the fall line air quality study (faqs) - georgia institute...

i

THE FALL LINE AIR QUALITY STUDY

FAQSPhase I Pilot Study

Acknowledgment

The Fall line Air Quality Study is a joint effort of the Center for Urban and Regional Ecology (CURE) and the Georgia Institute of Technology’s Schools of Earth and Atmospheric Sciences (EAS) and Civil and Environmental Engineering (CEE).

Principal Investigator: Dr. Michael E. Chang CURE/EAS

Senior Advisor: Dr. C. S. Kiang EAS

Co-Principal Investigators: Dr. Armistead Russell CEEDr. Karsten Baumann EAS Dr. William Chameides EAS

Investigators: Dr. Rodney Weber EASDr. Michael Bergin EAS/CEEDr. Doug Worsnop Aerodyne

Contributors: Dr. Michael BakerDr. Don BlakeDr. Carlos CardelinoDr. Manjula CarnagaratnaDr. Kip CarricoMr. Danny Dipasquale

Ms. Jennifer DreherMr. Roby GreenwaldMr. Frank IftMr. Brian JaynesDr. Jose JimenezMs. Yilin MaMr. Alberto Mendoza

Dr. Talat OdmanDr. Doug Orsini Mr. Kandarp PatelMs. Amy SullivanMr. Jin XuMr. Wes YoungerDr. Jing Zhao

The FAQS is directed by a Coordinating Council consisting of representatives from the Augusta, Macon, and Columbus, Georgia metropolitan areas, the Georgia Environmental Protection Division (EPD), the Georgia Regional Transportation Authority (GRTA), and the Department of Defense (DoD).

Chair: Mr. Ron Methier Chief, EPD Air Protection Branch

Members: Honorable Jack Ellis Mayor, City of MaconMr. Bob Fountain Bibb County EngineerMr. Billy Turner President, Columbus Water WorksDr. Art Cleveland Dean, CSU College of ScienceMr. Bill Price President, DSM Chemicals & Chair Augusta AQ task forceMr. Scott MacGregor VP Comm. Develop. Augusta Chamber of Commerce

Dr. Catherine Ross Executive Director, GRTAMr. George Carellas DoD (Army) Reg. Env. Coord. South. Reg. Env. Office

Ex Officio: Ms. Kay Prince US Environmental Protection Agency, Region IV(non-voting) Ms. Renee Shealy South Carolina Dept. of Health & Environmental Control Mr. Ken Barrett Alabama Department of Environmental Management

ii

Technical peer review of Phase I of the FAQS is provided by a Scientific Advisory Panel consisting of three members recommended by Georgia Tech and approved by the Coordinating Council:

Chair: Dr. Ellis B. Cowling North Carolina State University

Members: Mr. John Jansen Southern Company ServicesMr. Rafael Ballagas Georgia Environmental Protection Division

Special recognition and thanks are due to the following individuals for their work in securing funding for the study, in providing technical and logistical support to the science team, or in contributing to the continued operation and maintenance of the satellite sites.

Mr. Scott MacGregor Augusta Chamber of CommerceMs. Lisa Collins Columbus Chamber of Commerce

Mr. Charles Brooks Bibb County EngineeringMr. Dan Rothwell Bibb County EngineeringMr. Steve Willard Fort Gordon, Department of Public WorksMr. David Schulte Fort Gordon, Department of Public WorksMr. John Burnham Columbia County EngineeringMr. Aubrey Ewbanks Columbia County EngineeringMr. Steve Davis Columbus Water WorksMr. Frank Burch Columbus Water WorksDr. Art Cleveland Columbus State UniversityMr. Scott Southwick Georgia Environmental Protection Division

Mr. George Lee Georgia Tech, Economic Development Institute – MaconMr. John Mills Georgia Tech, Economic Development Institute – ColumbusMr. Elliot Price Georgia Tech, Economic Development Institute – Augusta

This report was prepared by Drs. Michael Chang, Karsten Baumann, Michael Bergin, Rodney Weber, Doug Worsnop, and Ted Russell.

Authorization

Under the direction of the Georgia General Assembly, the Georgia Environmental Protection Division administers funds to the Georgia Institute of Technology to plan and implement the Fall line Air Quality Study. For this purpose, a contract (#773-090108) by and between the EPD and Georgia Tech was established on 1 May 2000 that describes the scope of work, and schedules of delivery and payment. A copy of this contract is available on the FAQS website http://www.cure.gatech.edu/faqs.asp.

iii

http://www.cure.gatech.edu/faqs.asp

Table of Contents

Acknowledgment...........................................................................................................................................................ii

Authorization...............................................................................................................................................................iii

Table of Contents.........................................................................................................................................................iv

Table of Figures...........................................................................................................................................................vi

Table of Tables...........................................................................................................................................................viii

The Fall line Air Quality Study – Phase I...................................................................................................................1

I. EXECUTIVE SUMMARY......................................................................................................................................1II. INTRODUCTION..................................................................................................................................................4

A. Origins of the Fall line Air Quality study....................................................................................................4B. Project Initiation..........................................................................................................................................6C. Past and Present Air Quality in the Fall Line Cities...................................................................................6

1. Criteria data for determining nonattainment status........................................................................................................72. The Spatial Scale of Ozone Air Quality.........................................................................................................................93. Regional Particulate Matter Air Quality......................................................................................................................16

III. PHASE I PILOT STUDY DATA COLLECTION.....................................................................................................18A. Overview of site locations..........................................................................................................................18

1. Site Selection................................................................................................................................................................182. Proximity of major emitters to selected sites...............................................................................................................21

B. Instrumentation and Procedures...............................................................................................................221. Meteorology.................................................................................................................................................................232. Trace Gases..................................................................................................................................................................243. Aerosols........................................................................................................................................................................314. Volatile Organic Compounds (VOCs).........................................................................................................................35

C. Observations..............................................................................................................................................351. Meteorology.................................................................................................................................................................352. Trace Gases..................................................................................................................................................................463. Aerosols........................................................................................................................................................................574. Volatile Organic Compounds (VOCs).........................................................................................................................655. Observations of Ozone and PM2.5 at the Three Satellite Monitoring Stations...........................................................74

IV. AIR QUALITY MODELING AND EMISSIONS.....................................................................................................78V. OUTREACH AND EDUCATION..........................................................................................................................79VI. PEER REVIEW..................................................................................................................................................81

Appendices...................................................................................................................................................................82

A. Summary of scientific understanding of formation and accumulation of ground-level ozone and particulate matter pollution...............................................................................................................................A-1B. Ozone modeling for the Columbus, Georgia, Region: Preliminary analysis of the impact of local and regional emissions on ozone in columbus and the sensitivity to voc and nox emissions reductions................B-1C. 5 October 1999 Fall line Air Quality Study proposal.............................................................................C-129 October 1999 Review of Fall line Air Quality Study proposal by Harold Reheis, Director, Georgia Environmental Protection Division..................................................................................................................D-122 December 1999 Review of Fall line Air Quality Study proposal by Catherine Ross, Executive Director, Georgia Regional Transportation Authority.....................................................................................................E-1F. Policy statements governing the creation, role, and procedures of the FAQS Coordinating Council and Scientific Advisory Panel...................................................................................................................................F-1G. 30 June 2000 letter from Governor Roy Barnes toUS EPA Region IV Administrator John Hankinson recommending nonattainment areas in Georgia for the 8-hour ozone National Ambient Air Quality

iv

Standard............................................................................................................................................................G-1H. General criteria for identifying potential ambient air quality monitoring sites and specific criteria for deploying the mobile Air Quality Research lab................................................................................................H-1I. Description of sites proposed for use as locations to deploy the mobile Air Quality Research lab or to establish the stationary satellite monitoring sites in Columbus, Macon, and Augusta, GA..............................I-1J. Materials Used in the 1999 Georgia Emission Inventory Point Source Survey......................................J-1K. The Fall line Air Quality Study in the News”..........................................................................................K-1L. Peer Review Panel’s Critical Review of FAQS Phase I..........................................................................L-1

v

Table of Figures

Figure 1 Currently operating GA EPD ozone monitoring stations at: (a) Macon - Georgia Forestry Service, (b) Augusta - Bayvale ES, (c) Columbus - Airport, and (d) Columbus - Crime Lab.................................................7

Figure 2 Hourly averaged ozone concentrations in Macon, Augusta, and Columbus on 17 August 2000....................8

Figure 3 Peak daily 8-hour average ozone concentrations as a function of local resultant wind.................................11

Figure 4 Year 2000 correlation between peak 1-hour ozone concentrations at Atlanta (South Dekalb) and Macon.. 12

Figure 5 Number of monitoring sites observing an exceedance of the 8-hour ozone NAAQS in 2000......................14

Figure 6 Meteorological conditions of 17 August 2000, 1600 EDT. Infrared imagery shows position of clouds and relative temperatures of cloud tops. Also shown are positions of surface high and low pressure systems, and locations of surface warm, cold, and stationary fronts........................................................................................15

Figure 7 Diagram of proposed spatial ozone hypothesis: super-regional airshed with nested weak local airsheds in metropolitan Columbus and Augusta, and nested stonger inter-metropolitan Atlanta-Macon airshed..............15

Figure 8 Preliminary 1999 PM2.5 Data presented by the US EPA at the Southern Governors’ Summit on Mountain Air Quality, May 2000, Stone Mountain, GA. Green = sites meeting the PM2.5 annual standard. Red = sites not meeting the PM2.5 annual standard..............................................................................................................16

Figure 9 Fall line Air Quality Study primary domain...................................................................................................19

Figure 10 FAQS monitoring sites and major point sources across north-central Georgia...........................................21

Figure 11 Southern Center for the Integrated Study of Secondary Air Pollution mobile Air Quality Research lab....22

Figure 12 FAQS satellite ambient monitoring stations.................................................................................................23

Figure 13 AQR lab trace gas teflon sampling line extending along tower...................................................................24

Figure 14 Ozone analyzer and calibrator......................................................................................................................24

Figure 15 Carbon monoxide analyzer...........................................................................................................................25

Figure 16 Flow schematic of modified TEI 48C-TL CO analyzer...............................................................................26

Figure 17 Sulfur dioxide analyzer.................................................................................................................................26

Figure 18 Flow schematic of modified TEI 43C-TL SO2 analyzer..............................................................................27

Figure 19 NO / NOy analyzers.....................................................................................................................................28

Figure 20 Fine particualte matter mass, light scattering, and absorption detection in the AQR lab............................31

Figure 21 Schematic Diagram of the aerosol optical property measurements in the mobile AQR lab........................32

Figure 22 Particle Into Liquid Collector.......................................................................................................................33

Figure 23 Aerodyne Aerosol Mass Spectrometer.........................................................................................................33

Figure 24 Schematic of Aerodyne Aerosol Mass Spectrometer (AMS).......................................................................34

Figure 25 Estimated 850mb temperature (C), height (m), and wind (m/s) fields derived from the ETA model initial gridded data at 2000 LDT on each day during the FAQS Phase I pilot study (10 June to 26 July 2000)..........37

Figure 26 Wind frequency distributions (wind roses) from Columbus WW (left), Macon SBP (center), and Augusta FtG (right) for separate daytime periods: morning 0500-1000 LT (thin orange), midday 1000-1800 LT (thick red), nighttime 1800-0500 LT (dotted blue).......................................................................................................46

Figure 27 Trace gas wind roses for same locations and time categories as Figure 26.................................................48

Figure 28 Time series of main met parameters and trace gas species for Macon SBP, 30 minute averages...............50

vi

Figure 29 Time series of main met parameters and trace gas species for Augusta FtG, 30 minute averages..............51

Figure 30 Time series of main met parameters and trace gas species for Columbus WW, 30 minute averages.........52

Figure 31 Diurnal variations of temperature lapse rate (dT/dz), ozone (O3), and NO:NOy color coded according to wind direction, and size coded according to wind speed as presented for Macon, Augusta, and Columbus.....56

Figure 32 Hourly averaged PM2.5 mass concentrations and visibility measurements in Macon................................57

Figure 33 Hourly averaged PM2.5 mass concentrations and visibility measurements in Augusta..............................58

Figure 34 Hourly averaged PM2.5 mass concentrations and visibility measurements in Columbus...........................58

Figure 35 Comparison of total ion mass measured by PILS to total PM2.5 mass measured with a TEOM in Augusta..............................................................................................................................................................................60

Figure 36 Relative ion concentrations measured by the PILS in Macon......................................................................61

Figure 37 Relative ion concentrations measured by the PILS in Augusta...................................................................61

Figure 38 Relative ion concentrations measured by the PILSs in Columbus...............................................................62

Figure 39 Comparison of AMS and collected filter mass loadings in FAQS...............................................................63

Figure 40 Correlation of AMS and filter species in Augusta and Columbus...............................................................64

Figure 41 Representative AMS mass spectrum during FAQS.....................................................................................64

Figure 42 Ozone concentrations (1-hour average) observed in Macon and Leslie, Georgia, and wind direction observed in Macon on 15-19 July 2000..............................................................................................................75

Figure 43 Ozone concentrations (1-hour average) observed in Macon and Leslie, Georgia, and wind direction observed in Macon on 15-18 August 2000.........................................................................................................75

vii

Table of Tables

Table 1 Location and operational status of the GA EPD ozone monitoring network in the Fall Line cities.................6

Table 2 Peak daily 1-hour average ozone concentrations observed at the GA EPD Macon ozone monitor that are greater than 0.12 ppmv over the period 1997-1999..............................................................................................7

Table 3 Number of days with peak 8-hour averaged ozone concentrations exceeding 0.08 ppmv, 1997-1999............8

Table 4 Rank ordering of 5 highest peak daily 8-hour ozone concentrations observed in Augusta, Macon, and Columbus, 1997-1999, and 3-year average of 4th highest value...........................................................................9

Table 5 Local National Weather Service climatological monitoring stations near GA EPD ozone monitors.............10

Table 6 Correlation of daily peak 1-hour average ozone concentrations among selected GA EPD ozone monitoring stations, 1 March – 16 October 2000..................................................................................................................12

Table 7 Sites selected for FAQS Phase I pilot study....................................................................................................20

Table 8 Meteorological parameters measured at Georgia Tech's AQR trailer.............................................................23

Table 9 Conversion efficiencies and differences relative to NO2 for the 3 measurement periods...............................30

Table 10 Average trace gas DQIs valid for all three measurement periods.................................................................30

Table 11 Chemical compounds analyzed from 4-times daily canister samples collected in Macon: 11 – 21 June, Augusta: 29 June – 10 July, and Columbus: 17 – 29 July 2000..........................................................................35

Table 12 Statistical summary of chief meteorological parameters measured via the AQR lab at Macon SBP, Augusta FtG, and Columbus WW separated for different periods of the day...................................................................45

Table 13 Statistical summary of trace gas species measured via the AQR lab at Macon SBP, Augusta FtG, and Columbus WW separated for different periods of the day..................................................................................47

Table 14 Summary of FAQS Aerosol Optical Property Measurements.......................................................................59

Table 15 Summary of the mixing ratio (pptv except where noted) of volatile organic and other chemical compounds in air samples collected 4-times per day in Macon from 11 – 21 June 2000......................................................66

Table 16 Summary of the mixing ratio (pptv except where noted) of volatile organic and other chemical compounds in air samples collected 4-times per day in Augusta from 29 June – 10 July 2000............................................68

Table 17 Summary of the mixing ratio (pptv except where noted) of volatile organic and other chemical compounds in air samples collected 4-times per day in Columbus from 17 – 29 July 2000.................................................70

Table 18 Comparison of average VOCs from FAQS pilot study with VOCs from two 1997 Atlanta PAMS stations (all concentrations in ppbC)................................................................................................................................72

viii

The Fall line Air Quality Study – Phase I

I. Executive Summary

The Fall line Air Quality Study (FAQS) is a three-year project to assess urban and regional air pollution, identify the sources of pollutants and pollutant precursors, and recommend solutions to realized and potential poor air quality in the Augusta, Macon, and Columbus, Georgia metropolitan areas. The study will primarily address ground-level ozone but ancillary results will also provide better understanding of the mechanisms contributing to other pollutants such as fine particulate matter. This report documents the origins of the FAQS and the results from the first phase of this study – a preliminary assessment using existing available information and information gathered through a pilot field study. While the findings presented here are tentative and subject to change pending implementation of the remainder of the study, they do represent the science team’s consensus “initial impression of air quality in Augusta, Macon, and Columbus, Georgia.”

Following concerns that the Augusta, Macon, and Columbus metropolitan areas soon would be designated nonattainment for the 8-hour ozone National Ambient Air Quality Standard (NAAQS), the Fall line Air Quality Study was kicked-off in March 2000. After creating a

coordinating council to oversee the study, researchers at Georgia Tech immediately began reviewing existing bases of information for clues about the factors that contribute to poor air quality in the three cities. They also designed and implemented a pilot study to gain a further initial understanding about the

fundamental characteristics of air pollution in each metropolitan area. A secondary purpose of the pilot study was to provide state and local officials with additional information that may help enable them immediately to begin planning and implementing measures to mitigate the occurrences of poor air quality. The pilot study consisted of deploying an advanced mobile Air Quality Research laboratory to each of the three cities for a two-week intensive monitoring period, and establishing additional stationary satellite monitoring sites in each metropolitan area that would continue to measure ozone and particulate matter after the intensive monitoring period concluded.

Results from the initial assessment of existing data and other studies, suggests that there are multiple temporal and spatial scales involved in the formation and accumulation of ground-level ozone in the three FAQS cities and across the state. All locations are affected by a regional tide of elevated ozone concentrations that extend across much of the southeast. Ozone concentrations in Augusta and Columbus in particular seem to be securely coupled to the region, experiencing ozone concentrations greater than the 8-hour ozone NAAQS allows only on days when ozone concentrations across the whole region are elevated. Ozone concentrations observed at the GA EPD site in Macon however also seem to have a strong statistical relationship with ozone concentrations observed in Atlanta. This may imply that, under certain conditions, Macon and Atlanta may share all or part of a small local airshed, thereby providing a mechanism for

1

Number of days with peak 8-hour averaged ozone concentrations exceeding 0.08 ppmv, 1997-1999.

Monitor 1997 1998 1999Augusta 5 14 8Macon 12 18 18Columbus – Airport 1 8 9Columbus – Crime Lab 2 8 13

Macon to experience exceedance level ozone concentrations on any day that Atlanta has elevated ozone concentrations. Causality however, cannot be established through this statistical approach.

In the Spring of 2000 and prior to the pilot study, the FAQS science team worked directly with representatives from each of the three communities to select appropriate sites to deploy a mobile Air Quality Research (AQR) lab and to establish stationary satellite monitors. The mobile AQR lab houses ~$1.5 million worth of state-of-the-art air quality and meteorological monitoring instrumentation. From this platform, continuous measurements of certain trace gas species and meteorological parameters were made during the summer of 2000 at Macon, Sandy Beach Park (June 11 – 21), Augusta, Fort Gordon (June 25 – July 10), and Columbus, North Water Works facility (July 13 – 23). The satellite sites that were deployed at Sandy Beach Park in Macon, Lakeside High School near Augusta, and Oxbow Environmental Learning Center in Columbus began long-term continuous monitoring of O3 and PM2.5 mass concentrations.

Analysis of the hydrocarbon data collected during the pilot study, seemed to suggest that, like Atlanta, the local atmospheres in Macon, Augusta, and Columbus are characterized by an abundance of hydrocarbons, most notably isoprene, that typically originate only from natural sources. If this result is validated in subsequent phases of the FAQS, it could mean that in order to control ozone, which is a product of photochemical reactions involving hydrocarbons and nitrogen oxides, it might be more effective to try to control the anthropogenic (man-made) emissions of nitrogen oxides. Controlling anthropogenic emissions of hydrocarbons might be less effective because even after they are eliminated, there are still ample sources of natural hydrocarbon emissions to fuel the production of ozone.

Other findings from the pilot study include the identification of chemical markers in the form of ratios of NO:NOy, CO:NOy, and SO2:NOy that can indicate the age and origins of air masses encountered. This analysis seemed to indicate that the monitoring site at the North Columbus Water Works was significantly impacted by a transportation source – probably US80, the J.R. Allen Parkway. While this may indicate that this site is ill suited as an “upwind” site, capable of representing the air prior to influence by the local Columbus area, it does appear to be

2

a reasonable location from which to observe the transient in-situ photochemical transformations occurring locally. The Fort Gordon site in Augusta, and Sandy Beach site in Macon did not appear to be as influenced by nearby sources, and therefore may be more representative of the larger metropolitan area. Finally during the pilot study, fine particulate matter was observed to be composed largely of organic carbon and sulfates. While the impact of this finding is not yet evident, it may become more important as the study continues to seek clues about the origins of fine particulate matter in Georgia.

Within Phase I of the FAQS, efforts were also started to develop a current and comprehensive emission inventory of NOx, VOCs, CO, SO2, NH3, and primary organic aerosols. Two major activities were begun, and are continuing. First was initiation of updating the point source emissions inventory. This was done, in conjunction with EPD, by sending questionnaires to the industries in the counties around the three cities being studied. The questionnaire was to be filled out by all industries, estimated to emit more than 25 tons per year of ozone and PM precursors (including ammonia). With assistance from the relevant regional offices of the Georgia Tech Economic Development Institute, workshops were held in October in each city to assist those receiving a survey to correctly fill it out, to brief these stakeholders about nonattainment and FAQS related issues, and offer them an opportunity to informally ask questions and comment. A second activity during this period was to do a top-down inventory assessment using measurements from the 1999 Atlanta Supersite study and the current emissions inventory (grown to 1999). This assessment suggests that there are a number of areas for possible improvement since the measurements suggest biases in the inventory. In particular, the ammonia, anthropogenic area source VOC, and primary organic PM emissions are suggested to have significant biases. This focuses further attention in the coming inventory development activities to clear up these apparent biases.

Throughout Phase I, extensive efforts were made to keep all the FAQS partners, the public, and the media apprised of the plans, activities, and findings from the study. These included several public planning meetings in Macon, Augusta, and Columbus, open house tours of the mobile Air Quality Research laboratory in each city, and science and policy workshops in Atlanta to present preliminary findings. An email message listserv ([email protected]) was created to facilitate two-way mass communication, and a website (www.cure.gatech.edu/faqs.asp) established to provide additional information and document archive for the study. Reflecting this open communication framework, the activities of the FAQS have received substantial and fair media coverage further disseminating the results of the study.

Finally an external peer review committee was selected by Georgia Tech and approved by the Coordinating Council to attend the science and policy workshops that were held in October 2000. The peer review committee was asked to critically comment on the study’s design, methodology, execution, and conclusions. Though identifying some areas of the study that must be improved, and expressing caution to all about drawing premature conclusions, the review was generally favorable and validating of the study through Phase I.

3

http://www.cure.gatech.edu/faqs.asp

mailto:[email protected]

II. Introduction

The Fall line Air Quality Study is a three-year project to assess urban and regional air pollution, identify the sources of pollutants and pollutant precursors, and recommend solutions to realized and potential poor air quality in the Augusta, Macon, and Columbus, Georgia metropolitan areas. The study will primarily address ground-level ozone but ancillary results will also provide better understanding of the mechanisms contributing to other pollutants such as fine particulate matter. For a summary of the current scientific understanding regarding the formation and accumulation of ground-level ozone and particulate matter pollution, see Appendix A.

The study consists of four primary components: 1) enhanced monitoring; 2) emission inventory development; 3) scenario modeling; and 4) analysis, assessment, and recommendation. These four activities are to be performed concurrently, but staged over four phases. Phase I consists of a preliminary assessment using existing available information and information gathered through a pilot field study. Phase II requires that baseline emission inventories are developed, and a second, more extensive monitoring campaign is designed. Phase III encompasses the most critical elements of the study. In this Phase, the second field study will be completed and information from this investigation will be used to validate the emission inventories and subsequent air quality models that are developed. In turn, the air quality models will be used to isolate and examine the factors contributing to poor air quality in the urban and regional study areas, and to explore multiple scenarios for effecting change. In Phase IV, the final phase, the research team will transfer the technologies implemented in the three urban areas to local or state authorities and develop comprehensive recommendations for improving air quality in the short and long term.

This report documents the origins of the study and the results from Phase I of the study. While the findings presented here are tentative and subject to change pending the implementation of Phases II, III, and IV, they do represent the science team’s consensus “initial impression of air quality in Augusta, Macon, and Columbus, Georgia.” The report is organized into several distinct sections that may be of varying interest to the reader. In this introductory section (II), the regulatory and policy oriented origins of the study are documented and the findings from the preliminary analysis are provided. This section may be of general interest to all readers. The next section (III) is a highly technical description of the procedures and instruments employed in the Phase I pilot field study, and a discussion of the resultant observations. The last three sections (IV, V, and VI) describe other ongoing activities associated with the study: a status report on the development of the emission inventories, a description of several outreach and education activities that have been completed, and comments from an external peer review panel.

A. ORIGINS OF THE FALL LINE AIR QUALITY STUDYIn July of 1997, the United States Environmental Protection Agency (US EPA)

introduced a new National Ambient Air Quality Standard (NAAQS) for ozone. This new standard requires an area to be designated “nonattainment” if, over a three-year period the average of the fourth highest annual 8-hour average ozone concentration is greater than 0.08 ppmv. The US EPA’s interim implementation policy for the new standard indicated that data collected at local monitors in the years 1997, 1998, and 1999 would be used to decide by early 2000 which areas in the United States should be designated nonattainment. Based on this schedule and a preliminary assessment of data collected in 1997 and 1998, the Georgia Environmental Protection Division (GA EPD) began warning local representatives in the Spring

4

of 1999 that all or part of the Augusta, Macon, and Columbus metropolitan areas could be designated as nonattainment. In response, each of these three communities organized task forces to advise them about the Clean Air Act, to gain understanding about the process of designating nonattainment areas, to investigate the consequences of a nonattainment designation, to explore options, and to develop clean air strategies.

