Additional Analyses of Ozone Metrics Related to Consideration of the Ozone Secondary Standard


UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

OFFICE OF AIR QUALITY PLANNING AND STANDARDS
RESEARCH TRIANGLE PARK, NC 27711

December 16, 2020
MEMORANDUM

SUBJECT: Additional Analyses of Ozone Metrics Related to Consideration of the Ozone
Secondary Standard

FROM: Benjamin Wells (EPA, OAQPS)

TO:	Ozone NAAQS Review Docket (EPA-HQ-OAR-2018-0279)

Table of Contents

1.	Overview 	2

2.	Data Handling	2

2.1	Data Retrieval and Preparation	2

2.2	Derivation of the Metrics	2

2.3	Assignment of Monitoring Sites to NOAA Climate Regions	4

3.	Results		5

3.1	National Analysis Using Recent Air Quality Data	5

3.2	National Analysis Using Historical Air Quality Data	18

4.	Summary 	21

5.	References 	22

1

-------
1. OVERVIEW

This technical memorandum presents various analyses of ambient air monitoring data for
ozone (O3) concentrations in the U.S. relating to the form and averaging time of the current
secondary standard and some metrics reported in environmental assessments. These metrics
include the W126-based cumulative exposure index, the N100 (number of hours at or above 100
ppb), and D100 (number of days with one or more hours at or above 100 ppb). The calculation of
these metrics is described in Section 2 below. These analyses describe relationships between the
three environmental metrics and the design values for the current standard (the annual 4th
highest daily maximum 8-hour O3 concentration, averaged over 3 consecutive years; hereafter
referred to as the "4th max metric"). The analyses presented here are an extension of analyses
that are presented Section 2.4.5, Appendix 2A, and Appendix 4D of the Policy Assessment for
the review (U.S. EPA, 2020).

2. DATA HANDLING

2.1 Data Retrieval and Preparation

Hourly O3 concentration data were retrieved from the EPA's Air Quality System (AQS,
https://www.epa.gov/aqs) database for 1,982 ambient air monitoring sites which operated
between 2000 and 2018. These data were used to calculate W126 and 4th max metric values for
each 3-year period from 2000-2002 to 2016-2018. Before calculating these metrics, some initial
processing was done on the hourly data. First, data collected using monitoring methods other
than federal reference or equivalent methods and data collected at monitoring sites not meeting
EPA's quality assurance or other criteria in 40 CFR part 58 were removed from the analysis.
Second, data collected by multiple monitoring instruments operating at the same location were
combined according to Appendix U to 40 CFR Part 50. Finally, data were combined across pairs
of monitoring sites approved for such combination by the EPA Regional Offices. The final
hourly O3 concentration dataset contained 1,779 monitoring sites.

2.2 Derivation of the Metrics

The 4th max metric values were calculated according to the data handling procedures in
Appendix U to 40 CFR part 50. First, moving 8-hour averages were calculated from the hourly
O3 concentration data for each site. For each 8-hour period, an 8-hour average value was
calculated if there were at least 6 hourly O3 concentrations available. Each 8-hour average was
stored in the first hour of the period (e.g., the 8-hour average from 12:00 PM to 8:00 PM is
stored in the 12:00 PM hour). Daily maximum 8-hour average values were found using the 8-
hour periods beginning from 7:00 AM to 11:00 PM each day. These daily maximum values were

-------
used if at least 13 of the 17 possible 8-hour averages were available, or if the daily maximum
value was greater than 70 parts per billion (ppb). Finally, the annual 4th highest daily maximum
value was found for each year, then averaged across each consecutive 3-year period to obtain the
final set of 4th max metric values in units of ppb. Any decimal digits in these values were
truncated for applications requiring direct comparison to a 4th max level (e.g., Table 1),
otherwise, all decimal digits were retained. The 4th max metric values were considered valid if
daily maximum values were available for at least 90% of the days in the O3 monitoring season
(defined in Appendix D to 40 CFR part 58) on average across the three years, with a minimum of
75% of the days in the O3 monitoring season in any calendar year. In addition, 4th max metric
values were considered valid if they were greater than the 4th max levels to which they were
being compared.

The W126 metric values were calculated using the hourly O3 concentration data in parts
per million (80 FR 65374, October 26, 2015). For daytime hours (defined as the 12-hour period
from 8:00 AM to 8:00 PM Local Standard Time each day), the hourly concentration values at
each O3 monitoring site were weighted using the following equation:

Weighted 03 = 03 / (1 + 4403*exp(-126 * 03)).

