APPENDIX
C
C.
1
PROBABILISTIC
EXPOSURE/
RESPONSE
RELATIONSHIPS
Figures
C.
1
and
C.
2
(
which
are
Figures
B.
28
and
B.
29,
respectively,
from
Whitfield
et
al.,
1996)
show
the
probabilistic
exposure­
response
relationships
for
FEV
1
decrement

15%
and

20%,
derived
from
Folinsbee
et
al.
(
1988),
Horstman
et
al.
(
1990),
and
McDonnell
et
al.
(
1991)
for
8
hr
exposures
and
individuals
engaged
in
moderate
exertion.
The
small
squares
indicate
the
response
rates
at
the
ozone
concentrations
at
which
subjects
were
exposed
in
these
studies
(
0.08,
0.10,
and
0.12
ppm),
and
the
short
horizontal
line
segments
above
and
below
the
data
point,
which
are
connected
by
a
line
segment,
indicate
the
90%
Credible
Interval
(
CI).
For
example,
for
FEV
1
decrement

15%,
the
response
rate
at
0.08
ppm
was
about
18%,
and
the
90%
CI
about
this
value
is
about
10­
28%;
the
derived
90%
CI
compares
favorably
to
this
range.
The
"
derived"
90%
CI
and
the
"
experimental"
CIs
compare
less
favorably
at
0.12
ppm
because
the
linear
regression
used
to
fit
the
data
does
not
capture
the
nonlinear
characteristics
of
the
data.
The
derived
and
experimental
CIs
for
other
endpoints
compare
more
favorably
at
all
ozone
concentrations
for
the
FEV
1

20%
endpoint.
Figures
for
31
other
endpoints
are
given
in
Appendix
B
of
Whitfield
et
al.
(
1996).
Characteristics
(
i.
e.,
parameters
of
functions
fit
to
the
data,
and
regression
r2
values)
of
functions
representing
exposure­
response
relationships
for
33
endpoints
are
listed
in
Table
C.
1
and
C.
2
(
Tables
3.7
and
3.8,
respectively,
from
Whitfield
et
al.,
1996).
C­
2
0.04
0.08
0.12
0.16
0.00
0.20
20
40
60
80
0
100
O3
Concentration
(
ppm)
.05
frac
.25
frac
.50
frac
.75
frac
.95
frac
data
90%
CI
Legend:

0.04
0.08
0.12
0.16
0.00
0.20
20
40
60
80
0
100
O3
Concentration
(
ppm)
.05
frac
.25
frac
.50
frac
.75
frac
.95
frac
data
90%
CI
Legend:
FIGURE
C.
1
Probabilistic
Exposure­
Response
Relationship
for
FEV1
Decrement

15%
Derived
from
Folinsbee
et
al.
(
1988),
Horstman
et
al.
(
1990),
and
McDonnell
et
al.
(
1991)
[
for
8
hr
exposures,
heavy
exertion;
includes
data,
medians,
and
90%
CIs
for
data
and
relationship]

FIGURE
C.
2
Probabilistic
Exposure­
Response
Relationship
for
FEV1
Decrement

20%
Derived
from
Folinsbee
et
al.
(
1988),
Horstman
et
al.
(
1990),
and
McDonnell
et
al.
(
1991)
[
for
8
hr
exposures,
heavy
exertion;
includes
data,
medians,
and
90%
CIs
for
data
and
relationship]
C­
3
TABLE
C.
1
Summary
of
Functions
Fit
to
Experimental
Data
C
FEV1
Decrement
C
Corrected
for
Exercise
in
Clean
Air
Study
Endpointa
Function
a
b
d
e
r2
Avol
et
al.,
1984
(
1.33
hrs)
DFEV1

10%

DFEV1

15%

DFEV1

20%
Linear
Linear
Linear
­
0.2395
­
0.2400
­
0.2395
3.4388
2.9713
2.6825
0.98
0.99
0.99
Kulle
et
al.,
1985
(
2
hrs)
DFEV1

10%

DFEV1

15%

DFEV1

20%
Linear
Linear
Linear
­
0.3225
­
0.2600
­
0.2375
2.3500
1.6000
1.2500
0.95
0.93
0.89
McDonnell
et
al.,

1983
(
2.5
hrs)
DFEV1

10%

DFEV1

15%

DFEV1

20%
Logistic
Logistic
Logistic
0.6420
0.4968
0.3347
5.5996
9.4948
12.0073
­
27.2927
­
45.3838
­
60.4547
0.99
1.00
1.00
Seal
et
al.,
1993
(
2.33
hrs)
DFEV1

