United States Environmental Protection Agency Atmospheric Sciences Research Laboratory SI Research Triangle Park NC 27711 Research and Development EPA/600/S3-85/013 Apr. 1985 &ER& Project Summary Ozone Production and Transport in the Atlanta, Georgia Region Hal Westberg and Brian Lamb An air quality monitoring program was conducted in the Atlanta region during Jury, 1981. This field study in- cluded data collection at several ground stations as well as airborne measurements using an instrumented aircraft. Measurements included ozone, oxides of nitrogen (NOX). peroxyacetyl nitrate (PAN), individual hydrocarbons (C, to C10), halocarbons, condensation nuclei, and visual distance plus numerous meteorologi- cal parameters. Isoprene and mono- terpene emission rates were de- termined for the most abundant na- tural vegetation species in the Atlanta region. Three tracer (SF,) experiments were performed in order to better understand dispersion patterns in the Atlanta area. Specific objectives of this research were (1) to quantify oxi- dant transport and production in the Atlanta plume; (2) to characterize an- thropogenic and biogenic ambient hydrocarbon (HC) concentrations; and (3) to determine the contribution of natural hydrocarbon emissions to oxi- dant production in the Atlanta area. A well-defined pollutant plume was observed downwind of Atlanta on most afternoons during the study period. Ozone levels in the 130 to 160 ppb range were commonly observed at distances of 35 to 45 miles from the urban core. Elevated b,cat values were also observed in the Atlanta plume. Nonmethane HC concentrations rang- ing from 113 to 1677 ppbC were recorded at Atlanta sampling sites during the 6 to 9 a.m. time period. A mean concentration of 491 ppbC was calculated from about 50 samples. Mean NOX concentrations varied from 44 to 68 ppb at three urban monitoring sites. A mean HC/NOX ratio of 8.6 was observed during the 6 to 9 a.m. period in Atlanta. Biogenic HC including isoprene, a-pinene, 0-pinene, A'-carene and myrcene were routinely detected in air samples collected dur- ing the 6 to 9 a.m. period. The natural HC burden in these samples varied from 2 to 7% at four sampling sites in the Atlanta area. Ozone concentrations predicted by city-specific Empirical Kinetic Model- ing Approach (EKMA) Ozone Isopleth Plotting Package, (OZIPP) agreed reasonably well with observed con- centrations. This modeling effort con- sisted of assigning measured precur- sor and meteorological parameters to the 8 a.m. Atlanta air parcel, documenting the air parcel trajectory, and comparing ozone concentrations measured in the air parcel later in the day with model predictions. The OZIPP ozone concentrations were generally about 10% below the observed levels. The trajectory analysis used to assign locations to the predicted max- imum ozone concentrations was tested against the results of the SF, tracer experiments. At distances out to 45 miles, the calculated trajectories and the paths of the tracer plume were in good agreement. An analysis of a recent HC and NOX anthropogenic emissions inventory for the Atlanta area indicated that the HC/NO* ratio calculated from the emissions data was 1.5 while the mean ratio observed in ambient samples was 8.6. Application of a sim- ple box model using the emissions data yielded good agreement between predicted and observed NOX concen- ------- nations, but the predicted an- thropogenic HC concentration (99 ppbC) was approximately five times less than the mean observed concen- tration (491 ppbC). These results sug- gest that the anthropogenic HC emis- sion inventory may be severely underestimated. An emission inventory of blogenic HC developed from a series of vegeta- tion enclosure samples collected dur- ing the study yielded area-wide average fluxes of 2500 ng/m2 h for isoprene and 90 fig/m2 h for a-pinene. These levels are similar to those reported for other areas In the U.S. Applications of the box model using these rates provided an upper bound on the observed biogenic HC concen- trations. On a 3.2- x 3.2-km (2- x 2-mi) grid average basis, the max- imum gridded anthropogenic hydrocarbon was 26,600 ^g/m2 h. The 97- x 97-km (60- x 60-mi) area average emission flux of an- thropogenic HC was 1360 jtg/m2 h. This Project Summary was devel- oped by EPA's Atmospheric Sciences Research Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back). Introduction It has become apparent in recent years that photochemical smog problems exist in many sections of the United States. Ambient ozone concentrations commonly exceed the National Ambient Air Quality Standard (NAAQS) around the major ur- ban centers in California, Texas, and throughout much of the midwestern and eastern United States. Oxidant levels in excess of the standard are recorded in rural as well as urban regions of the coun- try. Many studies have been undertaken to determine the source(s) of the high ozone concentrations in these urban and rural areas. The production of oxidants in urban plumes and subsequent transport into rural regions is now well documented. However, considerable con- troversy still exists regarding photochemical ozone production from naturally emitted hydrocarbons. In many cases, natural hydrocarbon emission estimates appear to be comparable to an- thropogenic