These weighted values were summed over each calendar month, then adjusted for
missing data (e.g.; if 80% of the daytime hourly concentrations were available, the sum would be
multiplied by 1/0.8 = 1.25) to obtain the monthly W126 index values. Monthly W126 index
values were not calculated for months where fewer than 75% of the possible daytime hourly
concentrations were available. Next, moving 3-month sums were calculated from the monthly
index values, and the highest of these 3-month sums was determined to be the annual W126
index. Three-month periods spanning multiple years (e.g., November to January, December to
February) were not considered in these calculations. The annual W126 index values were
averaged across each consecutive 3-year period to obtain the final W126 metric values, with
units in parts per million-hours (ppm-hrs). The W126 metric values were rounded to the nearest
unit ppm-hr for applications requiring direct comparison to a W126 level (e.g., Table 1),
otherwise, all decimal digits were retained. For consistency with the 4th max metric calculations,
the W126 metric values were considered valid if hourly O3 concentration values were available
for at least 90% of the daytime hours during the O3 monitoring season on average across the
three years, with a minimum of 75% of the daytime hours during the O3 monitoring season in
any calendar year. For consistency with the 4th max metric calculations, the W126 metric values
were considered valid if they were greater than the W126 levels to which they were being
compared.

-------
The N100 metric was calculated as the maximum number of hours with an hourly O3
concentration of 100 ppb or greater in the three consecutive calendar months yielding the highest
number in a given year. Similarly, the D100 metric was calculated as the maximum number of
days with at least one hourly O3 concentration of 100 ppb or greater in the three consecutive
calendar months yielding the highest number in a given year. These metrics were considered
valid if the annual data completeness rate for the O3 monitoring season was at least 75 percent

In the final dataset, 1,723 of the 1,779 03 monitoring sites had sufficient data to calculate
valid annual 4th max, W126, N100 and D100 values for at least one year between 2000 and 2018.
The number of sites with valid annual metric values ranged from 1,055 in 2000 to 1,226 in 2014,
and 608 sites had valid annual metric values in all 19 years. Additionally, 1,550 of the 1,779 O3
monitoring sites had sufficient data to calculate valid 4th max and W126 metric values for at
least one 3-year period between 2000-2002 and 2016-2018. The number of sites with valid 4th
max and W126 metric values ranged from a low of 953 in 2000-2002 to a high of 1,122 in 2015-
2017, and 543 sites had valid 4th max and W126 metric values for all seventeen 3-year periods.

2.3 Assignment of Monitoring Sites to NOAA Climate Regions

In order to examine regional differences, many of the further analyses were stratified into
the nine NOAA climate regions (Karl and Koss, 1984), which are shown in Figure 1. Since the
NOAA climate regions only cover the contiguous U.S., Alaska was added to the Northwest
region, Hawaii was added to the West region, and Puerto Rico was added to the Southeast
region.

¦ NorthWest ¦ WestNorthCentral ~ EastNorthCentral ¦ Central ¦ NorthEast

¦ West ~ SouthWest ~ South ~ SouthEast

Figure 1. Map of the nine NOAA climate regions.

-------
3. RESULTS

3.1 National Analysis Using Recent Air Quality Data

This section presents various results based on the annual 4th max, W126, N100, and D100
metrics as well as the 3-year average 4th max and W126 metrics1 for the 2016-2018 period.
Figure 2 and Figure 3 show maps of the average annual N100 and D100 values, respectively, at
sites with valid 4th max metric values (design values) for 2016-2018. About 72% of the O3
monitoring sites did not have any hourly concentrations at or above 100 ppb in 2016-2018, and
an additional 15% of the sites had an average of one day or less per year where hourly O3
concentrations reached 100 ppb or more. Sites with more than one day per year where hourly O3
concentrations reached 100 ppb or more were generally located near large urban areas, with the
most extreme values located downwind of Los Angeles, CA.

For 2016-2018, Figure 4 and Figure 5 show scatter plots comparing sites of different 4th
max metric values (x-axis) with regard to N100 and D100 values (y-axis), respectively.

Similarly, Figure 6 and Figure 7 compare sites of different W126 metric values (x-axis) with
regard to N100 and D100 values (y-axis), respectively, for the same period. For sites meeting the
current standard (i.e., 4th max metric value < 70 ppb), the hourly O3 concentrations reached 100
ppb or more for at most ten hours on up to three distinct days. By contrast, sites with W126
metric values as low as 5 ppm-hrs had up to ten hours with concentrations of 100 ppb or greater
on up to four distinct days. Focusing on sites with W126 metric values below 20 ppm-hrs,
several sites had N100 values of ten or greater and D100 values of five or greater, with
individual sites having as many as 38 hours on up to seven distinct days with concentrations of
100 ppb or greater.