10%

DFEV1

15%

DFEV1

20%
Probit
Probit
Probit
­
1.0276
­
0.6639
­
0.3259
0.7917
0.8401
0.9192
0.99
0.99
0.97
Folinsbee
et
al.,

1988;
Horstman
et
al.,
1990;
McDonnell
et
al.,

1991
(
8
hrs)
DFEV1

10%

DFEV1

15%

DFEV1

20%
Linear
Linear
Linear
­
0.0980
­
0.2087
­
0.1462
5.0000
4.9000
2.9250
1.00
1.00
0.98
aDFEV1
means
forced
expiratory
volume
(
in
1
sec.)
decrement.
C­
4
TABLE
C.
2
Summary
of
Functions
Fit
to
Experimental
Data
C
Symptoms
C
Corrected
for
Exercise
in
Clean
Air
Study
Endpointa
Function
a
b
d
e
r2
Avol
et
al.,
1984
(
1.33
hrs)
Any
lower
respiratory
M/
S
lower
resp
Linear
Linear
­
0.2084
­
0.0902
2.6824
0.5206
0.99
0.94
Kulle
et
al.,
1985
(
2
hrs)
Any
cough
Any
PDI
M/
S
cough
M/
S
PDI
Linear
Linear
Linear
Linear
­
0.2650
­
0.4550
­
0.1626
­
0.5250
3.0000
3.8000
0.8675
3.0000
0.97
0.79
­
0.33b
0.72
McDonnell
et
al.,

1983
(
2.5
hrs)
Any
cough
Any
PDI
M/
S
cough
M/
S
PDI
Probit
Probit
Linear
Linear
­
2.0954
­
1.6071
0.0062
­
0.0427
1.2098
1.5124
1.2604
1.1512
0.99
0.96
0.70
0.96
Seal
et
al.,
1993
(
2.33
hrs)
Any
cough
Any
PDI
M/
S
cough
M/
S
PDI
Lognormal
Lognormal
Linear
Probit
0.2469
0.2464
­
0.1445
­
0.3209
1.9248
2.3641
1.3704
0.9317
0.97
0.99
0.97
0.96
Folinsbee
et
al.,

1988;
Horstman
et
al.,
1990;
McDonnell
et
al.,

1991
(
8
hrs)
Any
cough
Any
PDI
M/
S
cough
M/
S
PDI
Linear
Linear
Linear
Linear
­
0.2928
0.7372
­
0.1747
­
0.3087
5.0750
10.1750
2.3000
3.7000
0.54
1.00
0.88
0.93
aInitializations:
M/
S
means
moderate
or
severe,
PDI
means
pain
on
deep
inspiration.

bThe
data
do
not
support
a
meaningful
exposure­
response
relationship
for
this
health
endpoint.
The
negative
r2
value
flags
this
situation.
C­
5
1The
representative
distribution
is
obtained
by
computing
the
average
cumulative
probability
at
selected
points
along
the
X­
axis.
This
calculation,
like
the
risk
calculations
described
earlier,
implicitly
assumes
that
the
distributions
are
perfectly
correlated.
It
may
be
argued
that
perfect
correlation,
while
not
correct,
is
more
reasonable
that
perfect
independence,
and
there
is
no
basis
for
choosing
any
other
degree
of
correlation
between
these
two
extremes.
C.
2
HEADCOUNT
RISK
RESULTS
Risk
results
for
each
endpoint
are
available
in
the
form
of
10
probability
distributions
for
each
air
quality
scenario.
Since
there
are
nine
scenarios,
it
is
not
practical
to
plot
all
of
the
distributions
on
one
figure
because
the
figure
would
be
quite
messy.
The
nine
air
quality
scenarios
include:
one
for
baseline
("
As
Is")
conditions;
two
for
1­
hour,
1­
exceedance
standards
(
1112
and
1110);
four
for
8­
hour,
1
exceedance
standards
(
8110,
8109,
8108,
8107);
and
two
for
8­
hour,
5­
exceedance
standards
(
8508,
8509).
The
first
digit
in
the
code
for
each
standard
indicates
the
averaging
time,
the
second
digit
specifies
the
number
of
exceedances,
and
the
last
two
digits
specify
the
standard
expressed
in
pphm.
For
example,
8508
indicates
an
8­
hour,
5­
exceedances
standard
with
O
3
concentration
set
at
8
pphm
or
0.08
ppm.