On more than one occasion in the spring of 1999, each of these community based clean air task forces sought guidance from the legally designated state environmental protection authority, the GA EPD. Unfortunately, the agency’s Air Protection Branch was and remains consumed by the ongoing air quality crisis in the Atlanta area. Representatives from Augusta, Macon, and Columbus were told that the GA EPD could not divert resources at this time to address their emerging air pollution problems. Then in June of 1999, a serendipitous encounter occurred in Atlanta. Researchers at the Georgia Institute of Technology met separately with representatives from the Augusta and Columbus Air Quality Task Forces at the Georgia Chamber of Commerce 25th Annual Environmental Conference. Afterwards, dialog continued between researchers at Georgia Tech and representatives from the Augusta Air Quality Task Force, primarily through facilitation by the Augusta Metro Chamber of Commerce. Ensuing discussions in Atlanta and in Augusta centered on the identification of the information needed to make effective and efficient decisions that would lead to improved air quality in the Augusta area, and strategies to collect this information.

Independent of these discussions, the Columbus Environmental Task Force contracted with other Georgia Tech researchers to complete a preliminary analysis of the factors contributing to Columbus’ air quality problems. This study was completed in the summer of 1999 and consisted of a reanalysis of results from a regional air quality model developed for the Southern Appalachian Mountain Initiative. While not conclusive, this study did suggest that air quality in the Columbus area might be significantly impacted by pollutants and pollutant precursor emissions from regions outside of the Columbus / Muscogee County area. See Appendix B.

Recognizing the common nature of the problem among all three Fall Line cities, an alliance was created in the summer of 1999 between Augusta, Macon, and Columbus for the practical purpose of securing sufficient funds for a comprehensive air quality study in the middle Georgia region. Researchers at Georgia Tech provided the alliance with an outline that identified the tasks and estimated budget for the study. This outline was approved by the alliance in the fall of 1999 and formally proposed as the Fall line Air Quality Study (FAQS). See Appendix C.

Representatives from Augusta, Macon, Columbus, and Georgia Tech carried the proposal to Governor Barnes on 21 October 1999 to explain their concerns, their willingness to accept responsibility for addressing the problem, their willingness to cooperate with all duly authorized state and federal agencies, their plan for an early intervention approach, and their need for assistance. In turn, the Governor asked the directors of the GA EPD and Georgia Regional Transportation Authority (GRTA) to review the proposal. See Appendices D and E. With favorable reviews, the alliance then turned to the Georgia General Assembly to secure funds for the first year of the study. In the 2000 session, the legislature provided $375,000 for the FAQS in the FY00 Supplemental Budget and another $375,000 in the FY01 Annual Budget. These funds were appropriated to the GA EPD and by contract to the Center for Urban and Regional Ecology, and Georgia Tech’s School of Earth and Atmospheric Sciences and School of Civil and Environmental Engineering.

5

B. PROJECT INITIATIONThe Fall line Air Quality Study (FAQS) was kicked off at a Stakeholders Meeting on

March 2, 2000. Among the first order of business at this meeting and two subsequent meetings on May 9 and June 1, was the creation of the Coordinating Council and the approval of a Scientific Advisory Panel. The Coordinating Council consists of two representatives from each of the three Fall Line cities, and one representative each from the GA EPD, GRTA, and the Department of Defense (DoD). Ex officio (non-voting) members include a representative from the US EPA Region IV, the South Carolina Department of Health and Environmental Control (SC DHEC), and the Alabama Department of Environmental Management (ADEM). The primary purpose of the Coordinating Council is to advise GA EPD in directing the efforts of the contractor (Georgia Tech). The Council also serves as a forum and a medium for information dissemination for FAQS-related concerns such as non-attainment designations, implications of non-attainment status, the State Implementation Plan (SIP) and attainment demonstration timelines and process. With members to be named by Georgia Tech and the Coordinating Council, a plan also was approved to create a Scientific Advisory Panel that would provide peer review of the study process, and upon request of the Coordinating Council, provide technical advice on monitoring, inventories, modeling and control strategy issues. Official FAQS policy statements describing the role and procedures of the Coordinating Council and the Scientific Advisory Panel are included as Appendix F.

At these initial meetings, the science team also presented its plan for implementing Phase I of the study. The primary goal of Phase I was to plan and implement a pilot air quality monitoring study for the FAQS cities of Macon, Augusta, and Columbus. The purpose of the pilot study is to gain understanding about the fundamental characteristics of air pollution in each metropolitan area that will enable researchers to design an effective comprehensive study (to be implemented in Phases II and III). A secondary purpose of the pilot study is to provide state and local officials with a preliminary assessment of past and present air quality that will enable them immediately to begin planning and implementing measures to mitigate the occurrences of poor air quality.

C. PAST AND PRESENT AIR QUALITY IN THE FALL LINE CITIESThe GA EPD has monitored concentrations of ozone in the Augusta metropolitan area

continuously since 1989, in Macon since 1997, and in Columbus since 1981. See Table 1 and Figure 1. Data from the monitors that are still operating are the official determinants of air quality in the metropolitan areas and may be used to designate “nonattainment” areas.

Table 1 Location and operational status of the GA EPD ozone monitoring network in the Fall Line cities.City Monitor Latitude

(degrees)Longitude(degrees)

Elevation(masl)

Start Date End Data

Augusta Bayvale ES 33.43333 82.02194 46 4/27/89 Still OperatingMacon GA Forestry 32.80306 83.54472 54 5/7/97 Still OperatingColumbus Airport 32.52139 84.94361 101 4/1/83 Still Operating

Crime Lab 32.53944 84.84333 122 1/1/81 Still OperatingColumbus 32.50389 84.94028 NA 1/1/81 10/31/82

6

Figure 1 Currently operating GA EPD ozone monitoring stations at: (a) Macon - Georgia Forestry Service, (b) Augusta - Bayvale ES, (c) Columbus - Airport, and (d) Columbus - Crime Lab.

1. Criteria data for determining nonattainment status

For the 1-hour ozone NAAQS, if over a 3-year period, there are more than 3 days on which hourly averaged ozone concentrations exceed 0.12 ppmv (standard rounding conventions

apply), then the area may be considered to be in “nonattainment” of the standard. For the 3-year period 1997-1999, hourly ozone concentrations greater than 0.12 ppmv were observed on 5 different days at the Macon ozone monitoring station. See Table 2. By this measure, the Macon area could qualify as a nonattainment area for the 1-hour ozone NAAQS with a “design value” of 0.134 ppmv (i.e. the 4th highest daily peak hourly average value over the

3-year period). In contrast, there were no days over 1997-1999 period in Augusta or Macon on which ozone concentrations were greater than 0.12 ppmv and thus, these areas would be considered to be in attainment of the 1-hour ozone NAAQS.

7

(a)(b)

(c) (d)

Table 2 Peak daily 1-hour average ozone concentrations observed at the GA EPD Macon ozone monitor that are greater than 0.12 ppmv over the period 1997-1999.

Rank Date Peak1 26-Aug-99 0.1382 26-Jun-98 0.1373 27-Aug-97 0.1344 27-Jul-99 0.1345 6-Aug-99 0.133

During the summer of 2000 however, peak hourly averaged ozone concentrations greater than 0.12 ppmv were observed at the Macon monitor on 17 August (0.154 ppmv), and 18 July (0.131 ppmv), at the Augusta monitor on 17 August (0.127 ppmv), and at the Columbus-Airport monitor on 17 August (0.130 ppmv). The events of 17 August, see Figure 2, represent the all-time highest 1-hour average ozone concentrations in each respective monitors’ recorded history.

Figure 2 Hourly averaged ozone concentrations in Macon, Augusta, and Columbus on 17 August 2000.

For the 8-hour ozone NAAQS, if the 3 year average of the 4th highest annual daily peak 8-hour ozone average concentration is greater than 0.08 ppmv (standard rounding conventions apply), then the area could be considered in nonattainment of that standard. For the 3-year period

1997-1999, there were multiple days on which peak daily 8-hour ozone concentrations exceeded the 0.08 ppmv threshold. See Table 3. With 3-year averages of the 4th highest annual values of 0.092 ppmv at Augusta, 0.104 ppmv at Macon, 0.086 ppmv at Columbus – Airport, and 0.089 ppmv at Columbus –

Crime Lab, all four monitors fail to meet the 8-hour ozone NAAQS. See Table 4. These data and others gathered in the Atlanta area were the basis for Governor Barnes to recommend to US EPA on 30 June 2000 that 21 counties in the metro Atlanta area along with Augusta / Richmond County, Bibb County (including Macon), and Columbus / Muscogee County be designated nonattainment for the 8-hour ozone NAAQS. See Appendix G.

As with the 1-hour averaged ozone concentrations, elevated concentrations of 8-hour averaged ozone concentrations were also observed in 2000. Ozone concentrations exceeded the 0.08 ppmv threshold at the Augusta monitor on 5 days, at the Macon monitor on 17 days, at the Columbus – Airport monitor on 5 days, and at the Columbus – Crime Lab monitor on 7 days. The 4th highest ozone concentration observed at each of these sites was 0.089 (Augusta), 0.097 (Macon), 0.089 (Columbus – Airport), and 0.093 (Columbus – Crime Lab). Thus with respect to

8

Table 3 Number of days with peak 8-hour averaged ozone concentrations exceeding 0.08 ppmv, 1997-1999.

Monitor 1997 1998 1999Augusta 5 14 8Macon 12 18 18Columbus – Airport 1 8 9Columbus – Crime Lab 2 8 13

determining nonattainment, if the 3-year period 1998-2000 were considered instead of the 3-year period 1997-1999, each monitor would still exceed the 8-hour ozone NAAQS.

2. The Spatial Scale of Ozone Air Quality

The preliminary study by Russell et al. (see Appendix B) suggested that there may be a connection between elevated ozone concentrations in Columbus and excessive pollutant concentrations in other parts of the state and region. This general phenomena was also one of the findings of the Ozone Transport Assessment Group (OTAG) study of the eastern US from the mid-1990’s. Specifically, OTAG concluded that (ECOS, 1998 and summarized here from GA EPD, 1999):

The southeast appears to be meteorologically decoupled from the midwest and northeast, indicating little transport either way to and from the southeast.

There does appear to be significant interstate transport, including within the southeast.

Reductions of VOC and NOx in urban areas has an impact on ozone reduction within those areas.

Reductions in NOx emissions in rural areas can have a significant impact on urban areas longer distances away.

Specific to the FAQS cities during periods of exceeding ozone air quality however, it remains to be determined the relative contribution, if any, of other areas to total ozone loads in Augusta, Macon, and Columbus, the location or source of these contributions, and the form of the contributions (e.g. ozone or ozone precursors). It must also be determined under what meteorological conditions, and the persistence of these conditions (i.e. climate), that contributions from other areas manifest in the form of elevated local ozone concentrations.

9

Table 4 Rank ordering of 5 highest peak daily 8-hour ozone concentrations observed in Augusta, Macon, and Columbus, 1997-1999, and 3-year average of 4th highest value.

Rank Augusta Macon1997 1998 1999 avg 1997 1998 1999 avg

1 0.097 0.116 0.100 0.114 0.110 0.1192 0.095 0.108 0.092 0.102 0.108 0.1183 0.092 0.100 0.091 0.095 0.106 0.1154 0.087 0.099 0.090 0.092 0.095 0.106 0.113 0.1045 0.085 0.095 0.090 0.095 0.099 0.1036…

Rank Columbus – Airport Columbus – Crime Lab1997 1998 1999 avg 1997 1998 1999 Avg

1 0.085 0.102 0.097 0.096 0.104 0.1012 0.084 0.097 0.091 0.090 0.095 0.0983 0.081 0.095 0.091 0.084 0.094 0.0974 0.080 0.091 0.089 0.086 0.081 0.089 0.097 0.0895 0.078 0.088 0.089 0.081 0.087 0.0966…

(a) LocalFigure 3 shows ozone concentrations in Augusta, Macon, and Columbus plotted as a

function of wind direction and speed. In this analysis, peak daily 8-hour average ozone concentrations for the years 1997 – 1999 from the GA EPD’s ozone monitoring network (see Table 1 and Figure 1) are combined with concurrent 24-hour resultant winds* from the nearest National Weather Service station (see Table 5) as reported in the Local Climatological Data reports from the NOAA, National Climatic Data Center. Ozone concentrations are classified by Air Quality Index categories for clean air: good, moderate, unhealthy for sensitive groups, and unhealthy. Any event with air quality classified as worse than moderate would fail to meet the 8-hour ozone NAAQS. In Figure 3, the daily peak 8-hour ozone concentration is plotted on the radial axis as the resultant wind speed and in the angular compass direction from which the resultant wind is blowing.

For the highest ozone concentrations, all four monitors appear to exhibit some unique directional characteristics. Disregarding wind speed, higher ozone concentrations at the Augusta monitor are observed most frequently when winds are blowing from the southeast. In Macon, the highest ozone concentrations are observed under a westerly wind flow pattern. In Columbus, both monitors show a tendency towards higher ozone concentrations with northwesterly winds. If wind speed is also considered however, the figures show that the highest ozone concentrations recorded at the monitors are most frequently associated with light or stagnant winds (less than 4 mph). Taken alone, this latter condition might indicate that transport of pollutants or pollutant precursors from other areas does not contribute significantly to elevated local concentrations of ozone. There are two primary caveats to this analysis however: space and time. First the meteorological monitoring stations are not co-located with the ozone monitoring stations. It is possible therefore, that the winds are not representative of the air parcel sampled by the ozone monitoring station. Further, the winds are only representative of the surface winds and fail to characterize the winds aloft. It is these winds aloft, detached from the retarding frictional effects of the earth’s surface, that are more likely involved in the long-range transport of pollutants and pollutant precursors. Second, it was assumed that ozone concentrations are affected primarily by concurrent winds. A viable scenario exists in which ozone or ozone precursors could have been deposited in the local area by winds during the previous or prior days. The analysis presented here does not account for this possibility.

Table 5 Local National Weather Service climatological monitoring stations near GA EPD ozone monitors.City Station Latitude Longitude Elevation

(masl)~distance to O3

monitor (km)Augusta Bush Field 33.3667 81.9500 41.5 10Macon Wilson Airport 32.6833 83.6500 107.9 17Columbus Metropolitan Airport 32.5000 84.9333 135.6 3 (Airport)

10 (Crime Lab)

* Resultant wind is the vector sum of the wind speeds and directions divided by the number of observations for the 24 hour period beginning at 00 LDT.

10

Figure 3 Peak daily 8-hour average ozone concentrations as a function of local resultant wind.

(a) Augusta (b) Macon

(c) Columbus Airport (d) Columbus Crime Lab

Good(O3 < 0.065 ppmv)

Moderate(0.065 ppmv O3 < 0.085 ppmv)

Unhealthy for Sensitive Groups(0.085 ppmv O3 < 0.105 ppmv)

Unhealthy(0.105 ppmv O3 < 0.125 ppmv)

11

(b) StatewideAn examination of ozone concentrations in Augusta, Macon, and Columbus relative to

concurrent ozone concentrations at other monitors across the state reveals additional clues. Figure 4 is an example comparing peak daily 1-hour average ozone concentrations at the South Dekalb ozone monitoring site in metropolitan Atlanta, with concurrent peak daily 1-hour average ozone concentrations at the ozone monitoring site in Macon. The figure suggests that there is a fairly strong relationship between ozone concentrations observed in Atlanta with those observed in Macon. For the year 2000, when ozone concentrations were high in Atlanta, they also tended to be high in Macon. Likewise, when ozone concentrations were low in Atlanta, they also tended to be low in Macon. The R2 value, also called the coefficient of understanding, is a statistical measure of the strength of this relationship. It may range from 0.0, no relationship, to 1.0, a perfect relationship. R, or the correlation coefficient and from which the coefficient of understanding is derived, is also a statistical measure of the strength of the relationship. It may range from -1.0, a perfect anti-relationship (i.e. when values are high at one station, they are low at the other and vice versa), to 0.0, no relationship, to 1.0, a perfect direct relationship. Table 6 shows the correlation coefficient of daily peak 1-hour average ozone concentrations from the 2000 ozone season among nine different stations in Georgia.

Table 6 Correlation of daily peak 1-hour average ozone concentrations among selected GA EPD ozone monitoring stations, 1 March – 16 October 2000.

Aug

usta

Mac

on

Col

umbu

s A

irpo

rt

Col

umbu

s C

rim

e L

ab

Atla

nta

Sout

h D

ekal

b

Les

lie

Sava

nnah

Bru

nsw

ick

Ft. M

ount

ain

12

Figure 4 Year 2000 correlation between peak 1-hour ozone concentrations at Atlanta (South Dekalb) and Macon.

Augusta 1.00 0.36 0.74 0.68 0.32 0.72 0.01 0.55 0.09Macon 0.36 1.00 0.34 0.33 0.77 0.29 0.36 0.41 0.03Columbus Airport 0.74 0.34 1.00 0.90 0.31 0.88 -0.04 0.54 -0.02Columbus Crime Lab 0.68 0.33 0.90 1.00 0.28 0.81 -0.03 0.55 -0.09Atlanta South Dekalb 0.32 0.77 0.31 0.28 1.00 0.27 0.25 0.31 0.03Leslie 0.72 0.29 0.88 0.81 0.27 1.00 -0.08 0.62 -0.08Savannah 0.01 0.36 -0.04 -0.03 0.25 -0.08 1.00 0.12 -0.03Brunswick 0.55 0.41 0.54 0.55 0.31 0.62 0.12 1.00 -0.05Ft. Mountain 0.09 0.03 -0.02 -0.09 0.03 -0.08 -0.03 -0.05 1.00

Referencing Table 6 and Figure 1, ozone concentrations in Augusta are most closely correlated (highest absolute correlation coefficient) with ozone concentrations at both monitors in Columbus and at another station in Leslie, about 75 miles southeast of Columbus in South Central Georgia. In reciprocal, the ozone concentrations observed at the Columbus monitors are most closely related to the values observed at Augusta and Leslie. The two monitors in Columbus are also highly correlated with each other, as one might expect. Somewhat unexpectedly given that Macon lies midway between Augusta and Columbus along the Fall Line, ozone concentrations in Macon more closely track those observed at the South Dekalb monitoring station in metropolitan Atlanta than they do either Augusta or Columbus. Rounding out the state, the Brunswick station in South Coastal Georgia seems to be moderately related to all the Fall Line stations, while the Savannah and Fort Mountain (in the North Georgia Mountains) sites do not appear to be correlated with any other site in Georgia.

(c) RegionalLooking at the even larger region, there appears to be concurrence between high ozone

concentrations in Georgia with high ozone concentrations in other states of the Southeast. Figure 5 shows the number of monitoring sites that recorded an exceedance of the 8-hour ozone NAAQS on each day between 1 May and 30 September 2000 in Georgia (source: GA EPD), South Carolina (SC Department of Health and Environmental Control), and Alabama (AL Department of Environmental Management). Events appear to occur nearly simultaneously across all three states. This result is consistent with the meteorological scale that largely controls the region’s weather. It is this so called “synoptic” scale of approximately 1000 miles that characterizes the principal weather features of high and low pressure systems, the advance of warm and cold fronts, and the location and strength of jet streams. For example, the meteorological conditions present at 1600 EDT on 17 August 2000 are shown in Figure 6. On this date, many sites in Georgia, South Carolina, and Alabama exceeded the 8-hour ozone NAAQS. The whole Southeast was under the influence of the high-pressure dome centered over the Ohio Valley. This system substantially prevented the movement of air as evidenced by the stationary front extending across the region from Illinois to South Carolina. The apparent result was stagnation and a region-wide buildup of pollutants.

(d) Forming a Spatial Hypothesis for OzoneWhile the preceding analyses were simple and limited, they offer hints about the structure

and organization of the regional, statewide, and local airsheds that collectively influence air quality in the Fall Line cities. Any proposed paradigm describing the air quality in the region must be consistent with this information (including the earlier report by Russell et al.). From this examination, one might hypothesize that a super-regional airshed exists across most of Georgia’s piedmont and coastal plain and may extend into parts of Alabama and South Carolina. The

13

super-regional airshed is primarily governed by the synoptic scale meteorology. When meteorological conditions are conducive to ozone formation and accumulation, all areas within the super-regional airshed may experience elevated ozone concentrations. This broad influence may be more significant for Augusta and Columbus than for Macon. The mountainous and extreme coastal regions of Georgia seem to be independent of this super-regional airshed altogether however. This may be because the terrain in these areas is significantly different from the terrain of the piedmont and coastal plain. The specific types of terrain that are found in these areas can strongly influence local meteorological conditions. Meteorologists categorize these local influences that have sway over areas only a few miles to a few hundreds of miles in size as “mesoscale.” Strong mesoscalic weather conditions can over-ride the synoptic scale influences.

In Macon, a strong influence on ozone concentrations beyond just the effects of the synoptic scale meteorology, may be found in another airshed that is nested within the super-regional airshed. This nested airshed is aligned roughly along I-75 and includes both metropolitan Atlanta and metropolitan Macon. Like the mountainous regions or coastal regions, this region may also have a unique mesoscale characteristic: urbanization. Augusta and Columbus may also have local airsheds nested within the larger regional airshed, but they do not appear as intense as the Atlanta-Macon airshed. The dominate influence on local air quality in these areas seems to be associated with the synoptic scale. This working spatial hypothesis for ozone is illustrated in Figure 7.

14

Figure 5 Number of monitoring sites observing an exceedance of the 8-hour ozone NAAQS in 2000.

16

Figure 6 Meteorological conditions of 17 August 2000, 1600 EDT. Infrared imagery shows position of clouds and relative temperatures of cloud tops. Also shown are positions of surface high and low pressure systems, and locations of surface warm, cold, and stationary fronts.

Figure 7 Diagram of proposed spatial ozone hypothesis: super-regional airshed with nested weak local airsheds in metropolitan Columbus and Augusta, and nested stonger inter-metropolitan Atlanta-Macon airshed.

3. Regional Particulate Matter Air Quality

For more than a decade, the GA EPD has monitored the atmospheric concentration of airborne particulate matter smaller than 10 microns in diameter (PM10) at many locations around the state. At this time, these monitors indicate that all areas in Georgia are meeting both the annual and 24-hour PM10 air quality standards. In response to new standards for fine particulate matter introduced in July 1997 by the US EPA, Georgia deployed in 1999 a new network of monitors to measure the atmospheric concentrations of airborne particles with diameters less than 2.5 microns (PM2.5). Along with other states, Georgia is collecting data that will be used in the future to determine the attainment status of areas with respect to PM2.5. In May 2000 at the Southern Governors’ Summit on Mountain Air Quality at Stone Mountain, GA (Seabrook, 2000), researchers at Georgia Tech and the US EPA presented a preliminary compilation of PM2.5 data collected at many monitors across the contiguous United States. The findings that were presented, see Figure 8, suggest that many urban areas of the northeast and midwest, many areas in California, and virtually all areas of the southeastern states may not meet the annual 15 g/m3 PM2.5 standard. Although there is no indication by US EPA that suggests that PM2.5 nonattainment designations are imminent, these preliminary results validate the secondary purpose of the FAQS regarding the need for additional PM data collection and analysis.

References for section II:

17

Figure 8 Preliminary 1999 PM2.5 Data presented by the US EPA at the Southern Governors’ Summit on Mountain Air Quality, May 2000, Stone Mountain, GA. Green = sites meeting the PM2.5 annual standard. Red = sites not meeting the PM2.5 annual standard.

Environmental Council of the States, Ozone Transport Assessment Group: Final Report, http://www.epa.gov/ttn/rto/otag/finalrpt/; 1/8/98.

Georgia Environmental Protection Division (GA EPD), State Implementation Plan for the Atlanta Ozone Non-attainment Area, June 7, 1999.

Seabrook, C. Atlanta Smog Rivals LA’s, Atlanta Journal and Constitution, 5/5/00.

18

http://www.epa.gov/ttn/rto/otag/finalrpt/;

III. Phase I Pilot Study Data Collection

In the summer of 2000 under Phase I of the FAQS, a pilot field study was conducted in Macon, Augusta, and Columbus, Georgia. The purpose of the pilot study was to gain understanding about the fundamental characteristics of air pollution in each metropolitan area that will enable researchers to design an effective comprehensive study (to be implemented in Phases II and III). A secondary purpose of the pilot study is to provide state and local officials with additional information that may help enable them immediately to begin planning and implementing measures to mitigate the occurrences of poor air quality. The pilot study consisted of deploying an advanced mobile Air Quality Research laboratory to each of the three cities for a two week intensive monitoring period, and establishing additional stationary satellite monitoring sites in each metropolitan area that would continue to measure ozone and particulate matter after the intensive monitoring period concluded.

A. OVERVIEW OF SITE LOCATIONSThe Environmental Protection Division monitors air quality in Georgia through a

network of 58 monitors at 46 locations in 23 counties. As Figure 9 shows, this network includes two ozone monitors in Columbus / Muscogee county, one in Augusta / Richmond county, and one in Bibb county near Macon. It is from data collected at these four monitors between 1997 and 1999 that Governor Barnes recommended to the US EPA on 30 June 2000 that Muscogee, Richmond, and Bibb counties be designated nonattainment for the 8-hour ozone National Ambient Air Quality Standard (NAAQS). See Appendix G and previous discussion on nonattainment. While these limited number of EPD sites deployed in or near Columbus, Augusta, and Macon may be sufficient to determine attainment / nonattainment, they are insufficient for diagnosing the causes of poor air quality in these three cities.