Figure 8 and Figure 9 show scatter plots (similar to Figure 4 and Figure 5) that compare
sites having different 2016, 2017, and 2018 annual 4th max values (x-axis) with regard to the
2016, 2017, and 2018 N100 and D100 values (y-axis), respectively. As can be seen from these
figures, sites where the annual 4th max value was at or below 70 ppb generally had at most five
hours on two distinct days where the O3 concentrations reached 100 ppb or more. Figure 10 and
Figure 11 show similar scatter plots comparing sites having different 2016, 2017, and 2018
annual W126 values (x-axis) with regard to the 2016, 2017, and 2018 N100 and D100 values (y-
axis), respectively. In contrast to the 4th max values, there were sites that had five or more hours
at or above 100 ppb on up to three distinct days at W126 levels as low as 5 ppm-hrs. Focusing on
sites where the annual W126 values were below 20 ppm-hrs, several sites had ten or more hours
on five or more distinct days where O3 concentrations reached 100 ppb or more.

1 As defined in section 2.2 above, the term "W126 metric" refers to the 3-year average W126 index. The term
"annual W126" is used in reference to single-year W126 index values.

-------
• 0 (821 sites)	O 1.1 - 5.0 (109 sites) • > 10.0 (30 sites)

O 0.1-1.0 (167 sites) © 5.1 - 10.0 (11 sites) A 4th Max Metric > 70 ppb

Figure 2. Map of 2016-2018 Average N100 Values at sites with valid design values.

• 0 (821 sites)	O 1.1-5.0 (61 sites) • > 10.0 (18 sites)

O 0.1-1.0 (229 sites) © 5.1-10.0 (9 sites) A 4th Max Metric > 70 ppb

Figure 3. Map of 2016-2018 Average D100 Values at sites with valid design values.

-------
o NorthWest
• West

A 2016
¦ 2017







A

©

• WestNorthCentral
o SouthWest
o EastNorthCentral

• 2018









o

o South
o Central
O SouthEast









~

©



o NorthEast







n

~













¦

~

@0

OA

























A

A

i.

A













~ i

~o a

c

o











D

1

A

~ •









~

~ o

Z3

X 2











mk rr, Jor ®





rfi

~ ~

J











40

50

60	70	80	90

4th Max Metric Value (ppb)

100

110

Figure 4. Scatter plot of annual N100 values (y-axis) versus 4th max metric values (design
values, x-axis) based on 2016-2018 monitoring data.

7

-------
o NorthWest

•	West

•	WestNorthCentral
o SouthWest

A 2016
¦ 2017
• 2018







£

k o
©

o EastNorthCentral
o South
o Central
O SouthEast









~



o NorthEast







n
~

n ©

¦ A











©
a

A

~

~ ©

>0 •

1 r> P- ©>
©>

•





















~ ©

AA • • i
~

©~

n rf*, A









©

• [Dio A

o ~ ©A

~ ~ CSQ]
Kj aXiD Gfi©

O ^fcfcl© OC2S ®
ijtontoAAA no&ofe

a a





A *

~ ~
¦ «~ Qi

~ O

~H CD ~!





40

50

60	70	80	90

4th Max Metric Value (ppb)

100

110

Figure 5. Scatter plot of annual D100 values (y-axis) versus 4th max metric values (design
values, x-axis) based on 2016-2018 monitoring data.

8

-------
o NorthWes
• West

t

A 2016
¦ 2017











A



©

• WestNorthCentral* 2018
O Southwest
o EastNorthCentral











©





o South
o Central
_ o SouthEas

t







~







©





o NorthEasI











~

~























~

•



~



















~

A

•

c























A

A
A A

A

















~

~

•

A

~o

©

©
A













~





•

A
•

A

~

%
A

©











~

~ O

A



~ #

~

o



A

©
A







- n

~ A



A

©

o

f

Va

&

~

~

A







¦ /

JapqA d

O

~

• ¦

°A

oa(

A ©

£	1

~

O







0

10

20

30

4

0

50

6

0

Figure 6.

W126 Metric Value (ppm-hrs)

Scatter plot of annual N100 values (y-axis) versus W126 metric values (x-axis)
based on 2016-2018 monitoring data.

