United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-454/R-97-014
November 1997
AIR
E PA ANALYSIS OF THE AFFECT OF
ASOS-DERIVED METEOROLOGICAL DATA
ON REFINED MODELING
-------
-------
EPA-454/R-97-014
Analysis of the Affect of ASOS-Derived
Meteorological Data on Refined Modeling
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emissions, Monitoring, and Analysis Division
Research Triangle Park, NC 27711
November 1997
-------
NOTICE
The information in this document has been reviewed by the U. S. Environmental
Protection Agency (EPA) and approved for publication as an EPA document. Mention of trade
names, products, or services does not convey, and should not be interpreted as conveying official
EPA approval, endorsement, or recommendation.
-------
PREFACE
The National Weather Service (NWS) is in the process of installing an automated surface
observing network throughout the United States, the Caribbean, and overseas military bases.
When the installations are completed, hourly observations of surface weather conditions will no
longer be performed by a human observer. The instrumentation used in these automated systems
have some limitations. It is the purpose of this study to determine the affect of these automated
observations, if any, on concentration estimates from the Industrial Source Complex Short Term
(ISCST3) model.
11
-------
ACKNOWLEDGMENTS
The study was performed by James O. Paumier and Roger W. Erode of Pacific
Environmental Services, Inc. (PES), Research Triangle Park, North Carolina. In addition, this
document was reviewed and commented upon by Desmond Bailey (EPA, OAQPS), Ellen Cooler
(EPA, ORD), and Joe Tikvart (EPA, OAQPS). This effort was funded by the U. S.
Environmental Protection Agency under Contract No. 68D30032, with Dennis Atkinson as Work
Assignment Manager.
in
-------
-------
TABLE OF CONTENTS
1.0 INTRODUCTION 1-1
2.0 COMPARISON OF ASOS AND CONVENTIONAL OBSERVATION METHODS ... 2-1
2.1 Cloud Cover and Height 2-1
2.2 Temperature 2-1
2.3 Wind 2-1
3.0 STATION AND DATE SELECTION 3-1
3.1 Station Selection 3-1
3.2 Date Selection 3-1
4.0 METEOROLOGICAL DATA PREPARATION 4-1
4.1 Hourly Surface Observations 4-1
4.2 Twice-daily Mixing Heights 4-5
4.3 Periods of Missing Data 4-5
4.4 Running PCRAMMET 4-7
4.5 Analysis of Meteorological Data 4-8
5.0 ISCST3 MODEL INPUT 5-1
5.1 Model Options 5-1
5.2 Source Characterization 5-1
5.3 Receptor Network 5-2
6.0 ISCST3 MODEL RESULTS 6-1
6.1 ASOS Cloud Cover vs Conventional Cloud Cover Observations 6-1
6.2 Full ASOS vs Conventional Observations 6-11
7.0 CONCLUSIONS 7-1
8.0 REFERENCES 8-1
APPENDIX A A-l
APPENDIX B B-l
-------
-------
LIST OF FIGURES
4.1 Difference in wind speed between ASOS observations and conventional
(human) observations for all six stations (# observations = 7,440) 4-9
4.2 Difference in wind direction between ASOS observations and conventional
(human) observations for all six stations (# observations = 7,440) 4-10
4.3 Difference in ambient temperature between ASOS observations and conventional
(human) observations for all six stations (# observations = 7,440) 4-11
4.4 Difference in PG stability category between mixed ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440) .... 4-21
4.5 Difference in PG stability category between full ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440) .... 4-22
4.6 Difference in rural mixing heights between mixed ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440) .... 4-23
4.7 Difference in urban mixing heights between mixed ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440) .... 4-24
4.8 Difference in rural mixing height between full ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440) .... 4-25
4.9 Difference in urban mixing heights between full ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440) .... 4-26
6.1 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 10-meter point source 6-3
6.2 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 35-meter point source 6-4
6.3 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 35-meter point source with building
downwash 6-5
6.4 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 55-meter point source with building
downwash 6-6
6.5 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 100-meter point source 6-7
6.6 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 200-meter point source 6-8
6.7 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the area source 6-9
6.8 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the volume source 6-10
6.9 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the 10-meter point source 6-13
6.10 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the 35-meter point source 6-14
vn
-------
LIST OF FIGURES (Concluded)
6.11 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the 35-meter point source with building
downwash 6-15
6.12 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the 55-meter point source with building
downwash 6-16
6.13 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the 100-meter point 6-17
6.14 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the 200-meter point source 6-18
6.15 Relative difference between the high-1 st-high concentrations using the full
ASOS data and conventional data for the area source 6-19
6.16 Relative difference between the high-lst-high concentrations using the full
ASOS data and conventional data for the volume source 6-20
Vlll
-------
LIST OF TABLES
Table
4.1 CLOUD COVER COMPARISON 4-3
4.2 SUMMARY OF MISSING PERIODS FOR ASOS DATA 4-6
4.3 SUMMARY OF MISSING PERIODS OF TWICE-DAILY MIXING
HEIGHTS 4-7
4.4 DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN
WIND SPEED BETWEEN ASOS AND CONVENTIONAL OBSERVATIONS
FOR ALL STATIONS AND FOR EACH STATION 4-12
4.5 DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN
WIND DIRECTION BETWEEN ASOS AND CONVENTIONAL
OBSERVATIONS FOR ALL STATIONS AND FOR EACH STATION 4-13
4.6 NUMBER OF CALM WINDS BY STATION AND DATA TYPE 4-14
4.7 DISTRIBUTION OF CONVENTIONALLY-OBSERVED CLOUD COVER
WHEN ASOS REPORTED CLEAR SKIES 4-14
4.8 DISTRIBUTION, IN PERCENT, BY STATION OF THE DIFFERENCE
IN PG CATEGORY OBTAINED FROM MIXED ASOS DATA AND
CONVENTIONAL OBSERVATIONS 4-15
4.9 PG CATEGORY OBTAINED WITH MIXED ASOS DATA COMPARED TO
PG CATEGORY OBTAINED WITH CONVENTIONAL OBSERVATIONS 4-16
4.10 DISTRIBUTION, IN PERCENT, BY STATION OF THE DIFFERENCE IN PG
CATEGORY OBTAINED FROM FULL ASOS DATA AND CONVENTIONAL
OBSERVATIONS 4-17
4.11 COMPARISON OF PG CATEGORY OBTAINED WITH FULL ASOS
DATA TO PG CATEGORIES OBTAINED WITH CONVENTIONAL
OBSERVATIONS 4-17
4.12 DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN
RURAL MIXING HEIGHTS BETWEEN MIXED ASOS DATA AND
CONVENTIONAL OBSERVATIONS FOR ALL STATIONS AND FOR
EACH STATION 4-19
4.13 DISTRIBUTION, IN PERCENT, OF THE AB SOLUTE DIFFERENCE IN
RURAL MIXING HEIGHTS BETWEEN FULL ASOS DATA AND
CONVENTIONAL OBSERVATIONS FOR ALL STATIONS AND FOR
EACH STATION 4-20
5.1 SOURCE CHARACTERIZATION 5-2
B. 1 COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR ALBANY, NY B-2
B.2 COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR ALBANY, NY B-4
IX
-------
LIST OF TABLES (Concluded)
Table
B.3 COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR KANSAS CITY, MO B-6
B.4 COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR KANSAS CITY, MO B-8
B. 5 COMPARISON OF HIGH-1 ST-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR MONTGOMERY, AL B-10
B.6 COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR MONTGOMERY, AL B-12
B.7 COMPARISON OF HIGH-1 ST-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR MILWAUKEE, WI B-14
B. 8 COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR MILWAUKEE, WI B-16
B.9 COMPARISON OF HIGH-1 ST-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR PENDLETON, OR B-18
B. 10 COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR PENDLETON, OR B-20
B. 11 COMPARISON OF HIGH-1 ST-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR TUCSON, AZ B-22
B. 12 COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES
WITH CONVENTIONAL OBSERVATIONS TO ESTIMATES USING
MIXED ASOS AND FULL ASOS FOR TUCSON AZ B-24
-------
1.0 INTRODUCTION
The Automated Surface Observing System (ASOS) is a real-time, automated weather
information system that replaces conventional human observations for recording near-surface
weather conditions. The first ASOS was installed in August 1991 at Topeka, Kansas. When all
the installations are complete, the automated systems will be operational at over 900 stations
throughout the United States, the Caribbean, and overseas military installations.
Comparisons between pre-ASOS and ASOS instrumentation and their differences are
discussed in length in the paper by Heim Jr. and Guttman (1997). In addition, NWS
instrumentation changed by varying degrees when stations switched from observed observations
to ASOS. For example, temperature was previously measured by the HO-83 hygrothermometer.
Studies show that the ASOS hygrothermometer measures cooler temperatures than the HO-83.
Also, the ASOS instruments also have a smaller diurnal temperature range. In a direct
comparison, there appears to be a non-linear relationship between the ASOS and non-ASOS
data. With respect to precipitation amount, the ASOS gage consistently undermeasured
precipitation during heavy rain and snow events. And, wind speed comparisons indicate that 5-
second averages of ASOS peak winds are lower than the pre-ASOS data.
While improving the efficiency in acquiring weather data, the ASOS system lacks the
observational ability of the human observer to spatially integrate some of the weather elements
over a large area. Two such elements are ceiling height and opaque cloud cover, which are
important in estimating atmospheric stability and mixing heights required for applications of
several regulatory and nonregulatory dispersion models.
This analysis examines the effect of changing from a conventional observer-based system
to an automated system on the concentration estimates from the Industrial Source Complex Short
Term (ISCST3) model. Section 2 discusses the methods used to report several of the weather
variables for both ASOS and conventional observations. Section 3 describes the station and time
periods used in the analysis. Section 4 discusses the preparation of the meteorological data for
ISCST3 using the meteorological preprocessor PCRAMMET, including special processing
required for this analysis, and the differences in the meteorological data sets for both the 'raw'
data and the processed data. Section 5 presents the ISCST3 model input (options, sources,
receptors) and Section 6 presents the effects on the concentration estimates from ISCST3.
Section 7 summarizes the findings and proposes additional studies to further quantify the
differences that are seen in this study.
1-1
-------
2.0 COMPARISON OF ASOS AND CONVENTIONAL OBSERVATION METHODS
In this section, characteristics of the ASOS data observations that have an affect on the
concentration estimates are presented and are compared to conventional observation methods.
2.1 Cloud Cover and Height
To determine cloud cover and height, ASOS utilizes a vertical-pointing ceilometer. The
ceilometer is pulsed about 770 times per second to create a profile every 12 seconds and is
sampled every 30 seconds. A cloud "hit" or "no hit" is stored along with one or two cloud base
heights. Since the ceilometer has a vertical range of only 12,000 feet, any clouds above this
height are not detected. The cloud cover is determined from the ratio of "hits" to "possible hits"
in the previous 30 minutes (with the last 10 minutes given twice as much weight). The system
then converts the ratio to descriptive terms and, in the process, loses the information on the
"hits" to "possible hits" ratio. Since the ceilometer currently cannot distinguish between
transparent and opaque clouds, the fraction derived is recorded as opaque cloud cover. Reported
cloud base heights are the most frequent and meteorologically significant during the period, with
up to three layers reported for each hour.
To obtain the ceiling height under the conventional method, the human observer
evaluates the trace on a chart that records the amount of light reflected by the cloud layer. The
observer then views the sky to obtain cloud layer and cloud amount which is an instantaneous,
areally integrated average over the celestial dome at the time of the observation. This view also
allows the observer to modify the ceiling height if there are high clouds, and to make a
determination on the transparency of the clouds.
2.2 Temperature
One-minute average temperature is computed from 30-second samples. The ASOS
ambient and dew point temperatures are five-minute averages computed from the one-minute
processed data. The conventional method is an instantaneous reading at the observation time,
(taken 5 to 10 minutes before the hour).
2.3 Wind
For ASOS, wind speed and direction are collected once per second and an average is
computed every five seconds. The five second data are rounded to the nearest degree and knot.
A running two-minute average is computed from the five-second data. The two-minute average
is stored each minute. This running two-minute average is reported at the observation time. The
conventional method for reporting the winds is to take a one-minute average at the time of the
observation.
2-1
-------
3.0 STATION AND DATE SELECTION
3.1 Station Selection
Six stations across the United States were selected for this analysis. The selection was
based on the availability of co-located ASOS and conventional observations and climatological
regime. Since there is only a two month period available for each station, a cross-section of
climates was chosen to determine if there is any seasonal trend in changing from one
observational system to another. The surface stations, latitude and longitude, and the
corresponding upper air stations used in this analysis are:
Surface Station (airport ID) Latitude. Longitude Upper Air Station
Albany, New York (ALB) 42.75N, 73.SOW Albany, New York
Montgomery, Alabama (MGM) 32.30N, 86.40W Birmingham, Alabama
Kansas City, Missouri (MCI) 39.32N, 94.72W Topeka, Kansas
Tucson, Arizona (TUS) 32.12N, 110.93W Tucson, Arizona
Milwaukee, Wisconsin (MKE) 42.95N, 87.90W Green Bay, Wisconsin
Pendleton, Oregon (PDT) 45.68N, 118.85W Spokane, Washington
3.2 Date Selection
Since the data provided for this analysis were in 8-day segments and due to the
complexity of the data structure, the analysis was limited to the following periods:
September 18, 1994 through October 18, 1994 (31 days)
March 6, 1995 through April 5, 1995 (31 days)
The 1995 data sets for Kansas City and Milwaukee contained extended periods of
missing data. The 1995 data sets for these two cities were not used in this analysis.