1. Site Selection

To address this deficiency, the FAQS proposed to deploy additional stationary and mobile monitors during the summers of 2000, 2001, and 2002 that would complement and augment the EPD network. In addition to ozone, this initiative is to include enhanced monitoring of particulate matter, meteorology, and many other chemical compounds that contribute to secondary air pollutant formation. Not only will this additional information directly yield clues about the causes of poor air quality in the three cities, it will also be used to evaluate the integrity of the study’s subsequent emission inventories and air quality models. It is these latter products that are the foundation of all air quality planning and thus it is important to ensure that they are reasonable representations by comparing their predicted outcomes with observations under similarly prescribed conditions.

19

In the Spring of 2000 and prior to the pilot study, the FAQS science team worked directly with representatives from each of the three communities to select appropriate sites to deploy the mobile AQR lab and to establish the stationary satellite monitors. Each community was provided with a list of descriptive characteristics that collectively would meet the requirements for site selection. This list is provided as Appendix H. In addition to these characteristics, the science team also asked the local representatives to consider only sites that are located a reasonable distance away from the existing EPD ozone monitor(s). This was to ensure that information provided by the new monitoring stations would spatially complement and augment the information provided by the existing monitor(s) rather than simply provide redundancy. With these criteria, the local partners selected several sites for the science team to tour and consider. A description of all the sites considered is included as Appendix I. After careful deliberation, the sites shown in Table 7 were selected for the pilot study. The geographical location of these mobile and stationary monitoring sites in relation to the existing EPD statewide ozone monitoring network are also shown in the previous Figure 9.

20

Figure 9 Fall line Air Quality Study primary domain.

Table 7 Sites selected for FAQS Phase I pilot study.City Name Abbrev Used for Latitude Longitude Elevation Start Date Stop Date

deg N Deg E Masl LT LTMacon Sandy Beach Park SBP AQR lab 32.8277 -83.789 126 6/10/00

10:506/22/00 15:15

Sandy Beach Park SBP Satellite 32.8277 -83.789 126 6/19/00 15:00

Still operating

Augusta Fort Gordon FtG AQR lab 33.419 -82.14 146 6/25/00 18:40

7/10/00 18:34

Lakeside High School LHS Satellite 33.538 -82.099 111 6/27/00 22:30

Still operating

Columbus North Water Works WW AQR lab 32.5184 -84.989 166 7/14/00 21:30

7/26/00 9:19

Oxbow Meadows Environmental Learning Center

OLC Satellite 32.3871 -84.959 74 7/15/00 15:20

Still operating

Sandy Beach Park is located on the north shore of Lake Tobesofkee in Bibb County and approximately 10 miles due west of downtown Macon. The mobile AQR lab and the Macon satellite monitor were stationed on the ball field adjacent to the tennis courts, approximately 500 yards from the public beach. The park is owned and operated by Bibb County.

Fort Gordon is a large Army military base in Richmond and parts of Columbia, McDuffie, and Jefferson counties. Its principal military function is signal corps. The mobile AQR lab was positioned on the southside of Barton Field, the main parade grounds. This site is approximately 12 miles southwest of downtown Augusta.

Lakeside High School is located in Evans, Georgia in Columbia County. Columbia County is primarily a residential and commercial services community. The high school is jointly located with Lakeside Middle School. The metro Augusta satellite monitor was initially placed adjacent to the parking lot separating the high school from the middle school on a narrow grassy knoll above the ball fields. Due to risk of collision with vehicles using the parking lot, the station was relocated approximately one month later to a more secure site at the north end of the high school. The site is approximately 12 miles northwest of downtown Augusta. Permission to use the site was facilitated by the Columbia County Engineering Department and granted by Lakeside High School and the Columbia County Board of Education.

The North Columbus Water Resource facility is owned and operated by the Columbus Water Works. The Water Resource facility draws water from the Chattahoochee River for municipal use. The mobile AQR lab was positioned in a grassy elevated area between the settling basins and a large water storage tank. This site is approximately 4 miles north of downtown Columbus.

The Oxbow Meadows Environmental Learning Center is located on property also owned by the Columbus Water Works. The Environmental Learning Center is a project of Columbus State University. Constructed wetlands and the southern terminus of the Columbus Riverwalk characterize the property. The Columbus satellite monitor was stationed on a large grassy field between the parking lot and the Environmental Learning Center. This site is approximately 5 miles south of downtown Columbus.

21

2. Proximity of major emitters to selected sites

shows a map with the greater metropolitan Atlanta area to the north, the FAQS cities to the south, and the selected monitoring sites. This figure also shows the locations and emission strengths for SO2, NOx (as NO2), and CO of major point sources based on 1990 EPA inventory data for GA, 1996 for SC, and 1997 for AL. The depicted sources are all steam generation plants run by electric utility companies. Also included are the locations of other point sources with CO:NOx emission ratios greater than 1. The biggest CO emitter with 47,660 tons/year (or 3237 moles/min) and a CO:NOx ratio of 361 is the Continental Carbon Black plant just inside AL, close to Columbus.

Among all three primary pollutants, CO has the longest lifetime in the lower troposphere of almost 2 months, whereas NOy lifetimes particularly in plumes are estimated to be less than 12 hours, chiefly due to removal of HNO3. Since NOy includes all NOx (NO and NO2) and its more photochemically stable oxidation products, NOy measurements made at a nearby receptor location can be considered to represent the initially emitted NOx after plume dispersion and dilution if the plume transport time is less than a few hours. Since the dispersion and dilution process acts equally on all three pollutants, the measured SO2:NOy and CO:NOy ratios can be good tracers for nearby emission sources also.

22

Figure 10 FAQS monitoring sites and major point sources across north-central Georgia.

latit

ude

34.2

34.0

33.8

33.6

33.4

33.2

33.0

32.8

32.6

32.4

32.2-85.5 -85.0 -84.5 -84.0 -83.5 -83.0 -82.5 -82.0 -81.5

Atlanta

FAQS measurement sites point source emissions of

SO2 NOx CO [moles/min] and SO2:NOx CO:NOx ratios

point sources w/ CO:NOx > 1

20x20 km

Wansley7385 1527 44.84 0.003

Yates3856 910 604.24 0.07

Bowen9068 2611 913.47 0.03

McDonough1975 430 33 4.59 0.08

Branch3006 1789 104 1.68 0.06

Scherer941 752 931.25 0.12 Arkwright

532 170 17 3.13 0.1

Urquhart319 285 45 1.12 0.16

Aug LHS 111 masl

Aug FtG 146 masl

Mac SBP 126 masl

Col OLC 74 masl

Col WW 166 masl 3.161.88

1.29

361

3.54

1.33

2.26

3.58

longitude

CO and NOy are also indicative of urban, i.e. traffic related emissions, caused by relatively incomplete combustion of automotive fuels at high temperatures. The amount of CO relative to NOx (as NO2) emitted by vehicles driven on highways ranges from 3 to 10 molec/molec [Hartley et al., 2000]. This ratio has decreased nationwide over the past 30 years by about a factor of two, and is thought to be mainly due to improved technology and applications of catalytic converters [Parrish et al., 2000]. Among all the sites, the Columbus WW site is most likely influenced by traffic emissions due to the close vicinity of a 6-lane interstate highway (US 80 – J.R. Allen Parkway) to the north and west.

B. INSTRUMENTATION AND PROCEDURESThe pilot study was centered around the deployment of the Southern Center for the

Integrated Study of Secondary Air Pollutants (SCISSAP) mobile Air Quality Research (AQR) laboratory – a 41' long semi-trailer with an attached 10 meter triangular tower mounted to the side. See Figure 11. This platform houses ~$1.5 million worth of state-of-the-art air quality and meteorological monitoring instrumentation, which SCISSAP allowed the FAQS to use for the pilot study at no cost. From this platform, continuous measurements of certain trace gas species and meteorological parameters were made at Macon, Sandy Beach Park (June 11 – 21), Augusta, Fort Gordon (June 25 – July 10), and Columbus, North Water Works facility (July 13 – 23). For all AQR lab data, the reported time is local time (LT), which is Eastern Daylight Savings Time (EDT). All electrical connections and substantial logistics support were professionally arranged and graciously provided by the Bibb County Engineering Department, Fort Gordon Department of Public Works, and the Columbus Water Works at the respective deployment sites.

Besides the deployment of the AQR lab, additional satellite sites for longer-term continuous monitoring of O3 and PM2.5 mass concentrations were established in each city. The satellite site in Macon was collocated with the AQR lab in order to facilitate more efficient start up of the pilot study. Time allowed the other sites to be located away from the AQR lab providing even greater spatial distribution of observations relative to the established air sampling network of the Georgia EPD. At Sandy Beach Park in Macon, the Bibb County Engineering Department provided an appropriate shelter, electrical connections, cooling capacity, and

23

Figure 11 Southern Center for the Integrated Study of Secondary Air Pollution mobile Air Quality Research lab

arranged for proper security of the site. In turn, the FAQS science team provided and installed all the air quality monitoring equipment. Personnel from the Bibb County Engineering Department were then trained and asked to operate and provide routine maintenance of the monitoring equipment for the remainder of the summer. The Columbia County Engineering Department and the Columbus Water Works likewise donated onsite support and similar tangibles to the Lakeside High School and Oxbow Meadows Environmental Learning Center satellite sites respectively. See Figure 12. For all data gathered at the satellite monitors, the reported time is local time (LT), which is Eastern Daylight Savings Time (EDT) until 10/29/00, and Eastern Standard Time (EST) thereafter.

1. Meteorology

Meteorological monitoring was only completed at the AQR lab sites. The parameters that were measured, their units, the sensor height in meters above ground (mag), the sensors’ specifications, and accuracies are listed in Table 8.

Table 8 Meteorological parameters measured at Georgia Tech's AQR trailer.Parameter Unit Height (mag) Sensor specifications Accuracy

Barometric pressure mbar 2 Vaisala PTB100A 800-1060 mbar 0.3 mbarRelative humidity % 9.5 Vaisala HMP45A 0-100 % 1 % RH (0-90%)Air temperature C 9.5 Aspirated RMYoung RTD1000 –40 - +60 C 0.05 CAir temperature C 3.5 Aspirated RMYoung RTD1000 –40 - +60 C 0.05 CVisible radiation W/m2 9.5 LICOR LI-200SA pyranometer 400-1100nm 5 %UV-B radiation W/m2 9.5 YES UVB-1 pyranometer 280-320 nm 5 % (0-60zenith)Wind direction deg N 9.5 RM Young 05305AQ 0-360 deg 3 degWind speed m/s 9.5 RM Young 05305AQ 0-40 m/s 0.2 m/s

24

Figure 12 FAQS satellite ambient monitoring stations.

(a) Macon – Sandy Beach Park (b) Metro Augusta / Columbia County – Lakeside Evans HS

(c) Columbus – Oxbow Meadows Environmental Learning Center

2. Trace Gases

The sample gas for the measurement of ambient levels of NO, NOy, O3, CO, and SO2 was drawn through Teflon (PFA) tubing with inlets mounted on top of the tower, about 9 meters above ground. See . Samples of O3, CO, and SO2 were tapped off a manifold inside the trailer at the base of the tower. The manifold was connected to an 11 m long, 9.5 mm ID PFA tube through which ambient air was sampled at an average flow rate of 25 slm (liters per min @STP). The average residence time of the sample gas in the line was 1.4 s. Based on earlier tests with addition of CO and SO2 standards to the inlet, chemical losses of these species due to adsorption to the internal surface of the tube wall were not occurring. The quality of the measured data is assessed below in terms of response time, detection limit, precision and accuracy. Values for detection limit and precision are expressed at 95% confidence level assuming normal distribution of zero and span check signals acquired during each city’s measurement period. All data were acquired at 1 Hz and reported as 1 min averages if not noted otherwise.

25

(a) Ozone

Ozone (O3) was measured using a pressure and temperature compensated commercial UV absorption instrument (model TEI 49-C, Thermo Environmental Instruments, Inc., Franklin, MA), being absolutely calibrated by the known absorption coefficient of O3 at 254 nm. See Figure 14. The signal was generated by the difference of frequently alternating measure and reference (zero) cycles, i.e. full transfer of

26

Figure 13 AQR lab trace gas teflon sampling line extending along tower.

Figure 14 Ozone analyzer and calibrator.

Sampleline

O3 through versus complete removal of O3 from the flow system. The linearity and precision of the analyzer was checked on average once every 22 hours. Precision check mixing ratios of 0, 40, 80, 120, and 160 ppbv were provided by a primary standard calibrator with active feedback control (model TEI 49C-PS). The calibrator was supplied with O3-free (zero) air from a cartridge of activated carbon that effectively removed O3 from the ambient air. Each precision check resulted in a 5 point linear regression. Assuming normal distribution of the regressions’ intercepts, the O3 analyzer’s detection limits for all 3 sites were 0.5 to 0.6 ppbv; whereas the slopes of the linear regressions yielded ±4 to ±7 % precision. The accuracy is estimated to be the same. A near identical setup for ozone was also deployed in each of the satellite monitoring stations.

(b) Carbon monoxideCO was measured by gas filter correlation, nondispersive infrared absorption (model TEI

48C-TL with a hand-selected PbSe detector matched with an optimal preamplifier, and an absorption cell with gold-plated mirrors). See Figure 15. The signal output was pressure compensated while the absorption cell temperature was controlled at 42 ±0.3 oC during the entire study. The instrument was modified according to Parrish et al. [1994] by introducing a zero trap of 0.5 % Pd on alumina catalyst bed (type E221 P/D, Degussa Corp.) kept at 180 oC that quantitatively oxidized CO to CO2 and allowed frequent switching between measure and zero modes. Ambient water vapor was previously found to alter the zero level due to the capacity of the zero trap for water absorption. This problem was largely eliminated here, since a small fraction of the sample stream (<1 %) was drawn through the trap (50 cm3 volume) during measure mode, as indicated in the flow schematic shown in Figure 16. This sufficiently conditioned the zero trap to the varying ambient water vapor levels. Nevertheless, the instrument was switched into zero mode every 11 min for 2 min. This allowed frequent determination of full 1 min zero averages, since the instrument’s response time was 20 s. Careful balance of the two flow schemes prevented any noticeable pressure differences in the absorption cell.

NIST traceable calibration gas of 405 ±4 ppmv CO in N2 (Scott-Marrin Inc., Riverside, CA) was introduced into the sample stream by mass flow controlled standard addition and dynamic dilution at the instrument inlet for 2 min approximately every 7 hours. The valve sequence was programmed such that standard addition coincided with a zero about once every day allowing quantification of the zero trap efficiency, which resulted in CO removal > 99 % at all times. The detection limit for a 1-minute average based on the 1 Hz data ranged between 94 and 106 ppbv, and between 6 and 9 ppbv for a 1-hour average. The instrument’s precision, determined from the standard addition span checks, ranged between 8 and 14 % at ~600 ppbv. The accuracy was estimated as the RMS error of uncertainties in the calibration tank concentration (2 %), the mass flow controllers (4 % each MFC), the background variation (4 %), and potential inaccuracies from interpolation of the measured ambient CO during span checks (15 %). Thus, the total uncertainty in the CO measurement is estimated at ±17 % for the entire measuring range. The instrument’s linearity within its 5000 ppbv range was checked with a 4 point calibration (zero excluded) at the beginning of the study, and revealed an r2 of 0.9982.

27

Figure 15 Carbon monoxide analyzer

(c) Sulfur dioxide SO2 was measured by use of a commercial, pulsed UV fluorescence instrument (model TEI 43C-TL) with pressure and temperature compensated signal output. See Figure 17. It’s response time was ~45 s and therefore, required longer zeroing and calibration periods compared

to the CO instrument: zero for 4 min once every 55 min; calibration - via mass flow controlled standard addition of 30.6 ±0.3 ppmv SO2 in N2 NIST traceable calibration gas (Scott-Marrin Inc.) and dynamic dilution at the instrument inlet - was performed for 4 min once every 330 min. A flow schematic of the instrument’s modifications is shown in Figure 18. The chosen sequence caused the standard addition to coincide with the zero about every 14 hours. Zero [SO2-free] air was produced by passing ambient air through a HEPA glass fiber in-line filter (Balston) impregnated with a 0.15 molar

Na2CO3 solution. At a flow rate of 0.9 slm, the filter removed 100 % of the SO2 in the sample. The instrument exhibited a relatively large sensitivity to ambient temperature variations inside the mobile laboratory, which were monitored via an RTD temperature sensor next to the instrument’s position, and corrected by linear regression with the background signal. Calibrations were performed and evaluated analogous to the CO measurements resulting in a detection limit of 0.19 to 0.27 ppbv, and a precision of ±4 to 9 % at ~60 ppbv. Since the instrument’s measurement principle is known to be sensitive to organic hydrocarbons (HC), the efficiency of the internal HC removal through a semi-permeable wall was enhanced by introducing an activated carbon trap into the flow of the low-[HC]-side of the wall, and thereby further increasing the [HC] gradient across the wall. NO is known to be another interferent, and its level of interference was examined by standard addition of NO calibration gas earlier before

28

Figure 16 Flow schematic of modified TEI 48C-TL CO analyzer.

Figure 17 Sulfur dioxide analyzer

the study, resulting in a 2-3 % increase of signal. The reported data were not corrected for this relatively small interference. The accuracy was estimated as the RMS error of uncertainties in the calibration tank concentration (2 %), the mass flow controllers (4 % each MFC), the background variation (12 %), the NO interference (2 %), and potential inaccuracies from interpolation of the measured ambient SO2 during span checks (10 %). Thus, the total uncertainty in the SO2 measurement is estimated at ±17 % for the entire measuring range. The instrument’s linearity within its 100 ppbv range was checked with a 4 point calibration at the beginning of the study, and revealed an r2 of 0.9998.

(d) Nitrogen oxide and sum of total reactive nitrogen oxidesA proto-type Air Quality Design (AQD, Golden, Colorado) NO/NOy analyzer was

deployed for the measurement of NO and total reactive nitrogen oxides (NOy) that include NO, NO2, NO3, N2O5, HONO, HNO3, aerosol nitrate, PAN and other organic nitrates. See Figure 19. These measurements were based on the principal method of metal-surface induced reduction of the more highly oxidized species to NO [Fahey et al. 1985; 1986; Fehsenfeld et al. 1987; Atlas et al. 1992; Parrish et al. 1993; etc.], and its subsequent chemiluminescence detection (CLD) with excess ozone [Ridley and Howlett 1974; Kley and McFarland 1980; Bollinger 1982; Fehsenfeld et al. 1990]. The metal surface here was a 35 cm long, 0.48 cm ID MoO tube (Rembar Co., Dobbs Ferry, NY), temperature controlled at 330 ±2 oC, and housed inside an inlet box mounted to the met tower ~8 m above ground. The sample air was drawn continuously through a 15 cm long 0.64 cm OD SS tube, which extended ~5 cm to the outside bottom of the box and was coupled to two SS crosses, where the flow was diverted to a MoO converter tube for the NOy and a bypass PFA tube of same length for the NO measurement, at 1 slm respectively. All SS components were Teflon coated and temperature controlled at 40 oC. Each flow was filtered by a Teflon membrane filter (Gelman-Teflo) with 2 m pore-size, and directed through a stream selector assembly with mass flow controllers (MFC), all mounted inside the box. The sample residence time inside the PFA tubing between the inlet box on the tower and the CLD unit inside

29

Figure 18 Flow schematic of modified TEI 43C-TL SO2 analyzer.

the mobile lab at the ground was ~0.1 s. The chemiluminescence occurred in a gold-plated reaction vessel in front of a single PMT (Hamamatsu R2257) at ~8.5 Torr. The integrated photon counts were recorded at 1 Hz. The instrument background (or “zero”) was measured every 15 minutes for 2 minutes, overlapping the last minute of an NO measure with the first minute of the following NOy measure mode, resulting in an NO and NOy zero respectively. NO and NOy measure modes were switched every 2 minutes. The 1 Hz data were averaged to 1 minute, and the 1 minute zeroes were interpolated over, and subtracted from the 15 minute measure mode periods. Automated calibrations were performed via a programmed set of NO, NO2, n-propyl nitrate (NPN), and HNO3 standard additions to the sample inlet on an average of 3 times per day in ambient air, and once per day in zero air. The calibrations allowed the determination of specific parameters that are relevant for the assessment of the overall instrument performance, such as sensitivity, artifacts, detection limits, and conversion efficiencies of the MoO tube.

The number of calibrations performed in ambient air during each city’s measurement period ranged between 26 and 44. The NO detection limit for a 1 min integration time was ~3 pptv (at S/N=2), for all three data sets equally. The average detector NO sensitivity (S_NO) varied between 3.91 ±0.10 and 3.56 ±0.11 Hz/pptv, indicating an NO measurement precision between ±8 and ±13 %. A difference in signal was present when sampling zero air in NO measure mode versus NO zero mode. This NO artifact (A_NO) was largest in Macon with 19 22 pptv when research grade oxygen had to be used instead of zero air; while A_NO was 11 8 pptv in Augusta, and 8 ±5 pptv in Columbus. A_NO was interpolated between calibrations and subtracted from the ambient NO measurements. Since the zero volume efficiency was less than 100 %, i.e. on average between 95 and 98 %, the instrument’s zero varied with ambient NO

30

Figure 19 NO / NOy analyzers

and NOy levels, respectively. Thus, during low level periods typically occurring at night, the NO_zero signal counts typically averaged 1400 Hz ±1 %. The final NO mixing ratios (ppbv) were determined as follows:

[NO] (ppbv) = (NO – NO_zeroipol) / S_NOipol – A_NOipol

with NO being the actual signal counts in NO measure mode.The accuracy of the NO measurements had uncertainty due to variations in instrument

zeroes, sensitivities, MFC calibrations, and the level of calibration standard used. The latter was a compressed, NIST traceable gas tank of 4.58 ±0.09 ppmv NO in O2-free N2 (Scott-Marrin Inc.), which had been cross calibrated prior to the study against 7 other similar standards showing a 2 % deviation from the nominal value as measured by LIF (S. Sandholm personal communication). The MFC calibrations before and after the study were within 2 %. The biggest contributor to the overall uncertainty was the variable level of ambient NO before and after the standard addition and the interpolation necessary for the S_NO determination, which is estimated here at 12 % as reflected by the total S_NO variation. Therefore, the overall uncertainty of the NO measurement is estimated at ±15 % as RMS error of all the above potential inaccuracies.

Each calibration cycle allowed the determination of the instrument’s sensitivity to NO2, NPN, and HNO3. The same NO standard that was used for determining the NO sensitivity was used in a gas phase titration (GPT) cell inside the mobile lab at the bottom of the tower. Titration of NO occurred with O3 produced from zero air in front of a Hg penray lamp, and was set between 98 and 100 % efficiency (TT_eff). The NO2 sensitivity (S_NO2) was calculated by subtracting the signal portion that was due to the untitrated NO cal gas (NO_nontit) from the delta signal increase due to standard addition to the ambient NOy sample (NOy_cal – NOy_amb), divided by the net level of NO2 calibration gas,

S_NO2 = ((NOy_cal – NOy_amb) – NO_nontit) / (TT_eff x [NO_cal])

The nominal NO mixing ratio resulting from the dynamic dilution process [NO_cal] was 18.8 ppbv. In ambient air, S_NO2 ranged between 1.33 and 2.27 0.55 Hz/pptv revealing a NO2 conversion efficiency Q_NO2 of 36 to 63 ±15 % (= S_NO2 / S_NO x 100). With each calibration cycle the conversion efficiencies for NPN and HNO3, species that are typically harder to convert than NO2, were also determined via standard addition and calculated as follows,

Q_I = (NOy_cal – NOy_amb) / (S_NO x [I_cal])

with I either NPN or HNO3. NPN cal gas was delivered mass flow controlled to the converter inlet from a NIST traceable compressed air tank of 3.88 ±0.19 ppmv NPN in O2-free N2 (Scott-Marrin Inc.). HNO3 was supplied from a permeation tube (Kin-Tek) inside an oven controlled at 40 ±0.1 oC via a critical orifice controlled zero air flow of ~10 sccm. The permeation rate was verified before and after the study via dissolution of HNO3 using a small scale impinger and subsequent IC analysis of NO3

-. The conversion efficiencies for both NPN and HNO3 in ambient air varied between 30 and 60 %, with HNO3 always being 2 to 19 % lower. During the entire study, the conversion efficiency for NO2 was on average 5 % higher than for NPN, with a few exceptions where they were about the same. Table 2 lists a summary of the individual NO2, NPN, and HNO3 conversion efficiencies Q and the Q-differences of the harder to convert species

31

NPN and HNO3 relative to NO2 (RD-Q) in percent. For each site, i.e. city, the average (Avg), standard deviation (StD) and number of determinations (n) are given. It can be seen that NPN was converted on average 13 % and HNO3 40 % less efficient than NO2, which adds uncertainty to the NOy measurement as considered in Table 9.

Table 9 Conversion efficiencies and differences relative to NO2 for the 3 measurement periods.Q_NO2 Q_NPN Q_HNO3 RD-Q RD-Q

(%) (%) (%) NPN/NO2 HNO3/NO2

Macon 46 39 37 12% 14% Avg11 5 12 17% 26% StD30 29 27 29 27 n

Augusta 36 30 14 13% 59% Avg9 9 7 19% 24% StD38 41 36 37 36 n

Columbus 63 58 39 14% 44% Avg15 17 16 15% 18% StD26 31 29 26 24 n

Conversion efficiencies for all 3 species were much lower than what had been achieved during the 1999 Atlanta Supersite study (e.g. Q_NO2 = 97 ±7 %). For FAQS 2000 a set of three brand new MoO converter tubes from the same manufacturer (Rembar Co., Dobbs Ferry, NY) was used in an alternating fashion, allowing regeneration of one tube while the other one was in operation. Despite frequent regeneration in N2 at 450 to 500 oC, the NO2 conversion efficiency of any of the three tubes never exceeded 82 %, and typically decayed to below 40 % within 2-3 days of sampling. Therefore, the interpolated NO2 sensitivities were used to calculate the final NOy mixing ratios analogous to NO:

[NOy] (ppbv) = (NOy – NOy_zeroipol) / S_NO2ipol – A_NOyipol.