9

-------
o NorthWest A 2016

•	West ¦ 2017

•	WestNorthCentral® 2018

_ O .^OI ithWpct







A



©



©

o EastNorthCentra
o South
o Central
O South East







~













o NorthEasI









¦ A

~
~

~



©

















A

A
•

A
•

A
A®
~

A

h

¦

©©

~ ®

M,

©A

c

©

A

~

©

















~















~

A

~

©

A

19 #0

~ A1

~

A

©

A
~









~

~
~

r~r~n /V\



•

O

M

Sa'

~ ~ i

Jl JiA

~ •

© m

A n

A
A







i-i—i

OQ&EAiOQ

~

£1od [
G±>Ca

oo A

<

~ ©

A ~

AA

A,

©

M
~
©







EEm



:WTltw«

m*w T



ztn

_ W r T ^sr.r,' m

0

10 20

30

*¦ |

4i

0

50

i

6'

0

W126 Metric Value (ppm-hrs)

Figure 7. Scatter plot of annual D100 values (y-axis) versus W126 metric values (x-axis)
based on 2016-2018 monitoring data.

10

-------
o Central

o EastNorthCentral

A 2016
¦ 2017







Am



• Nortntast
o NorthWest
© South



• 2018







•



o SouthEast
o SouthWest
• West













~

©

• WestNorthCentral









~

~

















~

~

» %















Ac















A
A

aa

4















o 9
•















&
*~.

~













A













~

A •









~ ~

a A ~

tlip





40	50	60	70	80	90

Annual 4th Max Value (ppb)

100

110

Figure 8. Scatter plot of annual N100 values (Y-axis) versus annual 4th max values (x-axis),
based on 2016-2018 monitoring data.

n

-------
o Central

o EastNorthCentral
o NorthEast

_ O NnrthWest

A 2016
¦ 2017
• 2018







©

©



o South
o SouthEast
o SouthWest
• West













~

• WestNorthCentral







A

~

~

~

~

•













A (

M

A

A o

A
A
on

©

A O®

~

~ ©

©

A

~ CD
~

5

©

~

~

©

6



























A

A

~
©











m oAc

O ~ ©A ~ E

AA:a*!s3Acs£a

~





A

~ ~
A Am d

A

0
-------
o Central
o EastNorthCentra

A 2016
¦ 2017











A

©

o Northtast
o NorthWes
o South

t

• 2018









C







o SouthEas
o SouthWes
• West

,t









~



©







• WestNorthCentrs

I









~

~























~

® •

©

~



















~

A























A

A
A

>
¦i



















~

©













~





J
•

•
•

©

~













A

¦



bA

i

~

©

A

A

•

A







~

I	I

O
~O L

k •

©
A*

A

ran £

E

•

L A.

A





A







o
~ ©

V:

L.JP

~



f -i



0

10

20

30

40

5

0 6

0

Annual W126 Value (ppm-hrs)

Figure 10. Scatter plot of annual N100 values (Y-axis) versus annual W126 values (x-axis),
based on 2016-2018 monitoring data.

13

-------
o Central
o EastNorthCentra
o NorthEasi
o NorthWest

A 2016
¦ 2017
• 2018









©



©

o South
o SouthEast
o SouthWest
• West









~











• WestNorthCentrs

I





A

:

~

~

©
~

















&



A ©
A ©

©

A

©A ©
~

~ A

A

A& ~

~

A

















i—

L

o

c

~

i













~

A A
•

~ o

L

A.

© ~

A A-i

¥







r-

~

~ o

m~iA. A /"s

©

MP C

© in
© II
~ ® «
# cA ®
l.m A
nn ©

ci

ca i

• i
n

A

k

A





A



c

1 1 I- ¦¦

o mbi

cm tfc o Mm
' " ' fM ^

o

~ cam

-Tiifrm-*

A
©

k. GO

~

















C®nti A

*- ¦— -

A A

—A—4	









10

20	30	40

Annual W126 Value (ppm-hrs)

50

60

Figure 11. Scatter plot of annual D100 values (Y-axis) versus annual W126 values (x-axis),
based on 2016-2018 monitoring data.

14

-------
Table 1 below shows the number of sites where the 2016-2018 4th max metric values meet
the current standard or the number of instances (i.e., site-years) where the 2016, 2017, and 2018
annual 4th max values are at or below the level of the current standard and the 2016, 2017, and
2018 N100 or D100 values are above various thresholds. The table also shows number of sites
where the 2016-2018 W126 metric values are at or below specific W126 levels or the number of
instances where the 2016, 2017, and 2018 annual W126 values are at or below specific W126
levels and the 2016, 2017, and 2018 N100 or D100 values are above various thresholds. The
number of sites or instances where the N100 and D100 values were nonzero are always equal,
because having at least one hour where the concentration is at or above 100 ppb guarantees
having at least one day where the maximum hourly concentration is at least 100 ppb. The
number of sites or instances where the D100 values exceeded 2 and 5 were generally similar to
the number of sites or instances where the N100 values exceeded 5 and 10, respectively.

