NET occupation codes to the broad
APEX/CHAD occupation groups, there were several instances where more than one broad
occupation group could be assigned to a single O*NET occupation. Professional judgement was
used to assign the most appropriate broad occupation category when there were multiple choices,
with decisions made to complete the processing of the 34 occupation titles given in Attachment 2
of this appendix. The final distribution of days per week spent performing outdoor work
considering the BLS/O*NET data set and stratified by APEX/CHAD occupation groups is
provided in Table 5G-8. This data set provides the target for developing an activity pattern data
set that reasonably reflects a distribution of outdoor workers properly weighted by occupation
groups along with assignment of an approximate number of days per week they spend outdoors.
Table 5G-8. Distribution of days per week spent performing outdoor work considering the
BLS/O*NET data set and stratified by APEX/CHAD occupation groups.
CHAD
Group
ADMIN
ADMSUP
FARM
LABOR
MACH
PREC
PROF
PROTEC
T
SALE
SERV
TECH
TRANS
TOTAL
BLS/O*NET Occupations
(n)
7
7
19
10
3
42
3
15
1
4
6
27
144
Number of Outdoor
Work Days per Week
(mean)
1.4
3.4
4.0
4.5
4.0
2.5
1.3
3
1.0
3.2
3.4
4.3
-
(min)
1
1
1
1
2
1
1
1
1
1
3
1
-
(max)
3
5
5
5
4
5
4
5
1
5
4
5
-
Percent of Employed Persons
Performing Outdoor Work
Outdoor
Workers only
5.1
4.6
18.3
9.3
0.5
25.1
0.2
7.7
2.4
1
3.1
22.7
100
All Workers
0.6
0.5
2.1
1.1
0.1
2.9
<0.1
0.9
0.3
0.1
0.4
2.7
11.7
Target for CHAD
Diary Outdoor Work
Days per Week
(based on
BLS/0*NET mean)
1
3
4
5
4
3
1
3
1
3
3
4
-
CHAD contains diary information from a number of persons who reported their time
spent at work, and is indicated by CHAD activity codes beginning with either '100' or ' 101'.
This, combined with outdoor location information (i.e., 60 location codes provided in
Attachment 3 of this appendix) and selected for where the total time working outdoors for the
day was > 2 hours, yielded 1,510 CHAD daily activity patterns potentially usable in representing
21 http://www.census.gov/people/eeotabulation/documentation/occcategories.pdf
5G-33
-------
outdoor workers. We evaluated these diaries across a number of personal attributes and
calculated the mean time spent outdoors for each diary day (Table 5G-9). As expected, very few
diaries were from persons that identified their occupation, though where that information was
present, these diaries retained that specific identifier and used to simulate persons having that
particular occupation group. Because we were interested in generating a typical work week, that
is, work was performed during weekday days, only weekday diaries were used to develop the
target diary pool for the weekday schedule, giving a final pool of 1,403 usable diaries.
Table 5G-9. Personal attributes and mean time spent working outdoors for CHAD diaries
reporting at least two hours of outdoor work.
Category
CHAD
Occupation
Group
Age Range
(years)
Sex
Day of Week
Attribute
ADMIN
ADMSUP
FARM
LABOR
MACH
PREC
PROF
PROTECT
SALE
SERV
TECH
TRANS
X
15-18
19-24
25-34
35-44
45-54
55-64
65-74
75-84
>84
F
M
FRI
MON
SAT
SUN
THU
TUE
WED
Outdoor Worker
Diary days (n)
13
4
11
14
10
52
8
4
13
5
1
12
1363
19
78
155
364
384
360
137
11
2
349
1161
123
174
66
41
280
357
469
Mean Outdoor Work
Time (minutes)
264
336
478
469
320
416
307
402
290
291
290
328
344
327
420
443
359
347
306
286
243
475
290
365
349
379
399
434
341
342
328
5G-34
-------
We assumed outdoor worker diaries having missing occupation information could be
used to represent outdoor worker diaries having an assigned occupation and, in the absence of
any additional information to suggest stronger alternative, simply assigned CHAD occupation
groups to these diaries equally, though weighted by the appropriate days per week outdoor work
was performed by a particular occupation group. This is in part because, in performing an APEX
simulation, the options for developing a multiday diary profile can be either controlled by a key
variable (such as time spent outdoors), use the same diary to represent every day of the simulated
person's exposure period, or be constructed by an entirely random sequence of diaries. Outdoor
time, a commonly used key variable to represent intra- and inter-personal variability activity
patterns over multiple days by APEX, in general is an unknown for each occupation group and
thus the key variable approach cannot be used to develop the longitudinal profiles. Rather than
use the same diary for each person day, we elected to use a random selection of diaries, though
having the selected diaries drawn from specific occupation groups developed from a large diary
pool weighted by the target number of days per week where outdoor work was performed (Table
5G-8). To clarify how this was done, an example follows using a single occupation group of
outdoor workers, i.e., those comprising the Transportation (TRANS) group.
According to information summarized in Table 5G-8, transportation-associated workers
(TRANS) were estimated to, on average, spend four days per week working outdoors. We
assigned the set of all available weekday outdoor worker diaries having 'missing' for their
occupation (n=l,363) as now having an occupation of 'TRANS'. We then replicated this new
set of outdoor worker diaries (including those few diaries having a known occupation, TRANS
or otherwise, to total 1,403 outdoor worker diaries) to generate a data set now having four
weekday days per person day. Thus from an APEX modeling perspective, all personal attributes
for persons in that pool are identical from one day to the next and have an equal likelihood of
selection. This four day by 1,403 person activity pattern data set, now principally comprised of
outdoor workers having 'TRANS' as their occupation, was then combined with a single weekday
by 1,403 person activity pattern data set, only that this particular one day data set, while still
derived from the same set of outdoor worker diaries, differs in that all of the paid work time
spent outdoors was changed to paid work time occurring within indoor locations for that one
work day. Then, when APEX constructs a longitudinal diary for any person with the 'TRANS'
occupation using completely random selection, all other personal attributes remaining the same,
the probability of selecting an outdoor work diary for any day is 0.8 while that of an indoor diary
is 0.2, appropriately reflecting, on average, the days per week that occupation group spends time
working outdoors. This process of building up the activity pattern diary pool was repeated for
each outdoor worker occupation group to reflect both the probability of performing either indoor
or outdoor work during weekdays. This collection of weekday diaries was then combined with
5G-35
-------
the pool of all remaining CHAD weekend diaries (n=13,953 person days), though where persons
had missing occupation information (n=12,561 person days), any one of the twelve occupation
groups was randomly assigned to a given weekend day. This complete outdoor worker activity
pattern set now totaled 98,133 person days22 having 3,656,560 activity events.
That said, it was soon apparent this input data set was too large for APEX to use when
error messages were generated upon model execution. For these outdoor worker simulations to
proceed, we determined the maximum size of the diary data set was approximately 42,000 diary
days. Our current response to this limitation was to first restrict the exposure simulations using a
tighter age range, thus permitting us to also limit the activity pattern input data set by similar age
ranges. Two age groups of outdoor workers were of interest for our exposure simulations: 19-35
and 36-55 years old. To maximize the number of diaries used to model the 19-35 year olds,23 we
increased the range for APEX usable diaries from the large outdoor worker activity pattern data
set initially developed to include ages from 16 to 42 years old (n=30,657 person days). As a
reminder, only the activity patterns of 16-18 year olds characterized as outdoor workers (i.e.,
persons having > 2 hours of paid work occurring outdoors) were available to simulate adults ages
19 or above. All of the anthropometric attributes of these simulated adults (e.g., body mass,
resting metabolic rate, ventilation rate) were derived from their age appropriate data,
distributions, and/or equations. There were adequate numbers of diaries available to model the
36-55 year olds such that the age range for inclusion in that data set was restricted to those
between the ages of 36-54 (n=41,736 person days).
When modeling exposures using occupation groups, two additional input files were
needed by APEX. The first was simply a file containing the CHAD ID and the specific
occupation group identified for that person day. The second, a profile factors file, contains the
probabilities a simulated person in the model domain will have a particular occupation, a profile
variable that can be stratified by age, sex, and/or census tract (US EPA 2012a, b). Based on the
information we developed in Table 5G-8, we only assigned specific probabilities for each
particular occupation group, i.e., the percent of employed persons performing outdoor work
using the outdoor worker proportions equally across ages, sexes and tracts. An additional
modification to the APEX employment probabilities input file (Employment2000 043003.txt)
was also needed to generate exposures and appropriate output summary tables only for employed
persons. And finally, because of the generally limited number of diaries available in developing
the diary pool for the 19-35 year olds and to use as many of the diaries available, we lengthened
the age selection range (AgeCutPct = 30.0 and Age2Probab = 0.15) and only included two
22 In summary, the weekday data set was developed assuming 12 occupations for 5 days per work week (or 35
outdoor days+ 25 indoor days) for each of the 1,403 diary days and added to the 13,953 weekend days.
23 Note from Table 5G-9, there are just over 250 diaries from persons aged 15-34.
5G-36
-------
temperature diary pools (<84 or >84) to have an adequate number of diaries available in each
diary pool for APEX to execute the desired simulation.