5-1
-------
4.0 METEOROLOGICAL DATA PREPARATION AND ANALYSIS
Prior to running the ISCST3 model, both the surface data and mixing height data were
manipulated and reformatted prior to use in PCRAMMET. The steps required to reformat the
data and the assumptions made in preparation for input to PCRAMMET are discussed. Handling
of missing data and the running of PCRAMMET are also discussed in this section.
4.1 Hourly Surface Observations
One of the surface observation formats that PCRAMMET can process is the TD-1440
format from the National Climatic Data Center (NCDC). In this format, the weather variables
for each hour are coded in an 80-character record. The ASOS hourly surface observations that
were provided for this analysis required considerable manipulation prior to getting the data into
the TD-1440 format for use in PCRAMMET. The effort to reformat the data is described next.
4.1.1 ASOS data
Each file consisted of one record without any carriage return or line feed characters and
contained both conventional and ASOS observations for many of the first-order stations in the
United States. Briefly, the steps that were followed to reformat the data to the TD-1440 format
were:
1) retrieving the data from NCDC via anonymous ftp (8 files);
2) using a 'hex' editor, search for and replace the Az (control/z) character in the body
of the file with a blank; this character indicates an end-of-file, however, several of
the files contained one or more occurrences of this character prior to the actual
end of the file;
3) a special character indicated the end of a character string (some multiple of 80);
this special character was replaced with a CR/LF to form individual records;
4) extract only ASOS data from this newly created file;
5) extract the ASOS data for those stations that are to be included in the analysis;
6) reformat the ASOS data to the TD-1440 format.
The ASOS reports as received from NCDC consist of descriptive sky conditions for up to
three cloud levels rather than explicitly reporting ceiling height and cloud cover (in tenths or
octals) as is done in conventional observations. These sky conditions include the height of the
cloud layer in hundreds of feet and a description of the cloud layer. The descriptions are SCT,
BKN, and OVC for scattered, broken, and overcast, respectively. A fourth description is CLR
BLO 120 for clear skies below 12,000 feet. Since ASOS observations are limited to 12,000 feet,
any clouds above this level are not detected and clear skies are reported for the hour. There can
be one, two, or three descriptive conditions per hourly report. For example, the sky conditions
might be reported as
43 SCT M90 BKN
4-1
-------
which is interpreted as "the first layer is a deck of scattered clouds at 4,300 feet and the second
layer is a broken cloud layer measured (the 'M' before the 90) at 9,000 feet." The following
sections describe the conversion of the ASOS report to numeric values.
4.1.1.1 Sky conditions and cloud cover. Four possible descriptive sky conditions are
defined by the National Weather Service as follows:
clear less than 10% sky cover
scattered 10% to 50% sky cover
broken 60% to 90% sky cover
overcast more than 90% sky cover
A cloud layer is defined as 'thin' if the ratio of transparent to total cloud coverage is 1A or more.
This designation is not utilized for ASOS data since the ceilometer currently cannot distinguish
between transparent and opaque clouds. Hence, any distinction between total cloud cover and
opaque cloud cover is lost when a station converts to ASOS.
For this analysis, these descriptive elements were converted to fractional cloud cover as
follows:
clear 0.0
scattered 0.3
broken 0.7
overcast 1.0
If there is only one descriptive sky condition in an ASOS report, then the fractional cloud cover
shown above will be used to construct the meteorological input to PCRAMMET. See Section
4.1.1.3 for a discussion on cloud cover when multiple sky conditions are reported.
In estimating the stability category at night, the stability changes when cloud cover
changes from 4/10 to 5/10. This point can cause a difference in the stability estimate depending
on whether ASOS or conventional data are used. Appendix A contains the complete discussion
for estimating stability (EPA, 1987).
To see how ASOS cloud cover and conventional opaque cloud cover observations
compare, conventional cloud cover expressed in tenths was converted to ASOS' descriptive
four-category scale so the two can be compared on a category basis. A cloud cover of 0/10 was
assigned to clear skies, cloud coverage of 1/10 to 5/10 was assigned to scattered clouds, cloud
coverage of 6/10 to 9/10 was assigned to broken clouds, and 10/10 was assigned to overcast
skies. The results of this grouping and comparison are shown in Table 4.1.
4-2
-------
TABLE 4.1
CLOUD COVER COMPARISON
Conventional
Cloud Cover
Clear (0/10)
Scattered (1/10-5/10)
Broken (6/10-9/10)
Overcast (10/10)
Total ASOS
ASOS Cloud Cover
Clear
2459
1424
451
181
4515
Scattered
36
340
224
116
716
Broken
2
83
,255
191
531
Overcast
2
22
278
1376
1678
Total
Conventional
2499
1869
1208
1864
7440
From Table 4.1, it can be seen that there is a tendency for ASOS to underestimate the
cloud cover when compared to conventional observer data. For example, when ASOS reports
clear skies, there are 1424 occurrences of scattered clouds (1/10 - 5/10) reported by the
conventional observer. There are four cases of ASOS reporting broken or overcast skies when
the observer reported clear skies. There was no pattern to these events and these differences
appear to be valid. Three of the cases occurred at Montgomery in March 1995, two of which
were shortly before sunrise, and the fourth case was at Pendleton in October 1994. In the ASOS
cases, clear skies are followed immediately by extensive cloud cover. The observer at Pendleton
did not report any significant clouds for several hours. For the two periods just before sunrise at
Montgomery, the observer reported extensive clouds two hours after ASOS began reporting
clouds. Similar events occurred on other days, but the ASOS and observer reports were in better
agreement. For the third hour at Montgomery, both ASOS and the observer reported periods of
clearing mixed with periods of clouds.
4.1.1.2 Ceiling height. Currently, the ASOS ceilometer used to measure cloud base
height can detect clouds at or below 12,000 feet. It is anticipated that satellites may provide data
above 12,000 feet, but this information is not routinely available. Therefore, any high level
clouds are not included in an ASOS observation.
The ceiling height is defined as the height of the lowest layer of clouds that covers more
than 50% of the sky. From the sky conditions in Section 4.1.1.1, this definition means only
'broken' or 'overcast' are used to determine ceiling height for single descriptive sky conditions.
A sky condition of 'scattered' and 'clear' indicates that the ceiling height is unlimited.
For example, if only one layer is reported and is 9,000 feet and scattered, then the cloud
cover was coded as 0.3 with an unlimited ceiling. If the report is 7,000 feet and broken, then the
cloud cover was coded as 0.7 and the ceiling height as 7,000 feet.
In estimating the stability category during the day, the stability changes for a ceiling
height between 7,000 feet and 16,000 feet. Due to the limitation of 12,000 feet, this range can
cause a difference in the stability estimate depending on whether ASOS or conventional data are
used. Appendix A contains the complete discussion for estimating stability (EPA, 1987).
-------
4.1.1.3 Multiple cloud layers. Many hourly observations contain multiple descriptive
sky conditions. Since the automated system converts the heights and ratio of "hits" to "possible
hits" to the descriptive elements, converting multiple descriptions back to ceiling height and
cloud cover for use in dispersion modeling is not without complications. For example, if two
broken layers completely cover the celestial dome, an observer may report cloud cover of 1.0,
but this information cannot be determined with certainty from an automated report. Without
extensive effort, any method to convert back to fractional cloud cover and ceiling height in the
presence of multiple layers, or even single layers, is subject to error. For this analysis, when
multiple cloud layers are present, the single element that yields the largest cloud cover and the
single element that yields the lowest ceiling height were used. By not attempting to integrate
multiple pieces of information, this procedure may compensate for the lack of total versus
opaque clouds that a conventional observer-based system can provide, but again only for clouds
below 12,000 feet.
For example, if the ASOS report indicates a broken layer at 6,000 feet and overcast at
8,000 feet, then 1.0 (10/10) is used for the cloud cover and the ceiling height is 6,000 feet
associated with the broken layer. For a report of 5,500 feet and scattered and 6,500 feet and
broken and 8,000 feet and broken, then the ceiling height is 6,500 feet with a cloud cover of 0.7.
As an aside, NCDC currently is using a similar method for single-level cloud coverage as
is used here. The cloud cover associated with the descriptive conditions are defined by NCDC
as follows (the coverage used in this analysis follows in parentheses):
clear 0.0 (0.0)
scattered 0.2 (0.3)
broken 0.7 (0.7)
overcast 1.0 (1.0)
NCDC has not made any attempt thus far to integrate multiple cloud layers into a fractional
coverage.
4.1.2 Conventional observer data
The Meteorological Database Management System (MDMS) is a data base of hourly
meteorological data for the first-order stations in the United States. It resides on the EPA's IBM
mainframe computer. An interface allows the user to easily retrieve data from the data base.
Before using the data in PCRAMMET, there are several steps to reformat the meteorological
data:
1) Run MDMS for each station and for each year (once for 1994 and a second time
for 1995) on the IBM mainframe computer;
2) Download the data from the IBM mainframe;
3) Reformat the data for use in PCRAMMET - the data from the IBM are in a format
unique to the MDMS system which includes periods (".") to indicate missing
data.
The data obtained from MDMS for this project are the conventionally observed (i.e.,
human) data. Minimal data manipulation was required to convert the data into the TD-1440
format.
4-4
-------
4.2 Twice-daily Mixing Heights
Upper air data for the six upper air stations listed in Section 2 for the 2 month period in
1994 and 1995 were retrieved from the Radiosonde Data of North America, 1990-1995 compact
discs (CDS). This data served as a basis for computing the twice-daily mixing heights. The
morning and afternoon mixing heights were estimated following the technique developed by
Holzworth (1967) using the soundings and the surface temperature and pressure obtained from
the MDMS hourly observations.
4.3 Periods of Missing Data
The ASOS data had several hours of missing data for each station. There were no
missing periods of MDMS data. There were a few twice-daily mixing heights that could not be
calculated and required interpolation. The number of missing data periods and how the data
were filled in are discussed below.
4.3.1 Surface data
As noted earlier, the data for both Kansas City and Milwaukee for 1995 period were
omitted from this analysis due to extended periods of missing data. The ASOS data for the other
stations for both years suffered from several missing periods of data, as shown in Table 4.2.
Most missing data were for one- or two-hour periods and were filled by linearly interpolating the
meteorological data between the nonmissing periods before and after the missing period.
The exception to the linear interpolation is cloud cover. Since there are only four values
corresponding to the four sky conditions, interpolation could yield a nonexistent cloud cover in
the system defined in Section 4.1.1.1 and create a problem with the ceiling height. In most
cases, the cloud cover before and after the period of missing data was identical, thus there was
no decision to be made. However, there were a few cases where the cloud cover was different
for the hour preceding the missing period from the hour after. If there was a one-category
change (e.g., cloud cover changed from 3 to 7), the larger cloud cover was used and the ceiling
height associated with that larger cloud cover was used. If there was a two-category change (i.e.,
0 to 7 or 3 to 10), then category between the two (i.e., 3 or 7) was used. For the former (cloud
cover=3), the ceiling height was defined as unlimited; for the latter, the ceiling for the overcast
condition (cloud cover=10) was used.
There was one 8-hour period of missing data for the monthly period at Albany in 1995.
The data for the same 8-hour period from the previous day was used to fill in for the missing
data. The temperatures, however, were adjusted to to be more in line with the temperatures
immediately before and after the 8 hours of missing data.
4-5
-------
TABLE 4.2
SUMMARY OF MISSING PERIODS FOR ASOS DATA.
Station
Albany
Kansas City
Montgomery
Milwaukee
Pendleton
Tucson
Length of missing
period
1-hr
2-hr
other
1-hr
2-hr
1-hr
2-hr
1-hr
2-hr
1-hr
2-hr
1-hr
2-hr
Number of missing periods
1994 (1 month)
1
1
0
14
0
4
0
2
1
5
1
6
0
1995 (1 month)
3
0
1 (8 hr)
4
0
7
0
2
0
4.3.2 Mixing heights
Any missing morning mixing heights were filled by linearly interpolating between
periods of nonmissing morning mixing heights. A similar procedure was used to fill in for any
missing afternoon mixing heights. Table 4.3 shows the number of twice-daily mixing heights
that could not be computed and were filled by linear interpolation.
4-6
-------
TABLE 4.3
SUMMARY OF MISSING PERIODS OF TWICE-DAILY MIXING HEIGHTS.
Station
Albany
Kansas City
Montgomery
Milwaukee
Pendleton
Tucson
Missing period
morning
afternoon
morning
afternoon
morning
afternoon
morning
afternoon
morning
afternoon
morning
afternoon
Number of missing mixing heights
1994 (1 month)
0
0
0
0
2
1
0
0
0
0
0
0
1995 (1 month)
0
1
1
0
0
0
0
0
4.4 Running PCRAMMET
With the meteorological data formatted to the requirements for PCRAMMET,
PCRAMMET was run in its simplest mode since this analysis was to run the ISCST3 model
without deposition or depletion (both wet and dry). Thus, the only information required by
PCRAMMET was the type of estimate to be performed by ISCST3, the output type, input file
names (hourly surface observations and twice-daily mixing heights), the format of the hourly
surface data, and the location of the station. An example of the input runstream for Albany is
shown below:
N No/Dry /Wet Deposition calculations
A Unform/ASCII output type
mix94m.alb Mixing height data file
asos-94.alb Hourly surface data file
CD 144 Surface data format
42.750 Station latitude (decimal degrees)
73.800 Station longitude (decimal degrees)
5 Station time zone
The 31-day period from each year was run separately and the results were concatenated
to form a single meteorological input file for ISCST3.
4.5 Analysis of Meteorological Data
4-7
-------
The meteorological data that are input to and output from PCRAMMET were compared.
Input data included wind speed, wind direction, cloud cover, ceiling height, and ambient
temperature. Output data included Pasquill-Gifford (PG) stability category and rural and urban
mixing heights. Two sets of input to PCRAMMET were examined: 1) cloud cover and ceiling
from ASOS and winds, temperature, and pressure from conventional observations (referred
hereafter as "mixed ASOS"); and 2) all meteorological variables from ASOS (referred to as "full
ASOS"). In the former, the effects of cloud cover and ceiling are isolated without the effects of
winds and temperature.