To
gain
insight
about
the
risk
implications
of
the
air
quality
scenarios,
we
developed
"
representative
distributions"
and
"
Box
plots."
These
are
shown
in
Figures
C.
3­
20
and
C.
21­
23,
respectively.
Figures
C.
24­
26
estimate
the
number
of
times
that
responders
(
i.
e.,
outdoor
children
who
experience
a
specific
condition,
such
as
having
FEV
1
decrements

15%)
respond.

C.
2.1
Representative
Risk
Distributions
Figure
C.
3,
which
contains
representative
distributions,
shows
two
sets
of
nine
plots
for
FEV
1
decrement

15%,
8
hr
exposures,
moderate
exertion,
Chicago,
and
children.
The
top
half
of
the
figure
shows
representative
distributions
over
the
number
of
children
experiencing
the
effect
one
or
more
times
(
i.
e.,
persons
basis),
and
the
bottom
half
shows
representative
distributions
over
the
number
of
times
any
child
experiences
the
effect
(
i.
e.,
person­
occurrences
basis).
Each
representative
distribution
gives
some
idea
of
the
range
of
results
among
each
set
of
10
distributions.
There
is
one
representative
distribution
for
each
of
nine
air
quality
scenarios.
Each
plot
is
"
representative"
of
the
10
distributions
for
a
particular
scenario.
Since
there
are
only
9
plots
instead
of
90,
it
is
easier
to
see
patterns.
Each
of
these
plots
is
a
valid
cumulative
probability
distribution.
1
The
plot
indicates,
for
example,
that
the
median
number
of
children
in
Chicago
who
may
experience
FEV
1
decrements

15%
under
as­
is
air
quality
is
around
65
thousand.
When
the
most
stringent
standard
(
8107)
is
just
attained,
the
median
estimate
is
about
15
thousand
children.

C.
2.2
Box
Plots
for
Risk
Results
Box
plots
provide
another
perspective
about
risk
results.
Each
Box
plot
displays
the
ranges
of
the
medians
(
or
0.5
fractile),
0.05
fractiles,
and
0.95
fractiles
of
10
risk
distributions
that
result
from
the
10
pNEM
exposure
distributions
that
are
available.
These
ranges
are
represented
by
rectangles
in
the
figures
(
unless
there
is
no
range,
in
which
case
the
rectangle
"
collapses"
into
a
horizontal
line).
There
are
3
rectangles
above
the
code
letter
for
each
standard.
The
top
rectangle
represents
the
range
of
the
0.95
fractiles,
the
middle
rectangle
represents
the
range
of
the
medians,
and
the
bottom
rectangle
represents
the
range
of
the
0.05
fractiles.
A
line
connects
the
bottom
of
the
0.95­
fractile
rectangle
and
the
top
of
the
0.05­
fractile
rectangle
and
passes
through
the
0.5­
fractile
rectangle.
With
this
format,
results
for
81
scenarios
(
9
scenarios
for
each
of
9
urban
areas)
can
be
displayed
in
one
figure.
For
these
plots,
however,
we
switched
from
numbers
of
persons
or
person­
occurrences
to
percentage
of
persons
responding.
As
shown
in
C­
6
Figure
C.
3,
under
as­
is
conditions
in
Chicago,
the
median
risk
estimates
for
the
percentage
of
children
having
FEV
1
decrements

15%
vary
from
13­
14%,
the
0.95
fractiles
vary
from
21­
22%,
and
the
0.05
fractiles
vary
from
7­
8%.
Box
plots
have
the
following
characteristics.
If
the
risk
distributions
for
a
particular
air
quality
scenario
are
quite
"
similar,"
the
rectangles
will
be
small.
If
the
variance
of
a
risk
distribution
is
small,
the
rectangles
will
be
close
together.
If
the
distributions
are
spaced
far
enough
apart
(
indicative
of
widely
varying
risk
estimates
for
different
pNEM
runs),
rectangles
will
overlap.
C­
7
20
40
60
80
100
0
120
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
100
200
300
400
500
600
700
800
0
900
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
3
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Chicago,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
8
5
10
15
20
25
0
30
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
50
100
150
200
250
0
300
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
4
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Denver,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
9
10
20
30
40
50
60
70
0
80
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
92
183
275
367
458
0
550
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
5
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Houston,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
10
50
100
150
200
250
300
0
350
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
500
1000
1500
2000
2500
3000
3500
0
4000
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
6
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Los
Angeles,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
11
5
10
15
20
25
30
0
35
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
50
100
150
200
250
300
0
350
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
7
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Miami,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
12
25
50
75
100
125
150
175
200
0
225
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
500
1000
0
1500
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
8
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
New
York,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
13
20
40
60
80
0
100
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
200
400
600
800
0
1000
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
9
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Philadelphia,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
14
5
10
15
20
25
0
30
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
50
100
150
200
250
0
300
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
10
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
St.
Louis,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
15
10
20
30
40
50
60
0
70
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
108
217
325
433
542
0
650
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

