SIMULATING ATMOSPHERIC EXPOSURE USING
   AN INNOVATIVE METEOROLOGICAL SAMPLING SCHEME
         D.B. Schwede1, W.B. Petersen1 and S.K. LeDuc1

         Atmospheric Sciences Modeling Division, Air Resources Laboratory, National
         Oceanic and Atmospheric Administration, Research Triangle Park, NC
   INTRODUCTION

      Multimedia risk assessments require the temporal integration of atmospheric concentration
   and deposition estimates with other media modules. However, providing an extended time
   series of estimates is computationally expensive. An alternative approach is to substitute long-
   term average atmospheric estimates, but traditional methods for calculating long-term averages
   (e.g. joint frequency function) are not amenable to estimating wet deposition.  In an effort to
   produce the required estimates without the computational burden, we developed an extension
   to the Sampled Chronological Input Model (SCIM) (Koch and Thayer, 1974) for use in U.S.
   Environmental Protection Agency's (USEPA) Industrial Source Complex  - Short  Term
   (ISCST3) model (USEPA, 1995). SCIM samples the long term meteorological record at
   regular, user-specified intervals. Since hourly meteorology is being used, the serial correlation
   between deposition and concentration is maintained. However, this simple sampling scheme
   significantly underestimates wet deposition, particularly at sites with infrequent precipitation.
   We were able to reduce  the uncertainty by introducing an additional sampling interval for
   hours with  precipitation into the original SCIM methodology. Using this revised technique,
   concentration and  dry deposition  are calculated using the "regular" SCIM  sampling;
   concentration and dry and wet deposition are calculated from hours sampled during  "wet"
   SCIM sampling. A composite, weighted average  is taken at the end of the  simulation to
   determine annual values.
  RESULTS AND DISCUSSION

      To analyze the impact on ISCST3 estimates of using the sampled meteorological data, we
  made model runs using five area sources and two point sources. The sources varied in size and
       'On assignment to the National Exposure Research Laboratory, U.S. Environmental
Protection Agency.

-------
particle size distribution. Each source was run with 5 years of meteorological data for four
stations: Lake Charles, LA; Pittsburgh, PA; Salem, OR; Tucson, AZ, The sites were selected
to provide a diversity of climatological regimes.  A polar grid of receptors along 16 evenly
spaced radials at distances ranging from the edge of the source to several kilometers was used.
    We compared the results of various combinations of sampling rates with the results from
using  the full  meteorological  database.   The  basic  SCIM approach worked  best for
meteorological stations with frequent precipitation (e.g. Salem), while wet deposition at sites
with infrequent precipitation (e.g. Tucson) was generally underestimated.  The inclusion of a
higher "wet"  sampling  frequency with a  fairly  low frequency for the "regular" sampling
improved the results at all sites. Point sources required a higher sampling frequency (regular =
every 25th hour; wet = every 8th hour)  than area sources (regular =193; wet = 8) to achieve
similar results.  We also made model  runs varying the start hour of both the "regular" and
"wet" sampling to characterize the variability of the results. Figure 1 illustrates the results for
an example site and indicates that the sampling introduces little bias. The lowest scatter was
observed for higher sampling rates and at locations  with frequent precipitation.  Ratios of
annual values paired in space were calculated and frequency distributions were developed to
assess the ability  of the sampling scheme to reproduce the spatial pattern of impact.  An
example plot is shown in Figure 2. These plots showed that the enhanced SCIM methodology
reproduced the spatial pattern of deposition.
DISCLAIMER

This paper has been reviewed in accordance with the U.S. Environmental Protection Agency's peer and
administrative review policies and approved for presentation and publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
REFERENCES

Koch, R.C. and S.D. Thayer, 1974. Validity of the multiple-source gaussian-plume diffusion model using
    hourly estimates of input; also, Sensitivity analysis of the multiple-source gaussian plume urban
    diffusion model in Proceedings of the fifth Meeting of Expert Panel on Air Pollution Modeling NATO
    Committee on the Challenges of Modem Society, Roskilde Denmark.

USEPA, 1995. User's Guide for the Industrial Source Complex (ISC3) Dispersion Models- Volume II -
    Description of Model Algorithms. EPA-454/B-95-003b, U.S. Environmental Protection Agency,
    Research Triangle Park, North Carolina, 27711.
  < v
                       Salem Land Application Unit
                       Regular Sampling - 193 hours
                       Wet Sampling « 8 hours
                   Distance fm)
Figure 1.  Total annual deposition as a function of
distance for an area source.  Solid and dashed lines
are NoSClM results. Bars represent replicate
samples using SCIM.
                                                                        Tucson Point Source
                                                                        Regular S ampling = 25
                                                                        Wet Sampling - 8
                                                                    JU
                                                   <0.25<0.5<0,75<1.0<1.2S<1,S<1.?S<2.0
                                                             SCIM to NoScim Ratio
Figure 2. Ratio of SCIM to NoSCIM values paired
in space for a 35 m point source. Data from all
replicate samples are included.

-------
NERL-R1P-AMD-00-062
1, REPORT NO.
EPA/600/A-00/015
TECHNICAL REPORT DATA
2.
4 . TITLE AND SOBTITLE
Simulating Atmospheric Exposure Using An Innovative Meteorological
Sampling Scheme
7, AUTHOR(S)
D.B. Schwede, W.B. Petersen, and S.K. LeDuc
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Same as block 12


12. SPONSORING AGENCY NAME AND ADDRESS
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 277 1 1

3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
6.PERFORMING ORGANIZATION CODE
8.PERFORMING ORGANIZATION REPORT NO.
1 0.PROGRAM ELEMENT NO.
1 1. CONTRACT/GRANT NO.
i 3.TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/9
15, SUPPLEMENTARY NOTES
16, ABSTRACT
Multimedia risk assessments require the temporal integration of atmospheric concentration and deposition estimates with
other media modules. However, providing an extended time series of estimates is computationally expensive. An alternative
approach is to substitute long-term average atmospheric estimates, but traditional methods for calculating long-term averages
(e.g. joint frequency function) are not amenable to estimating wet deposition. In an effort to produce the required estimates
without the computational burden, we developed an extension to the Sampled Chronological Input Model (SCIM) (Koch and
Thayer, 1974) for use in U.S. Environmental Protection Agency's (USEPA) Industrial Source Complex - Short Term
(ISCST3) model (USEPA, 1995). SCIM samples the long term meteorological record at regular, user-specified intervals.
Since hourly meteorology is being used, the serial correlation between deposition and concentration is maintained. However,
this simple sampling scheme significantly underestimates wet deposition, particularly at sites with infrequent precipitation.
We were able to reduce the uncertainty by introducing an additional sampling interval for hours with precipitation into the
original SCIM mediodology. Using this revised technique, concentration and dry deposition are calculated using the
"regular" SCIM sampling; concentration and dry and wet deposition are calculated from hours sampled during "wet" SCIM
sampling. A composite, weighted average is taken at the end of the simulation to determine annual values.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS b.IDENTIFIERS/ OPEN ENDED TERMS

18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
20. SECURITY CLASS (This Page)
UNCLASSIFIED

c.COSATI

21. NO. OF PAGES
22. PRICE
EPA-2220

-------