4.5.1 Input to PCRAMMET
The histograms in Figures 4.1 - 4.3 show the difference between conventionally observed
data and automated data for wind speed, wind direction, and temperature for all stations and
hours (N=7440). Each bin represents a range of values and the values on the x-axis correspond
to the endpoints of each bin, with the left endpoint included in the bin and the right endpoint
excluded. The frequency in percent is shown on the y-axis with the number of observations
shown above each bar. The number of exact matches (i.e., a difference of zero) is shown in
parentheses next to the bar containing 0. The mean difference for wind speed, wind direction,
and temperature is -0.24 m s"1, -3.5 degrees, and -0.63 K, respectively. In other words, the ASOS
observations tend to be less than the conventionally observed counterpart.
4-8
-------
Wind Speed
40
30
o
c
0)
3
O"
0>
20
10
2862^(1814)
1497
530
94
16
1256
907
240
27
-6-5-4-3-2-1012345
Full ASOS - Conv Obs (meters/second)
Figure 4.1 Difference in wind speed between ASOS observations and conventional (human)
observations for all six stations (Number of observations = 7,440).
4-9
-------
o
c
0)
3
O"
0)
Wind Direction
50
40
30
£ 20
10
3446 (1625)
2628
397
36 40
75
119
446
106
32 34
-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180
Full ASOS - Conv Obs (degrees)
Figure 4.2 Difference in wind direction between ASOS observations and conventional
(human) observations for all six stations (Number of observations = 7,440).
4-10
-------
Ambient Temperature
40
30
o
c
0)
3
O"
0>
20
10
2542 2522
233
20
1921 (1497)
162
29
-6-5-4-3-2-1012345
Full ASOS - Conv Obs (K)
Figure 4.3 Difference in ambient temperature between ASOS observations and conventional
(human) observations for all six stations (Number of observations = 7,440).
4-11
-------
Tables 4.4 and 4.5 show the distribution of the difference in wind speed and wind
direction by station, the station abbreviations can be found in Section 3.1. The columns labeled
ALL6 are the distributions, in percent, that appear in Figures 4.1 and 4.2 for all stations
combined. Overall, the distributions between stations are very similar, although there are some
small differences in the tails of the distributions.
TABLE 4.4
DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN WIND SPEED
BETWEEN ASOS AND CONVENTIONAL OBSERVATIONS FOR ALL STATIONS AND
FOR EACH STATION.
Range
-7 < AWS < -6
-6 < AWS < -5
-5 < AWS < -4
-4 < AWS < -3
-3 < AWS < -2
-2 < AWS < -1
-1 < AWS<-0
0 < AWS < 1
1 < AWS < 2
2 < AWS < 3
3 < AWS < 4
4 < AWS < 5
5 < AWS < 6
ALL6***
0.03
0.04
0.22
1.26
7.12
20.12
16.88
38.47
12.19
3.23
0.36
0.05
0.03
ALB*
0.07
0
0.07
1.75
13.04
30.71
15.73
31.12
5.04
2.02
0.34
0.07
0.07
MCI**
0
0.13
0.27
1.48
6.85
19.89
20.97
36.69
10.89
2.55
0.27
0
0
MKE**
0
0
0.13
1.08
4.57
19.62
19.76
41.67
11.69
1.48
0
0
0
MGM*
0
0
0.13
0.34
4.57
14.85
16.87
42.67
14.52
5.51
0.40
0.13
0
PDT*
0
0
0
0.34
2.15
9.68
12.03
47.92
22.24
4.84
0.67
0.07
0.07
TUS*
0.07
0.13
0.67
2.62
10.15
25.60
19.42
31.45
7.86
1.75
0.27
0
0
* M=
N=l,488 observations
** N=744 observations
***N=7,440 observations
4-12
-------
TABLE 4.5
DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN WIND DIRECTION
BETWEEN ASOS AND CONVENTIONAL OBSERVATIONS FOR ALL STATIONS AND
FOR EACH STATION.
Range
-180 < AWD<-150
-150 < AWD<-120
-120< AWD<-90
-90 < AWD < -60
-60 < AWD < -30
-30
-------
Another variable that is required by PCRAMMET is the opaque sky cover. With the
12,000-foot limitation of the ceilometer, it is possible that ASOS reports clear skies when a
nonzero cloud cover is reported by a human observer. Table 4.7 shows the distribution of the
cloud cover reported by the human observer for those periods when ASOS reported clear skies.
The human observer reported clear skies for 2,459 cases of the 4,495 cases (or 55%) when
ASOS reported clear skies below 12,000 feet, but the human observer reported nonzero cloud
cover for 2,036 cases (45%). In 181 cases, ASOS reported clear skies while the human observer
reported overcast (10/10) skies. Although the number of cases is not exceedingly large, the
highest concentration differences may be driven by these cases, as will be seen in Section 6.
TABLE 4.7
DISTRIBUTION OF CONVENTIONALLY-OBSERVED CLOUD COVER WHEN
ASOS REPORTED CLEAR SKIES.
Cloud Cover
(tenths)
ALB*
MCI**
MGM*
MKE**
PDT*
TUS*
Total***
0
365, .
291
374
147
568
714
2459
1
99
24
96
21
107
130
477
2
65
29
64
21
40
88
307
O
56
16
86
22
50
62
292
4
31
10
20
22
40
56
179
5
9
16
33
17
36
38
149
6
19
7
39
13
24
31
133
7
19
4
15
12
28
28
106
8
7
16
29
13
10
39
114
9
8
5
27
6
22
30
98
10
19
10
59
29
20
44
181
Total
Non-
zero
332
137
468
176
377
546
2036
*N=1,488 observations
**N=744 observations
***N=7,440 observations
4-14
-------
4.5.2 Output from PCRAMMET
The two basic output parameters generated by PCRAMMET are the Pasquill-Gifford
stability category and the rural and urban mixing heights. Figure 4.4 shows the difference
between PG categories obtained from mixed ASOS and conventional observations. There is no
change for nearly 90% of the hours. For the non-zero differences, ASOS cloud cover overall
tends to increase the stability. Table 4.8 shows the distribution of the difference in PG category
by station. The column labeled ALL6 is the distribution, in percent, that appears in Figure 4.4
for all stations combined. Overall, the distributions between stations are very similar.
TABLE 4.8
DISTRIBUTION, IN PERCENT, BY STATION OF THE DIFFERENCE IN PG CATEGORY
OBTAINED FROM MIXED ASOS DATA AND CONVENTIONAL DATA.
APG
-2
-1
0
1
2
ALB*
0.30
2.22
93.08
4.50
0
MCI"
0.13
4.17
88.58
7.12
0
MKE"
0
2.15
86.96
10.35
0.54
MGM*
0.94
3.90
84.61
10.48
0.07
PDT*
0.20
1.61
90.52
7.53
0.13
TUS*
0.20
2.15
90.19
7.46
0
ATT s~***
ALL6
0.32
2.61
89.23
7.74
0.09
* M=
N=l,488 observations
** M=
=744 observations
*** M=
N=7,440 observations
A two-way distribution by PG category for all stations and hours is shown in Table 4.9.
The shaded cells indicate agreement between ASOS-based and observer-based stability
categories. About 11% of the differences are nonzero and range from 1% for category 7 to 20%
for category 5.
4-15
-------
TABLE 4.9
PG CATEGORY OBTAINED WITH MIXED ASOS DATA COMPARED TO PG
CATEGORY OBTAINED WITH CONVENTIONAL DATA.
ConvObs
1
2
O
4
5
6
7
All
1
^i'-'^rftvi
i
i
23
2
1
36
19
521
3
11
S^&jSs
sopite
97
897
ASOS
4
2
41
41
4
3078
5
234
14
1247
6
4
207
5
1220
7
1
82
454
Total
Nonzero
1
14
78
354
249
100
5
801
All
22
479
867
3344
1248
1104
376
7440
Figure 4.5 shows the difference between PG categories obtained from conventional and
full ASOS observations. There is no change for about 70% of the hours. For the non-zero
differences, ASOS data tends to increase the stability. Table 4.10 shows the distribution of the
difference in PG category by station. The column labeled ALL6 is the distribution of all stations
combined, in percent, that appears in Figure 4.5. Overall, the distributions between stations are
very similar. Compared to the distributions in Table 4.8, there are 10-20% fewer zero
differences with a cooresponding increase in 1 -category differences, although Tuscon showed a
larger drop in zero differences.
TABLE 4.10
DISTRIBUTION, IN PERCENT, BY STATION OF THE DIFFERENCE IN PG CATEGORY
OBTAINED FROM FULL ASOS DATA AND CONVENTIONAL DATA.
APG
O
-2
-1
0
1
2
3
ALL6***
0.03
0.52
12.19
71.34
14.46
1.41
0.04
ALB*
0
0.20
7.86
77.55
13.84
0.54
0
MCI**
0
0.67
13.17
70.70
13.17
1.75
0
MKE**
0
0.13
5.78
75.81
16.53
1.34
0.40
MGM**
0.13
1.14
12.57
70.90
13.91
1.34
0
PDT*
0
0.34
16.06
73.99
9.07
0.54
0
TUS*
0
0.54
14.72
61.02
20.63
3.09
0
* M=
N=l,488 observations
** XT=
=744 observations
N=7,440 observations
4-16
-------
The distribution by PG category for all stations and hours is shown in Table 4.11. The
shaded cells indicate agreement between ASOS-based and observer-based stability categories.
The percentage of nonzero differences ranges from 20% for category 4 to 44% for category 5.
TABLE 4.11
COMPARISON OF PG CATEGORY OBTAINED WITH FULL ASOS DATA
TO PG CATEGORIES OBTAINED WITH CONVENTIONAL DATA.
ConvObs
1
2
O
4
5
6
7
All
1
43
6
2
60
2
13
177
25
565
3
80
^fKfmiW-;'
271
925
ASOS
4
6
107
;VV->' :>;Vj'.'-W -j^lf i%^%$
137
5
2930
5
3
311
188
3
1203
6
57
374
91
1242
7
3
39
191
') £ ';'•'• •::* "
-------
unlimited ceiling and clear skies. A similar situation existed on the 25th. The sky conditions
yielded a neutral atmosphere with the conventionally observed data and a stable atmosphere with
the ASOS observed data.
The large negative differences of urban mixing heights are also the result of stability, but
between sunset of the previous day and sunrise of the current day. If the atmosphere is stable for
a particular hour in this period, the interpolation is between the previous day's maximum and the
current day's minimum mixing height. However, if the atmosphere is neutral for the hour, then
the interpolation is between the previous day's maximum and the current day's maximum. If the
atmosphere switches stability from hour to hour, then the method of determining the stability
category (stable or neutral) switches also. Again, the largest negative differences (< -2500
meters) in Figure 4.6 are for Tucson, with many of those differences occurring on September 20.
Conventional observations reported the atmosphere as overcast with 12,000-13,000 foot ceilings
but the ASOS reported clear skies and unlimited ceiling for the hours between midnight and
sunrise.
Table 4.12 shows the distribution of the absolute difference in rural mixing heights
between mixed ASOS and conventional observations. There is a tendency for positive
differences in eastern U.S. climates and negative differences in western U.S. climates. For
Pendleton, however, there were no differences for the 1,488 hours of data and for Tucson, all the
differences are negative (ASOS-based estimates are smaller). The reasons for this trend is not
known and would require analyses with additional stations and periods of data to determine if
the trend is actual or coincidental.
4-18
-------
TABLE 4.12
DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN RURAL MIXING
HEIGHTS BETWEEN MIXED ASOS DATA AND CONVENTIONAL OBSERVATIONS
FOR ALL STATIONS AND FOR EACH STATION.
Range
-4000 < AZ: < -3500
-3500 < AZ: < -3000
-3000 < AZ: < -2500
-2500 < AZ: < -2000
-2000 < AZ:<-1500
-1500 < AZ:<-1000
-1000 < AZ: < -500
-500 < AZ: < 0
0 < AZ: < 500
500 < AZ:< 1000
1000 < AZ:< 1500
1500 < AZ:<2000
ALL***
0.01
0.01
0.03
0.05
0.11
0.30
0.59
0.71
97.98
0.11
0.08
0.01
ALB*
0
0
0
0
0
0.07
0.54
0.40
98.19
0.34
0.40
0.07
MCI**
0
0
0
0
0
0.27
0.94
0.67
97.72
0.40
0
0
MKE**
0
0
0
0.13
0.27
0.67
1.48
3.63
93.82
0
0
0
MGM*
0
0
0
0
0.13
0.67
0.94
0.87
97.38
0
0
0
PDT*
0
0
0
0
0
0
0
0
100.0
0
0
0
TUS*
0.07
0.07
0.13
0.20
0.27
0.27
0.27
0.13
98.59
0
0
0
* M=
N=l,488 observations
** M=
=744 observations
*** M=
N=7,440 observations
Table 4.13 is similar to Table 4.12 except that full ASOS data were used to estimate
mixing heights. There is less tendency for a trend between parts of the country, although the
differences for Tucson are still all negative and there only five nonzero differences for
Pendleton.
4-19
-------
TABLE 4.13
DISTRIBUTION, IN PERCENT, OF THE ABSOLUTE DIFFERENCE IN RURAL MIXING
HEIGHTS BETWEEN FULL ASOS DATA AND CONVENTIONAL OBSERVATIONS FOR
ALL STATIONS AND FOR EACH STATION.
Range
-4000 < AZ: < -3500
-3500 < AZ: < -3000
-3000 < AZ: < -2500
-2500 < AZ: < -2000
-2000 < AZ: < -1500
-1500
-------
PG Stability
100
o
c
0)
3
O"
0)
80
60
40
20
-3
Figure 4.4
-2
-1 0 1
Mixed ASOS - Conv Obs
Difference in PG stability category between mixed ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440).