The NOy zeroes averaged 2000 Hz 10 %, and an artifact A_NOy was present when sampling zero air. This artifact varied with time and level of converter decay, and was therefore considered in a time-dependent manner; it averaged 0.61 ppbv for Macon, 0.95 ppbv for Augusta, and 0.49 ppbv for Columbus. Based on measured variations in NOy over 2 – 3 h periods, the precision of our NOy measurements ranged between 10 and ±15 %. In addition to the potential uncertainties that contribute to the NO inaccuracies above, our estimate for the overall accuracy of the NOy measurements included the uncertainties in the GPT derived NO2 calibration gas, and the unequal MoO converter efficiencies for NO2, NPN, and HNO3 resulting in an RMS error of ±25 %.

(e) Trace Gas Instrument SummaryThe data quality indicators (DQIs) assessed above are summarized in Table 10, with Tau

being the instrument’s response time in reaching 90 % of its end value signal.

Table 10 Average trace gas DQIs valid for all three measurement periods.units O3 CO SO2 NO NOy

32

Tau (s) 10 20 45 <1 <1Detection Limit (1min) (ppbv) 0.5 100 1.2 0.003 0.5Detection Limit (30min) (ppbv) 0.1 18 0.2 0.0005 0.1Precision (%) 5 10 6 10 15Accuracy (%) 5 17 17 15 25

3. Aerosols

(a) Optical Property and MassThe aerosol light scattering, sp, and absorption, ap, coefficients are key parameters linked with the ability of aerosols to impair visibility as well as modify climate through the extinction of solar radiation. It has also been found that the aerosol scattering coefficient is highly correlated with the concentration of PM2.5 (Waggoner et al., 1981). During the FAQS study, both sp and ap were measured in all three cities at the mobile AQR lab. See . A Radiance Research nephelometer was used to measure the light scattering coefficient, sp, at 530 nm. The instrument was calibrated prior to the field experiment using clean (particle-free) air as well as HFC-134a for the span gas. The aerosol light absorption coefficient, ap, was measured at 565 nm using a Radiance Research Particle Soot Absorption Photometer (PSAP). The absorption coefficient was corrected for both light scattering and instrument overestimation as described by Bond et al. (1999). Air was sampled at a flowrate of 16.7 lpm through a URG Corporation cyclone inlet that removed particles greater than 2.5 m in order to sample the particulate fraction responsible for light extinction. The air then passed through 6 m of 3/8” i.d. black conducting tubing and into a Perma Pure Nafion Drier which reduced the RH in the air stream to less than 40%. The aerosol was conditioned in order to ensure that the aerosol was dry, and that the aerosol optical properties being measured were intrinsic to the aerosol and not dependent on water content (Ogren et al., 1996). The flow was split into three separate sample streams using a URG

Corporation flow splitter. A portion of the flow (1.0 lpm) was sampled by the PSAP, with an additional 3.0 lpm going to a Rupprecht and Patashnick Tapered Element Oscillating Microbalance (TEOM), and the remaining 12.7 lpm going to the nephelometer. In all cities the TEOM was operated at 50 deg. C, with the exception that during the first several days of the field study in Macon measurements were made at 30 deg. C. A schematic diagram of the experimental setup is shown in Figure 21. Measurements were recorded from each instrument every minute using a laptop computer. Several times during the field study a HEPA capsule filter was placed upstream of each sampling line in order to insure the instruments were working properly. Typically, zero air sp and ap values were < 1 Mm-1, with TEOM PM2.5 values less than 2 g m –3. A

33

Figure 20 Fine particualte matter mass, light scattering, and absorption detection in the AQR lab.

TEOM instrument to measure the PM2.5 mass was also installed and operated at each of the Satellite monitoring stations.

(b) Aerosol Composition - Particle Into Liquid CollectorAs part of FAQS, an instrument for on-line measurements of ambient aerosol chemical

composition was developed and tested. In this technique, fine particles (PM2.5) are collected into a flowing liquid stream that is continually analyzed on-line with an ion chromatograph (IC) to quantitatively measure the soluble aerosol ionic species (referred to as a Particle Into Liquid Collector; PILS-IC). See Figure 22. The instrument consists of four main components: 1) diffusion denuders to strip out interfering gases prior to collecting particles into the liquid. 2) A particle activation and growth component where all particles larger than approximately 50 nm diameter are grown by condensation of water vapor to droplets of nominally 2 µm diameter. This makes the particles easy to collect by inertial techniques. 3) An impactor/droplet collector component where the particles are impacted onto a vertical plate that is washed with dilution flow. 4) ICs for analysis of the soluble ionic components. The water flow containing the dissolved aerosol ionic components is split for analysis by two IC’s (cation and anion). The technique is essentially an automated version of the standard filter technique for measuring aerosol ionic species. A detailed description of the instrument is reported in Weber et al [2000]. The following ions were measured during FAQS: Anions – chloride, nitrate, and sulfate: Cations: sodium, ammonium, potassium, and calcium. The primary ionic species were nitrate, sulfate, and ammonium. Only these components are reported here. The system was designed

34

Figure 21 Schematic Diagram of the aerosol optical property measurements in the mobile AQR lab.

Inlet

Naphion Drier

PM 2.5 Cyclone

PSAP (ap)Neph. (sp)(RH<40%)

TEOM (PM2.5)

Crit.Orifices

Rotameters

Carbon Vane Pump HEPA filter

Computer System

RH MeasurementNephelometers (sp)

PSAP (ap) TEOMS (PM2.5)

Flow Splitter

(16.7 lpm)

RH Sensor

specifically for rapid measurements. This was achieved by minimizing the wetted area to avoid liquid hang-up and time smearing of the measurement, and by choosing appropriate IC columns and eluents to shorten the retention times for the analyte ions. During FAQS, measurements were made continuously at a rate of 1 measurement every five minutes. This instrument was developed with the goal to provide new insights into sources and chemical transformations of PM2.5 particles by providing continuous chemical data, 24 hours a day.

(c) Aerosol Composition - Aerosol Mass SpectrometerThe aerosol mass spectrometer (AMS) developed by Aerodyne Research, Inc. was

deployed in the mobile AQR lab for the entire FAQS campaign during June and July of 2000. See Figure 23. The AMS sampled aerosols 24 hours a day (recorded every 15 minutes) and was operational about 75% of the time during the FAQS deployment. Overall, the AMS performed well and provided size resolved chemical composition of sub-micron aerosol loading.

35

Figure 22 Particle Into Liquid Collector.

The AMS samples submicron aerosol (60 nm to 2m diameter) into a vacuum system where aerosol chemical composition is determined mass spectrometrically. A schematic of the AMS is shown in Figure 24. Ambient aerosol is sampled with a critical orifice into an aerodynamic lens, which then focuses submicron particles into a particle beam in vacuum. The particle beam passes through differentially pumped vacuum chambers over a ~50cm path into the detection chamber. There the particles impact a resistively heated surface (400-800C) where they flash vaporize (in <100sec) within an electron impact (EI) ionizer where the particle vapor plume is ionized and the ions mass analyzed with a quadrupole mass filter. Upon entering the vacuum system, the particle beam passes through a single slot chopper (see Figure 24) that provides a particle pulse in time that enables the determination of particle aerodynamic size (inferred by the particle velocity measured by its time-of-flight over the flight path). Thus, particle size and composition are both determined by correlation of this size measurement with the chemical analysis provided by the mass spectrometric detection.

The quadrupole mass spectrometry utilized in the AMS provides quantitative chemical analysis of the gas plume of the vaporized particles. EI provides a universal detector and is a standard mass spectral analysis technique with many thousands of cataloged mass spectra available for comparison with unknown chemical compounds. The overall challenge inherent in

36

Figure 23 Aerodyne Aerosol Mass Spectrometer.

Figure 24 Schematic of Aerodyne Aerosol Mass Spectrometer (AMS)

Particle Inlet(1 atm)

QuadrupoleMass

Spectrometer

ThermalVaporization &

ElectronImpact

IonizationAerodynam

ic Lens(2 Torr)

Chopper

TurboPump

TurboPump

TurboPump

TOFRegion

Particle Beam Generation

Aerodynamic Sizing Particle Composition

interpretation of AMS data is the complexity of mass spectra from multiple chemical species that are detected simultaneously from ambient particles. It is relatively easy to analyze inorganic species (sulfate, nitrate, ammonium) which have well known fragmentation patterns. Organic speciation is limited to classification into different groups and is an on-going process. Within this limitation, the FAQS deployment of the AMS has demonstrated its quantitative analysis capabilities for both size and chemical composition of ambient aerosol.

The AMS samples ambient aerosol contained in a 0.1 lpm (liter/min) flow that was taken from a larger 10 lpm air flow transported in an 8mm tube from ~3 meters above the roof of the mobile lab. A PM2.5 cyclone inlet filtered out large particles. There were two operational modes of the AMS. One utilized the slotted chopper (with 3% throughput), providing size analysis with the quadrupole mass filter switching between ~10 pre-selected ion fragments. In the other mode, the chopper was translated out of the beam and the mass spectrometer scanned from 0 to 300 amu (at 3Hz), providing full mass spectral analysis of the total aerosol loading. The AMS automatically switched between these two modes every minute or so. During FAQS mass spectra and size distributions were logged every 15 minutes.

As was noted above, the AMS was operational about 75% of the time during FAQS. Because this represented one of the first demonstrations of the AMS, calibrations were performed about every other day, typically for 1-2 hours each morning. The calibration protocol involves generating size selected nitrate and sulfate aerosols. Other maintenance issues required half-day shutdowns, particularly in Columbus where the AMS was down every third day or so. With the experience gained during FAQS, virtually all these problems have been corrected at this point and we have also modified calibration protocols, so that we expect improved operation in future deployments.

4. Volatile Organic Compounds (VOCs)

Samples of air were collected four times per day (~0000, 0800, 1200, and 1700 LDT) for ten days at the AQR lab site in evacuated canisters. The canisters were then shipped to the University of California, Irvine for analysis of volatile organic and other chemical compounds via gas chromatography / mass spectroscopy (GC/MS). Table 11 shows the 83 chemical compounds that were measured from the canister samples.

Table 11 Chemical compounds analyzed from 4-times daily canister samples collected in Macon: 11 – 21 June, Augusta: 29 June – 10 July, and Columbus: 17 – 29 July 2000.CH4 n-PrONO2 o-xylene n-heptaneCO 2-BuONO2 1,3-butadiene 2,5-dimethylhexaneF-12 CHBr3 t-2-butene 2,3,4-trimethylpentaneCH3Cl ethane cis-2-butene 2-methylheptaneF-114 ethene 1-butene 3-methylheptaneH-1211 ethyne 3-methyl-1-butene nonaneMeBr propane 1-pentene isopropylbenzeneF-11 propene 2-methyl-1-butene alpha-pineneF-113 i-butane t-2-pentene propylbenzeneCH2Cl2 n-butane c-2-pentene 3-ethyltolueneCHCl3 i-pentane 2-methyl-2-butene 4-ethyltoleneMeCCl3 n-pentane 2,2-dimethylbutane 1,3,5-trimethylbenzeneCCl4 hexane 2,3-dimethylbutane 2-ethyltolueneC2Cl4 isoprene 4-methylpentene 1,2,4-trimethylbenzeneMeI heptane 2-methylpentane 2,2,4-trimethylpentane

37

H-2402 octane 3-methylpentane limoneneMeONO2 benzene 2,4-dimethylpentane 1,3-diethylbenzeneEtONO2 toluene 2-methylhexane 1,4-diethylbenzeneI-PrONO2 ethylbenzene 2,3-dimethylpentane 1,2-diethybenzeneC2HCl3 m-xylene 3-methylhexane 2-methyl-1-penteneCH2Br2 p-xylene 1,2,3-trimethybenzene

C. OBSERVATIONS

1. Meteorology

Figure 25 is a representation of the synoptic meteorological conditions that were present at 2000 (8:00PM) EDT on each day and spanning the period 10 June to 26 July 2000, the entire duration of the pilot study. These figures show the ambient air temperatures (C) at the 850 mb pressure surface, the height (m) of the 850 mb surface, and the 850 mb wind direction and wind speed. These fields were derived from observations made at the time and were used to initialize the National Weather Service Early ETA meteorological forecasting model (Black, 1994). Approximately 150 m above the ground, temperatures at this level are free of the daily temporal cycles seen at the surface making the location of cold and warm fronts more visible. Free from the influences of the terrain, the winds are also generally representative of the broad regional flow patterns. Finally, the contours show lines of constant height of the 850 mb surface. These contours can be used to locate areas of high and low-pressure systems (lower heights mean higher pressure). It is also understood that winds tend to flow parallel to these contours with wind speed increasing as the contours become more tightly packed. Around high-pressure systems, the winds are observed to move in a clockwise or anti-cyclonic motion. Near the center of a high-pressure system, winds typically become weak and the air is said to stagnate. Around low-pressure systems, the winds are observed to move in a counterclockwise or cyclonic motion.

From Figure 25, it can be seen that meteorological conditions during the Macon portion of the pilot study, 10-22 June 2000, were largely controlled by a persistent high-pressure system sitting off the southeastern coast of the United States. With the center of the high just offshore on 10 June, winds in the Macon area and across much of the southeast were light. As the high moved further offshore beginning 11 June, the anti-cyclonic motion brought slightly increased southerly (i.e. from the south) and eventually westerly (i.e. from the west) winds that persisted for the remainder of the period. This system was also still influencing the southeast as the study moved on to Augusta for the period 25 June to 10 July 2000. By 29 June however, a new loosely structured high-pressure system seemed to coalesce over Georgia. Under this new system however, the winds were still light and generally from the southwest. By the time the study had moved to Columbus on 14 July, winds in the southeast were largely controlled by the cyclonic flow associated with a well-defined low-pressure system centered on the upper Ohio River Valley. This brought more northwesterly (from the northwest) flow to the region. This pattern was reinforced by a second low-pressure system moving across southeast Canada and northern New England before high-pressure again settled in the southeast US around 22 July.

Table 12 summarizes the 1 min meteorological data collected at each mobile AQR lab site separated into three periods of the day: morning between 500 and 1000 LT, midday from 1000 to 1800 LT, and evening to early morning from 1800 to 500 LT. This separation presumes that certain meteorological conditions recur daily but at different times, and govern or affect the dispersion of certain emissions differently. The separation also reflects the photochemical nature

38

of ozone. Highest ozone levels are reported during slow moving, high pressure summertime conditions under mostly clear skies, and usually peak mid to late afternoon. Traffic and industrial emissions however, follow a different diurnal pattern. Traffic emissions typically peak in the morning hours, before and after sunrise, when the lower atmosphere is in transition from a stable, stratified nocturnal surface layer to a well-mixed, convective boundary layer (CBL). Despite the crude time separations, one can see from Table 12 that all three sites are subject to the daily solar cycle: in the night and morning hours, the near-surface atmosphere is characterized by lower temperatures and winds, higher humidities, and a stratified surface layer (evident by the smaller absolute lapse rates, dT/dZ) that impedes vertical mixing.

Since air mass transport may influence the levels of measured pollutants, correlations with wind direction were performed for the same time periods as above. Wind frequency distributions are presented for Macon Sandy Beach Park (SBP), Augusta Fort Gordon (FtG), and Columbus Water Works (WW) in Figure 26. The wind roses are divided into 20 sectors of 18o in order to complete the circle. The integral sum of each 3 curves equals the total percentage of 1-min wind measurements made (see Table 12) minus the percentage of occurrences of calm conditions (wind speed < 0.5 m/s) for each site’s measurement period. As mentioned above, calm conditions occurred most often at Macon with 19%, and the least often at Columbus with 2.7% of all measurements; Augusta experienced 4.0%. Figure 26 shows that Columbus WW was predominantly influenced by northerly flow at night and more northwesterly flow during midday periods, while Augusta FtG was mainly under southwesterly flow, and Macon was under mostly southeasterly flow at night and southwesterlies in the daytime. This is consistent with the prevailing synoptic effects on regional flow patterns described earlier in Figure 25.

39

40

Figure 25 Estimated 850mb temperature (C), height (m), and wind (m/s) fields derived from the ETA model initial gridded data at 2000 LDT on each day during the FAQS Phase I pilot study (10 June to 26 July 2000).

10 June 2000 11 June 2000

12 June 2000 13 June 2000

14 June 2000 15 June 2000

41

Figure 25 continued.

16 June 2000 17 June 2000

18 June 2000 19 June 2000

20 June 2000 21 June 2000

43


22 June 2000 23 June 2000

24 June 2000 25 June 2000

26 June 2000 27 June 2000

44


28 June 2000 29 June 2000

30 June 2000 1 July 2000

2 July 2000 3 July 2000

45


4 July 2000 5 July 2000

6 July 2000 7 July 2000

8 July 2000 9 July 2000

46


10 July 2000 11 July 2000

12 July 2000 13 July 2000

14 July 2000 15 July 2000

47


16 July 2000 17 July 2000

18 July 2000 19 July 2000

20 July 2000 21 July 2000

49


22 July 2000 23 July 2000

24 July 2000 25 July 2000

26 July 2000

Table 12 Statistical summary of chief meteorological parameters measured via the AQR lab at Macon SBP, Augusta FtG, and Columbus WW separated for different periods of the day.Parameter Macon Augusta Columbus

AM Midday PM AM Midday PM AM midday PMEDT 5:00-

10:0010:00-18:00

18:00-5:00

5:00-10:00

10:00-18:00

18:00-5:00

5:00-10:00

10:00-18:00

18:00-5:00

UVB Rad coverage 97% 98% 92% 99% 94% 96% 91% 87% 89%W/m2 Avg 0.28 2.28 0.04 0.26 2.09 0.04 0.18 2.14 0.05

StD 0.39 1.04 0.11 0.36 0.92 0.11 0.27 0.94 0.14Min 0.00 0.20 0.00 0.00 0.01 0.00 0.00 0.08 0.00Max 1.53 4.42 0.83 1.57 4.20 0.71 1.19 3.85 0.93

Temp coverage 97% 98% 92% 99% 97% 96% 99% 94% 95%C Avg 23.0 30.4 25.5 23.3 29.8 25.7 24.0 31.2 26.6

StD 1.9 2.5 3.3 1.9 2.6 3.2 2.2 3.8 3.4Min 18.3 22.5 20.2 18.7 22.2 19.8 19.8 21.6 19.9Max 28.2 35.5 35.1 29.1 37.8 35.8 30.4 37.4 35.5

Pressure coverage 97% 98% 92% 99% 97% 96% 99% 94% 95%Mbar Avg 1005 1004 1004 1001 1000 1000 998 997 997

StD 2 3 2 3 3 3 3 3 2Min 1000 997 973 994 993 993 991 990 990Max 1009 1009 1008 1006 1006 1005 1004 1001 1001

Rel Hum coverage 97% 98% 92% 99% 97% 96% 91% 87% 89%% Avg 88 52 77 83 54 71 82 52 65

StD 9 15 20 9 13 15 11 18 19Min 62 27 28 58 27 28 48 19 25Max 100 91 99 100 97 99 99 96 100

H2O coverage 97% 98% 92% 99% 97% 96% 91% 87% 89%%v Avg 2.53 2.25 2.47 2.43 2.29 2.35 2.49 2.28 2.27

StD 0.22 0.40 0.37 0.19 0.35 0.26 0.15 0.35 0.35Min 1.86 1.46 1.45 2.03 1.43 1.49 2.03 1.12 1.05Max 2.92 3.16 3.16 2.92 3.32 3.19 2.88 2.98 3.16

W Speed coverage 97% 98% 92% 99% 97% 96% 99% 94% 95%M/s Avg 0.8 1.9 1.2 2.2 2.8 2.0 1.5 2.1 1.9

StD 0.6 0.9 1.0 1.0 1.1 0.9 0.8 1.0 1.1Min 0.1 0.1 0.1 0.1 0.2 0.1 0.1 0.2 0.2Max 3.0 7.2 7.6 6.2 7.8 7.1 5.3 7.9 3.2

DT/dz coverage 97% 98% 92% 99% 97% 96% 74% 64% 70%K/m Avg -0.04 -0.08 -0.01 -0.03 -0.06 -0.01 -0.03 -0.06 0.00

StD 0.06 0.07 0.05 0.07 0.10 0.07 0.10 0.11 0.09Min -0.28 -0.37 -0.37 -0.34 -0.48 -0.31 -0.45 -0.51 -0.41Max 0.12 0.19 0.24 0.25 0.49 0.42 0.34 0.30 0.62

50

2. Trace Gases

Table 13 statistically summarizes the trace gases observed at the AQR lab in each of the three FAQS cities. Ozone (O3) concentrations in Macon were generally observed to be lower relative to the other two cities and periods. Midday ozone levels seemed to increase with time and season, and therefore were highest in Columbus both on average (66 ppbv) and maximum reported 1-min value (107 ppbv). At all three sites, the influence of the morning rush hour seems to cause the highest average carbon monoxide (CO), nitric oxide (NO), and total oxides of nitrogen (NOy) levels, with absolute highest values reported at the North Columbus Water Works facility. Sulfur dioxide (SO2) was lowest at Macon Sandy Beach Park and generally appeared on few occasions at higher levels without any noticeable diurnal dependence.

(a) Spatial RelationshipsBased on the same daytime categories, correlations between the main trace gas species

and wind direction resulted in the wind roses depicted in Figure 27. The polar graphs are again divided into 20 sectors of 18o each, whereas the scales are now in units of mixing ratio (ppbv). For visual comparison the scales on all wind roses are the same, except for SO2 with Augusta’s scale being twice that of the other two. Exceptional occurrences of easterly flow that brought in SO2-rich air masses required this larger scale. In this respect, it is important to interpret Figure 27, the trace gas wind roses, in the context of Figure 26, the wind direction frequency, before general conclusions can be drawn. The averages based on the more frequent wind directions are statistically more significant and characterize more closely the general conditions at the site, whereas the values associated with less frequent directions may have more episodic character.

51

Figure 26 Wind frequency distributions (wind roses) from Columbus WW (left), Macon SBP (center), and Augusta FtG (right) for separate daytime periods: morning 0500-1000 LT (thin orange), midday 1000-1800 LT (thick red), nighttime 1800-0500 LT (dotted blue).

Table 13 Statistical summary of trace gas species measured via the AQR lab at Macon SBP, Augusta FtG, and Columbus WW separated for different periods of the day.

Parameter Macon Augusta ColumbusAM midday PM AM midday PM AM midday PM

EDT 5:00-10:00

10:00-18:00

18:00-5:00

5:00-10:00

10:00-18:00

18:00-5:00

5:00-10:00

10:00-18:00 18:00-5:00

O3 coverage 85% 93% 84% 96% 94% 94% 91% 91% 89%ppbv Avg 17 43 26 33 60 46 27 66 45

StD 9 11 14 11 14 13 11 13 15Min 1 18 0 11 14 17 1 19 2Max 43 86 58 63 103 92 60 107 114

CO coverage 63% 68% 62% 72% 71% 70% 72% 68% 69%ppbv Avg 190 168 183 239 208 230 331 248 291

StD 45 76 61 71 48 84 110 41 116Min 116 67 91 110 86 97 175 152 110Max 833 1984 1348 661 490 1622 1020 472 1838

SO2 coverage 83% 82% 80% 81% 82% 83% 86% 83% 85%ppbv Avg 0.4 0.7 0.3 1.6 1.3 1.2 1.1 2.0 1.1

StD 0.3 0.8 0.3 2.8 1.5 2.4 0.9 2.2 1.3Min 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Max 1.9 7.7 16.7 20.9 36.3 23.2 6.3 21.9 14.3

NO coverage 33% 29% 32% 35% 33% 32% 36% 33% 34%ppbv Avg 0.97 0.38 0.19 0.78 0.44 0.20 3.71 1.11 0.33

StD 1.19 2.31 0.75 1.01 0.92 1.91 7.14 0.96 1.02Min 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00Max 8.42 85.20 16.74 6.87 36.34 57.23 58.25 14.75 26.27

NOy coverage 34% 37% 32% 35% 33% 32% 35% 32% 34%ppbv Avg 10.5 2.9 7.7 10.2 4.9 6.3 19.0 7.8 11.8

StD 6.9 4.4 6.2 9.4 4.5 6.8 15.3 3.6 9.8Min 1.4 0.0 0.7 0.8 0.0 0.5 2.6 1.1 1.2Max 42.9 143.9 45.4 88.5 59.9 132.2 102.3 30.4 111.8

52

53

Figure 27 Trace gas wind roses for same locations and time categories as Figure 26.

As expected, the average ozone levels are highest for the midday periods at all three sites indicative of a more regional character. Again considering the wind frequency distributions for all sites, the highest levels were observed at Columbus WW and the lowest at Macon SBP. Also, the ozone levels averaged for the three different time categories are clearly distinct from each other with the morning averages being the lowest. This is mainly due to the way “nighttime” was defined including the evening hours starting at 1800. As will be shown later, titration of ozone became more dominant late at night and early mornings. The titration effect, however, can be seen here from the anti-correlation between the average morning hours’ [NOy] and [O3] at all three sites. For example, the most frequent northerly winds at Columbus WW carried the highest NOy and the lowest ozone mixing ratios during the morning hours, a typical feature. This is also true for the most frequent southerly flows received at Macon SBP and Augusta FtG. The diurnal pattern of the NOy wind roses therefore, shows NOy features that are opposite to ozone.