Finally, a 10,000 person exposure simulation was performed for each age group of
outdoor workers in Atlanta for Jun 1- Aug 30, 2006 using air quality just meeting the existing
standard. In addition, as a point of comparison for the longitudinal approaches developed here
and used for estimating general population-based exposures, identical simulations were
performed in Atlanta during the same time of year, air quality scenario, and age groups only
differing by using the approach described in our primary HREA exposure simulations, i.e., using
outdoor time as the key variable in developing longitudinal profiles, sampling from all available
CHAD diaries, and not explicitly addressing simulated worker occupations and work performed
(structured schedules and associated METs values).24
We first summarized the outdoor work time performed by each simulated outdoor worker
during weekdays, stratified by particular occupation, to ascertain whether or not the exposure
simulation met our defined goal. In comparing the results of Table 5G-10 with those provided in
Table 5G-8 we see that the goal was met for both age groups of outdoor workers, i.e., the
longitudinal approach was structured correctly to reproduce the distribution of the outdoor
worker occupation groups and the number of days persons in a particular group spent working
outdoors. Estimated exposures are presented in (Figure 5G-10) for each of the two outdoor
worker study groups and considering either a longitudinal scenario-based approach designed
specifically to reflect an outdoor worker weekday schedule or using our general population-
based modeling approach. It is clear that when accounting for a structured schedule that includes
repeated occurrences of time spent outdoors for a specified study group, all while more
consistently performing work tasks that may be at or above moderate or greater exertion levels, a
greater percent of the study group experiences exposures at or above the selected health effect
benchmark levels than that estimated using our general population-based approach. The
differences between exposures estimated for the two longitudinal approaches become much
greater when considering the percent of persons experiencing multiple exposure days at or above
benchmark levels. For example, < 2% of the general population-based exposure group was
estimated to have two or more exposures at or above 60 ppb-8hr, while >17% of specifically
simulated outdoor workers were estimated to experience exposures at or above that same level.
In general, there was little difference in exposures estimated for the two age groups of outdoor
workers.
24 Because most CHAD IDs have unknown occupations, METs are sampled from a 'composite' triangular
distribution min, peak, max [1.2, 1.9, 5.6] developed from the METs distributions used for all occupation groups.
As such, exertion levels achieved by laborers, service, and transportation industries are not well represented using
a general population approach given their respective METs distributions triangular[3.6, 8.1, 13.8], triangular[1.6,
5.6, 8.4], and lognormal geometric mean, standard deviation [3.0, 1.5] truncated to min, max [1.3, 8.4].
5G-37
-------
Table 5G-10. Distribution of days per week spent performing outdoor work considering
the APEX simulated population and stratified by APEX/CHAD occupation and age
groups.
Age
Group
19-35
36-55
CHAD Occupation
Group
ADMIN
ADMSUP
FARM
LABOR
MACH
PREC
PROF
PROTECT
SALE
SERV
TECH
TRANS
ADMIN
ADMSUP
FARM
LABOR
MACH
PREC
PROF
PROTECT
SALE
SERV
TECH
TRANS
Percent of Simulated
Outdoor Workers
5.0
4.6
19.1
9.5
0.5
24.3
0.2
7.8
2.4
1.1
3.2
22.4
5.4
4.7
17.5
9.6
0.6
24.8
0.3
7.5
2.5
0.9
3.1
23.2
Number of Outdoor Work Days
per Week
(mean)
1.2
3.0
3.9
4.9
4.0
3.0
1.4
3.0
1.2
3.0
3.0
4.0
1.1
3
3.9
4.9
3.9
3
1
3
1.1
3
3
3.9
(min)
0
2
3
5
3
2
1
2
0
2
2
3
0
2
3
5
3
2
0
2
0
2
2
3
(max)
2
4
5
5
5
4
3
4
2
4
4
5
2
4
5
5
4
4
2
4
2
4
4
5
Additional context regarding the estimate of the number of persons exposed between the
two approaches can added with the following. Approximately 30% of our outdoor worker study
group ages 19-55 was estimated to experience at least one daily maximum 8-hr average exposure
at or above 60 ppb while at moderate or greater exertion. Assuming a 92% employment rate and
that outdoor workers constitute approximately 12% of the workforce (Table 5G-8), outdoor
workers experiencing at least one daily maximum 8-hr average exposure at or above 60 ppb
would comprise about 3.3% of a total simulated population in that study area. For the same air
quality scenario we estimated using the general population-based approach, about 5-8% of the
study group would experience exposures at or above the same benchmark. To some extent, the
general population-based approach will simulate exposure profiles of outdoor workers (persons
with frequent and above average time spent working outdoors experiencing instances of high
exposure concentrations at elevated exertion levels) by applying our standard longitudinal diary
5G-38
-------
selection method. Intuitively, outdoor workers probably constitute a significant portion of the
overall population-based study group that could exceed benchmark levels though at this time it is
unknown whether the portion estimated here of 40-60% is accurate. It is however reasonable to
conclude given the comparative exposure results that the general population-based approach
would tend to underestimate the multiday exposures at or above selected benchmark levels
experienced by persons adhering to a more rigid outdoor work schedule.
Outdoor Worker Scenario-based Approach (ages 19-35)
Atlanta, 2006, Just Meet Existing 75 ppb Standard
Number of Times Bench mark Exceeded (June-August 2006)
Outdoor Worker Scenario-based Approach (ages 36-55)
Atlanta, 2006, Just Meet Existing 75 ppb Standard
123456
Number of Times Bench mark Exceeded (June-August 2006)
General Population-based Approach (ages 19-35)
Atlanta, 2006, Just Meet Existing 75 ppb Standard
General Population-based Approach (ages 36-55)
Atlanta, 2006, Just Meet Existing 75 ppb Standard
123456
Number of Times Bench mark Exceeded (June-August 2007)
123456
Number of Times Bench mark Exceeded (June-August 2006)
Figure 5G-10. Percent of persons between age 19-35 (left) and 36-55 (right) experiencing
exposures at or above selected benchmark levels while at moderate or greater exertion
using an outdoor worker scenario-based approach (top) and a general population-based
approach (bottom) considering air quality adjusted to just meet the existing standard in
Atlanta, GA, Jun-Aug, 2006.
5G-4.3 Averting Behavior and Potential Impact to Exposure Estimates
A growing area of air pollution research involves evaluating the actions persons might
perform in response to high Os concentration days (ISA, section 4.1.1). Most commonly termed
averting behaviors, they can be broadly characterized as personal activities that either reduce
5G-39
-------
pollutant emissions or limit personal exposure levels. The latter topic is of particular interest in
this HREA due to the potential negative impact it could have on Os concentration-response (C-
R) functions used to estimate health risk and on time expenditure and activity exertion levels
recorded in the CHAD diaries used by APEX to estimate Os exposures. To this end, we have
performed an additional review of the available literature here beyond that summarized in the
ISA to include several recent technical reports that collected and/or evaluated averting behavior
data. Our purpose is to generate a few reasonable quantitative approximations that allow us to
better understand how averting behavior might affect our current population-based exposure and
risk estimates. We expect that the continued development and communication of air quality
information via all levels of environmental, health, and meteorological organizations will only
further increase awareness of air pollution, its associated health effects, and the recommended
actions to take to avoid exposure, thus making averting behaviors and participation rates an even
more important consideration in future Os exposure and risk assessments. The following is a
summary of our literature review, with details provided by Graham (2012). Later in this section,
preliminary results of an exposure simulation designed to account for averting behavior are
provided.
The first element considered in our evaluation is peoples' general perception of air
pollution and whether they were aware of alert notification systems. The prevalence of
awareness was variable; about 50% to 90% of survey study participants acknowledged or were
familiar with air quality systems (e.g., Blanken et al., 1991; KS DOH, 2006; Mansfield et al.,
2006; Semenza et al., 2008) and was dependent on several factors. In studies that considered a
persons' health status, e.g., asthmatics or parents of asthmatic children, there was a consistently
greater degree of awareness (approximately a few to 15 percentage points) when compared to
that of non-asthmatics. Residing in an urban area was also an important influential factor raising
awareness, as both the number of high air pollution events and their associated alerts are greater
when compared to rural areas. Of lesser importance, though remaining a statistically significant
influential variable, were several commonly correlated demographic attributes such as age,
education-level, and income-level, with each factor positively associated with awareness.
The second element considered in our evaluation was the type of averting behaviors
performed. For our purposes in this Os HREA, the most relevant studies were those evaluating
outdoor time expenditure, more specifically, the duration of outdoor events and the associated
exertion level of activities performed while outdoors. This is because both of these variables are
necessary to understanding Os exposure and associated adverse effects, and hence, in accurately
estimating human health risk.
As stated above regarding air quality awareness, asthmatics consistently indicated a
greater likelihood of performing averting behaviors compared to non-asthmatics - estimated to
5G-40
-------
differ by about a factor of two. This difference could be the combined effect of those persons
having been advised by health professional to avoid high air pollution events and them being
aware of alert notification systems. Based on the survey studies reviewed, we estimate that 30%
of asthmatics may reduce their outdoor activity level on alert days (e.g., KS DOH, 2006;
McDermott et al., 2006; Wen et al., 2009).25 An estimate of 15%, derived from reductions in
public attendance at outdoor events (Zivin and Neidell, 2009) would be consistent with our
estimate above when considering that the Zivin and Neidell (2009) study group is likely
comprised mainly of non-asthmatics. That said, both attenuation and the re-establishment of
averting behavior was apparent when considering a few to several days above high pollution
alert levels (either occurring over consecutive days or across an entire year) (McDermott et al.,
2006; Zivin and Neidell, 2009), suggesting that participation in averting behavior over a
multiday period for an individual is complex and likely best represented by a time and activity-
dependent function rather than a simple point estimate.
There were only a few studies offering quantitative estimates of durations of averting
behavior, either considering outdoor exertion level or outdoor time (Bresnahan et al., 1997;
Mansfield et.al, 2006, Neidell, 2010; Sexton, 2011). Each of these studies considered outdoor
time expenditure during the afternoon hours. Based on the studies reviewed, we estimate that
outdoor time/exertion during afternoon hours may be reduced by about 20-40 minutes in
response to an air quality alert notification. Generally requisite factors include: a high alert level
for the day (e.g., red or greater on the AQI), high Os concentrations (above the NAAQS), and
persons having a compromised health condition (e.g., asthmatic or elderly).