With regard to sites at or below specific annual 4th max and W126 values in any of the
three years, according to Table 1, there were only four instances out of over 2,600 site-years
(0.2%) where the N100 value exceeded 5 for sites during a year where the annual 4th max value
was at or below the level of the current standard. Additionally, there were only seven sites out of
over 800 (less than 1%) that met the current standard based on 2016-2018 data and also had
N100 values exceeding 5 in one or more years. By contrast, there were 47 instances out of over
3,300 (1.4%) where the N100 value exceeded 5 for sites that had an annual W126 value at or
below 19 ppm-hrs; and additionally 37 sites out of over 1,000 (more than 3%) that had a 2016-
2018 W126 metric value was at or below 17 ppm-hrs and a N100 value exceeding 5 in one or
more years. Even when looking at sites at or below a W126 level of 7 ppm-hrs, there were twice
as many sites (14) with N100 values exceeding 5 than for sites meeting the current standard (7).

Table 2 shows the same statistics as in Table 1 for the annual 4th max and annual W126
values broken out into individual years, with the maximum annual value across the three years
for each combination of 4th max/W 126 and N100/D100 thresholds highlighted in light blue. This
table shows that while there is considerable inter-annual variation in the 4th max and W126
values across years, the annual W126 values always have a higher proportion of sites below the
threshold and above the N100 or D100 thresholds compared to those of the annual 4th max
values. Further, during the highest year for the different N100 and D100 thresholds, the
proportion of sites exceeding those thresholds is greater for the sites at/below the different annual
W126 levels than it is for sites with design values at/below 70 ppb. This is also evident in
comparing Figure 5 to Figure 11 and Figure 4 to Figure 10.

15

-------
Table 1. Number of instances where 4th max or W126 values are at or below various

thresholds and N100 or D100 values are above various thresholds based on O3
monitoring data from recent years (2016-2018).



Total*

Number
N100 > 0

of instance
N100 > 5

s where:
N100 >10

Number
D100 > 0

of instance:
D100 > 2

where:
D100 > 5

Number of sites exceeding threshold in one or more years

3-year Total**

1,113

315
(28%)

86
(8%)

39
(4%)

315
(28%)

87
(8%)

39
(4%)

3-year 4th Max < 70

853

121
(14%)

7

(0.8%)

0
(0%)

121
(14%)

3

(0.4%)

0
(0%)

3-year W126 < 19

1,043

245
(23%)

38
(4%)

6

(0.6%)

245
(23%)

36
(3%)

3

(0.3%)

3-year W126 < 17

1,031

237
(23%)

37
(4%)

6

(0.6%)

237
(23%)

34
(3%)

3

(0.3%)

3-year W126 < 15

1,002

225
(22%)

34
(3%)

6

(0.6%)

225
(23%)

30
(3%)

3

(0.3%)

3-year W126 < 7

667

114
(17%)

14
(2%)

2

(0.3%)

114
(17%)

10

(2%)

1

(0.1%)



Total number of instances (si

e/years) exceeding threshold

Annual Total***

3,607

533
(15%)

154
(4%)

00 es

533
(15%)

168
(5%)

^ 00

^ 00

Annual 4th Max < 70

2,630

114
(4%)

4

(0.2%)

0
(0%)

114
(4%)

2

(0.1%)

0
(0%)

Annual W126<25

3,479

416
(12%)

57
(2%)

16
(0.5%)

416
(12%)

64
(2%)

9

(0.3%)

Annual W126 < 19

3,370

366
(11%)

47

(1%)

10
(0.3%)

366
(11%)

49
(2%)

4

(0.1%)

Annual W126<17

3,318

349
(1 1 %)

44
(1%)

9

(0.3%)

349
(1 1 %)

45
(1%)

4

(0.1%)

Annual W126 < 15

3,235

331
(10%)

44
(1%)

9

(0.3%)

331
(10%)

43
(1%)

4

(0.1%)

Annual W126<7

2,134

141
(7%)

14
(0.7%)

0
(0%)

141
(7%)

11
(0.5%)

0
(0%)

* Total number of sites where the 3-year 4th max or W126 value is at or below the threshold, or the total number of instances
(i.e., site/years) where the annual 4th max or W126 value is at or below the threshold.

** First column shows the number of sites with sufficient data to calculate valid 3-year 4th max and W126 values. Subsequent
columns tally the subset of those sites where the N100 or D100 value exceeds the threshold in one or more years.

*** First column shows the number of instances where a site had sufficient data to calculate valid annual 4th max and W126
values. Subsequent columns tally the subset of those instances where the N100 or D100 value exceeds the threshold.