4-21
-------
PG Stability
o
c
0)
3
O"
0)
100
80
60
40
20
-3
39
907
5308
1076
105
-2-101 2
Full ASOS - Conv Obs
Figure 4.5 Difference in PG stability category between full ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440).
4-22
-------
Rural Mixing Ht
0)
100
80
60
40
20
1 1
J7290 46639)^
8 22 44 53
-4000 -3000 -2000 -1000 0 1000
Mixed ASOS - Conv Obs (meters)
2000
Figure 4.6 Difference in rural mixing heights between mixed ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440).
4-23
-------
o
c
0)
3
O"
0)
Urban Mixing Ht
100
80
60
40
20
7113 (7018)
6 4 2 11 5 15 35 70
158
15 3 2 1
-5000 -4000 -3000 -2000 -1000 0 1000 2000 3000
Mixed ASOS - Conv Obs (meters)
Figure 4.7 Difference in urban mixing heights between mixed ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440).
4-24
-------
o
c
0)
3
O"
0)
Rural Mixing Ht
100
80
60
40
20
7193 (7166)
2 2 3 9 17 36 72 81
14 9 2
-4000 -3000 -2000 -1000 0 1000 2000
Full ASOS - Conv Obs (meters)
Figure 4.8 Difference in rural mixing height between full ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440).
4-25
-------
Urban Mixing Ht
100
80
60
o
c
0)
3
O"
0)
40
20
6932 (6741)
6 6 5 16 8 25 55 97
235
33 9 6 4 0 1 0 1 0 1
-5000-4000-3000-2000-1000 0 1000 2000 3000 4000 5000 6000
Full ASOS - Conv Obs (meters)
Figure 4.9 Difference in urban mixing heights between full ASOS observations and
conventional (human) observations for all six stations (# observations = 7,440).
4-26
-------
5.0 ISCST3 MODEL INPUT
The model input options, the sources data, and the receptor network are presented in this
section.
5.1 Model Options
For all the ISCST3 runs, the following model options were implemented:
DFAULT - implements all the regulatory default options
RURAL - utilize rural dispersion coefficients
FLAT - run without the effects of terrain elevation
Only ambient concentrations were estimated; there was no dry or wet deposition or depletion
options in effect.
Concentrations were estimated for 1-hr, 3-hr, 8-hr, and 24-hr averages. Period averages
were also computed. For Albany, Montgomery, Pendleton, and Tucson, the period was about
two months. For Kansas City and Milwaukee, the period is one month because the data for 1995
were not used due to extended periods of missing meteorological data.
5.2 Source Characterization
A standard set of source parameters was used in this analysis that have been used in
previous sensitivity tests by the EPA. These sources and parameters are shown in Table 5.1. All
sources were located at or centered on the origin (0,0) of the receptor network described in the
next section.
Building downwash effects were modeled for two point sources. For the 55-meter point
source, the building dimensions were specified to utilize the Huber-Snyder algorithms. For the
35-meter point source, the building dimensions were specified to exercise the Schulman-Scire
algorithms.
5-1
-------
TABLE 5.1
SOURCE CHARACTERIZATION
Point Sources
Source
ID
PI
P2
P3
P4
P5
P6
Stack
height
(m)
10.0
35.0
35.0
55.0
100.0
200.0
Stack diameter
(m)
2.4
2.4
2.4
2.4
4.6
5.6
Exit velocity
(ms-1)
11.7
11.7
11.7
11.7
18.8
26.5
Exit temp.
(K)
432
432
432
432
416
425
Bldg. ht.
(m)
NM
NM
34
34
NM
NM
Bldg.
width (m)
NM
NM
60
60
NM
NM
Emission rate
(gs4)
100.0
100.0
100.0
100.0
100.0
100.0
Volume Source
Source
ID
VOL1
Release
height
(m)
35.0
Initial lateral
dimension (m)
14.0
Initial vertical
dimension (m)
16.0
Emission rate
(gs4)
100.0
Area Source
Source
ID
AREA1
Release
height
(m)
0.01
Area
(m2)
250,000
Length
(m)
500
Width
(m)
500
Emission rate
(g s'1 m-2)
0.0004
NM = not modeled
5.3 Receptor Network
A polar grid network was utilized for all sources. The network consisted of 36 radials,
from 10° to 360° every 10° at distances (in meters) of 100, 500, 1000, 1500, 2000, 2500, 3000,
4000, 5000, 10000. There are 360 receptors with this network. The receptors at 100 meters are
inside the area source.
5-2
-------
6.0 ISCST3 MODEL RESULTS
In analyzing the results from ISCST3, attention was focused on the high-lst-high (H1H)
concentrations, with the results unpaired in space and time. Two sets of comparisons were
made. The first, and the primary goal of this analysis, compares the design concentration
estimates using conventionally-observed meteorological data (the MDMS data) to the design
concentration estimates using ceiling height and cloud cover obtained from ASOS and all other
variables (winds and temperature) the same as the conventionally-observed data. The second
compares the design concentration estimates using conventionally-observed meteorological data
to the design concentration estimates obtained using all ASOS data, including temperature and
winds.
6.1 ASOS Cloud Cover vs Conventional Cloud Cover Observations
When the H1H concentration estimates using mixed ASOS data (ASOS cloud
information with conventionally observed winds and temperature) are compared to estimates
based entirely on conventionally observed data, nearly 60% of the comparisons for all
combinations of averaging period/source/station show no difference. Figures 6.1-6.8 show the
relative difference (computed as the difference between the ASOS-based estimate and the
conventional estimate divided by the conventional estimate) with positive values indicating that
the ASOS-based H1H estimates are higher. The nonzero differences generally are less than
about 10%, with the H1H concentrations obtained with the mixed ASOS data generally higher
for each averaging period and source type except for the volume source. Of the nonzero
differences, 66 (27.5%) were associated with positive differences and 35 (14.5%) were
associated with negative differences. The tables in Appendix B show the H1H and high-2nd-
high (H2H) concentration estimates, absolute difference, and relative difference by station,
source, and averaging period.
There are, however, several cases where the difference is much larger: 50% or more, as
seen in Figures 6.2, 6.5, and 6.6. These differences occurred for different stations, but all
occurred under convective conditions. Why did changing to ASOS-based cloud data make such
a large difference? The 1-hour average for the 100-meter source (Point Source 5, Figure 6. le)
for Albany was 50% higher than the estimate using conventionally observed data and was
examined more closely to determine what factor(s) caused such a large difference. Since the
H1H results are unpaired, the date and time associated with the higher concentration determines
which data to pair together for the case comparison. Thus, the input meteorology and derived
parameters in ISCST3 (such as plume height and oz) for the date and time of the ASOS-based
result were compared to the conventionally observed meteorology and derived parameters for the
same date and time. The following table, for October 14, 1994 at 1200 Local Standard Time
(LST), shows the differences in the meteorology:
6-1
-------
ASOS
Conventional
Cloud
Cover
0/10
10/10
Ceiling
Height
(feet)
Unlimited
14,000
PG
Category
2
4
°z
(meters)
308
85
Plume
Height
(meters)
372
318
Mixing
Height
(meters)
378
378
Concentration
(H.g m"3)
60.6
0.326
The ASOS-based observation reported an unlimited ceiling with no cloud cover, whereas
the conventionally-based observation with the human observer indicated an overcast sky at
14,000 feet. Using the ASOS-based data resulted in a moderately unstable atmosphere (PG
category 2), whereas the use of the conventional data resulted in a neutral stability (PG category
4). The difference in stabilities resulted in a oz that was 3.5 times larger with the ASOS data
than with the conventional data, and resulted in a concentration estimate 200 times larger. Note
that the plume height using the ASOS-based data is only six meters below the mixing height.
Had the plume height been seven meters higher, ISCST3 would have predicted zero
concentration using the ASOS-based data, since the plume would have been above the mixing
height. If that had occurred, there would have been no difference between the two H1H
estimates. This comparison clearly demonstrates the potential effect of limited cloud
information on concentration estimates.
As noted above, for the volume source, there was a tendency for the concentration
estimates to be lower using the mixed ASOS observations than the concentration using
conventional observations. The 1-hour average, paired in time and in space, for Montgomery
(see Figure 6.8) was examined in more detail. The H1H concentration occurred at 6 a.m. and the
mixing height was 35.1 meters at Montgomery. The release height for the volume source was 35
meters. The PG stability category using the mixed ASOS data was 5. With this stability, there is
unlimited mixing in ISCST3. Using conventional observations, the stability was category 4,
which results in limited mixing in ISCST3. The concentrations using the conventional
observations was 13885 |ig m"3 and 4628 |ig m"3 for the ASOS data. With the release height just
below the mixing height, the result is higher concentrations using conventional observations.
6-2
-------
40
35
30
^p
^,25
$> 20
£
I15
JS
-------
6-4
-------
Point Source 2: H1H
Relative Difference (%)
->• N> CO -&. Ol CD
DOOOOOOC
1 1 1 1 1 1 1
U
Ps
\
\
/ \
/ \
\
v \
- ^ - Albany
- 0 ~ Kansas City
- A - Milwaukee
- ^ - Montgomery
- X ~ Pendleton
- ^ - Tucson
x \
v \
v \
^ \
^ \
\ \
\ \
\ \
\ \
\ ' \ ^
\ \ ^ ^[
' ^ — "
^ ^ -A ' - ' ' ' ^^"^v
H
ft
A
1-hr 3-hr 8-hr 24-hr Period
Averaging Period
Figure 6.2 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 35-meter point source.
6-5
-------
Point Source 3: H1H
HVJ
OC
OO
on
OU
o^ 95
^, ZO
0)
o
c
fl) Of>
»I t-v
£
Qin
o>15
"^
re
TIT -| r\
a) iu
a:
c
A
^
/ \
/ x
& . '
/ Kv"-":* ^
/ / x
/ / N
-;
x
/
/
- 3> ~ Albany
- E] ~ Kansas City
- A - Milwaukee
- ^ - Montgomery
- X - Pendleton
- ^ - Tucson
•N
/
/
/
/
/
N /
\A
*-""•'': 4
x "" ^ - 2
X •* ^. *""— — •!
-^
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.3 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 35-meter point source with building
down wash.
6-6
-------
Point Source 4: H1H
tu
oc
GO
Of)
OU
?9^
P_^ ^O
Q)
O
c
fl) or)
« ^U
£
° 15
G) 15
"«P
(0
"/!» 1 n
Q) IU
o:
c
IK"
V,
/ V
-X
/ v.
X
\ ' #
\ ' \
\ ' N
\ ' N
\ \ ' ^
\ \> ^
\ /\ A
X/ x N
/X v V
/ \ . ^
/ % , /* v
' V -' ^
/ * x
' i X
\ ''
/ \ '
X /
J&L nnt1 C^Sl
- <$> - Albany
- E] ~ Kansas City
- A ~ Milwaukee
- "w" - Montgomery
- X ~ Pendleton
— -^ - Tucson
x
/
/
/
X
^ / X
\ / ^ E
- . - . ^
x \ :.— ^ ^ - 4
- JK<- ~ "
^^
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.4 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 55-meter point source with building
down wash.
6-7
-------
Point Source 5: H1H
ou
-
oc
OO
on
/it OU
0
c
d)
1
O on
zu
0)
"J
M< *1 C
re 1 0
0)
o:
i n
1 U
z
^r
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
_ — ** \
^ - \
\ \
\ ^
\
\
\ X
>«•
- <6> - Albany
- 0 ~ Kansas City
- A ~ Milwaukee
- ^ - Montgomery
— "V — Ppnrllptnn
- ^ - Tucson
-> J5
•^
x ^
1
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.5 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 100-meter point source.
6-8
-------
Po^nt Source 6: H1H
uu
cc
OO
en
OU
AC
*tO
*""^ /in
^p 4U
0^
fi\
Q) oc
Q OO
c
0)
L. on
1
Qoc
ZD
0)
•TT on
+^ ZU
_re
0)
^> /) C
U. 1 0
i n
1 U
t;
\
\
\
\
V
\
\
\
\
\
\
\
\
\
\
A
^
w
\
\
\
\
\
\
\
\
\
\ ^ "
^ - "
- Q - Albany
- 0 - Kansas City
- & - Milwaukee
- ^ - Montgomery
- X ~ Pendleton
- ^ - Tucson
i
/
' J-
' /
/ /
/ /
/ '
/ ' J
'' ' '
«/'*** N
COfe — — — — 7
,,-Mr ----- ^
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.6 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the 200-meter point source.
6-9
-------
Area Source: H1H
35
30
<1>
o
c
0)
a,
15
10
0
- ^ - Albany
- 0 - Kansas City
- A ~ Milwaukee
- "w" - Montgomery
- X ~ Pendleton
- Jk - Tucson
A
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.7 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the area source.
6-10
-------
Volume Source: H1H
40
35
30
T 25
o 20
g
it 15
0)
~ 10
JS
0)
0
-5
-10
- A ~
- X ~
Albany
Kansas City
Milwaukee
Montgomery
Pendleton
Tucson
A
", ' ^X
N. X
X /
1-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.8 Relative difference between the high-lst-high concentrations using the mixed
ASOS data and conventional data for the volume source.
6-11
-------
6.2 Full ASOS vs Conventional Observations
When the concentration estimates using the full ASOS observation (clouds, winds, and
temperature) are compared to the concentration estimates using the conventional data, a different
picture emerges. The relative difference increases dramatically. Only 13 (5.5%) of the
comparisons for all combinations of averaging period/source/station show no difference, while
125 (52%) show that the ASOS-derived H1H estimates are higher and 102 (42.5%) of the
conventional-derived estimates are higher. There is no discernible trend by station or by
averaging period, although the 1-hour average for the volume source (Figure 6.16) at
Montgomery shows a large difference just as in the mixed ASOS results. The concentration
estimates by station, averaging period, and source are shown in the tables in Appendix B.