The third element considered in our preliminary evaluation was how to further define the
impact of averting behavior on modeled exposure estimates.26 As described in HREA section
5.2.5, APEX uses time location activity data (diaries) from CHAD to estimate population-based
exposures. These diaries originate from a number of differing studies; some were generated as
part of an air pollution research study, some were collected during a summer/ozone season, while
some diary days may have corresponded with high Os concentration and air quality alert days.
At this time, none of the diary days used by APEX have been specifically identified as
representing days where a person did or did not adjust their activity pattern reduce their
exposure. In considering the above discussion regarding the potential rate of participation and
averting actions performed, it is possible that some of the CHAD diary days express instances
where that selected individual may have reduced their time spent outdoors or reduced their
exertion level while outdoors. Currently, without having a personal identifier for averting
25 Many of these studies do not account for one important factor when using a recall questionnaire design: whether
the participant's stated response to air pollution is the same as the action they performed.
26 The discussion of another important effect of averting behavior is on concentration-response functions (more
relevant to the risk assessment in chapter 8). This is presented in the ISA (section 4.1.2).
5G-41
-------
behavior in CHAD, the diaries are assigned to a simulated persons' day without directly
considering ambient Os concentration levels27. Therefore, it is possible that there are instances
where, on a given APEX simulation day, the simulated person may use a diary day from a person
that did engage in one or more types of averting behavior (e.g., a diary having less time than
usual spent outdoors in the afternoon), while for most other persons simulated on the same day
(or the same person on a different high concentration day) the diaries used are from persons that
did not actively engage in averting behavior. As a result, the effect of averting behavior may
already be incorporated into our exposure modeling, albeit to an unknown though likely small
degree,28 though definitely generating low-biased estimates of exposures (and reduced number of
persons at or above selected 8-h Os benchmark concentrations) that would occur in the complete
absence of averting behavior.
With this in mind, we performed an APEX simulation to reflect the instance that a
fraction of a selected study group spends less time outdoors on high concentration ozone days.
First, a general APEX simulation was performed during June-August 2007 in Detroit to identify
a short time period where a high number of children/asthmatic children were estimated to be
exposed at or above the 8-hour Os benchmark levels of interest. To maintain a degree of
tractability in the simulations, the development of the new CHAD input data, and the analysis of
the exposure results, we restricted the exposure simulations to 5,000 total persons. One such
high exposure event occurred over a two-day period considering base year air quality - August
1-2, 2007. Because conditions in APEX simulations can be controlled by using an identical
random number seed, APEX daily files were output to explicitly identify all of the CHAD diaries
used for this two-day simulation.
The activity pattern data from the identified CHAD diaries used to simulate the two-day
exposure period were then used to generate a new activity pattern input data set, one adjusted for
the above estimated parameters used to reflect averting performed by the two exposure study
groups of interest. We did this after determining the following:
1) There were a total of 1,988 diaries used to simulate the maximum exposure day for
each person, obviously some CHAD diaries were used more than once to simulate
different people on their maximum exposure day. Note also, 48 diaries were used to
simulate both days for the same person, 37 of these occasions were for unique
individuals while for the remaining 11 instances the same diary was also used in
either two or three persons two-day simulation.
27 APEX uses maximum temperature in assigning diaries for a select day in an area, capturing some relevant
variability in O3 concentrations.
28 Neither the participation rate nor the duration of averting for simulated persons is being strictly controlled for by
APEX when simulating exposures.
5G-42
-------
2) We calculated the total time spent outdoors by hour of day for the CHAD diaries used
by APEX for each persons' maximum exposure day, estimated the mean outdoor time
and number of persons by hour of day, and then stratified these results by the number
of times a diary was used in the simulation. Ideally, the most effective adjustment to
outdoor time considering averting would be applied to the afternoon hours. We
observed that at hour 15 (3 PM-4 PM), there were both a reasonable number of
CHAD diaries and diaries having an appropriate total time spent outdoors to achieve
our desired adjustment of 15% and 30% participation for children and asthmatics,
respectively, for on average 40 minutes in a day. This was determined using the
results in the following table. To simulate averting for children, we selected the
collection of diaries from the '2' and '3' times a diary was used category (n=766 total
diary days or 15.3% of all diaries used for each person's maximum exposure day) and
reduced all outdoor time events to 0 minutes spent outdoors during 3 PM-4 PM. To
simulate averting for asthmatic children, we selected the collection of diaries from the
' 1' through '5' times a diary was used category (n=l,518 total diary days or 30.4% of
all diaries used for each persons' maximum exposure day) and reduced all outdoor
time events to 0 minutes spent outdoors during 3 PM-4 PM.
Number of
times a select
diary was used
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
17
Number
of
persons
181
182
134
84
47
20
13
7
4
5
5
2
1
1
3
1
Mean
Outdoor
Time
43.2
43.4
44.6
45.6
40.6
48.4
44.1
45.4
39.5
48.0
40.6
35.0
5.0
45.0
60.0
60.0
Total
Diary
Days
181
364
402
336
235
120
91
56
36
50
55
24
13
14
45
17
3) Following the selection of the hour where outdoor time would be adjusted to reflect
averting, we evaluated whether diaries used for both simulation days per person and
when used for multiple persons for both simulation days would affect the targeted
reduction in outdoor time. Even considering the small number of instances identified
5G-43
-------
in #1) above (48 single diaries used for both days of the two day simulation per
person out of 5,000 possible cases), only two diaries had time spent outdoors during
the hours of 3 PM-4 PM where both days would be adjusted for averting, thus likely
having a negligible effect the meeting of our approximate averting goals.
Both the outdoor time adjusted CHAD and the standard CHAD input files were
separately used to simulate exposures to children and asthmatic children. Exposure results for
the four simulations (simulated averting vs. no averting for both children and asthmatic children)
are found in the main body of the HREA.
5G-5 COMPARISON OF PERSONAL EXPOSURE MEASUREMENT AND APEX
MODELED EXPOSURES
A new evaluation of APEX was performed using a subset of personal Os exposure
measurements obtained from the Detroit Exposure and Aerosol Research Study (DEARS) (Meng
et. al, 2012). For five consecutive days, personal Os outdoor concentrations along with daily
time-location activity diaries were collected from 36 study participants in Wayne County
Michigan during July and August 2006. The majority of participants were female (80%) having
a mean age of 40.6 (min 20, max 72); mean age for males was 41.4 (min 22, max 65). Rather
than using daily personal exposures estimated below the reported detection limit of 3 ppb (i.e., 0,
1 and 2 ppb), we approximated those falling below this level using a random assignment of
concentrations of 1 and 2 ppb.
An APEX simulation was performed considering these same geographic and temporal
features, followed with the sub-setting of APEX output data according to important personal
attributes of the DEARS study participants (specific 5-day collection study periods, age/sex
distributions, outdoor time, ambient concentrations, and air exchange rate). For both data sets
and considering the output variables independently, the median daily values for each study
participant attribute was generated using each individual's 5 person-days of data, then ranked
median values were plotted along with each individual's corresponding minimum and maximum
value. Distributions for four of these variables of (personal Os exposure, outdoor time, ambient
Os concentrations, and air exchange rate from each of the two data sets are presented in Figure
5G-11.
Distributions of time spent outdoors and ambient concentrations were similar by design
of the APEX population-based sample selection method. The upper percentiles the DEARS
participant AER distribution was greater than that of AER of APEX simulated persons. For
example, 40% of DEARS participants had a median value of two air exchanges per hour, while
the same rate was only observed for 5% of APEX simulated individuals. In contrast, over 50%
of APEX simulated individuals had median daily Os exposure concentrations above 10 ppb,
5G-44
-------
while only 3% of DEARS participants' median values exceeded 10 ppb. This difference in
exposure is surprising given the sharply higher residential indoor air exchange rate for the
DEARS participants (i.e., indoor microenvironmental exposures would be expected to have a
greater influence on total DEARS exposure compared with the APEX simulated exposures) all
while holding all other potential influential variables the same between the two data sets and is
subject to further investigation.
5G-45
-------
DEARS DATA
+j
a, 60
Q.
0) ,-
.a 40
™30 -
320
X
*,y
x x iK"
,/,
/
I
j + + +
XT'
' +
+ +
+
+
X mm
+ max
10 15 20 25 30
Daily Mean 03 Exposure(ppb)
U
Q.
>40 -
™30 -
Eno
O
n -
X
XXX
x ?
,x/,
/ +
+
* X
r_--
7
+ +
+
+ + +
,
/
y"
I
+
/-
+
+
+
4
-p50
™
60 120 180 240 300 360 420 480 540 600
Daily Total OutdoorTime(minutes)
5? 70
c fin
320
O
0 5 10 15 20 25 30 35 40 45 50 55 60
Daily Mean Ambient 03 Concentration (ppb)
U
Q.
«,„
O
x
X
X
y +
i+
f
l+'
X
x x
K j£
%/+
L +
+
X
„ x
J
/+
^
f
*
+
t
+
p50
X mm
2345678
DailyAir Exchange Rate(hour1)
100 •
4-1
a, 60
Q.
r°
320
10 •
APEX DATA
x
x
x x
X
X
'
x *x >
x
.X
x
< x
x ..x"
X ,
x" I
x f
2:
/ +
x s
* /
\/ |
r +
/ \
' + +
+
t+
+
-^
+
+
+
+
f + +
+
t +4
+
+
+
*
*
t
t
p50
+ max
10 15 20 25 30
Daily Mean 03 Exposure (ppb)
5? 70
c fin
320
O
10
0
0 60 120 180 240 300 360 420 480 540 600
Daily Total Outdoor Time (minutes)
5? 70
= 20
O
10
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Daily Mean Am blent 03 Concentration (ppb)
5? 70
I 60
£40
I 30
320
O
10
0
345678
DailyAir Exchange Rate (hour1)
Figure 5G-11. Distribution of daily personal Os exposures (top row), outdoor time (2nd row
from top), ambient Os concentrations (3rd row from top), and air exchange rate (bottom
row) for DEARS study participants (left column) and APEX simulated individuals (right
column) in Wayne County, MI, July-August 2006.