hydrocarbon emission inven- tories. However, the direct comparison of natural and anthropogenic emission inven- tories is probably not a valid method for apportioning ozone contributions. Uncer- tainties in anthropogenic hydrocarbon emission inventories and in the fate of naturally emitted organics make it difficult to predict the impact that each of these categories will have upon the develop- ment of oxidants in a particular region. In order to improve our knowledge of the factors governing oxidant production and transport in different regions of the United States, a comprehensive air quality monitoring program was conducted in the Southeast during the summer of 1981. The Atlanta, GA area was selected for study because it provided a large, isolated city surrounded by natural vegetation. Specific goals of the research program in- cluded: (1) measurement of an- thropogenic ozone precursor concentra- tions in Atlanta, (2) development of a natural hydrocarbon emissions inventory for a 100-by-100-km area around Atlanta, (3) documentation of ozone production in and around Atlanta, and (4) acquisition of a meteorological data base sufficient for air quality modeling. All of these objec- tives were achieved, although the data base is somewhat limited in that the field study spanned only three weeks. Air quality data were collected between July 13 and August 3, 1981. The purpose of this report is to provide a tabulation and discussion of the data collected during the study period. Procedure The data collection phase of this pro- gram consisted of a joint effort involving research teams from Washington State University and the Environmental Protec- tion Agency (Atmospheric Sciences Research Laboratory, Research Triangle Park). Two special field study laboratories and an instrumented aircraft were used to obtain air quality information in the Atlan- ta region. Parameters that were measured include ozone, oxides of nitrogen (NOJ, halocarbons, Peroxyacetyl Nitrate (PAN), individual HC in the C2 to C,0 molecular weight range, and numerous meteorologi- cal parameters. Additional air quality in- formation was obtained from existing monitoring stations operated by federal, state, and local agencies. Figure 1 shows the location of surface stations and lists the pollutants that were measured at each site. Aircraft sampling missions were con- ducted on 14 days during the study period. Two flights were flown on most of the 14 days. An early morning flight was designed to establish boundary con- ditions on the upwind side of Atlanta. The morning mixing height as well as ozone and oxidant precursor transport aloft were monitored during the morning flight. Afternoon flights were conducted in the region downwind of Atlanta to establish dimensions of the urban plume and to locate the region of maximum ozone buildup. Results and Discussion Specific components of the intensive monitoring program conducted in Atlanta during July, 1981, included measurement of ozone precursor concentrations in Atlanta, chemical characterization of the upwind air flow, and monitoring of oxi- dant production and transport downwind of Atlanta. The results of this sampling ef- fort were used for evaluation of the city- specific Empirical Kinetic Modeling Ap- proach (EKMA) model Ozone Isopleth Plotting Package (OZIPP). Trajectories calculated from wind observations during the study were used to locate predicted maximum concentrations, and several SFe urban tracer releases were performed to aid in the analysis of pollutant transport originating in Atlanta. This work also in- cluded the documentation of available an- thropogenic HC and NOX emission inven- tories and the estimation of a similar biogenic HC emission inventory based upon series of vegetation enclosure samples collected during the field study. The various tasks directly related to the investigation of ozone formation were ac- complished through the operation of an instrumented aircraft during 14 days along with the use of two special field monitor- ing laboratories and several existing air quality monitoring sites during a total of 22 sampling days. Oxidant Precursors Early morning (6 to 9 a.m.) total nonmethane hydrocarbon (NMHC) con- centrations were obtained from gas chromatographic analyses of individual species. The average concentrations varied between 113 and 1677 ppbC with a mean equal to 491 ppbC at four ur- ban/suburban sites. These values are somewhat low compared to similar measurements from a number of other U.S. cities. The mean distribution of HC was approximately 55% paraffin, 15% olefin, and 30% aromatic, but this distribution varied widely from day to day. Biogenic HC were routinely detected in the early morning samples at the DeKalb CC site and on most days at Georgia Tech and DeKalb-Peachtree. These in- cluded isoprene, a-pinene, 0-pinene, ------- Dekalb-Peachtree HC, PAN, NO* Halocarbon, 03 Georgia Tech HC, PAN, NOX, 03 Fulton Co. Health HC, NO, • South Dekalb C. C. Chattahoochee River Figure 1. Location of ground level sampling sites in Atlanta. A3-carene, and myrcene. The total biogenic contribution to ambient NMHC varied from 2% to 7%. Isoprene typically constituted 1% of the total NMHC. Average concentrations of NOX measured during the 6 to 9 a.m. period ranged from 10 to 200 ppb, and the mean