The CO wind roses tend to correlate with NOy and therefore anti-correlate with ozone for most times but this characteristic is not as clearly evident. The reason has to be seen mainly in the CO lifetime, which is ~2 months and by far the longest of all the species represented here. With that, it is less reactive and does not participate in ozone titration. The longer lifetime also causes a higher background level, i.e. the lowest level reported here is 110 ppbv, which might well represent the more regional CO background for southern GA. In contrast, minimum [NOy] ranged much closer to zero at all three sites, which of course is affecting the visual appearance of the wind rose graphs where variations in [NOy] are enhanced relative to [CO]. The lifetime of NOy is governed by dry deposition with the loss of HNO3 being the most efficient.

(b) Time series and special occurrencesFigure 28, Figure 29, and Figure 30 depict the time series of the 30 min averaged data

collected at Macon SBP, Augusta FtG, and Columbus WW, respectively. The parameters shown in the five panels from top to bottom are: 1) the air temperatures from the 3 and 10 m above ground level and the barometric pressure; 2) the UV-B radiation of the photo-chemically important 280-320 nm wavelength range and the water vapor mixing ratio; 3) the wind direction, RH, and wind speed; 4) the CO and SO2 mixing ratios on log scales; and 5) ozone mixing ratio on a linear scale, and NO and NOy mixing ratios on log scales. The vertical lines mark midnight of each day. The numbers in the CO-SO2 plots mark linear regression slopes of CO versus NOy (brown) and SO2 versus NOy (blue) of certain plume encounters where r2 > 0.5. By comparing the slopes with emission ratios of nearby sources, information can be gained regarding the air mass transport. Assuming constant emission rates, the relationship between the slopes and the absolute magnitudes of species’ mixing ratios for re-occurring plumes of same origin provides insight on the mixing depth and stratification of the boundary layer.

54

55

Figure 28 Time series of main met parameters and trace gas species for Macon SBP, 30 minute averages.

1010

1005

1000

995

990

Pressure (m

bar)

35

30

25

20Tem

pera

ture

(C

)

T_3m T_10m

Press

4

3

2

1

0

UV

B R

ad.

(W/m2 )

3

2

1

H2 O

(%v)

UVB H2O

300

200

100

0RH

(%),

Win

d D

ir. (d

egN

)

6

4

2

0

Wind S

peed (m/s)

WDRH

WS

100

2x102

3

4

567

CO

(p

pbv)

0.1

1

10

SO

2 (ppbv)

COSO2

0.014

4.27.2

21.8 4111.4

15

2.32.4

3.1 7.13.1

100

80

60

40

20

0

O3

(ppb

v)

06/11/00 06/13/00 06/15/00 06/17/00 06/19/00 06/21/00Time [EDT]

0.01

0.1

1

10 NO

, NO

y (ppbv)

O3

NOyNO

56

Figure 29 Time series of main met parameters and trace gas species for Augusta FtG, 30 minute averages.

1010

1005

1000

995

990

Pressure (m

bar)

35

30

25

20Tem

pera

ture

(C

) T_3m T_10m Press

4

3

2

1

0

UV

B R

ad.

(W/m2 )

3

2

1

H2 O

(%v)

UVB H2O

300

200

100

0RH

(%),

Win

d D

ir. (d

egN

)

6

4

2

0

Wind S

peed (m/s)

WDRH

WS

100

2x102

3

4

567

C

O

(ppb

v)

0.1

1

10

SO

2 (ppbv)

CO SO2

5.1

4.2 5 4.3

147.2

14.43.6

1118

7.6

130.19

0.20.6

0.55 0.66

100

80

60

40

20

0

O3

(ppb

v)

06/27/00 06/29/00 07/01/00 07/03/00 07/05/00 07/07/00 07/09/00Time [EDT]

0.01

0.1

1

10 NO

, NO

y (ppbv)

O3 NOyNO

(c)

57

Figure 30 Time series of main met parameters and trace gas species for Columbus WW, 30 minute averages.

1010

1005

1000

995

990

Pressure (m

bar)

35

30

25

20Tem

pera

ture

(C

)

T_3m T_10m

Press

300

200

100

0RH

(%),

Win

d D

ir. (d

egN

)

6

4

2

0

Wind S

peed (m/s)

WDRH

WS

100

2x102

3

4

567

C

O

(ppb

v)

0.1

1

10

SO

2 (ppbv)

CO SO27.5 9.2 8 8.6 9.7

8.5

7.3

6.6 14 11 12

100

80

60

40

20

0

O3

(ppb

v)

07/15/00 07/17/00 07/19/00 07/21/00 07/23/00 07/25/00Time [EDT]

0.01

0.1

1

10 NO

, NO

y (ppbv)

O3 NOy NO

(d) City Specific Observations

Macon Sandy Beach Park, 10-22 June 2000As already mentioned, the Macon period was characterized by mostly calm and low wind

conditions. The highest ozone mixing ratios of about 80-85 ppbv were registered on the first day of the measurement period. The half-hour maxima of the following days ranged between 40 and 60 ppbv. The only night when ozone became completely titrated out of the ambient air was between 6/16 and 6/17 under southerly and easterly flow and near-saturated conditions with RH close to 100 % and water vapor mixing ratios close to 2.8 %v. These air masses were relatively high in all air pollutants with [NO] being up to 20 % of [NOy] indicating relatively fresh emissions from nearby sources. By means of the before-mentioned regression method, this and other plume encounters are investigated below.

The first two labeled plumes in Figure 28 with CO:NOy slopes of 11.4 and 15 impacted the site under low southerly flow. Both appeared between 2100 and 2200 LT and were relatively low in [NO] (< 0.1 ppbv) while [NOy] ~10 ppbv, indicating well-processed air of older emissions (since nitric acid, a main NOy component is removed via surface deposition within 6 to 12 hours, ratios larger than 10 can indicate “aged” traffic emissions). The next four plumes labeled in Figure 28 with ratios ranging between 2 and 4 all had a distinctive easterly component, therefore suggesting the influence from the point sources to the east and southeast (see), which are mainly wood and paper processing factories. Sunday 18 June was Father’s Day and brought increased recreational activity to the Park. The wind direction shifted from southerly to westerly flow between 0400 and 0700 LT. Much of the open space surrounding the AQR lab was used for parking. The first slope of 7.2 was registered in the morning between 0700 and 1000 when the park filled with vehicles, and [NO] was ~ 18 % of [NOy]. The higher values of 22 and 41 occurred in the late afternoon and evening and are concurrent to the times when patrons were observed to be using barbeques and open fires. The plume with CO:NOy ratio 7.1 impacted when the wind veered from west over north to east and south in the evening of 20 June, suggesting a traffic source, while the 3.1 ratio occurred under strong east-southeasterly flow in the early afternoon of 21 June, again pointing to the industrial sources.Augusta Fort Gordon, 25 June – 10 July 2000

Figure 29 shows that light winds prevailed on the first day of monitoring in Augusta, but that this was followed by a 3-day period of strong south-southwesterly flow and maximum ozone levels of 40 to 45 pbbv. Oscillating winds then caused concentrations of CO and NOy to build from Friday 30 June through Sunday 2 July. The highest 8-hour ozone average of the Augusta period was 84 ppbv observed between 1230 and 2030 on 30 June. On top of the elevated CO “background” level of ~200 ppbv, the data show excursions of both [CO] and [NOy] indicating emission ratios between 7.2 at midday and 14.4 at night on this day. The 3.6 CO:NOy ratio on the next day (1 July) was registered under easterly flow, suggesting the same sources that caused the ratios between 4 and 5 during the first three days of the Augusta monitoring period.

An event of increased traffic amounts in the immediate vicinity of the AQR lab occurred on the eve of July 4th. This was similar to the Father’s Day event in Macon. On 3 July, fireworks were displayed on the Parade Ground next to the AQR lab between 2100 and 2200 LT, and spectators parked cars in the vicinity. The regression of the CO versus NOy correlation plot resulted in a slope of 7.6, while NO was ~14 ppbv and 30 % of NOy. Since the emission source was so close, [NO] was expected to make up for most of [NOy] but consistent winds with hourly

58

averages of ~2 m/s caused rapid dilution and mixing, which also prevented ozone from being titrated out completely.

After the passage of a low-pressure front, the winds slowly veered from north over east to south on 7 and 8 July. This episode was accompanied by the absolute highest SO2 mixing ratios of the entire measurement period being correlated with easterly wind directions as depicted in Figure 27. Interestingly, the largest SO2 source in the region, with a 1996 SO2:NOx emission ratio of 1.12, is the Urquarth power plant east of Augusta and just across the state-line in SC. Thus, while this event was not related to a high ozone episode, there may be a chemical signal (e.g. SO2:NOx ratios of 0.5 to 0.7 ) indicating when emissions from the Urquarth power plant may be directly influencing local air quality. This needs to be examined in more detail once the FAQS regional emission inventory is completed.Columbus Water Works, 14-26 July 2000

Clear skies and the buildup of ozone concentrations characterized the first few days of the Columbus period. The 1-hour peak ozone concentration increased from 73 ppbv on 16 July at 1300 LT, to 87 ppbv on 17 July at 1500 LT, to 98 ppbv on 18 July at 1800 LT. On these same days, the 8-hour peak ozone concentration increased from 70, to 77, to 89 ppbv. The period 1300 – 2100 LT on 18 July was the only period of the entire three-city study where the 8-hour ozone NAAQS was exceeded.

As noted earlier, the Columbus Water Works site was most notably influenced by traffic emissions. The reason was probably a combination of meteorological conditions (persistent northwesterly winds) and source location (US80 – the J.R. Allen Parkway – to the north and west of the monitoring site). Figure 30 supports this in a more detailed manner. Unlike Macon and Augusta where CO:NOy ratios were spread across a broad range, the CO:NOy ratios in Columbus are more narrowly confined. Further, all CO:NOy ratios determined for northerly and westerly flow conditions lie within the range of 7.0 to 10.0. These are similar to the ratios observed on Father’s Day in Macon (7.2), and on 3 July in Augusta (7.6) when the monitors were clearly under the influence of nearby automobiles. If the site in Columbus is dominated by mobile source emissions however, some uncertainty remains about why the CO:NOy ratios that were observed cannot be more narrowly defined than the 7.0 to 10.0 range. Although the 7.0 to 10.0 range of values observed in Columbus could be due to the combined uncertainty of both measurements (CO and NOy), it also could be that the differences are more likely due to differences in plume transport times and time of day. Nitric acid (HNO3) concentrations usually maximize at noon and decline rapidly with declining OH concentrations. Also surface deposition becomes more effective in shallow surface layers (e.g. morning and night), therefore removing HNO3 from the NOy “ensemble” and shifting the CO:NOy ratio to higher values. Indeed the observations show that the ratios tend to be higher when plumes arrive in the evenings: e.g. 14 and 12 occurred at 2100 each.

In summary, all three sites were mainly influenced by emissions from mobile sources, although industrial emissions impacted the sites at certain times. CO and NOx emission ratios from mobile sources could be verified from two independent occasions at Macon (Father’s Day) and Augusta (3 July fireworks display) that allowed the measurement of nearby automotive exhausts. In both cases the emission ratio was determined to be between 7.2 and 7.6.

59

Interurban comparisonsThe top panel of Figure 31 shows the correlation of wind speed versus wind direction

providing the key for the three remaining panels, in that the size of the symbol increases with wind speed, and the color changes with direction: light blue and green signify southerly (from the south) winds, and red and purple denote northerly (from the north) winds. This key is then maintained for all the other graphs in Figure 31 that correlate the various parameters with the time of day, therefore revealing features that are re-occurring daily. One is the temperature “lapse” rate between 3 and 10 meters above ground (mag) indicating the diurnal variation of atmospheric stability within the surface layer, i.e. stably stratified conditions at night and convectively labile or neutral conditions during midday. Augusta was least stable at night probably due to the relatively strong winds that helped induce enough shear and mixing at night to prevent stable stratification.

Figure 31 also shows how the ozone levels increased as the summer season progressed from the earlier Macon period through Augusta to the Columbus period last. Due to the lack of NO2 measurements, the photochemical age of the probed air masses has to be assessed by means of the NO:NOy ratio. The morning rush hour emissions affected the NO:NOy ratio at all threes sites, in that the almost negligible fraction of NO at night increased to 10 – 20 % between 0500 and 1000 LT. Low to medium strength northerly winds at Columbus caused more scatter and occasional NO fractions up to 45 % during these early morning periods. During the afternoons, the NO:NOy ratios were smallest at Macon, ranging between 5 and 12 % for most values (and close to 70 % during the Father’s Day exception), while Augusta showed more scatter but with the bulk between 3 and 15 %. The afternoon ratios were highest at Columbus with a bulk 10 to 20 % range. It can also be seen that fresh NO was continuously fed into the sampled air masses, especially the ones coming from the nearby J.R. Allen Parkway to the north and west of the monitoring site.

At all three sites, the lowest O3 concentrations occurred during relatively calm nightime hours due to the absence of photochemical production and titration effects from primary emissions. In contrast, clear sky daytime conditions were associated with the highest ozone levels. The ozone diurnal profile shows ‘tight’ transients during the morning hours, which can be explained by the recurring effect of downmixing from the nocturnal reservoir layers. As the rising sun induces surface heating, the stratified nocturnal boundary layer breaks up and mixing from aloft sets in. Several intensive measurement campaigns in the southeastern U.S. carried out as part of the Southern Oxidants Study (SOS) have revealed the fact that entrainment of O3 from aloft can provide a large proportion of surface ozone. It was shown that O3 produced throughout the CBL on the previous day (or days) contributes to the levels measured at the surface [Baumann et al., 2000], but since nocturnal boundary layers generally strongly decouple the surface from the free troposphere, movement of these layers at night around large regions must be taken into account. Therefore, high [O3] measured near the ground may not only be due to emissions being imported during the day that drive photochemical production but those high surface O3 levels may also be due to O3 being imported into the region during the night.

60

.

61

Figure 31 Diurnal variations of temperature lapse rate (dT/dz), ozone (O3), and NO:NOy color coded according to wind direction, and size coded according to wind speed as presented for Macon, Augusta, and Columbus.

5

4

3

2

1Win

d S

peed

(m

/s)

3002001000Wind Direction (degN)

Macon


5.4Augusta


6.9

Columbus

-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

dT/d

z (

K/m

)

Macon Augusta Columbus

100

80

60

40

20

0

O3

(ppb

v)

Macon Augusta Columbus

0.5

0.4

0.3

0.2

0.1

0.0

NO

/ N

O y

00:00 10:00 20:00

Time (EDT)

0.67

0.69

Macon

00:00 10:00 20:00

Time (EDT)

Augusta

00:00 10:00 20:00

Time (EDT)

Columbus

3. Aerosols

(a) Optical Properties and Mass ConcentrationFigure 32, Figure 33, and Figure 34 show hourly averaged PM2.5 concentrations and

visibility estimates for Macon, Augusta and Columbus respectively. The visibility, or visual length (xv), defined as the maximum distance that an object can be discerned against the background sky is estimated using the equation xv = 1.9 / (sp + ap + g) as described by Husar et al. (1994), where g is the extinction due to the scattering of gas molecules ( ~13 Mm-1), and sp and ap are the measurements made by the nephelometer and the PSAP, respectively. As shown in the figures, there is a significant amount of variability in both the PM2.5 concentration as well as the visibility in all of the FAQS cities on timescales ranging from hours to days. The variability on timescales of ~ hours is likely linked with several factors including shifts in wind direction and hence source regions, the timing of emissions of aerosols and their precursors (for example changes in traffic patterns throughout the day), and the diurnal variability in the atmospheric mixing height. The day-to-day variability in aerosol optical properties and PM2.5 is linked with changes in source regions and the amount of precipitation, which influences the scavenging of aerosols from the atmosphere. In all cases the hourly-average PM2.5 concentrations never exceed the proposed EPA daily-averaged PM2.5 standard of 65 gm-3. There is a clear relationship between visibility and PM2.5 with higher values of PM2.5 associated with lower visibilities. This is due to the fact that particles less than 2.5 m in diameter efficiently scatter and absorb visible solar radiation (Waggoner et al., 1981).

62

Figure 32 Hourly averaged PM2.5 mass concentrations and visibility measurements in Macon.

63

Figure 33 Hourly averaged PM2.5 mass concentrations and visibility measurements in Augusta.

Figure 34 Hourly averaged PM2.5 mass concentrations and visibility measurements in Columbus.

Table 14 shows mean values (with standard deviations given in parenthesis) of hourly-averaged aerosol properties for each of the cities during the FAQS study. The mean PM2.5 value in Macon is 10 g m-3, which is roughly a factor of two lower than the mean values in Augusta and Columbus of 19 g m-3, and 18 g m-3, respectively. The aerosol extinction (sum of sp and ap, which is ep) is dominated by light scattering for each city. The single scattering albedo, (which is the ratio of sp to the aerosol extinction coefficient, ep) ranges from 0.82 to 0.86 in the FAQS cities showing that more than 80% of the light extinction is due to light scattering by aerosols, with ~15-20% contributed by light absorption. As mentioned above there is a strong correlation between aerosol extinction and PM2.5, since it is the aerosol particles in the PM2.5 size range that efficiently scatter light. This is shown by the relatively high r2 values for the linear regression of sp (which dominates light extinction) versus PM2.5 for Macon, Augusta, and Columbus of 0.56, 0.84, and 0.73. The lowest r2 value occurs at Macon, and is likely due to the fact that the TEOM was run for several days at 30 deg. C, which resulted in a significant amount of variability in TEOM measurements. Nonetheless, the comparison of the light scattering and TEOM PM2.5 data shows a clear link between aerosol extinction and PM2.5. Therefore, a strong relationship also exists between PM2.5 and visibility, since the extinction measurements are used to estimate the visual length. The greatest mean visibility during the study is 47 km in Macon. The mean visibility in Augusta and Columbus are roughly a factor of 2 less than Macon, which is directly related to the PM2.5 values measured in these cities. Also shown in Table 14 are aerosol properties measured during the EPA funded Atlanta SuperSite study which took place during August, 1999. In general both the aerosol extinction and PM2.5 values are a factor of 2 greater than measurements made during the Fall Line Air Quality Study. In addition, the visual length in Atlanta during the SuperSite study is significantly less than the values measured during the FAQS study.

64

Table 14 Summary of FAQS Aerosol Optical Property Measurements.

City PM2.5 (g m-3) sp ap Visibility (km)

Macon 10 (3) 26 (11) 6 (4) 0.82 (0.06) 47 (14)

Augusta 19 (8) 67 (36) 10 (5) 0.85 (0.05) 27 (14)

Columbus 18 (6) 65 (23) 11 (5) 0.86 (0.06) 24 (7)

Atlanta1 32 (13) 120 (48) 16 (12) 0.87 (0.08) 18 (13)

*Values in paranthesis represent standard deviations1 Atlanta Supersite Data from August 1999 (M.H. Bergin et al.)

(b)Aerosol Chemical Composition (PILS)As the field study progressed, it became clear that the instrument was severely under

measuring the ambient concentrations of PM2.5 ionic species. This could not readily be determined during the study since the standard filter data was not available for comparison. However, comparisons with total PM2.5 mass (TEOM) suggested that the PILS was under measuring PM2.5. For example, Figure 35 shows the PILS measurement of total ion mass at Augusta compared to the total PM2.5 mass measured by the TEOM. Other studies have shown that in the Southeast, ionic species contribute typically about 50% to the total PM2.5. In Augusta the ionic mass was often only 20 to 30% of the total PM2.5. During the study, an attempt was also made to assess the PILS performance by performing filter measurements of sulfate. This comparison was not perfected until late in the study due to our inability to extract the filters in the field. Analysis of the field study data and subsequent laboratory testing has shown that the problems were primarily due to poor mixing of ambient air and steam in the PILS inlet. These problems have been corrected.

The poor performance of the PILS during FAQS limits the usefulness of a detailed analysis using the PILS data set. However, since the ICs operated correctly, the systematic error can be eliminated by reporting relative ion concentrations, that is the concentration of a specific ionic component relative to the total PILS measurement of all ions. Figure 36 shows the relative concentrations of sulfate and nitrate in Macon (no ammonium was available during this period). Figure 37 and Figure 38 show relative concentrations of sulfate, nitrate and ammonium during the Augusta and Columbus phases of the project.

65

Figure 35 Comparison of total ion mass measured by PILS to total PM2.5 mass measured with a TEOM in Augusta.

66

Figure 36 Relative ion concentrations measured by the PILS in Macon.

Figure 37 Relative ion concentrations measured by the PILS in Augusta.

Figure 38 Relative ion concentrations measured by the PILSs in Columbus.

In summary, although the PILS did not contribute to the first phase of FAQS, the study provided us the chance to test and further refine the instrument. Based on the findings from FAQS, the PILS has been modified and further testing has shown that it now operates reliably. We believe that the investments in developing this instrument during the first phase of FAQS will pay off in the upcoming FAQS deployments by providing a unique tool to help assess the sources of PM2.5 in the Southeast.

(c) Aerosol Chemical Composition and Size Distribution (Aerosol Mass Spectrometer)Presented here are preliminary results giving a general picture of aerosol composition

observed during FAQS, including intercomparisons with chemical analysis from collected filters. Only mass loading results are presented; size distributions are still being analyzed.The AMS obtained aerosol data in all three cities. Unfortunately, during the initial deployment in Macon, a misalignment of the particle beam significantly reduced the overall detection sensitivity. That low sensitivity, combined with the low ambient aerosol loading that was observed during the Macon deployment, limited the usefulness of the AMS data acquired in Macon, other than confirming the low mass loading. The beam alignment was properly adjusted before the start of the Augusta deployment and results are presented for mass loading during Augusta (26 June to 10 July 2000) and Columbus (15-27 July 2000).

Figure 39 compares measured AMS mass loading with that measured via filter analysis. As labeled, the AMS is separated into inorganic and organic fractions, shaded in dark green and blue, respectively. It must be emphasized that these AMS results are still preliminary, since final calibrations have not been applied; in particular, daily fluctuations of the quadrupole detector have not yet been accounted for. Based on this FAQS deployment, these procedures have been improved and will be applied in near real time in the future. There are two traces plotting filter results in Figure 39. The top trace (in black) represents the total mass reported while the lower (red) trace sums inorganic species and organic carbon (OC). The data points represent 12 or 24-hour collection samples. The lower trace likely matches the semivolatile and volatile components that are detected by the AMS. In fact the lower filter trace and the total AMS mass loadings track fairly well, following similar trends during Augusta and Columbus. The difference between the two filter traces corresponds to unidentified composition, likely representing insoluble inorganic or dust (e.g. crustal oxides) which are not speciated in the filter analysis nor thermally vaporized in the AMS detector.

Correlations of AMS and filter derived mass loadings are further plotted in Figure 40, with ammonium, nitrate, sulfate and organic loadings as labeled. For both Augusta and Columbus, the agreement between AMS and filter results is reasonable. As plotted, the organic and sulfate AMS loadings are somewhat low, though those absolute values may change upon final calibration and analysis. It should also be noted that the AMS was not always operating over the full 12 or 24-hour collection period for the filter samples, likely contributing to some of the differences (including outlying points). The picture of ambient aerosol given by both the AMS and filter results is similar: mass loadings of 20 to 30 gm-3, roughly equal contributions of inorganic and organic species, with roughly 20-30% not identified (based on total filter samples). Inorganic components were slightly larger than organic in Augusta and vice versa in Columbus. Overall the AMS and filter analyses are complementary, with the filter giving a total mass loading and integrated analysis of a number of ionic species, while the AMS provides a measure of short term variability. Upon final calibration of the AMS results, detailed correlations of the AMS results with meteorological and gas species results will be analyzed on the 15-minute time

67

scale of the AMS samples. This also will include variations in aerosol size distributions, which have not yet been analyzed.

Finally, Figure 41 plots a representative mass spectrum (taken from the morning of 16 July 2000 in Columbus) showing the type of complex chemical analysis that has yet to be fully analyzed. As labeled, major inorganic species (ammonium, nitrate and sulfate) have clearly resolved fragmentation patters, while organic analysis is more complex. For example, the mass series [41, 55, 69, … ] and [43,57,71, …] represent alkyl and oxy-alkyl series (e.g. CnH2n-1

+ and CnH2n+1O+), respectively; while the mass 44 peak is largely due to CO2

+, a degradation product of carboxylic acids. Precise identification requires detailed investigation of multiple mass peaks as well as comparison chemical analysis using other techniques. Mass spectral identification of chemical classes is an on-going process that will be applicable to the FAQS data in the future.

68

Figure 40 Correlation of AMS and filter species in Augusta and Columbus.

Figure 41 Representative AMS mass spectrum during FAQS.

25

20

15

10

5

0

g m

-3 A

MS

2520151050

g m-3 Filters

Tota l (Inorg ani c + OC)

organicsulfateammoniumnitrate

Columbus

25

20

15

10

5

0

g m

-3 A

MS

2520151050

g m-3 Filters

Tota l (In orga nic + OC)organicsulfa teammoniumnitra te

Augusta

1.0

0.8

0.6

0.4

0.2

0.0200180160140120100806040200

AMU

ammonium (NH+,NH2+)

nitrate (NO+,NO2+)

sulfate (SO+,SO2+,SO3+,....)

organic

Columbus 7/16/00

16g m-3 6:00-9:00AM 43 44

55

5769

71169

41

4. Volatile Organic Compounds (VOCs)

For each of the 83 chemical compounds analyzed, Table 15, Table 16, and Table 17 show the mean concentration, median concentration, maximum concentration, minimum concentration, and standard deviation of the air samples collected ~4 times per day in canisters at the Macon, Augusta, and Columbus sites of the AQR lab. These samples were analyzed at UC, Irvine. In Augusta on 3 July 2000, the canister sample collected at 23:35 (11:35 PM) LDT showed unusual and extremely high concentrations of many compounds. While the reason for this anomaly has not been confirmed, likely these results may have been influenced by the fireworks display that was held adjacent to the monitoring site at Fort Gordon on this day beginning shortly after dusk. Thus, the statistical indicators for Augusta in Table 16 are shown both with and without this sample.