5G-46
-------
5G-6 REFERENCES
Ainsworth, B. E., Haskell, W. L., Leon, A. S., Jacobs, D. R., Jr., Montoye, H. J., Sallis, J. F.,
Paffenbarger, R. S., Jr. (1993). Compendium of physical activities: classification of energy
costs of human physical activities. Med Sci Sports Exer. 25:71-80.
Arcus-Arth, A. and Blaisdell, J. (2007). Statistical distributions of daily breathing rates for
narrow age groups of infants and children. Risk Analysis. 27(1):97-110.
Akland, G. G., Hartwell, T. D., Johnson, T. R., Whitmore, R. W. (1985). Measuring human
exposure to carbon monoxide in Washington, D. C. and Denver, Colorado during the
winter of 1982-83. Environ Sci Technol. 19:911-918.
Blanken, P. D., Dillon, J., Wismann, G. (2001). The impact of an air quality advisory program
on voluntary mobile source air pollution reduction. Atmos Environ. 35:2417-2421.
Bresnahan, B. W., Dickie, M., Gerking, S. (1997). Averting behavior and air pollution. Land
Econ. 73:340-357.
Brochu, P., Ducre-Robitaille, J.-F., Brodeur, J. (2006a). Physiological daily inhalation rates for
free-living individuals aged 1 month to 96 years, using data from doubly labeled water
measurements: a proposal for air quality criteria, standard calculations and health risk
assessment. Human and Ecological Risk Assessment. 12:675-701.
Brochu, P., Ducre-Robitaille, J.-F., Brodeur, J. (2006b). Supplemental Material for
Physiological daily inhalation rates for free-living individuals aged 1 month to 96 years,
using data from doubly labeled water measurements: a proposal for air quality criteria,
standard calculations and health risk assessment. Human and Ecological Risk Assessment.
12:1-12.
EPRI. (1988). A Study of Activity Patterns Among a Group of Los Angeles Asthmatics.
Research Project 940-5. Electric Power Research Institute. Prepared by Roth Associates.
November 1988.
EPRI. (1992). A Survey of Daily Asthmatic Activity Patterns in Cincinnati. TR-101396.
Research Project 940-05. Electric Power Research Institute. Prepared by Roth Associates.
November 1992.
Ford, E. S., Heath, G. W., Mannino, D. M., Redd, S. C. (2003). Leisure-time physical activity
patterns among US adults with asthma. Chest. 124:432-437.
Freedson, P. S. (1989). Field monitoring of physical activity in children. Pediatr Exerc Sci.
1:8-18.
George, B. J., McCurdy, T. (2009). Investigating the American Time Use Survey from an
exposure modeling perspective. JESEE. 21:92-105.
Geyh, A. S., Xue, J., Ozkaynak, H., and Spengler, J. D. (2000). The Harvard Southern
California chronic ozone exposure study: assessing ozone exposure of grade-school-age
children in two Southern California communities. Environ Health Persp. 108:265-270.
Glen, G., Smith, L., Isaacs, K., McCurdy, T., Langstaff, J. (2008). A new method of longitudinal
diary assembly for human exposure modeling. J Expos Sci Environ Epidem. 18:299-311.
Graham, S. E., McCurdy, T. (2004). Developing meaningful cohorts for human exposure
models. J Expo Anal Environ Epidemiol. 14:23-43.
5G-47
-------
Graham, S. E., McCurdy, T. (2005). Revised ventilation rate (VE) equations for use in
inhalation-oriented exposure models. Report no. EPA/600/X-05/008 is Appendix A of US
EPA (2009b). Available at: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=202543.
Graham, S. (2009). Response to peer-review comments on Appendix A, prepared by S. Graham
(US EPA). Response is Appendix D of US EPA (2009b). Available at:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=202543.
Graham, S. E. (2012). Comprehensive review of published averting behavior studies and
available technical documents. Memo to Bryan Hubbell, Group Leader, Risk and Benefits
Group, Office of Air Quality Planning and Standards. Docket # EPA-HQ-OAR-2008-
0699-0085.
Hartwell, T. D., Clayton, C. A., Ritchie, R. M., Whitmore, R. W., Zelon, H. S., Jones, S. M.,
Whitehurst, D. A. (1984). Study of Carbon Monoxide Exposure of Residents of
Washington, DC and Denver, Colorado. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Research and Development, Environmental Monitoring
Systems Laboratory. EPA-600/4-84-031.
Houtven, G. V., Johnson, F. R., Mansfiled, C., Yang, J.-C., Pyles, A. (2003). Parental Averting
Behavior With Respect to Ozone Alerts. For presentation at AEA/ASSA Meeting, San
Diego CA, January 2004.
Isaacs, K., Glen, G., McCurdy, T., Smith, L. (2008). Modeling energy expenditure and oxygen
consumption in human exposure models: accounting for fatigue and EPOC. JExpos Sci
Environ Epidemiol. 18: 289-298.
Johnson, T. (1984). A Study of Personal Exposure to Carbon Monoxide in Denver, Colorado.
Research Triangle Park, NC: U.S. Environmental Protection Agency, Environmental
Monitoring Systems Laboratory. EPA-600/4-84-014.
Johnson, T. (1989). Human Activity Patterns in Cincinnati, Ohio. Palo Alto, CA: Electric
Power Research Institute. EPRI EN-6204.
Johnson, T., Capel, J., McCoy, M. (1996a). Estimation of Ozone Exposures Experienced by
Urban Residents Using a Probabilistic Version of NEM and 1990 Population Data.
Prepared by International Technology Air Quality Services for Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park,
NC. Contract no. 63-D-30094. April 1996. Available at:
http://www.epa.gOv/ttn/naaqs/standards/ozone/s o3_prtd.html.
Johnson, T., Capel, J., McCoy, M., Warnasch, J. (1996b). Estimation of Ozone Exposures
Experienced by Outdoor Children in Nine Urban Areas Using a Probabilistic Version of
NEM. Prepared by International Technology Air Quality Services for Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, NC. Contract no. 63-D-30094. April 1996. Available at:
http://www.epa.gov/ttn/naaqs/standards/ozone/s_o3_pr_td.html
Johnson, T., Capel, J., McCoy, M., Warnasch, J. (1996c). Estimation of Ozone Exposures
Experienced by Outdoor Workers in Nine Urban Areas Using a Probabilistic Version of
NEM. 1996. Prepared by International Technology Air Quality Services for Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, NC. Contract no. 63-D-30094. April, 1996. Available at:
http://www.epa.gOv/ttn/naaqs/standards/ozone/s o3 pr td.html
5G-48
-------
Klepeis, N. E., Tsang, A. M., Behar, J. V. (1996). Analysis of the National Human Activity
Pattern Survey (NHAPS) Respondents from a Standpoint of Exposure Assessment.
Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Development. EPA/600/R-96/074.
Knowledge Networks. (2009). Field Report: National Scale Activity Survey (NSAS).
Conducted for Research Triangle Institute. Submitted to Carol Mansfield November 13,
2009.
KS DOH. (2006). Environmental Factors, Outdoor Air Quality, and Activity Level. Results
from 2005 Kansas Behavioral Risk Factor Surveillance System. Office of Health
Promotion, Kansas Department of Health and Environment. Available at:
http ://www.kdheks. gov^rfss/PDF/cste_report_final .pdf
Langstaff, J. E. (2007). OAQPS Staff Memorandum to Ozone NAAQS Review Docket (OAR-
2005-0172). Subject: Analysis of Uncertainty in Ozone Population Exposure Modeling.
[January 31,2007]. Available at:
http://www.epa.gOv/ttn/naaqs/standards/ozone/s o3crtd.html
Linn, W. S., Avol, E. L., Peng, R. C., Shamoo, D. A., Hackney, J. D. (1987). Replicated dose-
response study of sulfur dioxide effects in normal, atopic, and asthmatic volunteers. Am
RevRespirDis. 136:1127-1134.
Lioy, PJ. (1990). The analysis of total human exposure for exposure assessment: A multi-
discipline science for examining human contact with contaminants. Environ Sci Technol.
24: 938-945.
Liu, L.-J. S., Box, M., Kalman, D., Kaufman, J., Koenig, J., Larson, T., Lumley, T., Sheppard,
L., Wallace, L. (2003). Exposure assessment of particulate matter for susceptible
populations in Seattle. Environ Health Persp. Ill: 909-918.
Mansfield, C., Houtven, G. V., Johnson, F. R., Yang, J.-C. (2009). Environmental Risks and
Behavior: Do children spend less time outdoors when ozone pollution is high? ASS A
annual meeting, January 5, 2009. Update of Houtven et al. (2003), using the OAB CHAD
data set, and related to Mansfield et al. (2006).
Mansfield, C., Johnson, F. R., Van Houtven, G. (2006). The missing piece: averting behavior for
children's ozone exposures. Resource Energy Econ. 28:215-228.
Marino, A. J., Fletcher, E. N., Whitaker, R. C., Anderson, S. E. (2012). Amount and
environmental predictors of outdoor playtime at home and school: Across-sectional
analysis of a national sample of preschool-aged children attending Head Start. Health &
Place. 18: 1224-1230.
McCurdy, T. (2000). Conceptual basis for multi-route intake dose modeling using an energy
expenditure approach. J Expo Anal Environ Epidemiol. 10: 1-12.
McCurdy, T., Glen, G., Smith, L., Lakkadi, Y. (2000). The National Exposure Research
Laboratory's Consolidated Human Activity Database. J Expo Anal Environ Epidemiol. 10:
566-578.
McDermott, M., Srivastava, R., Croskell, S. (2006). Awareness of and compliance with air
pollution advisories: A comparison of parents of asthmatics with other parents. J Asthma.
43:235-239.
5G-49
-------
Meng, Q., Williams, R., Pinto, J. P. (2012). Determinants of the associations between ambient
concentrations and personal exposures to PM2.5, NCh, and Os during DEARS. Atmos
Environ. 63:109-116.