To determine what factors may contribute to large differences, two cases were examined
in more detail. The first is for Point Source 6 (200-meter release), 1-hr average for Montgomery
in which the ASOS-derived H1H estimate was 55% higher than the estimate using conventional
data. For this comparison, the input meteorology and derived parameters in ISCST3 for the date
and time of the ASOS-based result (April 1, 1995 at 1200 LST) were compared to the
conventionally observed meteorology and derived parameters for the same date and time.
ASOS
Conventional
Cloud
Cover
0/10
1/10
Ceiling
Height
(feet)
Unltd.
Unltd.
Wind
Speed
(knots)
3
12
PG
1
3
°z
(meters)
1982
127
Plume
Height
(meters)
1170
417
Mixing
Height
(meters)
1323
1323
Concentration
(Hg m"3)
35.61
0.0012
Unlike the case study with the mixed ASOS data, the ceiling is unlimited in both cases
and the cloud cover is very similar. However, the wind speed is four times larger for the
conventionally observed data than for the ASOS data, resulting in a two-category difference in
the PG stability. The PG stability category for a wind speed of 3 knots and a net radiation index
of 4 is category 1, but for a wind speed of 12 knots and net radiation index of 4, the stability
category is 3. (The only way to obtain these two categories with these wind speeds is with net
radiation index of 4 (EPA, 1987)). This wind speed difference, in turn, yields significantly
different plume heights, but more importantly, the oz obtained from the ASOS data is more than
an order of magnitude larger than the oz obtained from the conventional data.
The second case examined is for Point Source 5 (100-meter release), 1-hr average for
Milwaukee in which the ASOS-derived H1H estimate was 30% lower than the estimate using
conventional data. Since the H1H results are unpaired, the higher concentration determines the
date and hour to pair together for the case comparison. Thus, the input meteorology and derived
parameters in ISCST3 for the date and time of the conventionally-based concentration were
compared to the ASOS data for the same date and time. The meteorology for this case, which is
for October 13, 1994 at 1300 LST, is as follows:
6-12
-------
ASOS
Conventional
Cloud
Cover
0/10
5/10
Ceiling
Height
(feet)
Unltd.
Unltd.
Wind
Speed
(knots)
8
7
PG
2
3
°z
(meters)
178
373
Plume
Height
(meters)
317
368
Mixing
Height
(meters)
433
433
Concentration
(Hg m"3)
17.95
48.85
Here, the observed meteorology is about the same but the stability is one category
different. The 5/10 conventionally-observed cloud cover and 0/10 ASOS cloud cover do not
affect the stability determination since Turner's method (EPA, 1987) only modifies stability if
the cloud cover is greater than 5/10. The stability difference results from the wind speed
difference, only this time, the ASOS observed wind is larger. This 1-knot increase in the ASOS
wind speed produced a oz that is a factor of two smaller and a concentration that is about a factor
of three smaller compared to using the conventionally-observed data.
6-13
-------
Point Source 1: H1H
(Full ASOS)
~7f\
1 U
on
OU
en
OU
0
c
0)
™T on
fl) OU
fc
Q
^ ^U
JS
0)
U_ /) n
1 U
i n
- 1 U
X
\
A
n
- ^ - Albany
- 0 ~ Kansas City
- A - Milwaukee
- ^V - Montgomery
- X ~ Pendleton
- ^ - Tucson
\
/
^ s'
-'-'-Vl'
- - - ' s"
X
^^
\
\
/
/ x
/ \
/ ^
/ \
/ \
/ \
/ \
/ \
X X
' \ fi
' ^ *
/ t\
' ^ / V
/ £> '^
^ N
\ X X /
\ ' \ /
/ /
/ ^ \ ^ ' \
/ ' ^ / ^
^^ ^ ^_^__--A^ \x^^
•*• ^ " \x
^ *' ^-l®
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.9
Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the 10-meter point source.
6-14
-------
Point Source 2: H1H
(Full ASOS)
Relative Difference (%)
rO-li ->• INJ CO -&• Ol CO ~
oooooooooc
1 1 1 1 1 1 1 1 1
N
- <$> - Albany
- 0 ~ Kansas City
- A - Milwaukee
- ^ - Montgomery
- X ~ Pendleton
- -^ - Tucson
X
^
/ v
X
\
\
7
\
/
' \
/ X
, \ 1 x
^ / xA
' $•".!" "A x^ ' ' ""X
' ' X ~~ "~ 1ST , N f
/ / 1 X
\ \ /
\s / < ^ x
A \ ^ x /
\ / x^ XQ /
/N^ " X / ^_ - - ^X/-^ x
X " fe ~~ ^ /x-7
\xX-^" ' '^ ^ x - ^ ""A
x x x ""N^Rn
^ ^ s'- - - ^>' *
vO - x
0
1-hr
ZA
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.10 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the 35-meter point source.
6-15
-------
Point Source 3: H1H
(Full ASOS)
tu
on
OU
1 n
8 10
C
s
£ 0
M— U
Q
0)
"o>
a:
on
-OU
/in
/,
o-
/ A'
A '
x ^w /
/7 //x x/v
!// / N^
"==-"Si^-.-
- <3> - Albany
- 0 ~ Kansas City
- A - Milwaukee
- ^ - Montgomery
- X ~ Pendleton
- -^ - Tucson
/ \ / \
7 "* Mx / v
' \ ^ / \ /
v v . / v
\ v /N x \ / ^ A
/ \ / **^
^ /V\ SA''/'\ ^
Vr ' \ \ ' " " "^ ^
"/ ^ - ' ^ 'B ?'
^ \ >. \ ' X
\ \ /
\ 'W'
\ x
\ ^ X
\ ^
\ '
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.11 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the 35-meter point source with building
down wash.
6-16
-------
Point Source 4: H1H
(Full ASOS)
/ U
on
OU
en
OU
<*^
«^o
^
O
c
fl) on
2 OU
0) 2°
J
Tu in
0) 1 U
£
H n
- 1 U
on
\
& '
*:
X
- <$> - Albany
- EH ~ Kansas City
- A - Milwaukee
- ik - Montgomery
- X ~ Pendleton
- -^ — Tucson
__ - '
^ ^ "^
^ \- "
\
\
~~ -~ __ /
-. /
\ ^
' -x
**/ N^
""^ ^V^ 'C
\
><
\
/ ^
/ \
/ \
/ \
' v
-^r X x
' "* X N ^ ^
\ \ /
X N x x
'
/ \ ^
x x v
^ jf ^
— — A
^ ^ " " " ^""^V
*
- - A~ ~ ~ ~ -x-*
^ X - - "
^ X
1-hr
3-hr
8-hr
2^hr
Period
Averaging Period
Figure 6.12 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the 55-meter point source with building
down wash.
6-17
-------
50
40
30
8 20
E 10
Q
.1
IS o
-10
-20
-30
Point Source 5: H1H
[Full ASOS)
/N
Nr
\
\
\
\
- <$> - Albany
- El - Kansas City
- A - Milwaukee
- i^ - Montgomery
- X - Pendleton
~ ^ - Tucson
•*• x
\ / x
\ / ^.
- -yX
-<
sfa
1-hr
8-hr
24-hr
Perioc
Averaaina Period
Figure 6.13 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the 100-meter point.
6-18
-------
Point Source 6: H1H
(Full ASOS)
60
50
40
0)
O 20
£
£
o
> 0
0)
-30
p;
- A-
Albany
Kansas City
Milwaukee
Montgomery
Pendleton
Tucson
A
/ \
A
A
V -*£."
\
\
X- -
X
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.14 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the 200-meter point source.
6-19
-------
Area Source: H1H
en
DU
en
OU
yin
4-U
-*-». ^.n
*T» OU
vP
0^
<1>
O on
C 20
£
0)
H— in
M— 1 U
Q
0)
> n
._ u
•+j
JS
0)
f£ m
••^ - 1 u
on
-z\j
on
-ou
/in
(Full ASOS)
\
\
\
\
\
\
\
\
\ ^ - *•••-•
v _^ _"_"__- ^> ^
\s
' ' \ ^
^ x >
^ X \
~~ \^" " X
^
x X ^
/ A - -^
^
/ ^ ^
/
/ .
7 "" S<- - " "
/ x X
/ /
'iW —
~~ - ^
- ^> - Albany
- 0 ~ Kansas City
- A - Milwaukee
- ^ - Montgomery
^^ PanHlatnn
/\ rcllUlcluil
- -^ - Tucson
" ^D - - x '
^^ -t
-,^-r-""
'" V ------- "
2^S "- -
>
X' ~ "
£
*l
T^
^
i^
K
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.15 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the area source.
6-20
-------
Volume Source: H1H
(Full ASOS)
60
50
O 20
s
£
0)
£ °
JS
0)
-40
- ^> - Albany
- 0 ~ Kansas City
- A - Milwaukee
- ^ - Montgomery
- X ~ Pendleton
- lucson
%
w
x
^
XX
-x
- X
-x-
X
1-hr
3-hr
8-hr
24-hr
Period
Averaging Period
Figure 6.16 Relative difference between the high-lst-high concentrations using the full ASOS
data and conventional data for the volume source.
6-21
-------
7.0 CONCLUSIONS
This study demonstrates that the weather variables observed with an automated system
can produce very different concentration estimates from the ISCST3 dispersion model when
compared with concentrations derived with conventional (human observer) data. Differences in
the meteorology, both the input to and output from PCRAMMET, were presented. General
differences that the air quality modeling community might expect from a steady-state Gaussian
model such as the ISCST3 were discussed, along with several case studies of extreme
differences.
The study was divided into two parts: 1) isolating the effect of ASOS-derived cloud data
on concentration estimates, and 2) the effect of a full ASOS observation (including winds,
temperature, and pressure) on concentration estimates. In the first part, the high-lst-high
concentrations were examined. More than half of the 240 comparisons [(6 stations) x (5
observation periods) x (8 sources)] produced no difference, while most of the nonzero
differences were 10% or less. Overall, the ASOS-based concentration was higher for all sources,
except the volume source. However, trends by station or averaging period could not be
discerned. There were several cases where the ASOS-based concentrations were 35% or more
larger than the conventionally-based concentration. In a case study for Albany, this difference
was attributed directly to the 12,000-foot limit of the ASOS ceilometer.
In the second part, there were only 13 pairs with no difference, 125 differences where the
ASOS-based concentration was larger than the conventional-based concentration, and 102
differences where the conventional-based concentration exceed the ASOS-based concentration.
No clear trend could be discerned in these results.
These results are based on 2 months (late summer/early fall, later winter/early spring) of
meteorological data, which was readily available. This data limitation impacts the amount of
information that can be gathered. Should more data should become available for longer periods
of time, additional conclusions may result.
7-1
-------
8.0 REFERENCES
Heim, Richard R., Jr. and N. B. Guttman, 1997: On Computing 1971-2000 Climate Normals in
the ASOS Era. Presented at the American Meteorological Society 10th Conference on Applied
Meteorology, October 20-23, 1997, Reno, Nevada.
Holzworth, G. C., 1967: Mixing Depths, Wind Speeds and Air Pollution Potential for Selected
Locations in the United States. J. Appl. Meteorol., 6, 1039-1044.
Turner, D. B., 1964: A Diffusion Model for an Urban Area. J. Appl. MeteoroL, 3, 83-91.
U.S. Environmental Protection Agency, 1977: User's Manual for Single-Source (CRSTER)
Model. EPA-450/2-77-013. Research Triangle Park, North Carolina, 27711.
U.S. Environmental Protection Agency, 1987: On-Site Meteorological Program Guidance for
Regulatory Modeling Applications (Revised). EPA-450/4-87-013. Research Triangle Park,
North Carolina, 27711.
3-1
-------
APPENDIX A
ESTIMATING ATMOSPHERIC STABILITY IN PCRAMMET
The Turner (1964) method for estimating atmospheric stability is implemented in
PCRAMMET. For those periods when the sun is above the horizon, the stability is a function of
wind speed and a net radiation index, which is a function of solar elevation, cloud cover (CC),
and ceiling height (CLHT). The net radiation index is determined from the following procedure
(this description is taken directly from On-Site Meteorological Program Guidance For
Regulatory Modeling Applications (EPA, 1987); the italics are added to identify direct
quotation):
1. If the total cloud' cover is 10/10 and the ceiling is less than 7000 feet, use net
radiation index equal to 0 (whether day or night).
2. For nighttime (from one hour before sunset to one hour after sunrise):
(a) If total cloud cover <_ 4/10, use net radiation index equal to -2.
(b) If total cloud cover > 4/10, use net radiation index equal to -1.
3. For daytime:
(a) Determine the insolation class number as a function of solar altitude from
Table 6-5.
(b) If total cloud cover <5/10, use the net radiation index in Table 6-4
corresponding to the insolation class number.
© If cloud cover >5/10, modify the insolation class number using the
following six steps.
(1) Ceiling < 7000ft, subtract 2.
(2) Ceiling >7000ft but <16000ft, subtract 1.
(3) total cloud cover equal 10/10, subtract 1. (This will only apply to
ceilings >7000ft since cases with 10/10 coverage below 7000ft
are considered in item 1 above.)
(4) If insolation class number has not been modified by steps (1), (2),
or (3) above, assume modified class number equal to insolation
class number.
Although Turner indicates total cloud cover, opaque cloud cover is implied by Pasquill and is
preferred.
A-l
-------
(5) If modified insolation class number is less than 1, let it equal 1.
(6) Use the net radiation index in Table 6-4 corresponding to the
modified insolation class number.