NOX concentrations measured at four sites were between 44 and 68 ppb. Mean HC/NOX ratios at these sites were similar with a mean of 8.6, which is quite com- parable to ratios observed in other urban areas. Oxidant Production and Transport The NAAQS for ozone of 120 ppb was exceeded on 3 different days at a site 24 miles downwind of urban Atlanta, and the standard was exceeded at urban sites dur- ing one of those days. The maximum hourly average ozone concentration was 172 ppb observed in the urban center. Daily maximum hourly average ozone concentrations were less than 60 ppb on seven days, in the range 80 to 110 ppb on fourteen days, and greater than 150 ppb on one day. Maximum hourly average concentrations of PAN closely tracked the ozone levels and varied from less than 1 ppb to 4 ppb. During most of the flight days, moderate winds from the northwest pro- duced a well-defined, broad ozone plume downwind of Atlanta by midafternoon. Ozone levels in the 130 to 160 ppb range were commonly observed within the mixed layer 35 to 45 mi from the urban core. These concentrations were generally 50 to 70 ppb above background concen- trations. Stagnant wind conditions produced by a stationary front over Atlanta created the worst-case ozone formation of all the study days. Winds were light and variable, and the vertical mixing height was limited to approximately 925 m during the day. Unlike the days with northwest- erly winds, maximum ozone concentra- tions were observed in the city at the sur- face and aloft. The maximum ozone con- centration recorded with the aircraft was approximately 190 ppb on July 24. The hourly average maximum at the surface on that day was 172 ppb. Outside of ur- ban Atlanta, the maximum surface con- centration was 120 ppb. EKMA Evaluation Ozone production during the two types of wind conditions was modeled using city-specific EKMA with the OZIPP com- puter code. Model input data were taken from surface and aircraft measurements during each modeled day. Maximum ozone concentrations predicted with OZIPP were approximately 10% less than those observed. In addition, a simple air trajectory analysis based upon available surface and upper-air data indicated that the predicted ozone maxima occurred very near the locations and times where ozone maxima were actually observed. Results from three SFe tracer tests also in- dicated that predicted trajectories correct- ly simulated gas transport paths originating in Atlanta and extending as far as 45 mi downwind. Precursor Emission Inventories Anthropogenic HC and NOX emissions in Atlanta, recently inventoried by the State of Georgia, are predominantly from mobile sources. Annual average emission rates from all sources for a 97- x 97-km (60- x 60-mi) area centered over Atlanta are 133,000 tons HC/day and 275,000 tons NOx/day. These values yield an HC/NOX ratio equal to 1.5/1, assuming average molecular weights of HC and NOX to be 14.5 and 46 g/m respectively. This ratio is significantly less than observed in the ambient concentrations. Application of a simple box model of the urban area predicted an NOX concentration essentially equal to the mean observed NOX concen- tration. However, predicted HC concen- trations for the same conditions were 5.0 times less than the mean observed con- centration. The results suggest that the discrepancy in the HC/NOX ratio is the result of severe underestimation of an- thropogenic HC emission rates. Biogenic HC emission rates from the predominate tree species in Atlanta were measured using a branch enclosure technique. Results from 20 samples were similar in magnitude and temperature response to measurements conducted elsewhere in the U.S. At 27°C, the isoprene emission rate from several deciduous species was 11 /ig/g/h, and the a-pinene emission rate from loblolly pine was 0.7 /xg/g/h. These rates were combined with biomass density factors to yield emission flux estimates for deciduous and coniferous forests. In turn, land-use distributions for Atlanta obtained from an EPA analysis of Landsat data were used with the estimated fluxes to estimate the daily average emission rate: • 330 tons/day for isoprene and 24 tons/ day for a-pinene. On a unit area basis, these rates become 2500 jtg/m2 h and 90 /tg/m2 h, respectively. For comparison, the average (3.2- x 3.2-km grid) flux of anthropogenic HC for the 97- x 97-km area was 1350 /ig/m2 h and the highest gridded flux of anthropogenic HC was 26,600 /*g/m2 h. Unlike the results for the anthropogenic emissions, biogenic HC ------- concentrations predicted with the box model represented an upper • bound on typical measured levels. Hal Westberg and Brian Lamb are with Washington State University, Pullman, WA 99164-2730. William Lonneman is the EPA Project Officer (see below). The complete report, entitled "Ozone Production and Transport in the Atlanta, Georgia Region," (Order No. PB 85-173 839/AS; Cost: $20.50, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Atmospheric Sciences Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 {, U.S. GOVERNMENT PRINTING OFFICE: 1985-559-016/27032 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 MAY 16 •85 " \ ^^PJ)STAGE & FEES /jsE«3oo| ~ n i ll^'-x szaa'stn i Official Business Penalty for Private Use $300 OCOC329 PS U S EHVIR *GEMCY CHICAGO ------- |