For just the nonmethane hydrocarbons, Table 18 shows the average VOC concentrations in ppbC that were collected over the ten day canister sampling period in each FAQS city, and compares it to the average VOC concentrations observed at two Photochemical Assessment Monitoring Sites (PAMS) in Atlanta during the summer of 1997. The PAMS sites are part of the GA EPD statewide monitoring network. Although none of these data are concurrent and the FAQS sampling is extremely limited, the data seem to suggest that a relatively consistent mix of hydrocarbons may be present across the state. In particular concentrations of isoprene, a highly reactive chemical compound that is primarily emitted from natural sources such as trees, crops, and other vegetation, is roughly about the same or a little higher in the FAQS cities than it is in the Atlanta area. Concentrations of all the other hydrocarbons (most of which are human-made) in the Fall Line cities run less than or equal to the hydrocarbon concentrtrations that were observed in Atlanta. This is a key finding since several studies (with most tracing back to Chameides et al. 1988 and Trainer et al. 1987) have confirmed that when one considers the presence of natural hydrocarbons and the relative reactivity of all the different hydrocarbons (meaning a measure of a hydrocarbon’s contribution to the formation of ozone relative to another hydrocarbon, Chameides et al. 1992), the atmosphere in Atlanta has been found to be saturated with isoprene. That is, there is more “fuel” for producing ozone in the form of naturally occurring isoprene than the recipe for ozone (VOC + NOx) can use. Ultimately what this suggests for Macon, Augusta, and Columbus, is that the amount of ozone that can be formed locally via the photochemical process that involves VOCs and NOx, may be limited by how much NOx is emitted from various local sources. While this analysis is preliminary and must be confirmed with more comprehensive data and more rigorous diagnosis, if upheld, this would likely mean that a NOx based emission control strategy would be the most effective approach to decreasing local ozone contributions.

69

Table 15 Summary of the mixing ratio (pptv except where noted) of volatile organic and other chemical compounds in air samples collected 4-times per day in Macon from 11 – 21 June 2000.Chemical Compound Average Median Maximum Minimum Standard

DeviationCH4 (ppmv) 1.8 1.8 1.9 1.7 0.0CO (ppbv) 196.7 184.5 519.0 126.0 66.2F-12 543.6 541.6 572.0 535.9 7.2CH3Cl 615.0 583.5 872.0 516.0 93.8F-114 16.8 16.9 18.8 15.0 0.8H-1211 4.1 4.0 5.0 3.8 0.2MeBr 12.3 12.1 16.5 9.4 1.6F-11 268.6 264.4 331.0 257.8 14.1F-113 117.8 108.9 210.8 84.9 34.0CH2Cl2 39.0 35.1 131.7 18.7 19.8CHCl3 13.0 12.0 28.9 7.4 4.9MeCCl3 52.4 52.1 56.9 50.7 1.3CCl4 98.6 98.4 104.5 96.8 1.3C2Cl4 11.3 6.5 116.3 3.5 18.2MeI 2.7 2.2 8.9 1.4 1.4H-2402 0.5 0.5 0.6 0.5 0.0MeONO2 2.5 2.5 4.3 1.6 0.4EtONO2 2.5 2.2 6.3 1.6 0.9I-PrONO2 4.1 3.5 13.3 2.1 1.9C2HCl3 0.6 0.4 4.3 0.2 0.8CH2Br2 1.6 1.6 2.3 0.8 0.4n-PrONO2 0.5 0.4 2.3 0.2 0.42-BuONO2 2.0 1.7 8.7 0.8 1.4CHBr3 3.8 3.9 6.6 0.9 1.5ethane 745.8 691.0 1539.0 444.0 255.9ethene 457.0 343.5 3545.0 49.0 557.5ethyne 299.8 247.0 856.0 85.0 191.1propane 669.2 416.5 2448.0 90.0 605.9propene 156.5 106.0 1286.0 23.0 201.5i-butane 224.9 78.5 4880.0 12.0 768.0n-Butane 974.6 216.5 26566.0 42.0 4169.7i-pentane 927.1 360.0 16231.0 53.0 2563.0n-pentane 329.5 138.0 5776.0 20.0 906.0hexane 95.6 57.0 710.0 17.0 141.0isoprene 2207.8 2153.5 8103.0 106.0 1710.7heptane 213.4 27.0 5960.0 10.0 1031.8octane 120.2 23.0 1906.0 13.0 420.5benzene 107.3 77.5 907.0 29.0 138.4toluene 442.9 222.5 5551.0 64.0 907.8ethylbenzene 67.2 29.0 932.0 10.0 149.7m-xylene 213.5 70.0 3873.0 21.0 611.8p-xylene 89.1 33.0 1343.0 10.0 215.9o-xylene 97.2 40.0 1220.0 16.0 198.31,3-butadiene 57.8 43.0 130.0 33.0 40.9t-2-butene 83.9 14.5 767.0 10.0 215.7

70

Table 15 continued.Chemical Compound Average Median Maximum Minimum Standard

Deviationcis-2-butene 85.1 12.0 704.0 10.0 206.0 1-butene 42.2 30.0 390.0 13.0 63.93-methyl-1-butene 92.3 34.0 233.0 10.0 122.41-pentene 56.0 18.0 606.0 11.0 124.52-methyl-1-butene 82.1 23.5 980.0 13.0 197.8t-2-pentene 97.7 17.5 1061.0 10.0 234.9c-2-pentene 77.5 19.0 618.0 10.0 164.82-methyl-2-butene 121.9 28.0 1523.0 12.0 313.32,2-dimethylbutane 48.9 23.0 389.0 10.0 82.12,3-dimethylbutane 160.0 69.5 2251.0 30.0 362.9 4met1pente 77.2 19.0 240.0 13.0 97.52-methylpentane 716.5 344.0 8267.0 124.0 1352.73-methylpentane 294.1 118.5 4537.0 43.0 725.22,4-dimethylpentane 57.9 45.0 180.0 22.0 47.92-methylhexane 218.7 79.5 1291.0 47.0 382.12,3-dimethylpentane 93.0 50.5 295.0 19.0 99.13-methylhexane 232.8 84.0 5394.0 36.0 849.82,2,4-trimethylpentane 239.7 116.0 1619.0 34.0 323.5n-heptane 498.0 55.0 16987.0 14.0 2710.12,5-dimethylhexane 270.3 84.0 1406.0 32.0 502.62,3,4-trimethylpentane 85.0 55.5 365.0 15.0 86.02-methylheptane 193.5 47.0 653.0 27.0 306.63-methylheptane 954.0 954.0 1896.0 12.0 1332.2nonane 100.5 100.5 168.0 33.0 95.5isopropylbenzene 38.0 38.0 38.0 38.0alpha-pinene 457.5 208.0 3415.0 50.0 620.3propylbenzene 58.8 57.5 100.0 20.0 33.63-etyltoluene 128.5 83.0 669.0 20.0 124.64-ethyltolene 70.1 43.0 219.0 16.0 65.3135-trimethylbenzene 51.3 38.5 158.0 11.0 44.42-ethyltoluene 60.5 39.5 232.0 21.0 60.2124-trimethylbenzene 87.5 48.0 557.0 23.0 107.5p-cymene/1,2,3-trimethylbenzene 52.6 45.0 119.0 10.0 28.2Limonene 138.4 92.5 750.0 18.0 130.21,3-diethylbenzene 47.4 26.5 154.0 18.0 45.51,4-diethylbenzene 38.5 22.5 94.0 14.0 34.21,2-diethybenzene 33.3 30.5 55.0 17.0 17.6 2-methyl-1-pentene 44.2 25.0 132.0 16.0 49.3

71

Table 16 Summary of the mixing ratio (pptv except where noted) of volatile organic and other chemical compounds in air samples collected 4-times per day in Augusta from 29 June – 10 July 2000.Chemical Compound

with 3 July 2000 23:35 Anomaly without 3 July 2000 23:35 AnomalyAvg Med Max Min Stdev Avg Med Max Min Stdev

CH4 (ppmv) 1.9 1.8 2.6 1.8 1.9 1.8 2.1 1.8 0.1CO (ppbv) 1381.7 229.0 44983.0 147.0 7072.4 263.7 226.0 982.0 147.0 153.3F-12 544.4 542.1 600.3 530.2 12.9 543.0 541.9 581.8 530.2 9.3CH3Cl 598.7 595.1 786.6 493.5 64.2 597.5 589.5 786.6 493.5 64.6F-114 15.8 15.7 17.7 14.1 0.9 15.8 15.7 17.7 14.1 0.9H-1211 4.3 4.3 5.4 4.0 0.3 4.3 4.2 5.4 4.0 0.3MeBr 14.7 13.0 35.0 9.7 5.1 14.6 12.9 35.0 9.7 5.2F-11 309.3 261.8 2013.4 251.5 276.7 309.7 261.7 2013.4 251.5 280.3F-113 147.1 140.3 240.0 80.7 33.7 146.9 139.7 240.0 80.7 34.1CH2Cl2 79.5 35.8 958.2 13.8 155.2 56.9 35.4 326.5 13.8 62.3CHCl3 20.2 16.1 57.1 11.1 11.3 19.4 16.1 57.1 11.1 10.4MeCCl3 54.3 53.6 71.2 50.8 3.5 53.9 53.5 62.9 50.8 2.2CCl4 101.0 100.7 114.6 96.8 2.9 100.7 100.5 103.9 96.8 1.9C2Cl4 17.4 15.6 47.8 6.9 7.5 17.2 15.0 47.8 6.9 7.5MeI 1.3 1.3 2.7 0.7 0.5 1.3 1.3 2.7 0.7 0.5H-2402 0.5 0.5 0.6 0.5 0.0 0.5 0.5 0.6 0.5 0.0MeONO2 4.5 3.7 35.4 2.0 5.0 3.7 3.7 5.7 2.0 0.6EtONO2 6.4 5.4 40.5 2.0 5.6 5.5 5.4 7.4 2.0 0.9I-PrONO2 13.0 11.3 74.7 4.9 10.4 11.5 11.3 19.8 4.9 2.8C2HCl3 1.4 1.0 6.1 0.4 1.2 1.4 0.9 6.1 0.4 1.2CH2Br2 1.0 1.0 1.7 0.9 0.1 1.0 1.0 1.7 0.9 0.1n-PrONO2 1.7 1.5 4.8 0.1 0.7 1.6 1.5 2.7 0.1 0.52-BuONO2 8.8 6.8 56.6 0.7 8.3 7.6 6.6 15.9 0.7 3.0CHBr3 1.2 1.1 4.0 0.0 0.7 1.2 1.1 4.0 0.0 0.7ethane 6220.2 1387.0 185987.0 759.0 29547.0 1489.4 1376.5 2957.0 759.0 477.9ethene 37298.0 318.0 1469530.0 94.0 232264.1 574.1 317.0 5340.0 94.0 929.9ethyne 9139.3 279.5 344777.0 114.0 54435.7 533.2 276.0 5098.0 114.0 819.4propane 1234.5 602.0 18490.0 189.0 2869.2 792.0 587.0 3844.0 189.0 642.1propene 10499.7 111.5 413096.0 32.0 65288.7 176.7 109.0 1535.0 32.0 249.3i-butane 766.9 88.5 24892.0 22.0 3916.2 148.3 88.0 956.0 22.0 177.3n-Butane 3753.1 215.5 136540.0 61.0 21544.9 348.3 199.0 4479.0 61.0 699.8i-pentane 24807.6 276.0 951569.0 80.0 150332.9 1044.5 258.0 22596.0 80.0 3588.7n-pentane 6347.3 122.5 239245.0 32.0 37789.5 375.6 119.0 8062.0 32.0 1273.6hexane 1969.9 36.0 61300.0 17.0 10494.7 172.0 34.0 2262.0 17.0 495.4isoprene 1737.5 1101.5 12222.0 104.0 2015.6 1468.7 1101.0 4762.0 104.0 1096.6heptane 176.5 27.0 4485.0 10.0 787.1 37.5 27.0 212.0 10.0 39.1octane 197.5 18.0 1982.0 10.0 517.1 98.4 17.0 1259.0 10.0 292.3benzene 1115.6 81.0 37961.0 35.0 5982.1 170.8 79.0 1779.0 35.0 292.3toluene 1324.8 239.0 37073.0 40.0 5833.3 408.1 229.0 4067.0 40.0 656.5ethylbenzene 120.4 24.0 2408.0 10.0 418.9 48.9 22.5 472.0 10.0 83.9m-xylene 303.5 50.0 7304.0 10.0 1168.8 119.3 49.5 1250.0 10.0 209.1p-xylene 109.1 22.0 2220.0 5.0 361.2 53.5 21.5 615.0 5.0 101.9o-xylene 144.8 34.0 2964.0 11.0 483.0 68.6 33.0 667.0 11.0 114.01,3-butadiene 6680.0 332.0 19596.0 112.0 11186.1 222.0 222.0 332.0 112.0 155.6t-2-butene 5143.3 196.5 20169.0 11.0 10017.9 134.7 90.0 303.0 11.0 151.0

72

Table 16 continued.Chemical Compound

with 3 July 2000 23:35 Anomaly without 3 July 2000 23:35 AnomalyAvg Med Max Min Stdev Avg Med Max Min Stdev

cis-2-butene 4111.8 95.0 20144.0 11.0 8963.0 103.8 55.0 294.0 11.0 132.6 1-butene 2912.5 28.0 86234.0 15.0 15737.0 39.3 27.0 279.0 15.0 50.33-methyl-1-butene 4800.0 197.0 14179.0 24.0 8122.9 110.5 110.5 197.0 24.0 122.31-pentene 1802.5 22.0 29537.0 12.0 7149.1 69.1 22.0 731.0 12.0 176.82-methyl-1-butene 1843.3 21.0 19166.0 10.0 5750.6 111.0 21.0 849.0 10.0 260.2t-2-pentene 5324.3 58.5 30403.0 10.0 12298.2 308.6 28.0 1403.0 10.0 612.6c-2-pentene 4572.5 423.0 17423.0 21.0 8574.3 289.0 58.0 788.0 21.0 432.52-methyl-2-butene 3665.2 18.0 48592.0 11.0 12944.8 209.3 18.0 2289.0 11.0 626.32,2-dimethylbutane 1464.8 17.0 31429.0 10.0 6536.6 102.8 16.5 954.0 10.0 251.82,3-dimethylbutane 1656.2 63.0 47853.0 24.0 8098.6 297.5 62.5 5808.0 24.0 1000.7 4methy1pentene 0.0 0.0 0.0 0.02-methylpentane 5334.1 328.5 162152.0 76.0 25850.0 1313.2 313.0 29255.0 76.0 4696.63-methylpentane 3210.9 72.0 96526.0 11.0 15937.9 618.8 72.0 14053.0 11.0 2360.12,4-dimethylpentane 886.1 26.0 13096.0 10.0 3258.5 72.1 25.0 468.0 10.0 131.82-methylhexane 1481.0 87.0 31000.0 16.0 6007.4 345.6 86.5 5951.0 16.0 1156.42,3-dimethylpentane 685.3 46.0 15151.0 17.0 3016.4 82.5 44.5 619.0 17.0 130.33-methylhexane 1319.2 106.0 33967.0 30.0 5705.6 436.8 106.0 10714.0 30.0 1746.42,2,4-trimethylpentane

1427.8 55.5 40408.0 14.0 7126.3 170.4 55.0 2369.0 14.0 440.8

n-heptane 1813.2 65.0 28673.0 11.0 6160.4 1275.8 62.0 28673.0 11.0 5599.92,5-dimethylhexane 841.5 43.0 6704.0 16.0 2001.4 352.9 42.5 3497.0 16.0 992.32,3,4-trimethylpentane

886.4 37.0 8125.0 10.0 2545.5 82.1 32.0 283.0 10.0 109.5

2-methylheptane 902.9 119.0 4817.0 18.0 1769.1 868.3 87.0 4817.0 18.0 1935.43-methylheptane 956.4 109.0 2593.0 69.0 1209.0 716.5 102.0 2593.0 69.0 1251.1nonane 74.2 40.0 209.0 21.0 78.3 40.5 33.0 75.0 21.0 24.4isopropylbenzene 71.0 71.0 119.0 23.0 67.9 23.0 23.0 23.0 23.0alpha-pinene 155.6 101.0 1645.0 27.0 255.5 156.2 100.0 1645.0 27.0 258.9propylbenzene 109.5 75.0 275.0 13.0 119.7 54.3 31.0 119.0 13.0 56.73-etyltoluene 139.4 64.0 1124.0 22.0 234.2 94.6 63.0 455.0 22.0 95.94-ethyltolene 54.4 21.0 436.0 11.0 97.7 35.3 20.5 199.0 11.0 44.7135-trimethylbenzene

104.0 34.0 460.0 17.0 155.0 53.1 31.0 190.0 17.0 62.2

2-ethyltoluene 69.5 25.0 413.0 14.0 112.0 40.9 24.0 176.0 14.0 45.5124-trimethylbenzene

118.2 52.0 1164.0 22.0 226.7 80.8 52.0 563.0 22.0 106.5

p-cymene/1,2,3-trimethylbenzene

59.6 36.0 319.0 11.0 74.7 50.9 34.5 319.0 11.0 61.9

Limonene 62.5 44.5 280.0 18.0 59.0 56.4 44.0 280.0 18.0 51.01,3-diethylbenzene 78.3 51.0 231.0 10.0 82.9 72.4 36.0 231.0 10.0 91.21,4-diethylbenzene 105.4 49.0 299.0 13.0 108.4 89.3 47.0 299.0 13.0 109.11,2-diethybenzene 89.5 89.5 168.0 11.0 111.0 89.5 89.5 168.0 11.0 111.0 2-methyl-1-pentene 2728.7 280.0 7882.0 24.0 4464.8 152.0 152.0 280.0 24.0 181.0

73

Table 17 Summary of the mixing ratio (pptv except where noted) of volatile organic and other chemical compounds in air samples collected 4-times per day in Columbus from 17 – 29 July 2000.Chemical Compound Average Median Maximum Minimum Standard

DeviationCH4 (ppmv) 1.9 1.8 2.0 1.6 0.1CO (ppbv) 356.0 336.5 766.0 225.0 108.4F-12 544.5 543.0 578.4 529.0 9.3CH3Cl 627.3 619.5 825.9 478.6 73.3F-114 16.5 16.4 18.2 15.1 0.6H-1211 4.3 4.3 4.8 3.9 0.2MeBr 14.8 12.9 35.2 9.7 5.2F-11 268.4 265.0 393.6 249.0 21.3F-113 172.2 94.2 2970.2 83.2 454.1CH2Cl2 46.2 41.6 144.1 20.0 25.5CHCl3 25.5 17.3 100.6 10.6 20.3MeCCl3 52.1 52.3 55.0 48.1 1.3CCl4 98.8 98.8 106.0 95.2 1.7C2Cl4 14.4 12.4 54.3 4.2 8.4MeI 2.2 2.1 5.8 1.0 0.9H-2402 0.5 0.5 0.5 0.5 0.0MeONO2 3.9 3.7 5.5 2.1 0.7EtONO2 5.5 5.4 10.1 2.0 1.4I-PrONO2 11.9 11.6 22.1 3.6 3.3C2HCl3 2.6 2.0 13.0 0.2 2.7CH2Br2 0.8 0.8 1.3 0.7 0.1n-PrONO2 1.8 1.7 3.6 0.4 0.62-BuONO2 8.3 8.0 19.3 1.5 3.0CHBr3 1.5 1.1 7.6 0.5 1.3ethane 1606.8 1416.0 3665.0 698.0 616.6ethene 596.1 485.5 1829.0 76.0 340.2ethyne 483.0 400.0 1083.0 148.0 231.0propane 912.9 784.0 2362.0 311.0 464.1propene 182.2 146.5 580.0 25.0 114.6i-butane 158.0 123.5 472.0 47.0 97.2n-Butane 401.4 305.5 1757.0 108.0 300.7i-pentane 591.4 446.5 3026.0 98.0 508.7n-pentane 207.0 144.0 874.0 43.0 158.0hexane 61.1 42.5 171.0 15.0 39.6isoprene 1771.9 1265.0 7829.0 55.0 1826.0heptane 21.9 14.0 102.0 10.0 19.5octane 20.4 17.0 65.0 12.0 12.3benzene 146.3 123.0 371.0 34.0 69.8toluene 309.5 219.0 1547.0 48.0 272.2ethylbenzene 32.5 30.0 102.0 12.0 17.9m-xylene 82.5 57.5 339.0 16.0 65.6p-xylene 34.1 26.0 119.0 8.0 23.9o-xylene 43.7 32.5 125.0 11.0 26.81,3-butadiene 46.5 46.5 100.0 17.0 29.8t-2-butene 15.8 14.5 25.0 10.0 5.1

74

Table 17 continued.Chemical Compound Average Median Maximum Minimum Standard

Deviationcis-2-butene 16.4 14.0 43.0 10.0 9.2 1-butene 39.7 37.0 73.0 17.0 15.53-methyl-1-butene 20.5 15.0 41.0 11.0 14.21-pentene 30.1 23.5 99.0 10.0 19.42-methyl-1-butene 31.4 25.0 142.0 10.0 26.8t-2-pentene 26.6 18.0 79.0 10.0 17.8c-2-pentene 23.6 20.5 58.0 10.0 12.32-methyl-2-butene 29.7 23.5 80.0 13.0 17.12,2-dimethylbutane 26.4 22.0 80.0 11.0 17.92,3-dimethylbutane 62.1 54.0 148.0 12.0 28.8 4methy1pentene 14.5 14.5 15.0 14.0 0.72-methylpentane 405.4 365.0 973.0 104.0 218.33-methylpentane 101.8 86.0 236.0 50.0 44.72,4-dimethylpentane 40.7 43.0 63.0 16.0 23.62-methylhexane 59.3 45.0 105.0 28.0 40.52,3-dimethylpentane 41.0 41.0 41.0 41.03-methylhexane 80.1 72.0 188.0 29.0 37.72,2,4-trimethylpentane 77.9 68.0 197.0 15.0 44.8n-heptane 49.2 41.5 221.0 19.0 38.92,5-dimethylhexane 28.4 26.0 45.0 15.0 10.52,3,4-trimethylpentane 39.9 36.5 57.0 30.0 10.12-methylheptane 30.5 30.5 34.0 27.0 4.93-methylheptane 0.0 0.0nonane 0.0 0.0isopropylbenzene 0.0 0.0alpha-pinene 243.1 162.5 844.0 44.0 200.9propylbenzene 0.0 0.03-etyltoluene 74.4 54.5 242.0 19.0 54.54-ethyltolene 30.8 29.0 53.0 14.0 13.0135-trimethylbenzene 24.8 23.0 44.0 12.0 11.02-ethyltoluene 29.1 29.5 48.0 11.0 11.0124-trimethylbenzene 51.0 45.5 147.0 16.0 29.7p-cymene/1,2,3-trimethtlbenzene 42.2 29.0 134.0 10.0 32.2Limonene 76.2 72.0 287.0 14.0 51.41,3-diethylbenzene 17.0 17.0 17.0 17.01,4-diethylbenzene 30.5 30.5 32.0 29.0 2.11,2-diethybenzene 0.0 0.0 2-methyl-1-pentene 15.5 15.5 17.0 14.0 2.1

75

Table 18 Comparison of average VOCs from FAQS pilot study with VOCs from two 1997 Atlanta PAMS stations (all concentrations in ppbC).Chemical Compound South Dekalb

(Atlanta)Conyers (Atlanta)

Macon Augusta (w/o 7/3 anomoly)

Columbus

ethane 5.62 4.602 1.492 2.979 3.214ethene 4.442 1.08 0.914 1.148 1.192ethyne 4.266 1.589 0.600 1.066 0.966propane 6.371 3.494 2.007 2.376 2.739propene 2.402 0.721 0.469 0.530 0.5471-butene 0.651 0.297 0.169 0.157 0.1591,3-butadiene 0.231 0.888 0.186cis-2-butene 0.235 0.045 0.340 0.415 0.066i-butane 2.165 0.745 0.900 0.593 0.632n-butane 4.266 1.589 3.898 1.393 1.606t-2-butene 0.388 0.077 0.336 0.539 0.0631-pentene 0.533 0.101 0.280 0.346 0.1512-methyl-1-butene 0.411 0.555 0.1572-methyl-2-butene 0.609 1.047 0.1483-methyl-1-butene 0.462 0.553 0.103c-2-pentene 0.29 0.023 0.388 1.445 0.118i-pentane 7.425 3.343 4.635 5.222 2.957isoprene 7.308 5.407 11.039 7.343 8.859n-pentane 3.877 1.373 1.647 1.878 1.035t-2-pentene 0.558 0.053 0.489 1.543 0.1332-methyl-1-pentene 0.082 0.077 0.265 0.912 0.0934-methyl-1-pentene 0.463 0.0872,2-dimethylbutane 0.293 0.617 0.1582,3-dimethylbutane 1.113 0.175 0.960 1.785 0.3722-methylpentane 2.707 0.472 4.299 7.879 2.4323-methylpentane 1.953 0.36 1.765 3.713 0.611benzene 2.398 1.02 0.644 1.025 0.878hexane 1.637 0.446 0.574 1.032 0.3672,3-dimethylpentane 0.771 0.194 0.651 0.578 0.2872,4-dimethylpentane 0.494 0.112 0.405 0.504 0.2852-methylhexane 0.671 0.28 1.531 2.419 0.4153-methylhexane 0.952 0.514 1.630 3.057 0.561heptane 1.494 0.263 0.153n-heptane 0.55 0.225 3.486 8.931 0.344toluene 6.797 2.305 3.100 2.857 2.1662,2,4-trimethylpentane 2.998 0.836 1.917 1.363 0.6232,3,4-trimethylpentane 1.046 0.285 0.680 0.657 0.3192,5-dimethylhexane 2.162 2.823 0.2272-methylheptane 0.165 0.092 1.548 6.947 0.2443-methylheptane 0.202 0.094 7.632 5.732ethylbenzene 1.028 0.401 0.538 0.391 0.260m-xylene 3.474 1.04 1.708 0.955 0.660octane 0.376 0.162 0.962 0.787 0.164o-xylene 1.422 0.444 0.778 0.549 0.349p-xylene 0.712 0.428 0.2731,2,4-trimethylbenzene 1.614 0.588 0.788 0.727 0.4591,3,5-trimethylbenzene 0.441 0.169 0.461 0.478 0.2232-ethyltoluene 0.376 0.156 0.545 0.368 0.2623-ethyltoluene 0.604 0.021 1.156 0.852 0.669

76

Table 18 continued.Chemical Compound South Dekalb

(Atlanta)Conyers (Atlanta)

Macon Augusta (w/o 7/3 anomoly)

Columbus

4-ethyltoluene 1.251 1.556 0.631 0.318 0.278isopropylbenzene 0.044 0.028 0.342 0.207nonane 0.176 0.156 0.905 0.3651,2,3-trimethylbenzene 1.242 0.905 0.473 0.458 0.380propylbenzene 0.132 0.096 0.529 0.4891,2-diethylbenzene 0.333 0.8951,3-diethylbenzene 0.474 0.724 0.1701,4-diethylbenzene 0.322 0.114 0.385 0.893 0.305alpha-pinene 1.255 1.254 4.575 1.562 2.431limonene 1.384 0.564 0.762

77

5. Observations of Ozone and PM2.5 at the Three Satellite Monitoring Stations

As described earlier in section III.A.1, ozone and fine particulate mass monitors were added to new stations in each of the three cities. These new stations were first made operational during the pilot study and continued to record data after the pilot study was completed. Due to a lack of high ozone concentrations during the winter months, the ozone monitors were shutdown in November. These will be restored to operation in the Spring of 2001. The TEOM instruments for measuring PM2.5 mass continued to collect data through the winter. The site at Macon Sandy Beach Park in Bibb County proved to be the most stable over the startup period and subsequent monitoring campaign. In contrast, the other monitors at Lakeside-Evans High School in Columbia County near Augusta, and Columbus Oxbow Environmental Learning Center in Muscogee County suffered from unavoidable downtime. Once school operations began in earnest in mid August, it became apparent that the monitoring station at Lakeside High School was susceptible to deliberate and accidental damage if left at its initial location. As a result the monitor was down for much of August while the shelter was relocated to a more secure location at the school. Almost concurrently, monitoring operations at the Oxbow site were interrupted for repairs that were necessitated by a lightening strike that damaged equipment.