Table 6-4
Turner's Key to P-G Stability Categories
Wind Speed
(knots) (m/s)
0,1
2,3
4,5
6
7
8,9
10
11
> 12
Table 6-5
0-0.7
0.8-1.8
1.9-2.8
2.9-3.3
3.4-3.8
3.9-4.8
4.9-5.4
5.5-5.9
>6.0
Insolation Class as a Function
Solar Altitude
60 < 0
35 < 0 < 60
15 < 0 < 35
< 15
0 (degrees)
Net Radiation Index
43210-1 -2
112
122
123
223
223
233
334
334
344
of Solar Altitude
Insolation
strong
moderate
slight
weak
3467
3467
4456
4456
4445
4445
4445
4444
4444
Insolation Class Number
4
3
2
1
A-2
-------
APPENDIX B
COMPARISONS OF THE H1H AND H2H CONCENTRATION ESTIMATES
This appendix contains the comparisons of the high-lst-high (H1H) and high-2nd-high
(H2H) concentrations for each station. There two tables for each station: the first contains the
H1H concentrations, and the second contains the H2H concentrations. The data are grouped in
each table by averaging period. For each averaging period, the data are reported by source
category. Source identifiers can be found in Table 5.1. For each averaging period, the ASOS
data are grouped according to whether the estimate used the mixed ASOS or full ASOS data.
The following definitions apply in these tables:
Conv Obs = concentration estimated using conventional data (|ig m"3)
Mixed ASOS = concentration estimated using mixed ASOS data (|ig m"3)
Abs Diff = Mixed ASOS - Conv Obs (jig m'3)
Rel Diff = ((Mixed ASOS - Conv Obs) / Conv Obs) * 100 (percent)
Full ASOS = concentration estimated using full ASOS data (|ig m"3)
Abs Diff = Full ASOS - Conv Obs ftig m'3)
Rel Diff = ((Full ASOS - Conv Obs) / Conv Obs ) * 100 (percent).
B-l
-------
TABLE B.I
COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR ALBANY, NY
Source
Conv Obs
(H.§ m 3)
Mixed ASOS Abs Diff Rel Diff
Gigm3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Oigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
484.15
205.66
3094.12
327.29
40.32
21.36
9080.98
48505.05
484.15 0.00 0.0
205.66 0.00 0.0
3094.12 0.00 0.0
327.29 0.00 0.0
60.59 20.27 50.3
21.36 0.00 0.0
9080.98 0.00 0.0
48505.05 0.00 0.0
484.15 0.00 0.0
206.23 0.57 0.3
3096.04 1.92 0.1
438.96 111.68 34.1
60.58 20.27 50.3
22.78 1.43 6.7
9067.07 -13.91 -0.2
48505.05 0.00 0.0
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
451.65
202.23
1521.23
249.14
27.41
11.84
5890.81
34759.97
451.65 0.00 0.0
202.23 0.00 0.0
1521.23 0.00 0.0
249.14 0.00 0.0
27.41 0.00 0.0
11.84 0.00 0.0
5890.81 0.00 0.0
34759.97 0.00 0.0
432.08 -19.58 -4.3
160.78 -41.45 -20.5
2082.01 560.78 36.9
259.97 10.83 4.3
21.61 -5.80 -21.2
8.51 -3.32 -28.1
5933.61 42.80 0.7
39383.71 4623.74 13.3
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
265.29
114.99
908.43
181.20
14.43
5.25
3328.60
30549.24
265.29 0.00 0.0
114.99 0.00 0.0
908.43 0.00 0.0
181.20 0.00 0.0
14.43 0.00 0.0
5.25 0.00 0.0
3328.60 0.00 0.0
30549.24 0.00 0.0
261.35 -3.94 -1.5
93.52 -21.47 -18.7
1252.55 344.12 37.9
158.17 -23.03 -12.7
12.74 -1.69 -11.7
4.59 -0.66 -12.5
4173.47 844.87 25.4
35197.43 4648.19 15.2
B-2
-------
TABLE B.I, CONCLUDED
Source
Conv Obs
(H.§ m 3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Oigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
140.94
66.44
876.93
129.10
6.57
2.33
2148.34
20899.02
140.94 0.00 0.0
66.44 0.00 0.0
870.67 -6.26 -0.7
128.94 -0.16 -0.1
6.57 0.00 0.0
2.33 0.00 0.0
2077.97 -70.37 -3.3
21565.50 666.49 3.2
169.20 28.26 20.1
63.84 -2.60 -3.9
559.39 -317.54 -36.2
113.00 -16.10 -12.5
5.67 -0.90 -13.7
1.77 -0.56 -24.1
2966.77 818.43 38.1
21351.89 452.87 2.2
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
20.03
10.22
99.77
16.91
0.78
0.14
531.32
9031.12
20.01 -0.02 -0.1
10.10 -0.12 -1.2
102.04 2.27 2.3
17.14 0.23 1.3
0.77 -0.01 -0.7
0.14 0.00 -0.8
525.15 -6.18 -1.2
9095.42 64.30 0.7
17.33 -2.70 -13.5
7.84 -2.38 -23.3
81.21 -18.56 -18.6
14.82 -2.10 -12.4
0.59 -0.19 -24.2
0.09 -0.05 -36.1
614.48 83.15 15.7
10579.92 1548.80 17.1
-------
TABLE B.2
COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR ALBANY, NY
Source
Conv Obs
(H.§ m 3)
Mixed ASOS Abs Diff Rel Diff
Gigm3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Oigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
472.88
201.04
2751.82
305.80
34.95
15.02
9025.45
48505.05
472.88 0.00 0.0
201.04 0.00 0.0
2751.82 0.00 0.0
316.27 10.47 3.4
31.06 -3.89 -11.1
15.02 0.00 0.0
9025.45 0.00 0.0
48505.05 0.00 0.0
475.57 2.69 0.6
201.61 0.57 0.3
2750.13 -1.69 -0.1
318.33 12.53 4.1
39.20 4.26 12.2
12.25 -2.77 -18.4
8956.48 -68.97 -0.8
47726.78 -778.27 -1.6
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
350.33
141.86
1185.85
212.66
17.27
9.00
5047.89
32466.96
350.33 0.00 0.0
141.86 0.00 0.0
1185.85 0.00 0.0
212.66 0.00 0.0
17.27 0.00 0.0
9.00 0.00 0.0
4806.61 -241.28 -4.8
32466.96 0.00 0.0
370.94 20.61 5.9
152.94 11.08 7.8
1147.84 -38.01 -3.2
245.17 32.51 15.3
14.07 -3.19 -18.5
6.25 -2.74 -30.5
5184.10 136.21 2.7
36726.76 4259.80 13.1
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
218.11
109.27
904.65
146.38
8.36
4.47
2795.84
26388.61
218.11 0.00 0.0
109.27 0.00 0.0
904.65 0.00 0.0
146.38 0.00 0.0
8.37 0.01 0.1
4.47 0.00 0.0
2795.84 0.00 0.0
26388.61 0.00 0.0
212.34 -5.77 -2.6
78.11 -31.16 -28.5
662.73 -241.92 -26.7
139.52 -6.86 -4.7
8.51 0.15 1.8
2.90 -1.57 -35.1
3763.83 967.99 34.6
26662.97 274.36 1.0
B-4
-------
TABLE B.2, CONCLUDED
Source
Conv Obs
(H.§ m 3)
Mixed ASOS Abs Diff Rel Diff
Gigm3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Oigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
131.27
55.93
518.52
98.93
3.79
1.89
1873.87
16461.07
129.81 -1.46 -1.1
55.93 0.00 0.0
518.52 0.00 0.0
98.93 0.00 0.0
3.79 0.00 0.0
1.89 0.00 0.0
1832.61 -41.25 -2.2
16084.50 -376.58 -2.3
105.43 -25.84 -19.7
39.57 -16.37 -29.3
451.65 -66.87 -12.9
69.79 -29.14 -29.5
3.79 0.01 0.2
1.07 -0.81 -43.1
1870.63 -3.24 -0.2
20066.87 3605.79 21.9
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
18.97
10.05
88.44
16.76
0.63
0.10
481.22
9023.04
18.97 0.00 0.0
9.89 -0.16 -1.6
91.77 3.33 3.8
17.11 0.35 2.1
0.65 0.01 2.4
0.10 0.00 -0.1
474.73 -6.49 -1.3
9080.30 57.26 0.6
16.54 -2.43 -12.8
7.84 -2.22 -22.1
80.62 -7.82 -8.8
14.46 -2.30 -13.7
0.54 -0.09 -13.8
0.08 -0.02 -18.1
560.60 79.38 16.5
10499.23 1476.19 16.4
B-5
-------
TABLE B.3
COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR KANSAS CITY, MO
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
542.29
240.41
3131.58
275.53
51.32
23.14
9370.34
47301.91
542.29 0.00 0.0
357.10 116.69 48.5
3131.58 0.00 0.0
329.51 53.99 19.6
51.32 0.00 0.0
23.14 0.00 0.0
9420.04 49.70 0.5
47301.91 0.00 0.0
463.39 -78.90 -14.5
325.26 84.86 35.3
3113.11 -18.47 -0.6
329.51 53.99 19.6
47.79 -3.53 -6.9
21.70 -1.43 -6.2
9442.90 72.57 0.8
48608.82 1306.91 2.8
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
371.43
141.72
2096.40
212.54
19.24
7.86
5019.56
40844.07
371.43 0.00 0.0
233.38 91.65 64.7
2096.40 0.00 0.0
212.54 0.00 0.0
19.24 0.00 0.0
7.86 0.00 0.0
5019.56 0.00 0.0
40844.07 0.00 0.0
328.68 -42.75 -11.5
172.07 30.35 21.4
2035.17 -61.22 -2.9
245.95 33.41 15.7
21.68 2.44 12.7
11.69 3.82 48.6
5555.32 535.77 10.7
46245.97 5401.90 13.2
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
208.32
91.41
1144.78
132.89
12.28
4.48
3635.21
32626.49
208.32 0.00 0.0
133.86 42.45 46.4
1144.78 0.00 0.0
132.89 0.00 0.0
12.28 0.00 0.0
4.48 0.00 0.0
3492.36 -142.85 -3.9
32626.49 0.00 0.0
209.78 1.45 0.7
111.97 20.56 22.5
1075.15 -69.63 -6.1
139.53 6.64 5.0
12.97 0.69 5.6
5.69 1.21 26.9
3721.59 86.38 2.4
39774.34 7147.84 21.9
B-6
-------
TABLE B.3, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
109.90
57.92
577.13
82.13
4.40
1.56
2197.01
22058.43
109.90 0.00 0.0
57.92 0.00 0.0
577.13 0.00 0.0
82.13 0.00 0.0
4.40 0.00 0.0
1.56 0.00 0.0
2108.67 -88.33 -4.0
22611.53 553.10 2.5
121.08 11.18 10.2
59.90 1.98 3.4
564.63 -12.50 -2.2
92.13 10.00 12.2
4.82 0.42 9.5
2.07 0.51 32.7
2384.19 187.19 8.5
25739.81 3681.39 16.7
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
13.83
6.85
69.65
10.48
0.52
0.12
418.84
9796.43
14.55 0.73 5.3
7.18 0.33 4.8
72.19 2.54 3.6
10.93 0.46 4.3
0.54 0.02 4.6
0.12 0.01 5.2
418.08 -0.75 -0.2
9914.39 117.95 1.2
12.09 -1.73 -12.5
5.84 -1.01 -14.8
71.77 2.12 3.0
10.87 0.39 3.8
0.46 -0.05 -10.5
0.13 0.01 8.9
457.92 39.08 9.3
10886.16 1089.72 11.1
B-7
-------
TABLE B.4
COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR KANSAS CITY, MO
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
451.52
220.37
3060.63
268.39
33.43
17.59
7142.66
47272.51
451.52 0.00 0.0
319.76 99.39 45.1
3060.63 0.00 0.0
268.39 0.00 0.0
33.43 0.00 0.0
15.10 -2.49 -14.2
7142.66 0.00 0.0
47272.51 0.00 0.0
459.83 8.31 1.8
214.61 -5.76 -2.6
3050.21 -10.41 -0.3
268.48 0.09 0.0
37.24 3.81 11.4
16.84 -0.75 -4.3
9239.04 2096.38 29.4
48353.98 1081.47 2.3
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
293.55
132.21
1278.03
204.60
16.20
6.32
3943.42
38089.55
293.55 0.00 0.0
141.72 9.51 7.2
1278.03 0.00 0.0
204.60 0.00 0.0
16.20 0.00 0.0
6.11 -0.21 -3.3
3943.42 0.00 0.0
38089.55 0.00 0.0
267.64 -25.90 -8.8
126.33 -5.88 -4.4
1547.94 269.92 21.1
192.25 -12.34 -6.0
18.24 2.04 12.6
7.10 0.78 12.4
4274.32 330.90 8.4
44979.18 6889.63 18.1
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
144.29
67.21
706.18
126.40
7.83
2.46
2686.65
27066.23
158.78 14.49 10.0
70.86 3.65 5.4
706.18 0.00 0.0
126.40 0.00 0.0
7.83 0.00 0.0
2.46 0.00 0.0
2497.58 -189.07 -7.0
28687.34 1621.11 6.0
149.25 4.96 3.4
85.99 18.78 27.9
954.38 248.20 35.1
117.56 -8.83 -7.0
11.07 3.24 41.3
3.12 0.66 27.0
2976.26 289.61 10.8
32292.20 5225.97 19.3
-------
TABLE B.4, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
93.86
38.24
438.23
66.46
2.79
0.95
1653.60
18198.37
93.86 0.00 0.0