15%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
11
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

15%,
Washington
DC,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
16
10
20
30
40
50
60
0
70
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
50
100
150
200
250
300
0
350
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
12
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Chicago,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
17
5
10
15
0
20
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
25
50
75
100
0
125
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
13
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Denver,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
18
10
20
30
40
0
50
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
46
92
137
183
229
0
275
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
14
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Houston,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
19
50
100
150
0
200
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
500
1000
1500
0
2000
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
15
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Los
Angeles,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
20
5
10
15
0
20
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
25
50
75
100
0
125
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
16
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Miami,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
21
25
50
75
100
0
125
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
100
200
300
400
500
600
0
700
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
17
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
New
York,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
22
10
20
30
40
50
0
60
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
500
1000
1500
0
2000
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
17
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Philadelphia,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
23
5
10
15
0
20
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
25
50
75
100
0
125
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
19
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
St.
Louis,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
24
10
20
30
0
40
0.2
0.4
0.6
0.8
0.0
1.0
Children
Having
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
46
92
137
183
229
0
275
0.2
0.4
0.6
0.8
0.0
1.0
Occurrences
of
FEV1
Decrements

20%
(
thousands)
ASIS
8110
1112
8109
1110
8509
8508
8108
8107
FIGURE
C.
20
Representative
Risk
Distributions
for
Alternative
Air
Quality
Scenarios
(
FEV1
Decrements

20%,
Washington,
D.
C.,
Outdoor
Children,
8
Hr
Exposures,
Moderate
Exertion)
C­
25
5
10
15
20
25
30
35
0
40
Scenarios:

Chicago
ZDABHJFCG
Denver
ZDABHJFCG
Houston
ZDABHJFCG
Los
Ang
ZDABHJFCG
Miami
ZDABHJFCG
New
York
ZDABHJFCG
Phila
ZDABHJFCG
St
Louis
ZDABHJFCG
WashDC
ZDABHJFCG
Z=
AsIs
D=
8110
A=
1112
B=
8109
H=
1110
J=
8509
F=
8508
C=
8108
G=
8107
Key:

0.05
fractiles
medians
0.95
fractiles
FIGURE
C.
21
Headcount
Risk
Results
for
the
Percentage
of
Children
Responding
for
the
Eight­
Hr,
Moderate
Exertion,
FEV1
Decrements

15%
Endpoint
C­
26
5
10
15
20
0
25
Scenarios:

Chicago
ZDABHJFCG
Denver
ZDABHJFCG
Houston
ZDABHJFCG
Los
Ang
ZDABHJFCG
Miami
ZDABHJFCG
New
York
ZDABHJFCG
Phila
ZDABHJFCG
St
Louis
ZDABHJFCG
WashDC
ZDABHJFCG
Z=
AsIs
D=
8110
A=
1112
B=
8109
H=
1110
J=
8509
F=
8508
C=
8108
G=
8107
Key:

0.05
fractiles
medians
0.95
fractiles
FIGURE
C.
22
Headcount
Risk
Results
for
the
Percentage
of
Children
Responding
for
the
Eight­
Hr,
Moderate
Exertion,
FEV1
Decrements

20%
Endpoint
C­
27
5
10
0
15
Scenarios:

Chicago
ZDABHJFCG
Denver
ZDABHJFCG
Houston
ZDABHJFCG
Los
Ang
ZDABHJFCG
Miami
ZDABHJFCG
New
York
ZDABHJFCG
Phila
ZDABHJFCG
St
Louis
ZDABHJFCG
WashDC
ZDABHJFCG
Z=
AsIs
D=
8110
A=
1112
B=
8109
H=
1110
J=
8509
F=
8508
C=
8108
G=
8107
Key:

0.05
fractiles
medians
0.95
fractiles
FIGURE
C.
23
Headcount
Risk
Results
for
the
Percentage
of
Children
Responding
for
the
One­
Hr,
Moderate
Exertion,
Moderate­
to­