To date, scant ozone and no particulate analysis has been completed using the satellite data. The greatest value afforded by these sites however, is their spatial separation from the GA EPD’s monitors. The intent of locating these sites opposite to the existing EPD monitors, was potentially to be able to discern local pollutant contributions from regional or transported pollutant contributions by observing the differences between an upwind monitoring site and a downwind monitoring site. For example, Figure 42 shows the ozone concentrations for 15-19 July 2000 that were observed at the GA EPD site at the Georgia Forestry Commission east of Macon, the FAQS site at Sandy Beach Park west of Macon, at the GA EPD rural site in Leslie, GA in South Central Georgia, and the wind direction as measured by the National Weather Service at Wilson Airport south of downtown Macon. In general, the two ozone monitors in Macon track each other fairly well, but show considerably higher concentrations than what was observed in rural Leslie. Contrast this with Figure 43 that shows corresponding ozone concentrations and wind directions for 15-18 August 2000. For the first two days, ozone concentrations at both monitors in Macon seem to track the values recorded by the rural Leslie monitor. On the third day however, 17 August, ozone concentrations observed at the GA EPD monitor dramatically separate from values seen at both the rural Leslie monitor and the cross town FAQS monitor. Although the peak ozone concentration observed at the Georgia Forestry site on 17 August is the all time highest value ever recorded in Macon, the cross town site was not similarly impacted.

While it is not possible to draw any conclusions from this limited analysis about why the two monitors in Macon diverged, these observations do raise at least two interesting questions for which answers will be pursued actively in subsequent phases of the FAQS, and passively via the continued and long-term operation of the satellite sites in all three cities. Is the 17 August event indicative of the incremental contribution (~0.030 ppmv) that the Macon metropolitan area can add to the regional background air? Or is the location of the GA EPD monitor at the Georgia Forestry Commission, somehow influenced or biased by highly local sources of some kind as suggested by several stakeholders in Macon? Similar questions are also relevant to Augusta and Columbus.

78

79

Figure 42 Ozone concentrations (1-hour average) observed in Macon and Leslie, Georgia, and wind direction observed in Macon on 15-19 July 2000.

Figure 43 Ozone concentrations (1-hour average) observed in Macon and Leslie, Georgia, and wind direction observed in Macon on 15-18 August 2000.

References for Section III

Atlas, E.L., et al., Partitioning and budget of NOy species during the Mauna Loa Observatory Photochemistry Experiment, J. Geophys. Res., 97, 10,449-10,462, 1992.

Baumann, K., et al., Ozone production and transport near Nashville, Tennessee: results from the 1994 study at New Hendersonville, J. Geophys. Res.,105, 9137-9153, 2000.

Black, T. “The new NMC mesoscale ETA model: description and forecast examples,” Wea. Forecast. 9, 265-278, 1994.

Bollinger, M.J., Chemiluminescent measurements of the oxides of nitrogen in the clean troposphere and atmospheric chemistry implications, Ph.D. thesis, Chem. Dep., Univ. of Colo., Boulder, 1982.

Bond, T.C., T.L. Anderson, D. Campbell, Calibration intercomparison of filter-based measurements of visible light absorption by aerosols, Aerosol Sci. Technol., 30, 582-600, 1999.

Chameides, W.L., R.W. Lindsay, J. Richardson, and C.S. Kiang, The role of biogenic hydrocarbons in urban photochemical smog: Atlanta as a case study, Science, 241, 1473-1475, 1988.

Chameides, W.L., F. Fehsenfeld, M.O. Rodgers, C.Cardelino, J. Martinez, D. Parrish, W. Lonneman, D.R. Lawson, R.A. Rasmussen, P. Zimmerman, J. Greenberg, P. Middleton, and T.Wang, Ozone Precursor Relationships in the Ambient Atmosphere, J. Geophys. Res., 97, 6037-6055, 1992.

Fahey, D.W., C.S. Eubank, G. Huebler,, and F.C. Fehsenfeld, Evaluation of a catalytic reduction technique for the measurement of total reactive odd-nitrogen (NOy) in the atmosphere, J. Atmos. Chem., 3, 435-468, 1985.

Fahey, D.W., et al., Reactive nitrogen species in the troposphere: Measurements of NO, NO2, HNO3, particulate nitrate, peroxyacetyl nitrate (PAN), O3, and total reactive odd nitrogen (NOy) at Niwot Ridge, Colorado, J. Geophys. Res., 91, 9781-9793, 1986.

Fehsenfeld, F.C., et al., A study of ozone in the Colorado mountains, J. Atmos. Chem., 1, 87-105, 1983.

Fehsenfeld, F.C., et al., Intercomparison of NO2 measurement techniques, J. Geophys. Res., 95, 3579-3597, 1990.

Harley, R.A., S.A. McKeen, M.O. Rodgers, J. Pearson, W.A. Lonneman, Analysis of motor vehicle emissions during the Nashville/Middle Tennessee ozone study, 2000 Fall Meeting, Eos, Transactions, American Geophysical Union 81, No.48, F104, 2000.

Husar, R.B., Elkins, J.B., Wilson, W.E., U.S. visibility trends, 1960-1992, CAPITA report, Washington University, St. Louis, MO, 1994.

Kley, D., and M. McFarland, Chemiluminescence detector for NO and NO2, Atmos. Technol.,12, 63-69, 1980.

Ogren, J.A., Sheridan, P.A., Vertical and horizontal variability of aerosol single scattering albedo and hemispheric backscatter fraction over the United States, Proceedings of the Conference on Nucleation and Atmospheric Aerosols, University of Helsinki, Helsinki, Finland, Aug. 26-30, 780-783, 1996.

Parrish, D.D., et al., The total reative oxidized nitrogen levels and their partitioning between the individual speciesat six rural sites in Eastern North America, J. Geophys. Res., 98, 2927-2939, 1993.

Parrish, D.D., J.S. Holloway, and F.C. Fehsenfeld, Routine, continuous measurement of carbon monoxide with parts per billion precision, Environ. Sci. Technol., 28, 1615-1618, 1994.

80

Parrish, D.D., et al., Decrease in emission ratios of carbon monoxide to nitrogen oxides in two urban U.S. areas over the past decade, 2000 Fall Meeting, Eos, Transactions, American Geophysical Union 81, No.48, F104, 2000.

Ridley, B A and L C Howlett, An instrument for nitric oxide measurements in the stratosphere, Rev. Sci. Instrum., 45, 742-746, 1974.

Trainer, M., E.J. Williams, D.D. Parrish, M.P. Buhr, E.J. Allwine, H.H. Westburg, F.C. Fehsenfeld, and S.C. Liu, Models and observations of the impact of natural hydrocarbons on rural ozone, Nature, 329, 705-707, 1987.

Waggoner, A.,P., R.E. Weiss, N.C. Ahlquist, D.S. Covert, S. Will, R.J. Charlson, Optical characteristics of atmospheric aerosols, Atmos. Environ., 15, 1891-1909, 1981.

Weber, R.J., D. Orsini, Y. Daun, Y.-N. Lee, P. Klotz, and F. Brechtel, A new particle-in-liquid collector for rapid measurements of aerosol chemical composition, J. Aerosol Sci. Tech., in press, 2000.

81

IV. Air Quality Modeling and Emissions

On the emissions inventory and air quality modeling side, the first year of FAQS was primarily devoted to emissions inventory development. Two major activities were begun, and are continuing. First was initiation of updating the point source emissions inventory. This is being done, in conjunction with EPD, by sending questionnaires to the industries in the counties around the three cities being studied. The questionnaire is to be filled out by all industries, estimated to be emitting more than 25 tons per year of ozone and PM precursors (including ammonia). Smaller emitters do not need to include emissions estimates. However, identification of all of the various possible industrial emitters, no matter how small, will aid in the development of the area source inventory. With assistance from the relevant regional offices of the Georgia Tech Economic Development Institute, workshops were held in each city to assist those receiving a survey to correctly fill it out, to brief these stakeholders about nonattainment and FAQS related issues, and offer them an opportunity to informally ask questions and comment:

30 October 2000 (9:00AM to 11:30AM) – Columbus Water Works 30 October 2000 (2:30PM to 5:00PM) – Middle Georgia Regional Development Center 31 October 2000 (1:00PM to 4:00PM) – Augusta Chamber of Commerce

Copies of the survey materials (excluding the spreadsheet worksheet) are included as Appendix JThe first round of surveys has been returned and is being processed. Industries

overlooked by the first mailing or those not responding to the first survey, are being identified and sent a second survey. Early results suggest that many of these are out of business. This activity will be followed up by a further assessment making sure that all major emitters have been identified and appropriately inventoried.

A second activity during this period was to do a top-down inventory assessment using measurements from the 1999 Atlanta Supersite study and the current emissions inventory (grown to 1999). This assessment suggests that there are a number of areas for possible improvement since the measurements suggest biases in the inventory. In particular, the ammonia, anthropogenic area source VOC and primary organic PM emissions are suggested to have significant biases. This focuses further attention in the coming inventory development activities to clear up these apparent biases.

Finally, a review of the air quality measurements for the summer of 2000 was also completed. The review shows a period that may be of interest for air quality modeling. This period surrounds August 17-18 (and thus, for modeling, would include two or more ramp-up days). There were widespread exceedences during this period.

82

V. Outreach and Education

The Fall line Air Quality Study (FAQS) was kicked off at a Stakeholders Meeting on March 2, 2000. Two subsequent planning meetings of the Coordinating Council were held at the Bibb County (Macon) Engineering Department on May 9 and the Columbus Water Works, on June 1. At these first three meetings stakeholders, regulators, and research staff were briefed on the status of nonattainment recommendations and continuously updated on the progress and plans for the 2000 FAQS Phase I pilot study. A first look at results from the pilot study were provided at a fourth meeting of the Coordinating Council at the Georgia Tech Economic Development Institute’s Augusta Regional Office on August 31. A FAQS Science Workshop was held on the Georgia Tech campus in Atlanta on October 19-20. At this workshop, the research team, peer review committee and other stakeholders met to discuss the findings from the pilot study and to draft initial impressions about air quality in Augusta, Macon, and Columbus. In turn, the results of the science workshop were immediately presented to the Coordinating Council and other stakeholders at the FAQS Policy Workshop at the offices of the Georgia Regional Transportation Authority in Atlanta in the afternoon of October 20. After further detailed deliberation, these findings also constitute the primary content of this report.

In addition to the stakeholder and Coordinating Council meetings and science and policy workshops, three open house tours were held at the site of the mobile Air Quality Research lab:

June 15 at Sandy Beach Park in Macon, July 6 at Fort Gordon in Augusta, and July 20 at Columbus Water Works.

State and local government officials, representatives from many industries, military and other DoD personnel, the media, and the general public attended these open house tours. The media in particular have focused on the FAQS throughout its campaign and have provided significant coverage of each event. A reference listing of all the media stories addressing the Fall line Air Quality Study is included in Appendix K.

To foster efficient communication, an email listserv ([email protected]) and website (http://www.cure.gatech.edu/faqs.asp) have also been created to distribute all FAQS related information and announcements. All meeting minutes and presentation material is archived and freely available on the website. .

83

http://www.cure.gatech.edu/FAQS/inthenews/Default.asp


Finally, the FAQS science team has been active in reaching out to the local communities through a series of invited presentations. These include:

Southern Center for Advanced Transportation Electric and Hybrid Electric Buses Workshop, Macon, GA 10/6/2000.

Macon Rotary Club, Macon, GA 10/9/2000. Georgia Transit Association Annual Meeting, Columbus, GA 11/2/2000. Briefing to the staff of Georgia's congressional delegation, Atlanta, GA 11/9/2000. Environmental Action!, Columbus, GA 11/14/2000. Strategic Environmental Research and Development Program Ecosystem

Management Project (SEMP - Ft. Benning) Research Coordination Meeting, Columbus, GA 11/14/2000.

84

VI. Peer Review

An external three-member peer review panel was convened for the FAQS science and policy workshops that were held in Atlanta on October 19 and 20, 2000. The panel consisted of Professor Ellis Cowling, Director Southern Oxidants Study, North Carolina State University, Mr. John Jansen, Principal Scientist, Southern Company Services, and Mr. Rafael Ballagas, Program Manager Ambient Monitoring, Georgia Environmental Protection Division. The panel was asked to observe the workshop proceedings and review all relevant technical information generated or considered by the research team. Generally favorable, the peer review panel’s critical review is included here as Appendix

Appendices

A. SUMMARY OF SCIENTIFIC UNDERSTANDING OF FORMATION AND ACCUMULATION OF GROUND-LEVEL OZONE AND PARTICULATE MATTER POLLUTION.

K. Baumann

1. Introduction The following is a description of the general processes that are involved in ozone

formation in the lower troposphere according to the current scientific understanding. As the

scientific understanding of the underlying processes improved, regulatory strategies followed in

the form of National Ambient Air Quality Standards (NAAQS) promulgated by the United States

Environmental Protection Agency (EPA), which is reviewed briefly in section 1.2. As a

response to a growing number of scientific studies linking elevated fine-particle concentrations

with adverse health effects, the EPA released NAAQS also for PM2.5 (particulate matter with

aerodynamic diameter less than 2.5 microns). Processes that seem to play the most important

roles in the PM2.5 problematic are presented in section 1.3 with emphasis on their relationship to

ozone pollution.

1.1 Ozone Pollution in the Atmospheric Boundary LayerOzone (O3) production in the lower troposphere occurs via the free-radical

initiated oxidation of volatile organic compounds (VOC) or CO in the presence of nitrogen

oxides (NOx = NO + NO2) and sunlight. The path on which the hydroxyl radical (OH) becomes

recycled is of critical importance. OH is essential for the initiation of nearly all atmospheric

oxidation processes. Under “clean” conditions it directly oxidizes carbon monoxide (CO) and

methane (CH4) to convert into HO2 which in turn is recycled back to OH via

HO2 + NO OH + NO2 (R1)

HO2 + O3 OH + 2O2 (R2).

Model calculations [Liu and Trainer 1988] show that for [NO]/[O3] ratios > 10-3,

therefore rather polluted conditions, recycling path R1 is favored forming nitrogen dioxide

(NO2). NO2 is known to be the only source for ozone production via its photolysis:

NO2 + hv320-430nm NO + O (R3)

O + O2 + M O3 + M (R4)

Due to the relatively high partial pressure of O2 in the troposphere R4 is rapid and not

rate limiting. But R3 and R4 are balanced by:

A-1

NO + O3 ---> NO2 + O2 (R5)

so that this cycle (R3 - R5) is in equilibrium and neutral, i.e., overall no ozone production nor

destruction occurs in this so-called photostationary state (PSS).

The oxidation path of OH with alkanes (e.g. CH4) and most unsaturated, biogenic or

anthropogenic non-methane hydrocarbons (NMHC, e.g. alkenes, aromatics, alkynes) leads to

alcyl radicals Rn with n-2 C atoms which in turn react explicitely with atmospheric oxygen to

form organic peroxy radicals RnO2 (written here as RO2, and sometimes referred to as ROx = RO2

+ HO2 as the total peroxy radicals). If R is a NMHC chain, e.g. CH3CH2 for RH being ethane,

and M is a third body (N2, O2), then

RH + OH R + H2O (R6)

R + O2 + M RO2 + M (R7)

Under conditions of sufficient [NOx], another NO oxidation alternative to R5 becomes effective

and represents the source for a net ozone production since NO2 is formed without consuming O3:

NO + RO2 NO2 + RO (R8)

RO + O2 HO2 + carbonyl (R9).

Of note in above sequence is that VOCs (denoted RH) are consumed, while both OH/HO2 and

NOx act as catalysts. The main rate terminating step for ozone production is the removal of the

catalysts, often by one of two paths:

HO2 + HO2 + M H2O2 + O2 + M (R10)

or NO2 + OH + M HNO3 + M (R11)

forming nitric acid at approximately k11 = 1.3 x 10-11 cm3/molec/s [DeMore et al., 1994].

While the PSS (R3 - R5) is observed to be null with respect to [O3] change [Ridley et al.,

1992], RO2 and R8 represent the perturbation of PSS causing ozone production. Among other

methods, this PSS deviation method has been commonly used to quantify ozone production

[Parrish et al., 1986; Chameides et al., 1990; Ridley et al., 1992; Cantrell et al., 1993;

Kleinmann et al., 1995; Hauglustaine et al., 1996; to mention only a few]. Assuming steady

state conditions and, that besides ozone and RO2 no other oxidants are available for effectively

oxidizing NO (e.g. IO, [Parrish et al., 1986]), the PSS deviation implies

d[NO]/dt = -d[NO2]/dt = j(NO2)[NO2] - [NO] (k5[O3] + k8i[RO2]i = 0 (1)

leading to an instantaneous steady state ozone production P(O3)

P(O3) = [NO] k8i[RO2]i = j(NO2)[NO2] - k5[NO][O3] (2)

A-2

by use of measured mixing ratios and the first order photolytic rate constant for NO2, j(NO2).

Formerly, the rate constants k8i were believed to be similar for all “kinds” of peroxy

radicals RO2i (see e.g. Finnlayson-Pitts and Pitts [1986]), but more recent studies revealed that

rate constants for alkenes are temperature dependent (e.g Atkinson et al. [1992], Talukdar et al.

[1994]), and that aromatic RO2s can quickly shift to carbonyls which can photolyse, further

oxidize, or deposit and therefore become irrelevant for ozone production [Becker et al., 1994].

J(NO2) must be determined in situ by measuring the actinic flux or can be derived by in situ UV

measurements [Madronich et al., 1987]. The right hand side of Equ. 2 is applicable for

homogeneous air masses and j(NO2) values larger 0.001 s-1 that is about 10% of mid day clear

sky maximum [e.g. Volz et al., 1988]. The rate constant k5 only depends on the ambient

temperature and can be adapted from laboratory experiments [DeMore et al., 1994]:

k5 = 2.10-12 . exp(-1400/T) (3).

Trainer et al. [1993] defined net ozone production by the product of the ozone production

efficiency d[O3]/d[NOx] and the rate of NOx oxidation d[NOx]/dt:

P(O3) = d[O3]/dt = d[O3]/d[NOx] . d[NOx]/dt (4).

Under high [NOx] conditions, however, the amount of ozone produced was found to depend

nonlinearly on [NOx] driven by the [NMHC] to [NOx] ratio (see also Liu et al., [1987],

Chameides et al., [1988]). This has been more recently specified as an enhancement of ozone

formed per NOx oxidized under higher [NMHC] to [NOx] ratios along with an enhancement of

[OH] and [RO2] (see e.g. McKeen et al. [1991]). While some model calculations could not show

differences in the effects of biogenic vs anthropogenic NMHCs [McKeen et al., 1991], in the

following denoted BHC and AHC respectively, Chameides et al. [1988] demonstarted the

importance of BHC compounds, in particular isoprene, and its role in the ozone formation

process.

In general, the O3 production rate can be limited by either VOC or NOx. The existence of

these two opposing regimes can be mechanistically understood in terms of the relative sources of

OH and NOx [Kleinman et al., 1994]. When the OH source is greater than the NOx source,

reflecting a greater abundance of OH relative to NOx, termination is dominated by R10. Under

these conditions, the formation of H2O2 is much more effective than that of HNO3, NOx is in

short supply, and as a result the O3 production rate is “NOx-limited”. Therefore, O3

concentrations [O3] are most effectively reduced by lowering the emissions, hence concentrations

A-3

of NOx [NOx] instead of VOC. On the other hand, when the OH source is less than the NOx

source, termination proceeds predominantly via R11, the formation of HNO3 exceeds the one of

H2O2, NOx is relatively abundant, and O3 production is “VOC-limited”, i.e. [O3] is most

effectively reduced by lowering [VOC].

The more stable NOx oxidation products can be measured indirectly by the difference of

the total reactive odd nitrogen NOy and NOx, in the following called NOz [e.g. Volz et al., 1993],

which in the lower troposphere is principally composed of peroxyacetic nitric anhydride (or

more colloquially called peroxyacetyl nitrate = PAN), nitric acid (HNO3), and aerosol nitrate

(NO3-) [Fahey et al., 1986]. Thus NOy includes all NOx and its more (photochemically) stable

oxidation products.

Lifetimes of ozone and nitric acid in the lower troposphere are greatly determined by

their rate of removal via surface deposition. Dry deposition velocities chiefly depend on land

surface cover and wind speed. Measurement based upper limits for nitric acid range between 3

and 10 cm/s [Hoefken et al., 1986; Dollard et al., 1987; Bennett, 1988; Meyers et al., 1989;

Hanson and Lindberg, 1991], while summertime midday ozone depositon velocities for grass,

corn, and forest canopies vary from 0.3 to 1.2 cm/s [Garland and Derwent, 1979; Weseley et al.,

1978; Lenschow et al., 1981; Duyzer et al., 1983; Baumann, 1992; Gao et al., 1993; Frost et al.

1998, McMillan et al., 2000]. Therefore, ozone’s lifetime (with respect to dry removal) is about

3 to 10 times larger than nitric acid’s. Due to the longer lifetime of ozone relative to nitric acid,

the [O3]/[NOz] ratio, i.e. the slope of their linear regression is regarded as an upper limit for

ozone production efficiency (OPE) estimates and has been reported on the order of 5 to 10

molecules of ozone formed per molecule of NOx oxidized [Trainer et al., 1993; Olszyna et al.,

1994; Chin et al. 1994; Daum et al., 1996; Hirsch et al. 1996]. This OPE is supposed to be an

estimate for how often NOx catalytically cycle to generate ozone, before termination occurs by

removal of NO2 via OH reaction (R7) to form nitric acid as an effective sink. It has been

emphasized, though, that great care must be taken on the selection and interpretation of measured

species, since their mixing ratios depend on, and are influenced by their individual regional

“background” levels, their removal rates (due to chemical losses, wet and/or dry deposition), the

mixture of emitted precursors, the rates of ozone formation d[O3]/dt and NOx oxidation

d[NOx]/dt, and the mixing and transport of air masses of different histories [Volz-Thomas et al.,

1993; Trainer et al., 1995]. Knowledge of all these factors is essential to use NOy that accurately

A-4

accounts for the initial NOx emissions; to use the ratio NOx/NOy, i.e. the fraction of NOx not

oxidized to more stable products to detemine the photochemical age of air masses; and to

estimate more realistic OPEs which help to better understand ozone photochemistry in the lower

troposphere and allow to implement more effective ozone control strategies.

Another recent study on regional tropospheric ozone, identified air masses in the rural

South-Eastern US that had contributions of BHC induced photochemically produced ozone up to

75 % of the “regional background” level [Williams et al. 1997]. Isoprene is currently considered

the major BHC emission accounting for 44% of the total global NMHC flux [Guenther et al.

1995]. The above reactions take also place when isoprene is oxidized in the presence of NO.

While most oxidation products (due to further reactions of RO mostly via carbonyls) like

ketones, aldehydes, and especially PAN can be formed by oxidation of a variety of NMHC

(BHC and AHC), the peroxymethacrylic nitric anhydride (MPAN) is specifically produced on

the isoprene-NOx oxidation path and hence is a unique indicator for BHC driven photochemistry

[Williams et al. 1997a]. Due to the joint involvement of NMHC and NOx in the production of

PAN and ozone, good correlations between the two can be anticipated [Roberts 1990; Roberts et

al. 1995] under certain conditions, which include well mixed homogeneous air preferably away

from surfaces, because both PAN and ozone tend to deposit on to surfaces. Roberts et al. [1998]

elaborate on these special relationships in their evaluation of the 1 July 1994 case at the

Hendersonville site in contrast to aircraft measurements, and discern a significant impact of BHC

photochemistry on NOx and AHC fueled ozone production within the Nashville urban plume.