38.47 0.23 0.6
459.86 21.63 4.9
67.00 0.54 0.8
2.79 0.00 0.0
0.95 0.00 0.0
1642.13 -11.47 -0.7
18198.37 0.00 0.0
95.82 1.96 2.1
43.48 5.24 13.7
382.95 -55.28 -12.6
71.76 5.31 8.0
3.69 0.91 32.6
1.09 0.14 14.6
1726.01 72.41 4.4
20332.68 2134.32 11.7
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
13.42
6.83
65.71
10.34
0.49
0.12
410.47
9727.76
14.03 0.62 4.6
7.05 0.22 3.2
67.19 1.49 2.3
10.73 0.39 3.8
0.51 0.03 5.4
0.12 0.00 1.0
413.90 3.43 0.8
9838.43 110.67 1.1
11.96 -1.45 -10.8
5.82 -1.01 -14.9
68.03 2.33 3.5
10.21 -0.13 -1.3
0.45 -0.04 -7.7
0.11 0.00 -2.5
457.45 46.98 11.4
10678.65 950.89 9.8
B-9
-------
TABLEB.5
COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR MONTGOMERY, AL
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
402.07
264.68
3111.64
404.71
55.89
22.81
13885.39
70534.76
402.07 0.00 0.0
264.68 0.00 0.0
3120.33 8.68 0.3
404.71 0.00 0.0
60.51 4.62 8.3
37.03 14.22 62.4
13421.34 -464.05 -3.3
70534.76 0.00 0.0
400.54 -1.53 -0.4
240.13 -24.55 -9.3
3117.22 5.58 0.2
496.46 91.75 22.7
54.78 -1.11 -2.0
35.61 12.81 56.2
9742.98 -4142.41 -29.8
48592.17 -21942.59 -31.1
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
271.89
164.22
1925.89
197.46
21.98
10.13
7120.59
70227.18
271.89 0.00 0.0
164.22 0.00 0.0
2127.38 201.49 10.5
262.42 64.96 32.9
24.50 2.51 11.4
12.34 2.21 21.8
6302.63 -817.96 -11.5
70227.18 0.00 0.0
295.37 23.48 8.6
194.01 29.78 18.1
2260.34 334.44 17.4
254.89 57.44 29.1
27.45 5.47 24.9
11.87 1.74 17.1
5785.95 -1334.63 -18.7
45927.36 -24299.82 -34.6
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
176.13
108.11
1094.12
116.45
14.46
4.98
3541.93
38356.22
168.35 -7.78 -4.4
103.35 -4.76 -4.4
1179.21 85.09 7.8
146.97 30.52 26.2
13.89 -0.57 -3.9
4.73 -0.24 -4.9
4021.97 480.04 13.6
38356.22 0.00 0.0
217.02 40.89 23.2
122.13 14.02 13.0
1090.34 -3.78 -0.3
157.51 41.06 35.3
14.82 0.36 2.5
5.90 0.92 18.5
4356.89 814.96 23.0
34338.97 -4017.25 -10.5
B-10
-------
TABLE B.5, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
113.46
57.93
714.25
83.08
5.45
1.76
2464.76
21893.69
113.46 0.00 0.0
57.93 0.00 0.0
714.25 0.00 0.0
83.08 0.00 0.0
5.45 0.00 0.0
1.76 0.00 0.0
2535.72 70.96 2.9
22260.57 366.88 1.7
112.13 -1.33 -1.2
52.24 -5.69 -9.8
628.48 -85.77 -12.0
92.70 9.62 11.6
6.09 0.64 11.8
2.40 0.63 36.0
2554.53 89.77 3.6
23576.49 1682.80 7.7
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
13.11
6.74
76.72
11.96
0.58
0.16
678.91
12426.46
13.81 0.70 5.4
6.62 -0.12 -1.8
89.56 12.84 16.7
13.44 1.48 12.3
0.64 0.06 11.2
0.19 0.03 18.1
665.83 -13.08 -1.9
12685.42 258.96 2.1
16.71 3.59 27.4
8.49 1.75 26.0
98.92 22.20 28.9
15.73 3.77 31.5
0.74 0.16 27.9
0.22 0.05 32.3
719.06 40.16 5.9
12302.00 -124.45 -1.0
B-ll
-------
TABLE B.6
COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR MONTGOMERY, AL
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
358.54
216.16
3058.88
335.62
33.41
17.53
9044.54
70296.98
358.54 0.00 0.0
216.16 0.00 0.0
3058.88 0.00 0.0
335.62 0.00 0.0
42.79 9.38 28.1
20.86 3.33 19.0
9114.18 69.65 0.8
70296.98 0.00 0.0
391.47 32.93 9.2
223.13 6.96 3.2
3074.25 15.38 0.5
336.34 0.72 0.2
36.73 3.31 9.9
17.72 0.19 1.1
9114.03 69.49 0.8
48303.98 -21993.00 -31.3
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
229.96
143.71
1431.74
162.87
19.26
9.67
4642.91
49794.24
229.96 0.00 0.0
146.04 2.33 1.6
1511.42 79.68 5.6
167.56 4.69 2.9
20.30 1.04 5.4
9.67 0.00 0.0
4698.30 55.39 1.2
49794.24 0.00 0.0
252.00 22.05 9.6
151.69 7.98 5.6
1732.57 300.84 21.0
192.31 29.44 18.1
21.10 1.83 9.5
8.00 -1.67 -17.3
5244.81 601.90 13.0
40791.47 -9002.77 -18.1
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
162.88
103.09
850.32
107.51
13.16
4.25
3040.70
34940.96
162.88 0.00 0.0
101.87 -1.22 -1.2
958.57 108.25 12.7
113.14 5.63 5.2
13.16 0.00 0.0
4.25 0.00 0.0
3278.41 237.71 7.8
34940.96 0.00 0.0
156.38 -6.50 -4.0
101.56 -1.53 -1.5
1062.86 212.54 25.0
146.94 39.43 36.7
14.45 1.29 9.8
5.65 1.40 32.9
3863.80 823.10 27.1
32488.06 -2452.90 -7.0
B-12
-------
TABLE B.6, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
78.76
46.44
516.58
65.09
5.19
1.70
2429.36
20870.34
78.76 0.00 0.0
46.44 0.00 0.0
514.98 -1.59 -0.3
67.64 2.55 3.9
5.19 0.00 0.0
1.70 0.00 0.0
2464.76 35.40 1.5
21219.30 348.96 1.7
91.63 12.86 16.3
46.44 0.00 0.0
456.83 -59.74 -11.6
81.97 16.87 25.9
5.17 -0.03 -0.5
1.98 0.28 16.4
2171.26 -258.10 -10.6
20246.15 -624.19 -3.0
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
12.91
6.71
71.85
10.48
0.56
0.15
659.81
12352.01
13.62 0.71 5.5
6.53 -0.18 -2.6
74.32 2.47 3.4
11.16 0.68 6.5
0.64 0.08 13.4
0.18 0.03 19.8
642.08 -17.73 -2.7
12627.85 275.83 2.2
16.27 3.37 26.1
8.48 1.78 26.5
95.31 23.46 32.7
14.70 4.21 40.2
0.73 0.17 29.4
0.21 0.05 34.1
667.02 7.21 1.1
12165.54 -186.47 -1.5
B-13
-------
TABLE B.7
COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR MILWAUKEE, WI
Source
Conv Obs
(H.§ m'3)
Mixed ASOS Abs Diff Rel Diff
Gigrrr3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigrrr3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
488.17
233.78
3121.51
368.38
48.85
17.26
8772.27
48303.98
488.17 0.00 0.0
352.32 118.54 50.7
3134.49 12.97 0.4
368.38 0.00 0.0
48.85 0.00 0.0
17.26 0.00 0.0
8347.83 -424.44 -4.8
47189.77 -1114.21 -2.3
499.51 11.34 2.3
232.00 -1.77 -0.8
3116.64 -4.87 -0.2
369.36 0.98 0.3
34.41 -14.44 -29.6
17.82 0.56 3.2
8860.73 88.46 1.0
48641.17 337.19 0.7
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
349.70
191.14
1875.05
229.18
26.00
8.09
4053.01
45006.59
349.70 0.00 0.0
193.17 2.03 1.1
1875.05 0.00 0.0
229.18 0.00 0.0
26.00 0.00 0.0
8.09 0.00 0.0
3652.17 -400.85 -9.9
45006.59 0.00 0.0
360.38 10.67 3.1
143.14 -48.00 -25.1
2310.10 435.05 23.2
244.08 14.89 6.5
17.80 -8.20 -31.5
7.60 -0.49 -6.1
5102.92 1049.91 25.9
44031.21 -975.39 -2.2
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
306.06
124.85
788.92
200.99
13.62
4.47
2776.83
37113.01
328.80 22.75 7.4
138.99 14.14 11.3
908.43 119.51 15.1
201.22 0.23 0.1
13.62 0.00 0.0
4.47 0.00 0.0
2776.83 0.00 0.0
37113.01 0.00 0.0
288.85 -17.20 -5.6
115.81 -9.04 -7.2
1011.93 223.00 28.3
185.64 -15.35 -7.6
11.80 -1.82 -13.4
4.47 0.00 0.0
3009.17 232.34 8.4
37529.57 416.56 1.1
TABLE B.7, CONCLUDED
B-14
-------
Source
Conv Obs
(H.§ m'3)
Mixed ASOS Abs Diff Rel Diff
Gigrrr3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigrrr3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
134.23
54.95
517.45
89.05
4.75
1.55
1683.33
20058.30
141.81 7.58 5.6
59.66 4.71 8.6
541.90 24.45 4.7
89.11 0.07 0.1
4.75 0.00 0.0
1.55 0.00 0.0
1534.21 -149.11 -8.9
20058.30 0.00 0.0
129.44 -4.79 -3.6
53.05 -1.90 -3.5
551.69 34.24 6.6
86.78 -2.27 -2.5
4.59 -0.16 -3.4
1.74 0.18 11.8
1623.94 -59.39 -3.5
20009.35 -48.95 -0.2
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
12.27
5.62
66.44
11.70
0.40
0.10
306.90
9943.92
12.23 -0.03 -0.3
5.38 -0.24 -4.2
67.73 1.29 1.9
11.84 0.14 1.2
0.39 -0.01 -2.3
0.12 0.02 14.5
309.01 2.11 0.7
10325.45 381.52 3.8
10.32 -1.95 -15.9
4.95 -0.68 -12.0
76.29 9.86 14.8
11.22 -0.48 -4.1
0.47 0.06 15.7
0.11 0.00 3.4
337.81 30.91 10.1
10265.46 321.53 3.2
B-15
-------
TABLEB.8
COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR MILWAUKEE, WI
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
441.63
207.92
2773.37
338.36
29.22
11.84
5353.37
47189.77
441.63 0.00 0.0
210.14 2.22 1.1
2800.54 27.17 1.0
338.36 0.00 0.0
29.22 0.00 0.0
11.84 0.00 0.0
5316.37 -36.99 -0.7
46522.03 -667.73 -1.4
458.24 16.61 3.8
209.33 1.41 0.7
2776.88 3.52 0.1
338.36 0.00 0.0
31.84 2.62 9.0
11.74 -0.10 -0.9
6410.32 1056.95 19.7
48303.98 1114.21 2.4
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
343.59
143.51
1273.43
228.65
13.48
5.56
3264.00
41362.13
343.59 0.00 0.0
143.51 0.00 0.0
1273.43 0.00 0.0
228.65 0.00 0.0
14.12 0.65 4.8
5.56 0.00 0.0
3264.00 0.00 0.0
41362.13 0.00 0.0
338.05 -5.54 -1.6
141.26 -2.25 -1.6
1534.70 261.28 20.5
226.44 -2.20 -1.0
15.97 2.49 18.5
5.06 -0.50 -9.0
3436.52 172.52 5.3
35083.26 -6278.87 -15.2
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
161.67
81.00
751.57
128.02
7.30
2.55
2150.25
35945.66
161.67 0.00 0.0
81.00 0.00 0.0
749.11 -2.47 -0.3
128.02 0.00 0.0
7.32 0.02 0.3
2.55 0.00 0.0
2087.27 -62.99 -2.9
35945.66 0.00 0.0
127.95 -33.72 -20.9
56.31 -24.70 -30.5
987.31 235.73 31.4
106.14 -21.88 -17.1
7.11 -0.19 -2.6
2.52 -0.03 -1.3
2487.24 336.99 15.7
37410.01 1464.34 4.1
B-16
-------
TABLE B.8, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
101.64
52.31
359.77
83.49
3.72
1.13
1260.58
17558.32
98.42 -3.22 -3.2
49.37 -2.94 -5.6
346.48 -13.30 -3.7
84.35 0.86 1.0
3.62 -0.11 -2.9
1.13 0.00 0.0
1151.92 -108.66 -8.6
17558.32 0.00 0.0
61.35 -40.30 -39.6
26.76 -25.55 -48.8
428.88 69.11 19.2
63.72 -19.78 -23.7
2.48 -1.25 -33.5
1.01 -0.13 -11.3
1439.85 179.27 14.2
18237.12 678.79 3.9
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
12.11
5.50
65.57
10.36
0.34
0.10
300.58
9908.17
12.20 0.09 0.7
5.19 -0.31 -5.7
67.14 1.56 2.4
10.74 0.38 3.7
0.35 0.01 2.9
0.11 0.01 11.6
287.65 -12.93 -4.3
10255.73 347.56 3.5
10.29 -1.82 -15.0
4.81 -0.69 -12.6
75.66 10.09 15.4
11.06 0.70 6.7
0.45 0.11 30.9
0.10 0.00 1.8
314.58 14.00 4.7
10209.32 301.14 3.0
B-17
-------
TABLE B.9
COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR PENDLETON, OR
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
598.21
216.25
3126.51
369.36
39.54
20.50
13537.74
69369.86
598.21 0.00 0.0
216.25 0.00 0.0
3126.51 0.00 0.0
471.68 102.32 27.7
39.54 0.00 0.0
20.50 0.00 0.0
13513.59 -24.15 -0.2
69369.86 0.00 0.0
756.33 158.12 26.4
233.98 17.73 8.2
3134.49 7.98 0.3
362.12 -7.24 -2.0
41.87 2.33 5.9
20.50 0.00 0.0
9080.98 -4456.76 -32.9
110589.66 41219.80 59.4