Severe
Pain
on
Deep
Inspiration
Endpoint
C­
28
C.
2.3
Ratios
of
Mean
Numbers
of
Occurrences
and
Mean
Numbers
of
Responders
Figures
C.
24­
26
show
the
ratios
for
the
mean
numbers
of
occurrences
and
mean
numbers
of
outdoor
children
responding
for
three
endpoints,
nine
urban
areas,
and
five
air
quality
scenarios.
The
endpoints
are
FEV1
decrements

15%
and

20%
for
8
hr
exposures
at
moderate
exertion,
and
moderate­
to­
severe
pain
on
deep
inspiration
for
1
hr
exposures
at
moderate
exertion.
The
following
letter
codes
were
used
to
identify
the
urban
areas:
CH
=
Chicago,
DE
=
Denver,
HO
=
Houston,
LA
=
Los
Angeles,
MI
=
Miami,
NY
=
New
York
City,
PH
=
Philadelphia,
SL
=
St.
Louis,
and
DC
=
Washington,
D.
C.

The
ratios
were
computed
in
the
following
way.
For
a
specific
endpoint,
available
risk
results
include
10
probability
distributions
(
one
for
each
of
10
pNEM
runs)
over
the
number
of
persons
who
respond
one
or
more
times,
and
10
probability
distributions
over
the
number
of
personoccurrences
(
which
allows
for
the
possibility
that
an
individual
may
respond
more
than
one
time).
The
ratio
of
interest
here
is
the
sum
of
the
expected
values
of
the
person­
occurrences
distributions
divided
by
the
sum
of
the
expected
values
of
the
persons
distributions.
The
ratio
is,
in
a
sense,
an
estimate
of
the
average
number
of
times
that
a
responder
responds
during
an
ozone
season.
C­
29
5
6
7
8
4
9
Urban
Area
8108
8508
8109
8509
1112
DE
HO
LA
MI
NY
PH
SL
CH
DC
Scenarios:

3
4
2
5
Urban
Area
8108
8508
8109
8509
1112
DE
HO
LA
MI
NY
PH
SL
CH
DC
Scenarios:
FIGURE
C.
24
RATIOS
OF
MEAN
NUMBER
OF
OCCURRENCES
AND
MEAN
NUMBER
OF
RESPONDERS
(
FEV1
DECREMENTS

15%,
9
URBAN
AREAS,
OUTDOOR
CHILDREN,
8
HR
EXPOSURES,
MODERATE
EXERTION)

FIGURE
C.
25
RATIOS
OF
MEAN
NUMBER
OF
OCCURRENCES
AND
MEAN
NUMBER
OF
RESPONDERS
(
FEV1
DECREMENTS

20%,
9
URBAN
AREAS,
OUTDOOR
CHILDREN,
8
HR
EXPOSURES,
MODERATE
EXERTION)
C­
30
10
12
14
16
18
8
20
Urban
Area
8108
8508
8109
8509
1112
Scenarios:

DE
HO
LA
MI
NY
PH
SL
CH
DC
FIGURE
C.
26
RATIOS
OF
MEAN
NUMBER
OF
OCCURRENCES
AND
MEAN
NUMBER
OF
RESPONDERS
(
MODERATE­
TO­
SEVERE
PAIN
ON
DEEP
INSPIRATION,
9
URBAN
AREAS,
OUTDOOR
CHILDREN,
1
HR
EXPOSURES,
MODERATE
EXERTION)
C­
31
C.
3
BENCHMARK
RISK
Figures
C.
27­
29
illustrate
benchmark
risk
results
for
the
probability
that
the
benchmark
response
will
be
exceeded
5
or
more
times
in
an
ozone
season.
Benchmark
response
is
r,
the
fraction
of
the
population
who
experience
the
specified
health
effect
upon
exposure
to
ozone.
Benchmark
risk
is
defined
as
the
probability
that
the
benchmark
response
is

r
n
or
more
times
in
a
given
period
of
time
(
1
ozone
season)
at
some
location
within
a
geographic
region,
given
a
specific
condition
of
air
quality
(
e.
g.,
that
standard
1112
is
just
attained).
In
this
report,
we
use
r
values
of
0.05
and
0.1
(
sometimes
referred
to
as
0.05
and
0.1
benchmarks,
or
5%
and
10%
benchmarks).