Montoye, H. J., Kemper, H. C. G., Saris, W.H.N., Washburn, R.A. (1996). Measuring Physical
Activity and Energy Expenditure. Champaign IL: Human Kinetics.
National Research Council (1991). Human Exposure Assessment for Airborne Pollutants:
Advances and Opportunities. Washington, DC: National Academy of Sciences.
Neidell, M. (2010). Air quality warnings and outdoor activities: evidence from Southern
California using a regression discontinuity approach design. JEpidemiol Community
Health. 64:921-926.
Richmond H., Palma, T, Langstaff, J., McCurdy, T., Glenn, G., Smith, L. (2002). Further
Refinements and Testing of APEX (3.0): EPA's population exposure model for criteria and
air toxic inhalation exposures. Joint meeting of the International Society of Exposure
Analysis and International Society of Environmental Epidemiology, Vancouver, CAN.
August 11-15, 2002.
Robinson, J. P., Wiley, J. A., Piazza, T., Garrett, K., and Cirksena, K. (1989). Activity Patterns
of California Residents and their Implications for Potential Exposure to Pollution.
California Air Resources Board, Sacramento, CA. CARB-A6-177-33.
Sallis, J. F., Bauman, A., Pratt, M. (1998). Environmental and policy interventions to promote
physical activity. American Journal of Preventive Medicine. 15(4):379-397.
Santuz, P., Baraldi, E., Filippone, M., Zacchello, F. (1997). Exercise performance in children
with asthma: Is it different from that of healthy controls? Eur Respir J. 10:1254-1260.
SAS (2012). SAS/STAT 9.2 User's Guide, Second Edition. Available at:
http://support.sas.com/documentation/cdl/en/statug/63033/PDF/default/statug.pdf
Semenza J. C., Wilson, D. J., Parra J., Bontempo, B. D., Hart, M., Sailor, D. J., George, L. A.
(2008). Public perception and behavior change in relationship to hot weather and air
pollution. Environ Res. 107:401-411.
Sexton, A. L. (2011). Responses to Air Quality Alerts: Do Americans Spend Less Time
Outdoors? Available at: http://www.apec.umn.edU/prod/groups/cfans/@pub/@cfans/
@apec/documents/asset/cfans asset 365645.pdf
Shamoo, D. A., Linn, W. S., Peng, R.-C., Solomon, J. C., Webb, T. L., Hackney, J. D., Hong, H.
(1994). Time-activity patterns and diurnal variation of respiratory status in a panel of
asthmatics: implications for short-term air pollution effects. J Expo Anal Environ
Epidemiol. 4(2): 133-148.
Spier, C. E., Little, D. E., Trim, S. C., Johnson, T. R., Linn, W. S., Hackney, J. D. (1992).
Activity patterns in elementary and high school students exposed to oxidant pollution. J
Expo Anal Environ Epidemiol. 2:277-293.
Suh, H. (2008). Personal Communication, August 27, 2008, regarding personal exposure data
for Atlanta.
Tsang, A. M., Klepeis, N. E. (1996). Descriptive Statistics Tables from a Detailed Analysis of
the National Human Activity Pattern Survey (NHAPS) Data. U.S. Environmental
Protection Agency. EPA/600/R-96/148.
5G-50
-------
US BLS. (2012a). American Time Use Survey User's Guide. Understanding ATUS 2003 to
2011. United States Bureau of Labor Statistics (BLS). August 2012. Data and
documentation available at: http://www.bls.gov/tus/
US BLS. (2012b). Employment outlook: 2010-2020. Overview of projections to 2020.
Available at: http://www.bls.gov/opub/mlr/2012/01/artlfull.pdf.
US Census Bureau. (2007). 2000 Census of Population and Housing. Summary File 3 (SF3)
Technical Documentation, available at: http://www.census.gov/prod/cen2000/doc/sf3.pdf
Individual SF3 files '30' (for income/poverty variables pctSO) for each state were
downloaded from: http://www2.census.gov/census_2000/datasets/Summary_File_3/.
USDA. (2000). Continuing Survey of Food Intake by Individuals (CSFII) 1994-96, 1998.
Agricultural Research Service, CD-ROM. Available at:
http://www.ars.usda.gov/Services/docs.htm? docid=14531
US DHHS. (1999). Promoting physical activity: a guide for community action. Champaign, IL:
Human Kinetics. Department of Health and Human Services, Public Health Service,
Centers for Disease Control and Prevention, National Center for Chronic Disease
Prevention and Health Promotion, Division of Nutrition and Physical Activity. See Table
2-1 available at
http://www.cdc.gov/nccdphp/dnpa/physical/pdf/PA Intensity table 21.pdf.
US EPA. (1986). Air Quality Criteria for Ozone and Other Photochemical Oxidants. Office of
Health and Environmental Assessment, Environmental Criteria and Assessment Office,
U.S. Environmental Protection Agency, Research Triangle Park, NC. EPA-600/8-84-
020aF-eF. Available from: NTIS, Springfield, VA., PB87-142949.
US EPA. (1996a). Review of National Ambient Air Quality Standards for Ozone: Assessment
of Scientific and Technical Information - OAQPS Staff Paper. Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park,
NC. EPA/452/R-96-007. Available at:
http ://www. epa. gov/ttn/naaqs/standards/ozone/s_o3_pr_sp. html
US EPA . (1996b). Air Quality Criteria for Ozone and Related Photochemical Oxidants. Office
of Research and Development, National Center for Environmental Assessment, U.S.
Environmental Protection Agency, Research Triangle Park, NC. EPA/600/P-93/004aF-cF.
Available at: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2831
US EPA. (2002). Consolidated Human Activities Database (CHAD) Users Guide. Database and
documentation available at: http://www.epa.gov/chadnet 1 /
US EPA . (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final).
National Center for Environmental Assessment, U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA/600/R-05/004aF-cF. Available at:
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
US EPA. (2007a). Review of National Ambient Air Quality Standards for Ozone: Policy
Assessment of Scientific and Technical Information - OAQPS Staff Paper. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park,NC. EPA-452/R-07-007. Available at:
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2007 07 ozone staff_paper.pdf
US EPA. (2007b). Ozone Population Exposure Analysis for Selected Urban Areas. Office of
Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research
5G-51
-------
Triangle Park, NC. Available at:
http://www.epa.gOv/ttn/naaqs/standards/ozone/s o3crtd.html
US EPA. (2008). Risk and Exposure Assessment to Support the Review of the NCh Primary
National Ambient Air Quality Standard. EPA-452/R-08-008a. November 2008. Available
at: http://www.epa.gov/ttn/naaqs/standards/nox/data/20081121 NO2 REA fmal.pdf.
US EPA . (2009a). Risk and Exposure Assessment to Support the Review of the SCh Primary
National Ambient Air Quality Standard. EPA-452/R-09-007. August 2009. Available at
http://www.epa. gov/ttn/naaq s/standards/so2/data/200908 SO2REAFinalReport.pdf.
US EPA. (2009b). Metabolically derived human ventilation rates: a revised approach based
upon oxygen consumption rates. National Center for Environmental Assessment,
Washington, DC; EPA/600/R-06/129F. Available at:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=202543
US EPA. (2010). Quantitative Risk and Exposure Assessment for Carbon Monoxide -
Amended. EPA Office of Air Quality Planning and Standards. EPA-452/R-10-009. July
2010. Available at: http://www.epa.gov/ttn/naaqs/standards/co/data/CO-REA-Amended-
Julv2010.pdf
US EPA. (2011). Exposure Factors Handbook: 2011 Edition. National Center for Environmental
Assessment, Washington, DC; EPA/600/R-09/052F. Available at:
http://www.epa.gov/ncea/efh/pdfs/efh-complete.pdf
US EPA. (2012a). Total Risk Integrated Methodology (TRIM) - Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4.4) Volume I: User's Guide. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park,NC. EPA-452/B-12-00la. Available at:
http://www.epa.gov/ttn/fera/human apex.html
US EPA. (2012b). Total Risk Integrated Methodology (TRIM) - Air Pollutants Exposure Model
Documentation (TRIM.Expo / APEX, Version 4.4) Volume II: Technical Support
Document. Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-452/B-12-001b. Available at:
http://www.epa.gov/ttn/fera/human apex.html
US EPA. (2013). Integrated Science Assessment of Ozone and Related Photochemical Oxidants.
U.S. Environmental Protection Agency, National Center for Environmental Assessment,
U.S. Environmental Protection Agency, Research Triangle Park, N. EPA 600/R-10/076F.
Available at: http://cfpub.epa.gOv/ncea/isa/recordisplay.cfm?deid=247492#Download
University of Michigan. (2012). The Panel Study of Income Dynamics (PSID). Data and
documentation available at http://psidonline.isr.umich.edu.
van Gent, R., van der Ent, K., van Essen-Zandvliet, L. E. M., Rovers, M. M., Kimpen, J. L. L.,
de Meer, G., Klijn, P. H. C. (2007). No Difference in physical activity in (un)diagnosed
asthma and healthy controls. Pediatric Pulmonology. 42:1018-1023.
van Veldhoven, N. H. M. J., Vermeer, A., Bogaard, J. M., Hessels, M. G. P., Wijnroks, L.,
Colland, V. T., van Essen-Zandvliet, E. E. M. (2001). Children with asthma and physical
exercise: effects of an exercise programme. Clinical Rehabilitation. 15:360-370.
Wen, X.-J., Balluz, L., Mokdad, A. (2009). Association between media alerts of Air Quality
Index and change of outdoor activity among adult asthma in six states, BRFSS, 2005. J
Comm Health. 34:40-46.
5G-52
-------
Wheeler, A., Zanobetti, A., Gold, D. R., Schwartz, 1, Stone, P., Suh, H. H. (2006). The
relationship between ambient air pollution and heart rate variability differs for individuals
with heart and pulmonary disease. Environ Health Perspect. 114(4):560-6.