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
386.22
147.73
1626.77
225.88
20.34
8.22
8570.55
47479.30
386.22 0.00 0.0
147.73 0.00 0.0
1626.77 0.00 0.0
225.88 0.00 0.0
20.34 0.00 0.0
8.22 0.00 0.0
8519.02 -51.53 -0.6
47479.30 0.00 0.0
388.35 2.13 0.6
193.78 46.05 31.2
1609.43 -17.34 -1.1
225.07 -0.81 -0.4
26.28 5.94 29.2
9.40 1.18 14.4
6225.58 -2344.97 -27.4
42907.42 -4571.88 -9.6
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
215.97
126.05
989.46
140.62
17.57
6.73
4525.16
36980.24
215.97 0.00 0.0
126.05 0.00 0.0
989.46 0.00 0.0
140.62 0.00 0.0
17.57 0.00 0.0
6.73 0.00 0.0
4525.16 0.00 0.0
36980.24 0.00 0.0
282.92 66.95 31.0
115.69 -10.36 -8.2
879.36 -110.11 -11.1
179.47 38.85 27.6
15.40 -2.16 -12.3
5.31 -1.41 -21.0
3508.63 -1016.52 -22.5
28507.95 -8472.29 -22.9
B-18
-------
TABLE B.9, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
125.88
52.94
508.78
96.74
5.27
2.02
1544.96
20943.94
125.88 0.00 0.0
52.94 0.00 0.0
508.78 0.00 0.0
96.74 0.00 0.0
5.27 0.00 0.0
2.02 0.00 0.0
1544.96 0.00 0.0
20943.94 0.00 0.0
214.78 88.90 70.6
90.05 37.11 70.1
688.15 179.37 35.3
157.99 61.24 63.3
5.92 0.65 12.3
1.77 -0.25 -12.2
1485.00 -59.97 -3.9
16902.25 -4041.69 -19.3
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
13.14
6.42
90.34
12.51
0.84
0.31
518.24
12469.97
12.84 -0.30 -2.3
6.30 -0.12 -1.9
93.05 2.72 3.0
12.90 0.38 3.0
0.85 0.02 1.9
0.31 0.00 0.6
501.65 -16.60 -3.2
12771.49 301.52 2.4
15.89 2.74 20.9
7.65 1.23 19.1
83.10 -7.24 -8.0
12.31 -0.20 -1.6
0.72 -0.11 -13.6
0.26 -0.05 -15.6
462.49 -55.75 -10.8
10507.61 -1962.36 -15.7
B-19
-------
TABLE B. 10
COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR PENDLETON, OR
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
443.47
213.92
3079.97
364.82
32.95
14.99
9067.07
67860.33
443.47 0.00 0.0
213.92 0.00 0.0
3094.12 14.15 0.5
364.82 0.00 0.0
32.95 0.00 0.0
14.99 0.00 0.0
9067.07 0.00 0.0
67860.33 0.00 0.0
504.03 60.55 13.7
213.92 0.00 0.0
3085.59 5.63 0.2
351.99 -12.84 -3.5
33.26 0.30 0.9
13.67 -1.32 -8.8
9025.45 -41.62 -0.5
48642.56 -19217.77 -28.3
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
288.90
135.09
1387.23
184.52
19.45
7.56
7563.98
42916.54
288.90 0.00 0.0
135.09 0.00 0.0
1387.23 0.00 0.0
184.52 0.00 0.0
19.45 0.00 0.0
7.56 0.00 0.0
7544.01 -19.97 -0.3
42916.54 0.00 0.0
302.20 13.30 4.6
163.99 28.90 21.4
1418.77 31.54 2.3
199.12 14.59 7.9
23.28 3.83 19.7
8.50 0.94 12.5
4676.19 -2887.80 -38.2
35815.73 -7100.81 -16.5
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
206.34
79.71
676.69
139.12
10.68
4.16
3990.63
34589.76
206.34 0.00 0.0
79.71 0.00 0.0
819.69 143.00 21.1
139.12 0.00 0.0
10.68 0.00 0.0
4.16 0.00 0.0
3990.63 0.00 0.0
34589.76 0.00 0.0
211.18 4.84 2.3
95.12 15.41 19.3
685.73 9.04 1.3
159.47 20.35 14.6
11.25 0.57 5.3
3.95 -0.21 -5.0
3159.75 -830.88 -20.8
27078.06 -7511.70 -21.7
B-20
-------
TABLE B. 10, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
111.35
45.01
379.85
78.57
3.50
1.44
1511.57
19168.63
111.35 0.00 0.0
45.01 0.00 0.0
379.85 0.00 0.0
78.57 0.00 0.0
3.50 0.00 0.0
1.44 0.00 0.0
1497.05 -14.52 -1.0
19599.36 430.73 2.2
145.53 34.17 30.7
61.61 16.60 36.9
395.05 15.19 4.0
95.50 16.94 21.6
4.43 0.93 26.5
1.51 0.07 5.1
1215.77 -295.80 -19.6
16543.30 -2625.32 -13.7
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
12.04
6.23
78.12
11.36
0.80
0.29
493.44
12405.17
11.72 -0.32 -2.7
6.19 -0.04 -0.7
79.77 1.65 2.1
11.45 0.09 0.8
0.81 0.01 1.2
0.30 0.00 1.4
494.60 1.16 0.2
12697.93 292.77 2.4
15.36 3.32 27.6
7.58 1.36 21.8
73.50 -4.62 -5.9
11.30 -0.05 -0.5
0.70 -0.11 -13.4
0.24 -0.06 -19.8
450.40 -43.04 -8.7
10442.54 -1962.63 -15.8
B-21
-------
TABLE B. 11
COMPARISON OF HIGH-1ST-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR TUCSON, AZ
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
588.06
250.58
3137.37
477.19
57.32
23.20
13513.59
70467.91
588.06 0.00 0.0
250.58 0.00 0.0
3137.37 0.00 0.0
477.19 0.00 0.0
57.32 0.00 0.0
23.20 0.00 0.0
13513.59 0.00 0.0
70467.91 0.00 0.0
525.75 -62.31 -10.6
250.58 0.00 0.0
3184.21 46.84 1.5
490.02 12.83 2.7
57.32 0.00 0.0
23.20 0.00 0.0
13290.68 -222.91 -1.6
48592.17 -21875.74 -31.0
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
424.79
186.58
2063.02
278.06
25.48
8.52
5407.82
40951.27
424.79 0.00 0.0
186.58 0.00 0.0
2063.02 0.00 0.0
278.06 0.00 0.0
25.48 0.00 0.0
8.52 0.00 0.0
5407.82 0.00 0.0
42881.61 1930.34 4.7
397.70 -27.09 -6.4
163.27 -23.31 -12.5
2063.02 0.00 0.0
278.06 0.00 0.0
21.82 -3.66 -14.4
8.31 -0.21 -2.5
5273.64 -134.18 -2.5
40951.27 0.00 0.0
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
234.22
99.09
1027.07
162.82
14.16
5.09
2915.85
32991.43
234.22 0.00 0.0
98.13 -0.97 -1.0
1027.07 0.00 0.0
177.84 15.01 9.2
14.16 0.00 0.0
5.09 0.00 0.0
2915.85 0.00 0.0
33646.06 654.63 2.0
234.22 0.00 0.0
97.74 -1.35 -1.4
1417.26 390.19 38.0
149.35 -13.47 -8.3
14.16 0.00 0.0
5.13 0.04 0.9
3502.45 586.60 20.1
32991.43 0.00 0.0
B-22
-------
TABLE B. 11, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
146.83
64.34
640.23
119.55
4.79
1.70
1415.62
17498.47
146.83 0.00 0.0
64.34 0.00 0.0
640.23 0.00 0.0
119.55 0.00 0.0
4.79 0.00 0.0
1.70 0.00 0.0
1415.62 0.00 0.0
17498.47 0.00 0.0
127.58 -19.26 -13.1
53.69 -10.65 -16.6
524.96 -115.27 -18.0
94.70 -24.85 -20.8
4.72 -0.07 -1.6
1.78 0.09 5.2
1415.62 0.00 0.0
18573.82 1075.35 6.1
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
14.14
6.27
109.17
16.02
0.65
0.22
549.97
10722.61
13.92 -0.22 -1.6
6.15 -0.13 -2.0
109.91 0.73 0.7
16.24 0.22 1.4
0.68 0.03 4.5
0.23 0.01 4.4
538.77 -11.20 -2.0
10967.82 245.21 2.3
13.69 -0.46 -3.2
5.68 -0.59 -9.4
111.42 2.25 2.1
16.16 0.14 0.8
0.71 0.06 8.5
0.24 0.02 6.8
556.30 6.33 1.2
11830.82 1108.21 10.3
B-23
-------
TABLE B. 12
COMPARISON OF HIGH-2ND-HIGH CONCENTRATION ESTIMATES WITH
CONVENTIONAL OBSERVATIONS TO ESTIMATES USING MIXED ASOS AND FULL
ASOS FOR TUCSON, AZ
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
1 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
503.35
225.92
2822.07
371.18
35.48
13.14
9080.98
58856.02
503.35 0.00 0.0
225.92 0.00 0.0
2822.07 0.00 0.0
371.18 0.00 0.0
35.48 0.00 0.0
13.14 0.00 0.0
9080.98 0.00 0.0
58856.02 0.00 0.0
480.65 -22.70 -4.5
221.26 -4.66 -2.1
3165.65 343.58 12.2
377.76 6.58 1.8
34.94 -0.54 -1.5
12.73 -0.41 -3.1
9080.98 0.00 0.0
48505.06 -10350.96 -17.6
3 -hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
335.24
137.60
1514.49
210.85
17.84
7.32
4850.03
36504.00
335.24 0.00 0.0
137.60 0.00 0.0
1514.49 0.00 0.0
219.22 8.37 4.0
17.84 0.00 0.0
7.32 0.00 0.0
4850.03 0.00 0.0
38227.32 1723.31 4.7
311.14 -24.10 -7.2
136.21 -1.39 -1.0
1514.49 0.00 0.0
203.59 -7.26 -3.4
17.21 -0.63 -3.6
6.92 -0.40 -5.5
4296.01 -554.02 -11.4
36271.45 -232.55 -0.6
8-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
207.06
90.67
876.39
149.35
11.51
3.80
2534.55
28221.59
203.62 -3.44 -1.7
86.98 -3.68 -4.1
876.39 0.00 0.0
149.35 0.00 0.0
11.51 0.00 0.0
3.80 0.00 0.0
2534.55 0.00 0.0
30380.88 2159.28 7.7
195.01 -12.05 -5.8
83.52 -7.15 -7.9
876.39 0.00 0.0
127.32 -22.03 -14.8
11.62 0.11 1.0
4.27 0.47 12.3
2702.28 167.73 6.6
29478.13 1256.54 4.5
B-24
-------
TABLE B.I2, CONCLUDED
Source
Conv Obs
(H.g m"3)
Mixed ASOS Abs Diff Rel Diff
Gigm-3) Oigm-3) (%)
Full ASOS Abs Diff Rel Diff
Oigm-3) Gigm-3) (%)
24-hour Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
127.18
53.69
474.95
94.70
3.99
1.29
1283.99
15910.70
123.46 -3.72 -2.9
50.23 -3.46 -6.4
528.14 53.18 11.2
101.14 6.44 6.8
3.99 0.00 0.0
1.29 0.00 0.0
1283.99 0.00 0.0
16449.66 538.97 3.4
85.17 -42.00 -33.0
38.09 -15.60 -29.1
440.61 -34.34 -7.2
67.42 -27.28 -28.8
3.91 -0.08 -2.0
1.29 0.00 0.0
1283.99 0.00 0.0
17341.93 1431.24 9.0
Period Average
PI
P2
P3
P4
P5
P6
VOL1
AREA1
13.84
6.25
89.14
14.22
0.65
0.21
520.84
10638.10
13.63 -0.21 -1.5
6.07 -0.18 -2.9
90.03 0.89 1.0
14.40 0.17 1.2
0.68 0.04 5.5
0.22 0.01 5.9
509.24 -11.61 -2.2
10876.40 238.30 2.2
13.56 -0.27 -2.0
5.46 -0.78 -12.5
87.75 -1.39 -1.6
13.86 -0.36 -2.5
0.67 0.03 4.4
0.22 0.01 3.2
531.74 10.90 2.1
11717.09 1079.00 10.1
B-25
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-454/R-97-014
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
November 1997
Analysis of the Affect of ASOS-Derived Meteorological Data
on Refined Modeling
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Pacific Environmental Services, Inc.
5001 South Miami Boulevard
P.O. Box 12077
Research Triangle Park, NC 27709-2077
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68D30032
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emissions, Monitoring, and Analysis Division
Research Triangle Park, NC 27711
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This document presents the results of a study which examined the affects of the Automated Surface
Observing System (ASOS) data on refined air dispersion modeling. This study compares Industrial
Source Complex Short Term 3 (ISCST3) output using ASOS and standard (manual) meteorological data
taken concurrently for 2 months (during 1994 and 1995) at 6 sites across the United States. The ASOS
station conversion, which began in 1991, lacks some significant information required in air dispersion
modeling, namely ceiling height and cloud cover. This document compiles results from ISCST3
execution runs for 6 point sources, 1 area source, and 1 volume source.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Meteorological Data
Air Dispersion Models
Meteorology
Dispersion
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
100
20. SECURITY CLASS (Page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77)PREVIOUS EDITION IS OBSOLETE
------- |