16

-------
Table 2. Number of instances where annual 4th max or W126 values are at or below

various thresholds and N100 or D100 values are above various thresholds based
on O3 monitoring data from 2016-2018



Total
Number
of Sites*

Num
N100 > 0

>er of sites wh
N100 > 5

ere:
N100 >10

Num
D100 > 0

ber of sites wl
D100 > 2

lere:
D100 > 5



Number of sites exceeding threshold in the maximum year of the three

3-year 4th Max < 70

853

53 (6%)

4 (0.5%)

0 (0%)

53 (6%)

2 (0.2%)

0 (0%)

Annual 4th Max < 70

See
Below

43 (5%)

3 (0.3%)

0 (0%)

43 (5%)

2 (0.2%)

0 (0%)

Annual W126<25

165(14%)

25 (2%)

6 (0.5%)

165(14%)

25 (2%)

4 (0.3%)

Annual W126 < 19

145(13%)

21 (2%)

4 (0.4%)

145(13%)

20 (28%)

3 (0.3%)

Annual W126 < 17

138(13%)

20 (2%)

4 (0.4%)

138(13%)

18(2%)

3 (0.3%)

Annual W126 < 15

130(13%)

20 (2%)

4 (0.4%)

130 (13%)

18(2%)

3 (0.3%)

Annual W126<7

59 (7%)

8(1%)

0 (0%)

59 (7%)

6(1.0%)

0 (0%)



Number of sites exceeding threshold in individual years

2018 Total**

1,187

208(18%)

62 (5%)

28 (2%)

208(18%)

64 (5%)

25(2%)

2017 Total**

1,205

179 (15%)

52 (4%)

29 (2%)

179 (15%)

59 (5%)

31 (3%)

2016 Total**

1,215

146 (12%)

40 (3%)

28 (2%)

146 (12%)

46 (4%)

27 (2%)

2018 4th Max <70

801

32 (4%)

0 (0%)

0 (0%)

32 (4%)

0 (0%)

0 (0%)

2017 4th Max <70

931

43 (5%)

1 (0.1%)

0 (0%)

43 (5%)

0 (0%)

0 (0%)

2016 4th Max <70

898

39 (4%)

3 (0.3%)

0 (0%)

39 (4%)

2 (0.2%)

0 (0%)

2018 W126 < 25

1,142

165(14%)

25 (2%)

6 (0.5%)

165(14%)

25 (2%)

2 (0.2%)

2017 W126 < 25

1,163

138(12%)

19(2%)

5 (0.4%)

138(12%)

23 (2%)

4 (0.3%)

2016 W126 < 25

1,174

113(10%)

13(1%)

5 (0.4%)

113(1%)

16(1%)

3 (0.3%)

2018 W126 < 19

1,090

145(13%)

21 (2%)

4 (0.4%)

145(13%)

20 (2%)

1 (0.1%)

2017 W126 < 19

1,128

120(11%)

17(2%)

4 (0.4%)

120(11%)

17 (2%)

3 (0.3%)

2016 W126 < 19

1,152

101 (9%)

9 (0.8%)

2 (0.2%)

101 (9%)

12(1%)

0 (0%)

2018 W126 < 17

1,066

138(13%)

20 (2%)

3 (0.3%)

138(13%)

18(2%)

1 (0.1%)

2017 W126 < 17

1,113

114(10%)

16(1%)

4 (0.4%)

114(10%)

16(1%)

3 (0.3%)

2016 W126 < 17

1,139

97 (9%)

8 (0.7%)

2 (0.2%)

97 (9%)

11 (1%)

0 (0%)

2018 W126 < 15

1,029

130(13%)

20 (2%)

3 (0.3%)

130 (13%)

18(2%)

1 (0.1%)

2017 W126 < 15

1,092

109 (10%)

16(1%)

4 (0.4%)

109 (10%)

15(1%)

3 (0.3%)

2016 W126 < 15

1,114

92 (8%)

8 (0.7%)

2 (0.2%)

92 (8%)

10 (0.9%)

0 (0%)

2018 W126 < 7

625

47 (8%)

8(1%)

0 (0%)

47 (8%)

6(1%)

0 (0%)

2017 W126 < 7

788

59 (7%)

5 (0.6%)

0 (0%)

59 (7%)

4 (0.5%)

0 (0%)

2016 W126 < 7

721

35 (5%)

1 (0.1%)

0 (0%)

35(5%)

1 (0.1%)

0(0%)

* Total number of sites where the annual 4th max or W126 value is at or below the threshold.

** First column represents the number of sites with sufficient data to calculate a valid annual 4th max value. Subsequent

columns tally the subset of those sites where the N100 or D100 value exceeds the threshold in one or more years.