Whitfield, R., Biller, W., Jusko, M., Keisler, J. (1996). A Probabilistic Assessment of Health
Risks Associated with Short- and Long-Term Exposure to Tropospheric Ozone. Argonne
National Laboratory, Argonne, IL.
Wiley, J. A., Robinson, J. P., Piazza, T., Garrett, K., Cirksena, K., Cheng, Y.-T., Martin, G.
(1991a). Activity Patterns of California Residents: Final Report. California Air Resources
Board, Sacramento, CA. ARB/R93/487. Available from: NTIS, Springfield, VA., PB94-
108719.
Wiley, J. A., Robinson, J. P., Cheng, Y.-T., Piazza, T., Stork, L., Pladsen, K. (1991b). Study of
Children's Activity Patterns: Final Report. California Air Resources Board, Sacramento,
CA. ARB-R-93/489.
Williams, R., Suggs, J., Creason, J., Rodes, C., Lawless, P., Kwok, R., Zweidinger, R., Sheldon,
L. (2000). The 1998 Baltimore particulate matter epidemiology-exposure study: Part 2.
Personal exposure associated with an elderly population. J Expo Anal Environ Epidemiol.
10(6): 533-543.
Williams, R., Suggs J., Rea, A., Leovic, K., Vette, A., Croghan, C., Sheldon, L., Rodes, C.,
Thornburg, J., Ejire, A., Herbst, M., Sanders Jr., W. (2003a). The Research Triangle
particulate panel study: PM mass concentrations relationships. Atmos Environ. 37:5349-
5363.
Williams, R., Suggs, J., Rea, A., Sheldon, L., Rodes, C., Thornburg, J. (2003b). The Research
Triangle particulate panel study: modeling ambient source contributions to personal and
residential PM mass concentrations. Atmos Environ. 37:5365-5378.
Xue, J., McCurdy, T., Spengler, J., Ozkaynak, H. (2004). Understanding variability in time
spent in selected locations for 7-12-year old children. J Expo Anal Environ Epidemiol.
14(3):222-33.
Zivin, J. G., Neidell M. (2009). Days of haze: Environmental information disclosure and
intertemporal avoidance behavior. J Environment Econ Manag. 5 8(2): 119-128.
5G-53
-------
Attachment 1. Occupations estimated to have at least one day per week where work is performed outdoors.
Major BLS
SOC code
11
11
11
11
13
13
13
13
17
17
17
19
19
19
29
33
33
33
33
33
33
33
33
Major BLS SOC name
Management
Occupations
Management
Occupations
Management
Occupations
Management
Occupations
Business and Financial
Operations Occupations
Business and Financial
Operations Occupations
Business and Financial
Operations Occupations
Business and Financial
Operations Occupations
Architecture and
Engineering Occupations
Architecture and
Engineering Occupations
Architecture and
Engineering Occupations
Life, Physical, and Social
Science Occupations
Life, Physical, and Social
Science Occupations
Life, Physical, and Social
Science Occupations
Healthcare Practitioners
and Technical
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
BLS SOC name
Industrial Production Managers
Farmers, Ranchers, and Other
Agricultural Managers
Farmers, Ranchers, and Other
Agricultural Managers
Construction Managers
Insurance Appraisers, Auto
Damage
Compliance Officers
Farm Labor Contractors
Appraisers and Assessors of
Real Estate
Surveyors
Surveyors
Surveying and Mapping
Technicians
Conservation Scientists
Conservation Scientists
Forest and Conservation
Technicians
Emergency Medical
Technicians and Paramedics
First-Line Supervisors of Police
and Detectives
First-Line Supervisors of Fire
Fighting and Prevention
Workers
First-Line Supervisors of Fire
Fighting and Prevention
Workers
Firefighters
Firefighters
Fire Inspectors and
Investigators
Fire Inspectors and
Investigators
Forest Fire Inspectors and
Prevention Specialists
BLS SOC
code
11-0000
11-9013
11-9013
11-9021
13-0000
13-1041
13-1074
13-2021
17-0000
17-1022
17-3031
19-0000
19-1031
19-4093
29-0000
33-0000
33-1021
33-1021
33-2011
33-2011
33-2021
33-2021
33-2022
Employed
(1000's)
150.3
400.8
400.8
523.1
10.6
216.6
0.3
77.8
25.6
25.6
56.9
7.8
7.8
36.5
226.5
106.1
30
30
155.2
155.2
6.8
6.8
1.6
O*NET Code
11-3051.02
11-9013.01
11-9013.03
11-9021.00
13-1032.00
13-1041.04
13-1074.00
13-2021.02
17-1022.00
17-1022.01
17-3031.01
19-1031.01
19-1031.03
19-4093.00
29-2041 .00
33-1012.00
33-1021.01
33-1021.02
33-2011.01
33-2011.02
33-2021.01
33-2021.02
33-2022.00
O*NET Name
Geothermal Production
Managers
Nursery and Greenhouse
Managers
Aquacultural Managers
Construction Managers
Insurance Appraisers, Auto
Damage
Government Property
Inspectors and
Investigators
Farm Labor Contractors
Appraisers, Real Estate
Surveyors
Geodetic Surveyors
Surveying Technicians
Soil and Water
Conservationists
Park Naturalists
Forest and Conservation
Technicians
Emergency Medical
Technicians and
Paramedics
First-Line Supervisors of
Police and Detectives
Municipal Fire Fighting and
Prevention Supervisors
Forest Fire Fighting and
Prevention Supervisors
Municipal Firefighters
Forest Firefighters
Fire Inspectors
Fire Investigators
Forest Fire Inspectors and
Prevention Specialists
Context
77
96
82
78
89
82
79
77
90
78
92
79
78
90
87
82
88
86
80
87
86
84
77
Out
days/wk
1
5
2
1
3
2
1
1
3
1
4
1
1
3
3
2
3
3
1
3
3
2
1
Comment
divided bis code 11-9013 (Farm/Ranch/Other Ag
Managers- 1202.5) by 3, possibly an
overestimate
divided bis code 11-9013 (Farm/Ranch/Other Ag
Managers- 1202.5) by 3, possibly an
overestimate
divided bis code 17-1022 (surveyors- 51.2) by 2,
estimate should be ok
divided bis code 17-1022 (surveyors- 51.2) by 2,
estimate should be ok
divided bis code 19-1031 (Conservation
Scientists- 23.4) by 3, possibly an overestimate
divided bis code 19-1031 (Conservation
Scientists- 23.4) by 3, possibly an overestimate
divided bis code 33-1021 (First-Line Supervisors
of Fire Fighting and Prevention Workers- 60.1) by
2, estimate should be ok
divided bis code 33-1021 (First-Line Supervisors
of Fire Fighting and Prevention Workers- 60.1) by
2, estimate should be ok
divided bis code 33-2011 (Firefighters- 310.4) by
2, estimate should be ok
divided bis code 33-2011 (Firefighters- 310.4) by
2, estimate should be ok
divided bis code 33-2021 (Fire Inspectors and
Investigators- 13.6) by 2, estimate should be ok
divided bis code 33-2021 (Fire Inspectors and
Investigators- 13.6) by 2, estimate should be ok
5G-54
-------
Major BLS
SOC code
33
33
33
33
33
33
33
33
33
33
37
37
37
37
37
39
39
43
43
43
45
Major BLS SOC name
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Protective Service
Occupations
Building and Grounds
Cleaning and
Maintenance Occupations
Building and Grounds
Cleaning and
Maintenance Occupations
Building and Grounds
Cleaning and
Maintenance Occupations
Building and Grounds
Cleaning and
Maintenance Occupations
Building and Grounds
Cleaning and
Maintenance Occupations
Personal Care and
Service Occupations
Personal Care and
Service Occupations
Office and Administrative
Support Occupations
Office and Administrative
Support Occupations
Office and Administrative
Support Occupations
Farming, Fishing, and
Forestry Occupations
BLS SOC name
Detectives and Criminal
Investigators
Detectives and Criminal
Investigators
Detectives and Criminal
Investigators
Fish and Game Wardens
Parking Enforcement Workers
Police and Sheriffs Patrol
Officers
Police and Sheriffs Patrol
Officers
Animal Control Workers
Private Detectives and
Investigators
Crossing Guards
First-Line Supervisors of
Landscaping, Lawn Service,
and Groundskeeping Workers
Pest Control Workers
Landscaping and
Groundskeeping Workers
Pesticide Handlers, Sprayers,
and Applicators, Vegetation
Tree Trimmers and Pruners
Baggage Porters and Bellhops
Couriers and Messengers
Meter Readers, Utilities
Postal Service Mail Carriers
First-Line Supervisors of
Farming, Fishing, and Forestry
Workers
BLS SOC
code
33-3021
33-3021
33-3021
33-3031
33-3041
33-3051
33-3051
33-901 1
33-9021
33-9091
37-0000
37-2021
37-3011
37-3012
37-3013
39-0000
39-6011
43-0000
43-5041
43-5052
45-0000
Employed
(1000's)
23.9
23.9
23.9
7.6
9.8
332
332
15.5
34.7
69.3
202.9
68.4
1151.5
29.5
50.6
29.3
46
116.2
40.5
316.7
11.8
O*NET Code
33-3021.01
33-3021.03
33-3021.05
33-3031 .00
33-3041 .00
33-3051.01
33-3051 .03
33-9011.00
33-9021.00
33-9091 .00
37-1012.00
37-2021.00
37-3011.00
37-3012.00
37-3013.00
39-4031 .00
39-6011.00
43-5021.00
43-5041 .00
43-5052.00
45-1011.05