Figure
C.
27
shows
the
graphical
format
used
to
display
benchmark
risk
results.
The
figure
is
for
the
probability
that
the
benchmark
response
for
the
FEV
1
decrement

15%,
8
hr
exposure,
moderate
exertion
endpoint
will
be
exceeded
5
or
more
times
in
an
ozone
season
for
0.05
and
0.1
benchmarks.

This
figure
includes
results
for
9
urban
areas
and
9
air
quality
scenarios
for
each
urban
area.
The
air
quality
scenarios
are
indicated
by
a
letter
code
above
the
name
of
each
area.
The
letter
code
is
explained
on
the
right
side
of
the
figure.
There
are
2
vertical
lines
for
each
air
quality
scenario:
the
one
on
the
left
(
a
solid
line)
is
for
the
0.05
benchmark,
and
the
one
on
the
right
(
a
dotted
line)
is
for
the
0.1
benchmark.
The
height
of
the
line
indicates
the
benchmark
risk.
The
benchmark
risk
for
the
0.05
benchmark
is,
logically,

the
benchmark
risk
for
the
0.1
benchmark.

For
example,
for
Miami,
scenario
G,
daily
maximum
8­
hr­
running­
average
ozone
concentrations,
and
using
the
distribution
for
the
highest
ozone
concentration,
the
benchmark
risk
for
the
0.05
benchmark
is
about
0.52;
and
the
benchmark
risk
for
the
0.1
benchmark
is
about
0.2.
In
other
words,
if
standard
8107
is
just
attained
in
Miami,
then
the
benchmark
risk
(
i.
e.,
probability)
is
0.52
that

5%
of
the
population
will
experience
FEV
1
decrements

15%
5
or
more
times
in
an
ozone
season.
C­
32
0.2
0.4
0.6
0.8
0.0
1.0
Benchmarks:

Scenarios:

Z=
AsIs
J=
8509
B=
8109
F=
8508
C=
8108
A=
1112
H=
1110
D=
8110
G=
8107
Chicago
ZJBFCAHDG
Denver
ZJBFCAHDG
Houston
ZJBFCAHDG
Los
Ang
ZJBFCAHDG
Miami
ZJBFCAHDG
New
York
ZJBFCAHDG
Philadel
ZJBFCAHDG
St
Louis
ZJBFCAHDG
Wash
DC
ZJBFCAHDG
0.05
0.10
FIGURE
C.
27
Probability
that
the
Benchmark
Response
for
the
Eight­
Hr,
Moderate
Exertion,
FEV1
Decrement

15%
Endpoint
Will
Be
Exceeded
5
or
More
Times
in
an
Ozone
Season
C­
33
0.2
0.4
0.6
0.8
0.0
1.0
Benchmarks:

Scenarios:

Z=
AsIs
J=
8509
B=
8109
F=
8508
C=
8108
A=
1112
H=
1110
D=
8110
G=
8107
Chicago
ZJBFCAHDG
Denver
ZJBFCAHDG
Houston
ZJBFCAHDG
Los
Ang
ZJBFCAHDG
Miami
ZJBFCAHDG
New
York
ZJBFCAHDG
Philadel
ZJBFCAHDG
St
Louis
ZJBFCAHDG
Wash
DC
ZJBFCAHDG
0.05
0.10
FIGURE
C.
28
Probability
that
the
Benchmark
Response
for
the
Eight­
Hr,
Moderate
Exertion,
FEV1
Decrement

20%
Endpoint
Will
Be
Exceeded
5
or
More
Times
in
an
Ozone
Season
C­
34
0.2
0.4
0.6
0.8
0.0
1.0
Benchmarks:

Scenarios:

Z=
AsIs
J=
8509
B=
8109
F=
8508
C=
8108
A=
1112
H=
1110
D=
8110
G=
8107
Chicago
ZJBFCAHDG
Denver
ZJBFCAHDG
Houston
ZJBFCAHDG
Los
Ang
ZJBFCAHDG
Miami
ZJBFCAHDG
New
York
ZJBFCAHDG
Philadel
ZJBFCAHDG
St
Louis
ZJBFCAHDG
Wash
DC
ZJBFCAHDG
0.05
0.10
FIGURE
C.
29
Probability
that
the
Benchmark
Response
for
the
One­
Hr,
Moderate
Exertion,
Moderate­
to­
Severe
Pain
on
Deep
Inspiration
Endpoint
Will
Be
Exceeded
5
or
More
Times
in
an
Ozone
Season