17

-------
3.2 National Analysis Using Historical Air Quality Data

Figure 12 and Figure 13 show the trend in national 10th percentile, median, 90th percentile
and mean N100 and D100 values, respectively, based on 808 U.S. O3 monitoring sites with
complete data for 2000 to 2018. A site must have 75% annual data completeness in terms of the
4th max m etric (see section 2.2) for at least 15 of the 19 years, with no more than two consecutive
years missing to be included in the trend. As can be seen from the figures, the median N100 and
D100 values in the U.S. have been zero since 2006, meaning over half of the monitoring sites
have N100 and D100 values of zero. The mean N100 value has decreased from more than ten in
2000-2002 to less than two in recent years, a decline of more than 80%. Similarly, the mean
D100 value has decreased from four or more in 2000-2002 to less than one in recent years, also a
decline of more than 80%. The 90th percentile values of both metrics have decreased at an even
faster rate.

10 -























c

50th Percentile



























































1

0th Perce

r.,































ean





























































































































V





































\



































\









\



































•Os















1

N











o
o
o

CM

O
O
CM

CM
O
O
CM

co
o
o

C\l

o
o

CM

10
o
o

CM

CD
O
O
CM

r-
o
o

CM

CO

o
o

CM

CD
O

o

CM

O
CM

O
CM

CNJ

O
CM

CO

O
CM

O
CM

lO

O
CM

CD

O
CM

O
CM

CO

O
CM

Figure 12. Trend in N100 values from 2000 to 2018 based on data from 808 U.S. O3
monitoring sites

18

-------
14

12

10

d)

3

« 8

o
o

5

4

2

90th Percentile

Median

10th Percentile

Mean

Figure 13. Trend in D100 values from 2000 to 2018 based on data from 808 U.S. O3
monitoring sites

Table 3 below shows the number of instances (site-years) where a site had an annual 4th
max value or 4th max metric value at or below the level of the current standard and an annual
N100 or D100 value above various thresholds based on the full dataset spanning years 2000 to
2018. The table also shows number of instances (site-years) where a site had an annual W126
value or W126 metric value at or below specific W126 levels and N100 or D100 values above
various thresholds based on the full 2000-2018 dataset. The numbers in Table 3 are generally
proportionally similar to those shown previously in Table 1.

According to Table 3, there were only 9 instances where the \ 100 value exceeded 5 at a
site with an annual 4ih max value at or below the level of the current standard, and only 89
instances out of over 8,000 (about 1%) that met the current standard and also had N100 values
exceeding 5 in one or more of the three years of the design value period. By contrast, there were
over 1,500 instances where the annual W126 value was less than or equal to 19 ppm-hrs and the
N100 value in that year exceeded 5, and nearly 2,500 instances (more than 16%) where the

19

-------
W126 metric value was at or below 17 ppm-hrs and the N100 value exceeded 5 in one or more
years of the 3-year period. Even when looking at sites at or below a W126 level of 7 ppm-hrs,
there were more instances with N100 values exceeding 5 (145) than for sites meeting the current
standard (89).

Table 3. Number of instances where 4th max or W126 values are at or below various

thresholds and N100 or D100 values are above various thresholds based on data
from all years (2000-2018)



Total*

Number
N100 > 0

of instance
N100 > 5

s where:
N100 >10

Number
D100 > 0

of instance;
D100 > 2

where:
D100 > 5

Number of instances where site exceeds threshold in one or more years

3-year Total**

17,879

9,306
(52%)

4,661
(26%)

3,063
(17%)

9,306
(52%)

4,658
(26%)

2,364
(13%)

3-year 4th Max < 70

8,116

1,371
(17%)

89
(1%)

14
(0.2%)

1,371
(17%)

74
(0.9%)

7

(0.1%)

3-year W126 <19

15,804

7,319
(46%)

2,987
(19%)

1,616
(10%)

7,319
(46%)

2,928
(19%)

1,002
(6%)

3-year W126 <17

15,000

6,621
(44%)

2,492
(17%)

1,263
(8%)

6,621
(44%)

2,432
(16%)

729
(5%)

3-year W126 <15

13,865

5,716
(41%)

1,926
(14%)

902
(7%)

5,716
(41%)

1,869
(13%)

495
(4%)

3-year W126 < 7

6,059

1,201
(20%)

145
(2%)

39
(0.6%)

1,201
(20%)

133
(2%)

23
(0.4%)



Total number of instances (site/years) exceeding threshold

Annual Total***

22,102

7,489
(34%)

3,500
(16%)

2,232
(10%)

7,489
(34%)

3,468
(16%)