O*NET Name
Police Detectives
Criminal Investigators and
Special Agents
Immigration and Customs
Inspectors
Fish and Game Wardens
Parking Enforcement
Workers
Police Patrol Officers
Sheriffs and Deputy
Sheriffs
Animal Control Workers
Private Detectives and
Investigators
Crossing Guards
First-Line Supervisors of
Landscaping, Lawn
Service, and
Groundskeeping Workers
Pest Control Workers
Landscaping and
Groundskeeping Workers
Pesticide Handlers,
Sprayers, and Applicators,
Vegetation
Tree Trimmers and
Pruners
Morticians, Undertakers,
and Funeral Directors
Baggage Porters and
Bellhops
Couriers and Messengers
Meter Readers, Utilities
Postal Service Mail
Carriers
First-Line Supervisors of
Logging Workers
Context
87
76
85
96
95
91
88
87
76
100
94
98
99
89
99
77
82
89
99
97
95
Out
days/wk
3
1
2
5
4
4
3
3
1
5
4
5
5
3
5
1
2
3
5
5
4
Comment
divided bis code 33-3021 (Detectives and
Criminal Investigators- 119.4) by 5, possibly an
overestimate
divided bis code 33-3021 (Detectives and
Criminal Investigators- 119.4) by 5, possibly an
overestimate
divided bis code 33-3021 (Detectives and
Criminal Investigators- 119.4) by 5, possibly an
overestimate
divided bis code 33-3051 (Police and Sheriff
Patrol Officers- 663.9) by 2, estimate should be
ok
divided bis code 33-3051 (Police and Sheriff
Patrol Officers- 663.9) by 2, estimate should be
ok
used employment data from bis code 39-4831
(funeral dir., etc), estimate should be ok
divided bis code 45-1011 (First-Line Supervisors
of Farming, Fishing, and Forestry Workers- 47)
by 4, estimate should be ok
5G-55
-------
Major BLS
SOC code
45
45
45
45
45
45
45
45
45
45
45
45
45
45
47
47
47
47
47
47
47
47
47
Major BLS SOC name
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Farming, Fishing, and
Forestry Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
BLS SOC name
First-Line Supervisors of
Farming, Fishing, and Forestry
Workers
First-Line Supervisors of
Farming, Fishing, and Forestry
Workers
First-Line Supervisors of
Farming, Fishing, and Forestry
Workers
Animal Breeders
Fishers and Related Fishing
Workers
Forest and Conservation
Workers
Fallers
Logging Equipment Operators
Log Graders and Sealers
First-Line Supervisors of
Construction Trades and
Extraction Workers
Brickmasons and Blockmasons
Stonemasons
Carpenters
Carpenters
Cement Masons and Concrete
Finishers
Construction Laborers
Paving, Surfacing, and
Tamping Equipment Operators
Pile-Driver Operators
BLS SOC
code
45-1011
45-1011
45-1011
45-2021
45-2091
45-2092
45-2092
45-2093
45-301 1
45-3021
45-401 1
45-4021
45-4022
45-4023
47-0000
47-2021
47-2022
47-2031
47-2031
47-2051
47-2061
47-2071
47-2072
Employed
(1000's)
11.8
11.8
11.8
11.5
186.6
186.6
186.6
186.6
32
0.6
13.7
9.6
35.1
3.8
558.5
89.2
15.6
500.8
500.8
144.7
998.8
51.6
4.1
O*NET Code
45-1011.06
45-1011.07
45-1011.08
45-2021.00
45-2091 .00
45-2092.01
45-2092.02
45-2093.00
45-301 1 .00
45-3021.00
45-401 1 .00
45-4021.00
45-4022.00
45-4023.00
47-1011.00
47-2021.00
47-2022.00
47-2031.01
47-2031 .02
47-2051.00
47-2061 .00
47-2071 .00
47-2072.00
O*NET Name
First-Line Supervisors of
Aquacultural Workers
First-Line Supervisors of
Agricultural Crop and
Horticultural Workers
First-Line Supervisors of
Animal Husbandry and
Animal Care Workers
Animal Breeders
Agricultural Equipment
Operators
Nursery Workers
Farmworkers and
Laborers, Crop
Farmworkers, Farm,
Ranch, and Aquacultural
Animals
Fishers and Related
Fishing Workers
Hunters and Trappers
Forest and Conservation
Workers
Fallers
Logging Equipment
Operators
Log Graders and Sealers
First-Line Supervisors of
Construction Trades and
Extraction Workers
Brickmasons and
Blockmasons
Stonemasons
Construction Carpenters
Rough Carpenters
Cement Masons and
Concrete Finishers
Construction Laborers
Paving, Surfacing, and
Tamping Equipment
Operators
Pile-Driver Operators
Context
87
87
84
91
95
76
98
83
93
99
83
99
95
81
83
99
92
78
83
100
97
94
100
Out
days/wk
3
3
2
4
4
1
5
2
4
5
2
5
4
2
2
5
4
1
2
5
5
4
5
Comment
divided bis code 45-1011 (First-Line Supervisors
of Farming, Fishing, and Forestry Workers- 47)
by 4, estimate should be ok
divided bis code 45-1011 (First-Line Supervisors
of Farming, Fishing, and Forestry Workers- 47)
by 4, estimate should be ok
divided bis code 45-1011 (First-Line Supervisors
of Farming, Fishing, and Forestry Workers- 47)
by 4, estimate should be ok
used employment data from bis code 45-2090
(misc ag- 746.4) divided by 4, possibly an
overestimate
used employment data from bis code 45-2090
(misc ag- 746.4) divided by 4, possibly an
overestimate
used employment data from bis code 45-2090
(misc ag- 746.4) divided by 4, possibly an
overestimate
used employment data from bis code 45-2090
(misc ag- 746.4) divided by 4, possibly an
overestimate
diff of parent bis 45-3000 (fish hunt- 32.6) and
sub 45-3011 (fishing- 32), estimate should be ok
divided bis code 47-2031 (Carpenters-1001 .7) by
2, estimate should be ok
divided bis code 47-2031 (Carpenters- 1001.7) by
2, estimate should be ok
5G-56
-------
Major BLS
SOC code
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
47
Major BLS SOC name
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
Extraction Occupations
Construction and
BLS SOC name
Operating Engineers and Other
Construction Equipment
Operators
Glaziers
Insulation Workers, Floor,
Ceiling, and Wall
Pipelayers
Plumbers, Pipefitters, and
Steamf liters
Reinforcing Iron and Rebar
Workers
Roofers
Structural Iron and Steel
Workers
Helpers-Brickmasons,
Blockmasons, Stonemasons,
and Tile and Marble Setters
Helpers-Pipelayers, Plumbers,
Pipefitters, and Steamfitters
Helpers— Roofers
Construction and Building
Inspectors
Fence Erectors
Highway Maintenance Workers
Rail-Track Laying and
Maintenance Equipment
Operators
Septic Tank Servicers and
Sewer Pipe Cleaners
Segmental Pavers
Derrick Operators, Oil and Gas
Rotary Drill Operators, Oil and
Gas
Service Unit Operators, Oil,
Gas, and Mining
Earth Drillers, Except Oil and
Gas
Explosives Workers, Ordnance
Handling Experts, and Blasters
Rock Splitters, Quarry
Roustabouts, Oil and Gas
BLS SOC
code
47-2073
47-2121
47-2131
47-2151
47-2152
47-2171
47-2181
47-2221
47-3011
47-3015
47-3016
47-4011
47-4031
47-4051
47-4061
47-4071
47-4091
47-5011
47-5012
47-5013
47-5021
47-5031
47-5051
47-5071
Employed
(1000's)
349.1
41.9
23.2
53.1
419.9
19.1
136.7
59.8
29.4
57.9
12.7
102.4
32.1
148.5
15
25.3
1.3
18.9
22.5
40.7
17.8
6.8
3.5
52.7
O*NET Code
47-2073.00
47-2121.00
47-2131.00
47-2151.00
47-2152.02
47-2171.00
47-2181.00
47-2221.00
47-3011.00
47-3015.00
47-3016.00
47-4011.00
47-4031 .00
47-4051.00
47-4061.00
47-4071 .00
47-4091 .00
47-5011.00
47-5012.00
47-5013.00
47-5021.00
47-5031 .00
47-5051 .00
47-5071 .00
O*NET Name
Operating Engineers and
Other Construction
Equipment Operators
Glaziers
Insulation Workers, Floor,
Ceiling, and Wall
Pipelayers
Plumbers
Reinforcing Iron and Rebar
Workers
Roofers
Structural Iron and Steel
Workers
Helpers— Brickmasons,
Blockmasons,
Stonemasons, and Tile and
Marble Setters
Helpers— Pipelayers,
Plumbers, Pipefitters, and
Steamfitters
Helpers— Roofers
Construction and Building
Inspectors
Fence Erectors
Highway Maintenance
Workers
Rail-Track Laying and
Maintenance Equipment
Operators
Septic Tank Servicers and
Sewer Pipe Cleaners
Segmental Pavers
Derrick Operators, Oil and
Gas
Rotary Drill Operators, Oil
and Gas
Service Unit Operators, Oil,
Gas, and Mining
Earth Drillers, Except Oil
and Gas
Explosives Workers,
Ordnance Handling
Experts, and Blasters
Rock Splitters, Quarry
Roustabouts, Oil and Gas
Context
99
91
87
97
79
94
100
97
86
87
88
88
99
84
93
92
84
100
99
99
95
98
94
100
Out
days/wk
5
4
3
5
1
4
5
5
3
3
3
3
5
2
4
4
2
5
5
5
4
5
4
5
Comment
5G-57
-------
Major BLS
SOC code
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
49
Major BLS SOC name
Extraction Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
and Repair Occupations
Installation, Maintenance,
BLS SOC name
Telecommunications
Equipment Installers and
Repairers, Except Line
Installers
Electrical and Electronics
Repairers, Powerhouse,
Substation, and Relay
Automotive Glass Installers and
Repairers
Bus and Truck Mechanics and
Diesel Engine Specialists
Farm Equipment Mechanics
and Service Technicians
Mobile Heavy Equipment
Mechanics, Except Engines
Rail Car Repairers
Motorboat Mechanics and
Service Technicians
Mechanical Door Repairers
Control and Valve Installers
and Repairers, Except
Mechanical Door
Heating, Air Conditioning, and
Refrigeration Mechanics and
Installers
Heating, Air Conditioning, and
Refrigeration Mechanics and
Installers
Millwrights
Electrical Power-Line Installers
and Repairers
Telecommunications Line
Installers and Repairers
Coin, Vending, and Amusement
Machine Servicers and
Repairers
Commercial Divers
Locksmiths and Safe Repairers
Manufactured Building and
Mobile Home Installers
Riggers
Signal and Track Switch
Repairers
BLS SOC
code
49-0000
49-2095
49-3022
49-3031
49-3041
49-3042
49-3043
49-3051
49-901 1
49-9012
49-9021
49-9021
49-9044
49-9051
49-9052
49-9091
49-9092
49-9094
49-9095
49-9096
49-9097
49-9099
Employed
(1000's)
194.9
23.4
18.1
242.2
32.9
124.6
21.7
20.8
12.8
43.8
133.9
133.9
36.5
108.4
160.6
39.1
3.8
25.7
7.8
15.2
7.1
143.6
O*NET Code
49-2022.00
49-2095.00
49-3022.00
49-3031 .00
49-3041 .00
49-3042.00
49-3043.00
49-3051 .00
49-901 1 .00
49-9012.00
49-9021.01
49-9021.02
49-9044.00
49-9051 .00
49-9052.00
49-9091 .00
49-9092.00
49-9094.00
49-9095.00
49-9096.00
49-9097.00
49-9099.01
O*NET Name
Telecommunications
Equipment Installers and
Repairers, Except Line
Installers
Electrical and Electronics
Repairers, Powerhouse,
Substation, and Relay
Automotive Glass Installers
and Repairers
Bus and Truck Mechanics
and Diesel Engine
Specialists
Farm Equipment
Mechanics and Service
Technicians
Mobile Heavy Equipment
Mechanics, Except
Engines
Rail Car Repairers
Motorboat Mechanics and
Service Technicians
Mechanical Door Repairers
Control and Valve Installers
and Repairers, Except
Mechanical Door
Heating and Air
Conditioning Mechanics
and Installers
Refrigeration Mechanics
and Installers
Millwrights
Electrical Power-Line
Installers and Repairers
Telecommunications Line
Installers and Repairers
Coin, Vending, and
Amusement Machine
Servicers and Repairers
Commercial Divers
Locksmiths and Safe
Repairers
Manufactured Building and
Mobile Home Installers
Riggers
Signal and Track Switch
Repairers
Geothermal Technicians
Context
79
92
88
83
83
78
81
86
93
88
84
94
77
95
94
85
92
87
86
86
96
94
Out
days/wk
1
4
3
2
2
1
2
3
4
3
2
4
1
4
4
2
4
3
3
3
5
4
Comment
divided bis code 49-9021 (Heating AC refrig
mech- 267.8) by 2, estimate should be ok
divided bis code 49-9021 (Heating AC refrig
mech- 267.8) by 2, estimate should be ok
used employment data from bis code 49-9799
5G-58
-------
Major BLS
SOC code
51
51
51
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
Major BLS SOC name
and Repair Occupations
Production Occupations
Production Occupations
Production Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
BLS SOC name
Water and Wastewater
Treatment Plant and System
Operators
Gas Plant Operators
Petroleum Pump System
Operators, Refinery Operators,
and Gaugers
Aircraft Cargo Handling
Supervisors
Commercial Pilots
Bus Drivers, Transit and
Intercity
Driver/Sales Workers
Heavy and Tractor-Trailer Truck
Drivers
Light Truck or Delivery Services
Drivers
Taxi Drivers and Chauffeurs
Locomotive Engineers
Locomotive Firers
Rail Yard Engineers, Dinkey
Operators, and Hostlers
Railroad Brake, Signal, and
Switch Operators
Railroad Conductors and
Yardmasters
Subway and Streetcar
Operators
Sailors and Marine Oilers
Captains, Mates, and Pilots of
Water Vessels
BLS SOC
code
51-0000
51-8092
51-8093
53-0000
53-2012
53-3021
53-3031
53-3032
53-3033
53-3041
53-401 1
53-4012
53-4013
53-4021
53-4031
53-4041
53-501 1
53-5021
Employed
(1000's)
110.7
13.7
44.2
6.3
32.7
186.3
406.6
1604.8
856
239.9
38.7
1.1
5.6
21.7
40.8
6.5
33.4
36.1
O*NET Code
51-8031.00
51-8092.00
51-8093.00
53-1011.00
53-2012.00
53-3021.00
53-3031 .00
53-3032.00
53-3033.00
53-3041 .00
53-401 1 .00
53-4012.00
53-4013.00
53-4021.00
53-4031 .00
53-4041 .00
53-501 1 .00
53-5021.02
O*NET Name
Water and Wastewater
Treatment Plant and
System Operators
Gas Plant Operators
Petroleum Pump System
Operators, Refinery
Operators, and Gaugers
Aircraft Cargo Handling
Supervisors
Commercial Pilots
Bus Drivers, Transit and
Intercity
Driver/Sales Workers
Heavy and Tractor-Trailer
Truck Drivers
Light Truck or Delivery
Services Drivers
Taxi Drivers and
Chauffeurs
Locomotive Engineers
Locomotive Firers
Rail Yard Engineers,
Dinkey Operators, and
Hostlers
Railroad Brake, Signal, and
Switch Operators
Railroad Conductors and
Yardmasters
Subway and Streetcar
Operators
Sailors and Marine Oilers
Mates- Ship, Boat, and
Barge
Context
93
92
93
77
92
83
76
100
100
78
86
86
84
100
87
77
99
95
Out
days/wk
4
4
4
1
4
2
1
5
5
1
3
3
2
5
3
1
5
4
Comment
(install, repair, other, etc.), possibly an
overestimate
5G-59
-------
Major BLS
SOC code
53
53
53
53
53
53
53
53
53
53
53
53
53
Major BLS SOC name
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
Transportation and
Material Moving
Occupations
BLS SOC name
Motorboat Operators
Ship Engineers
Parking Lot Attendants
Automotive and Watercraft
Service Attendants
Transportation Inspectors
Transportation Attendants,
Except Flight Attendants
Dredge Operators
Excavating and Loading
Machine and Dragline
Operators
Gas Compressor and Gas
Pumping Station Operators
Pump Operators, Except
Wellhead Pumpers
Wellhead Pumpers
Refuse and Recyclable Material
Collectors
Tank Car, Truck, and Ship
Loaders
BLS SOC
code
53-5022
53-5031
53-6021
53-6031
53-6051
53-6061
53-7031
53-7032
53-7071
53-7072
53-7073
53-7081
53-7121
Employed
(1000's)
3.1
10.1
125.1
86.3
27.4
24.8
2.1
61.5
4.5
10.8
15.1
139.9
10.4
O*NET Code
53-5022.00
53-5031 .00
53-6021.00
53-6031 .00
53-6051 .08
53-6061 .00
53-7031 .00
53-7032.00
53-7071.00
53-7072.00
53-7073.00
53-7081 .00
53-7121.00
O*NET Name
Motorboat Operators
Ship Engineers
Parking Lot Attendants
Automotive and Watercraft
Service Attendants
Freight and Cargo
Inspectors
Transportation Attendants,
Except Flight Attendants
Dredge Operators
Excavating and Loading
Machine and Dragline
Operators
Gas Compressor and Gas
Pumping Station Operators
Pump Operators, Except
Wellhead Pumpers
Wellhead Pumpers
Refuse and Recyclable
Material Collectors
Tank Car, Truck, and Ship
Loaders
Context
85
77
76
94
80
77
86
88
91
99
93
100
91
Out
days/wk
2
1
1
4
1
1
3
3
4
5
4
5
4
Comment
5G-60
-------
Attachment 2. Additional mapping of O*NET occupation codes to CHAD/APEX METs occupation activity codes.
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
if
oc
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
occ
z cod
code
~code
code
~code
code
code
code
code
~code
code
~code
code
code
code
code
code
code
~code
code
~code
code
code
code
code
code
code
~code
code
~code
code
code
code
code
5 onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
onet=
'13-1032
13-1041
13-1074
13-2021
17-3031
19-4093
33-9011
33-9021
33-9091
37-3013
43-5021
45-2021
45-2093
45-3011
45-4011
45-4023
47-2061
47-2071
47-2073
47-2121
47-4011
47-4051
47-4061
47-5071
49-3022
49-9091
49-9094
49-9096
49-9099
53-1011
53-5011
53-5031
53-6051
53-3032
. 00
04 '
00'
02 '
01'
00 '
00 '
00 '
00 '
00'
00'
00'
00'
00 '
00 '
00 '
00 '
00'
00'
00'
00'
00'
00 '
00'
00 '
00 '
00'
00'
01'
00'
00 '
00 '
08 '
00'
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
then
Oc
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
Occ
z nam
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
name
3 CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
CHAD=
'ADMSUP'
ADMSUP'
ADMSUP'
ADMSUP'
TECH' ;
TECH' ;
PROTECT
PROTECT
PROTECT
FARM ' ;
ADMSUP'
FARM ' ;
FARM ' ;
FARM ' ;
FARM ' ;
FARM ' ;
LABOR '
TRANS '
TRANS '
PREC' ;
ADMIN '
LABOR '
LABOR '
LABOR '
PREC' ;
PREC' ;
PREC' ;
TRANS '
TECH' ;
TRANS '
TRANS '
TRANS '
TRANS '
TRANS '
5G-61
-------
This page left intentionally blank
5G-62
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-452/R-14-004c
Environmental Protection Health and Environmental Impacts Division August 2014
Agency Research Triangle Park, NC
------- |