1,640
(7%)

Annual 4th Max < 70

10,265

501
(5%)

9

(0.1%)

0
(0%)

501
(5%)

5

(<0.1 %)

0
(0%)

Annual W126<25

20,684

6,165
(30%)

2,356
(11%)

1,234
(6%)

6,165
(30%)

2,295
(11%)

N^O OO

Annual W126 < 19

18,941

4,836
(26%)

1,529
(8%)

713
(4%)

4,836
(26%)

1,456
(8%)

334
(2%)

Annual W126<17

17,941

4,179
(23%)

1,183
(7%)

516
(3%)

4,179
(23%)

1,128
(6%)

228
(1%)

Annual W126 < 15

16,661

3,455

(21 %)

853
(5%)

315
(2%)

3,455
(21 %)

815
(5%)

140
(0.8%)

Annual W126<7

8,612

709
(8%)

61
(0.7%)

3

(<0.1 %)

709
(8%)

49
(0.6%)

2

(<0.1 %)

* Total number of sites where the 3-year 4th max or W126 value is at or below the threshold, or the total number of instances
(i.e., site/years) where the annual 4th max or W126 value is at or below the threshold.

** First column shows the number of sites with sufficient data to calculate valid 3-year 4th max and W126 values. Subsequent
columns tally the subset of those sites where the N100 or D100 value exceeds the threshold in one or more years.

*** First column shows the number of instances where a site had sufficient data to calculate valid annual 4th max and W126
values. Subsequent columns tally the subset of those instances where the N100 or D100 value exceeds the threshold.

20

-------
4. SUMMARY

The presentation here shows various analyses of ambient air monitoring data for O3
concentrations in the U.S. relating to the form and averaging time of the current secondary
standard, the W126-based cumulative exposure index, the N100 metric (number of hours at or
above 100 ppb) and D100 metric (number of days with one or more hours at or above 100 ppb).

•	About 72% of the O3 monitoring sites with valid design values in 2016-2018 did not have
any hourly concentrations at or above 100 ppb, and another 15% had only a single day
where hourly O3 concentrations reached 100 ppb or more (Figure 2 and Figure 3).

•	Based on data from 2016-2018, sites where the current standard was met (4th max metric
value was at or below 70 ppb) had a maximum annual N100 count of 10 and D100 count
of 3 (Figure 4 and Figure 5). Sites with W126 metric values as low as 5 ppm-hrs had a
maximum annual N100 count of 10 and D100 count of 4. At sites with W126 metric
values below 20 ppm-hrs, several sites had N100 values of ten or greater and D100
values of five or greater, with individual sites having as many as 38 hours on up to seven
distinct days with concentrations of 100 ppb or greater (Figure 6 and Figure 7).

•	In 2016-2018, sites where the annual 4th max value was at or below 70 ppb had a
maximum annual N100 count of 6 and D100 count of 3 (Figure 8 and Figure 9). Sites
with annual W126 values as low as 5 ppm-hrs had a maximum N100 count of 8 and
D100 count of 3. At sites with annual W126 values below 20 ppm-hrs, some sites had ten
or more hours on up to seven distinct days where O3 concentrations reached 100 ppb or
more (Figure 10 and Figure 11).

•	Based on data from 2016-2018, less than 1% of sites that met the current standard had an
N100 value exceeding 5 in one or more years. By comparison, more than 3% of sites
where the W126 metric value was at or below 17 ppm-hrs had an N100 value exceeding
5 (Table 1). While there was considerable variability across individual years, there were
always far fewer sites with N100 values exceeding 5 among sites with annual 4th max
values at or below the level of the current standard than among sites with annual W126
values at or below 19 ppm-hrs (Table 2).

•	Based on data from 2000-2018, about 1% of design values that met the current standard
had N100 values exceeding 5 in one or more years of the 3-year period. By comparison,
about 17%) of W126 metric values at or below 17 ppm-hrs had N100 values exceeding 5
in one or more years of the 3-year period (Table 3).

•	Since 2000-2002, the national mean N100 and D100 values have decreased by more than
80%o (Figure 12 and Figure 13).

21

-------
5. REFERENCES

Karl, T and Koss, WJ (1984). Regional and national monthly, seasonal, and annual temperature
weighted by area, 1895-1983. 4-3. National Environmental Satellite and Data
Information Service (NESDIS). Asheville, NC.

U.S. EPA (2020). Policy Assessment for the Review of National Ambient Air Quality Standards
for Ozone. Office of Air Quality Planning and Standards, Health and Environmental
Impacts Divison. Research Triangle Park, NC. U.S. EPA. EPA-452/R-20-001. May 2020.

22

-------