It has been recognized, that control strategies focusing on AHC reduction only are

ineffective when photochemical activity is dominated by BHC [Trainer et al., 1987; Chameides

et al., 1988]. Various studies of the photochemical production of ozone and its feedback

relationship to its precursors in the outflow of a large urban area during periods of high

photochemical activity and stagnant high pressure conditions, typical for the South-Eastern

United States, have been conducted in the framework of the Southern Oxidants Study [Cowling

et al., 1999]. These studies resulted in a large number of scientific publications that improved

our current understanding of ozone pollution substantially.

1.2 Review of Ozone and PM2.5 Control StrategiesThe 1967 Clean Air Act designated areas as out of compliance if more than three 1-hour

averages of [O3] exceeded 120 ppbv, the National Ambient Air Quality Standards (NAAQS) for

A-5

ozone mixing ratios measured at surface monitoring stations, within a 4-year period. Due to the

lack of ozone abatement progress, the United States Environmental Protection Agency (EPA)

recognized the need for an independent scientific assessment of the problem and co-sponsored a

study by the Committee on Tropospheric Ozone Formation and Measurement under the National

Research Council in 1989 [NRC, 1991]. In July 1997, the EPA has revised the previous standard

to that the annual fourth-highest daily maximum 8-hour ozone concentration (i.e., 8-hour running

means of 1-hour O3 averages) must not exceed 80 ppbv within a 3 year period. Although this

newer, proposed standard is yet to be legally instated and inforced, scientific studies already

show that it is more stringent than the older, existing standard [St. John and Chameides, 1997].

Atlanta is one of the fastest growing urban areas in the country and is currently categorized by

the EPA as a “serious” nonattainment area for O3 [EPA, 1998]. A major contributor to Atlanta’s

O3 problem is automobile exhaust, since the city has more daily vehicle miles traveled (VMT)

per capita than any other in the country [EPA, 2000]. Coincidentally, automobile exhaust also

plays a significant role in the production of fine particles [Cadle et al., 1998].

In 1997, NAAQS for PM2.5 (particulate matter with aerodynamic diameter less than 2.5

microns) were promulgated by the EPA as a response to a growing number of scientific studies

linking elevated fine-particle concentrations with adverse health effects [EPA, 1997]. These

NAAQS were set both as short-term (< 65 g/m3 for a 24-hour average) and long term (< 15

g/m3 for an annual average) standards, and served to further enhance the effectiveness of the

Clean Air Act and protect the public health within an adequate margin of safety. As the

constitutionality of EPA’s enforcement of these standards is politically disputed, there remains a

limited, although growing, amount of PM2.5 data, that seem to indicate that the air quality in a

significant number of areas in the country, and especially in the Atlanta metro area, will exceed

the proposed standards [Butler, 2000, Mulholland et al., 1998].

1.3 Current Understanding of PM2.5 Pollution 1.3.1. Primary PM2.5 Sources

The term “particulate matter”, or short “PM” is commonly used to describe the complex

mixture of solid particles and liquid droplets suspended in the lower atmosphere. The

heterogeneous mixture of gas-, liquid-, and solid-phase species defines the medium of “aerosol”.

In general, PM is classified as either “coarse” or “fine,” separated by a 2.5 micron aerodynamic

diameter cutoff. PM2.5 results from both anthropogenic and biogenic sources, which are further

A-6

classified as being either primary or secondary. Primary sources emit PM2.5 directly into the

atmosphere, while secondary sources typically emit gaseous precursor species that undergo

atmospheric transformation into particles.

Primary PM2.5 sources include both biogenic and anthropogenic emissions. Examples of

anthropogenic sources are combustion engines, residential fireplaces, meat cooking operations,

and many other industrial processes [Rogge et al., 1991, Huang et al., 1994, Lowenthal et al.,

1994, Schauer et al., 1996, McDonald et al., 1998]. These particles are generally less than 1 m

in diameter, with a significant fraction less than 0.3 m. Fine particles are emitted in the exhaust

of gasoline-burning automotive combustion engines and consist primarily of organic carbon and

metallic species [Mueller, 1994].

PM2.5 is also formed when crankcase oil leaks into the combustion chamber. The carbon

species often result from incomplete combustion of fuel and oil leaking through piston seals,

while metal-containing particles result from antiknock agents added to the fuel [see Seinfeld,

1998 and literature cited therein]. It has been impossible to predict the exact composition of

particulate exhaust products using chemical reaction models due to the complexity of the

combustion process and the variable constituency of gasoline.

Diesel engines, particularly heavy-duty trucks, emit large fractions of elemental carbon

(EC) that is ubiquitous to urban areas [Cass and Gray, 1995, Dreher and Harley, 1998]. The

EC, or black smoke, has a chemical structure similar to that of impure graphitic carbon, but it

also contains high molecular weight, dark-colored, non-volatile organic materials, such as tar and

coke. The concentrations of EC emitted are directly related to engine load, but the wide variety

of diesel engine sizes and gross vehicle weights makes it difficult to estimate diesel exhaust

emissions of fine PM [Singer and Harley, 1996].

In addition to motor vehicles, wood-burning fireplaces and meat charbroiling operations

are major sources of primary PM2.5 in certain parts of the country [Hawthorne et al., 1989]. In

western US cities during wintertime, wood stoves and fireplaces are estimated to emit fine

particles to the atmosphere at an approximate rate of 10 tons/day, comprising nearly 20% of the

urban carbonaceous aerosol [Watson et al., 1988]. Organic carbon (OC) is typically found to be

the largest fraction of PM2.5 emitted from residential wood combustion, with a significant amount

of EC, and small levels of inorganic species present as well. Similar to the emissions differences

A-7

observed across grades of coal and oil-based fuels, woods also release a wide variety of pollutant

combinations, mostly related to moisture content [McCrillis et al., 1989].

Meat charbroiling operations contribute approximately 10 tons/day of PM2.5 to California’s South

Coast Air Basin [Whynot et al., 1998], and OC has been observed to dominate the fine particle

emissions, averaging greater than 95% of the total measured [McDonald et al., 1998]. PM2.5

emissions from meat charbroiling operations also contribute small amounts of EC, potassium,

chloride, and sulfate to the atmosphere in urban environments [Rogge et al., 1991, Schauer et al.,

1996].

Primary biogenic emission sources include volcanoes, soil and sea spray. It is estimated

that volcanic activity contributes 25-150 Tg of PM to the ambient atmosphere every year

[Seinfeld, 1998]. The majority of these particles contain sulfur and are oxidized to H2SO4 in the

atmosphere. A second major primary biogenic emission source is soil debris, which contributes

an estimated 50-250 Tg/yr to the continental aerosol. The most common constituents of soil

debris are crustal elements, such as Fe, Ca and Si. The final significant primary biogenic source

is sea spray, consisting of ionic salt material, most of which resides in the PM10 size range;

however, the size range does extend toward smaller values [Warneck, 1988]. Other,

miscellaneous biogenic particles include those released from plants (seeds, spores, pollen, etc.)

and biomass burning, both controlled and uncontrolled, e.g. forest fires (organic and elemental

carbon).

1.3.2 Secondary PM2.5 Formation in the Atmosphere (adapted from Butler, 2000)

The mechanisms, pathways, and equilibria that are believed to exist between gas- and

particle-phase atmospheric chemistry resulting in PM production are extremely complex, and

only a brief overview of the major processes will be presented here. The primary PM species of

interest are sulfate (SO42-), nitrate (NO3

-), and ammonium (NH4+). Ammonium sulfate

[(NH4)2SO4], ammonium bisulfate (NH4HSO4), and sulfuric acid (H2SO4) are the significant

forms of PM sulfate resulting from atmospheric conversions. These are the most prevalent

species found in continental aerosols and primarily result from the gas-to-particle conversion of

SO2 that is emitted from coal and oil combustion processes and volcanoes [e.g. Warneck, 1988].

Based on reaction kinetics, the most significant SO2 conversion occurring in the atmosphere, is

observed to be [Calvert et al., 1977]

A-8

The reactions leading from the HOSO2 radical to H2SO4 are generally believed to be [Calvert

and Stockwell 1984]

The gas-phase reaction mechanism shown above for SO2 oxidation in the atmosphere occurs in

the daytime and results in a rate of H2SO4 formation of approximately one percent per hour.

Roughly one-half the sulfate present in urban and rural environments is the result of gas-phase

oxidation; the other half is produced by the following aqueous-phase reactions

Most aqueous-phase SO2 oxidation takes place in clouds (R14 is catalyzed by metal species), as

water droplets scavenge the pollutant and rapidly oxidize it to sulfate. Some of these acidic

water droplets rain out, but most are considered to evaporate at night leaving a sulfate residue

above the nocturnal mixing layer [Main and Roberts, 1999, Meagher et al., 1983]. The overall

SO2 oxidation rate (gas- and solution-phase reactions) is on the order of 1.2-13% per hour in

urban plumes [Alkezweeny and Powell, 1977; Chang, 1979].

Like SO2, NO and NO2 (NOx) result primarily from combustion processes and play a

significant role in atmospheric chemistry and the production of urban smog as described in

section 1.1 above. The focus here is on the NOx oxidation pathways producing nitric acid

(HNO3) and NO3- aerosol, as summarized for the gas phase in Figure 1.1. The oxidation of NO2

to HNO3 occurs either via reaction with OH radicals (R7) or through a series of reactions with O3

and formaldehyde, producing NO3 as an intermediate. NOx oxidation rates range from 5-50%

A-9

per hour, with more than 95% of the conversion accomplished through gas-phase reactions

[Spicer et al., 1981]. Some of the HNO3 condenses to aerosol NO3-, but a large portion remains

in the gas phase. The stability of aerosol NO3-depends on its pH, as well as its water content

[Tang, 1980]. The most common nitrate compound found in the atmosphere is ammonium

nitrate (NH4NO3), which presents measurement difficulties due to its tendency to evaporate when

changes in temperature and relative humidity occur.

Ammonia (NH3), which can be emitted as a side-product from large-scale, catalytically

controlled combustion processes, bacterial decomposition of animal and human wastes, and

emission from soils, is the primary trace gas capable of neutralizing H2SO4 and HNO3 [Seinfeld,

1998, Hildemann et al., 1984]. Ammonia reacts slowly in the gas phase, and the most important

reaction appears to be

which is generally viewed as an unimportant sink of NH3. However, the fate of the NH2 radical

in the atmosphere is unknown.

A-10

Figure 1.1 Atmospheric NOx chemistry reaction pathways, adapted from Russell et al. [1985]

and Warneck [1988].

Although the mechanisms leading from gaseous NH3 to particulate NH4+ are unclear, the results

of size distribution experiments indicate that NH4+ arises from gas-to-particle conversion,

specifically, NH3 neutralization of H2SO4 [Stelson and Seinfeld, 1981]. The amount of NH4+

produced is highly dependent upon the relative amounts of NH3 and H2SO4, and several studies

have noted aerosol acidity during the summer months in the eastern United States, where sulfates

may only be half neutralized via ammonium bisulfate, (NH4)HSO4 [Ferek et al., 1983, Malm, et

al., 1994].

A-11

References

Alkezweeny, A.J.; Powell, D.C.Estimation of transformation rate of SO2 to SO4 from atmospheric concentration data, Atmos. Environ. 11, 179-182, 1977.

Atkinson, R. et al., Evaluated kinetic and photochemical data for atmospheric chemistry - supplement 4, J. Phys. Chem. Ref. Data, 21, 1125pp, 1992.

Baumann, K., Ozoneintrag in Waldbestaende: Bestimmung mit verschiedenen experimentellen Methoden, Ph.D.thesis, IVD, University of Stuttgart, 1992.

Becker, K.-H. et al., OH initiated oxidation of NMVOC under variable NOx conditions, in: LACTOZ-Report 1993, EUROTRAC ISS, Garmisch-Partenkirchen, 1994.

Bennett, M., A simple physical model of dry deposition to a rough surface, Atmos. Environ., 22, 2701-2705, 1988.

Butler, A., Temporal and spatial analysis of PM2.5 mass and composition in Atlanta, Ph.D. thesis, Civil & Environ. Engin. Dept., Georgia Tech, Atlanta, 2000.

Cadle, S.H.; Mulawa, P.A.; Ball, J.; Donase, C.; Weibel, A.; Sagebiel, J.; Knapp, K.; Snow, R., Environ. Sci. Technol. 31, 3405-3412, 1998.

Calvert, J.G.; Su, F.; Bottenheim, J.W.; Strausz, O.P., Mechanism of the homogeneous oxidation of sulfur dioxide in the troposphere, Atmos Environ. 12, 197-226, 1977.

Calvert, J.G.; Stockwell, W.R. “The mechanism and rates of the gas phase oxidations of sulfur dioxide and nitrogen oxides in the atmosphere,” in SO2, NO, and NO2 Oxidation Mechanisms: Atmospheric Considerations, J.G. Calvert (Ed.), Butterworth, Boston, 1-62, 1984.

Cantrell, C.A. et al., Peroxy radicals as measured in ROSE and estimated from photostationary state deviations, J. Geophys. Res., 98, 18,355-18,366, 1993.

Cass, G.R., and H.A. Gray, In Diesel Exhaust: A Critical Analysis of Emissions, Exposure, and Health Effects, Health Effects Institute, Cambridge, MA, 127-137, 1995.

Chameides, W.L. et al., The role of biogenic hydrocarbons in urban photochemical smog: Atlanta as a case study, Science, 241, 4872, 1988.

Chameides, W.L. et al., Observed and model-calculated NO2/NO ratios in tropospheric air sampled during the NASA GTE/CITE-2 field study, J. Geophys. Res., 95, 10,235-10,247, 1990.

Chang, T.Y. Atmos. Environ. 13, 1663-1664, 1979.Chin, M. et al., Relationship of ozone and carbon monoxide over North America, J. Geophys.

Res., 99, 14,565-14,573, 1994.Cowling, E.B. et al., Introduction to special section: Southern Oxidants Study Nashville / Middle

Tennessee Ozone Study, J. Geophys. Res., 103, 22,209-22,212, 1999.Daum, P.H. et al., Chemical and physical properties of plumes of anthropogemic pollutants

transported over the North Atlantic during the North Atlantic Regional Experiment, J. Geophys. Res., 101, 29,029-29,042, 1996.

DeMore, W.B. et al., Chemical kinetics and photochemical data for use in stratospheric modeling, Eval. 11, JPL Publ.. 94-26, 1994.

Dollard, G.J.et al., Concentrations and dry deposition velocities of nitric acid, Nature, 326, 481-483, 1987.

Dreher, D.B., and Harley, R.A., A fuel-based inventory for heavy-duty diesel truck emissions, J. Air & Waste Manage. Assoc. 4, 352-358, 1998.

A-12

Duyzer, J.G., G.M. Meyer, and R.M. van Aalst, Measurement of dry deposition velocities of NO, NO2, and O3, and the influence of chemical reactions, Atmos. Environ., 17, 2117-2120, 1983.

EPA, U.S. Environmental Protection Agency, “National Ambient Air Quality Standards for Particulate Matter – Final Rule,” 40 CFR part 50, Federal Register, 62(138):38651-38760, July 18, 1997.

EPA, U.S. Environmental Protection Agency, Title 40, Code of Federal Regulations, part 81, 1998.

EPA, U.S. Environmental Protection Agency, Clean Air Initiative Web Page, http://www.epa.gov/region04/air/cai/, 2000.

Fahey, D.W., et al., Reactive nitrogen species in the troposphere: Measurements of NO, NO2, HNO3, particulate nitrate, peroxiacetyl nitrate (PAN), O3, and total reactive odd nitrogen (NOy) at Niwot Ridge, Colorado, J. Geophys. Res., 91, 9781-9793, 1986.

Ferek, R.J. et al., Strong and weak acidity of aerosols collected over the northeastern United States, Environ. Sci. Tech. 17, 315-324, 1983.

Finlayson-Pitts, B.J., and J.N. Pitts, Atmospheric chemistry: Fundamentals and experimental techniques, Wiley, New York, 1986.

Frost, G.J. et al., Photochemical ozone production in the rural southeastern United States during the 1990 Rural Oxidants in the Southern Environment (ROSE) program, J. Geophys. Res.103, 22,491-22,508, 1998.

Gao, W., M.L. Weseley, and P.V. Doskey, Numerical modeling of the turbulent diffusion and chemistry of NOx, O3, Isoprene, and other reactive trace gases in and above a forest canopy, J. Geophys. Res., 98, 18,339-18,353, 1993.

Garland, J.A., and R.G. Derwent, Destruction at the ground and the diurnal cycle of concentration of ozone and other gases, Quart. J. R. Met. Soc., 105, 169-183, 1979.

Guenther, A.B. et al., Isoprene fluxes measured by enclosures, relaxed eddy accumulation, surface layer gradient, mixed layer gradient, and mixed layer mass balance techniques, J. Geophys. Res. 101, 18,555-18,567, 1995.

Hanson, P.J., and S.E. Lindberg, Dry deposition of reactive nitrogen compounds: A review of leaf, canopy, and non-foliar measurements, Atmos. Environ., 25A, 1615-1634, 1991.

Hauglustaine, D.A. et al., Observed and model-calculated photostationary state at Mauna Loa Observatory during MLOPEX 2, J. Geophys. Res., 101, 14,681-14,696, 1996.

Hawthorne, S.B.; Krieger, M.S.; Miller, D.J.; Mathiason, M.B., Collection and quantitation of methoxylated phenol tracers for atmospheric pollution from residential wood stoves, Environ. Sci. Technol. 23, 470-475, 1989.

Hildemann, L.M.; Russell, A.G.; Cass, G.R., Ammonia and nitric acid concentrations in equilibrium with atmospheric aerosols: experiment vs theory, Atmos. Environ. 18, 1737-1750, 1984.

Hirsch, A.I. et al., Seasonal variation of the ozone production efficiency per unit NOx at Harvard Forest, Massachusetts, J. Geophys. Res., 99, 12,659-12,666, 1996.

Hoefken, K.-D.et al., Untersuchungen zur trockenen Deposition und Emission von atmosphaerischem NO, NO2 und HNO3 an natuerlichen Oberflaechen, Juel-2054, KFA, Forschungszentrum Juelich GmbH, 1986.

Huang, X.; Olmez, I.; Aras, N.K.; Gordon, G.E., Emissions of trace elements from motor vehicles: potential marker elements and source composition profile, Atmos. Environ. 28, 1385-1391, 1994.

A-13

http://www.epa.gov/region04/air/cai/

Kleinman, L.I. et al., Low and high NOx tropospheric photochemistry, J. Geophys. Res. 99, 16,831-16,838, 1994.

Kleinman, L.I. et al., Peroxy radical concentration and ozone formation rate at a rural site in the southeastern United States, J. Geophys. Res., 100, 7263-7273, 1995.

Lenschow, D.H., R. Pearson Jr., and B.B. Stankov, Estimating the ozone budget in the boundary layer by use of aircraft measurements of ozone eddy flux and mean concentration, J. Geophys. Res. 86, 7291-7297, 1981.

Liu, S.C. et al., Ozone production in the rural troposphere and the implications for regional and global ozone distributions, J. Geophys. Res., 92, 4191-4207, 1987.

Liu, S.C., and M. Trainer, Responses of the tropospheric ozone and hydrogen radicals to column ozone change, J. Atmos. Chem. 6, 221-233, 1988.

Lowenthal, D.H.; Zielinska, B.; Chow, J.C.; Watson, J.G.; Gautam, M.; Ferguson, D.H.; Neuroth, G.R.; Stevens, K.D., Characterization of heavy-duty diesel emissions, Atmos. Environ. 28, 731-744, 1994.

Madronich, S., Intercomparison of NO2 photodissociation and U.V. radiometer measurements, Atmos. Environ. 21, 569-578, 1987.

Main, H., and P.T. Roberts, PM2.5 Data Analysis Workbook (Draft), U.S. Environmental Protection Agency, Research Triangle Park, NC, 1999.

Malm, W.C. et al., Spatial and seasonal trends in particle concentration and optical extinction in the United States, J. Geophys. Res. 99, 1347-1370, 1994.

McCrillis, R.C.; Leese, K.E.; Harkins, S.M. Effects of Burn Rate, Wood Species, Moisture Content and Weight of Wood Loaded on Woodstove Emissions, Report No. 600/2-89/025, U.S. Environmental Protection Agency, 1989.

McDonald, J.; Zielinska, B.; Fujita, E.; Chow, J.; Watson, J.; Sagebiel, J. “The development of chemical emission profiles for residential wood combustion and meat cooking for use in the apportionment of atmospheric carbonaceous aerosol,” in PM2.5: A Fine Particle Standard, Proceedings of the Specialty International Conference, 645-660, 1998.

McKeen, S.A., E.-Y. Hsie, and S.C. Liu, A study of the dependence of rural ozone on ozone precursors in the Eastern United States, J. Geophys. Res., 96, 15,377-15,394, 1991.

McMillen, R.T., R.L. Gunter, W.T. Luke, and T.B. Watson, Ozone fluxes and flux divergences, obtained with the NOAA Twin Otter during the Southern Oxidant Study, J. Geophys. Res., 2000.

Meagher, J.F., and E.M. Bailey, The seasonal variation of the atmospheric SO2 to SO42-

conversion rate, J. Geophys. Res. 88, 1525-1527, 1983.Meyers, T.P., B.J. Huebert, and B.B. Hicks, HNO3 deposition to a deciduous forest, Bound.

Layer Meteor., 49, 395-410, 1989.Mueller, P.K. Atmos. Environ. 28, 248, 1994.Mulholland, J.A.; Butler, A.J.; Wilkinson, J.G.; Russell, A.G.; Tolbert, P.E. J. Air & Waste

Manage. Assoc. 5, 418-426, 1998.NRC, National Research Council, Committee on Tropospheric Ozone Formation and

Measurement, Rethinking the ozone problem in urban and regional air pollution, Natl. Academy

Press, Washington D.C., 1991

Olszyna, K.J. et al., O3 and NOy relationships at a rural site, J. Geophys. Res., 99, 14557-14,563, 1994.

A-14

Parrish, D.D. et al., Measurements of the NOx-O3 photostationary state at Niwot Ridge,

Colorado, J. Geophys. Res., 91, 5361-5370, 1986.

Ridley, B.A. et al., Measurements and model simulations of the photostationary state during the Mauna Loa Observatory Photochemistry Experiment: implications for radical concentrations and ozone production and loss rates, J. Geophys. Res., 97, 10,375-10,388, 1992.

Roberts, J.M., The atmospheric chemistry of organic nitrates, Atmos. Environ., 24A, 243-287, 1990.

Roberts, J.M., et al., Relationships between PAN and and ozone at sites in Eastern North America, J. Geophys. Res., 100, 22,821-22,830, 1995.

Roberts, J.M., et al., Measurements of PAN, PPN, and MPAN made during the 1994 and 1995 Nashville intensives of the Southern Oxidants Study: Implications for regional ozone production from biogenic hydrocarbons, J. Geophys. Res. 103, 22,473-22,490, 1998.

Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; Cass, G.R.; Simoneit, B.R.T., Sources of fine organic aerosol: 1. charbroilers and meat cooking operations, Environ. Sci. Technol. 25, 1112-1125, 1991.

Russell, A.G.; McRae, G.J.; Cass, G.R., Dynamics of nitric acid production and the fate of nitrogen oxides, Atmos. Environ. 19, 893-903, 1985.

Schauer, J.J.; Rogge, W.F.; Hildemann, L.M.; Mazurek, M.A.; Cass, G.R.. Source apportionment of airborne particulate matter using organic compounds as tracers, Atmos. Environ. 30, 3837-3855, 1996.

Seinfeld, J.H., and S.N. Pandis, Atmospheric Chemistry and Physics, John Wiley & Sons, New York, 1998.

Singer, B.C., and R.A. Harley, Fuel-based motor vehicle emission inventory, J. Air & Waste Manage. Assoc. 46, 581-593, 1996.

Spicer, C.W.; Koetz, J.R.; Keigley, G.W.; Sverdrup, G.M.; Ward, G.F. “A study of nitrogen oxides reactions within urban plumes transported over the ocean”, U.S. Environmental Protection Agency, Research Triangle Park, NC, 1981.

Stelson, A.W., and Seinfeld, J.H., Chemical mass accounting of urban aerosol, Environ. Sci. Technol. 15, 671-679, 1981.

St. John, J.C., and W.L. Chameides, Climatology of ozone exceedences in the Atlanta metropolitan area: 1-hour vs. 8-hour standard and the role of plume recirculation air pollution episodes, Environ. Sci. Technol., 31, 2797-2804, 1997.

Talukdar, R.K. et al., Kinetics of the reactions of OH with alkanes, Intern. J. Chem. Kinetics 26, 973-990, 1994.

Tang, I.N., Equilibrium partial pressure of nitric acid and ammonia in the atmosphere, Atmos. Environ. 14, 819-828, 1980.

Trainer, M. et al., Models and observations of the impact of natural hydrocarbons on rural ozone, Nature 329, 705-707, 1987.

Trainer, M. et al., Correlation of ozone with NOy in photochemically aged air, J. Geophys. Res. 98, 2917-2926, 1993.

Trainer, M. et al., Regional ozone and urban plumes in the southeastern United States: Birmingham, a case study, J. Geophys. Res., 100, 18,823-18,834, 1995.

Volz, A. et al., Ozone production in the Black Forest: direct measurements of RO2, NOx, and other relevant parameters, in: Tropospheric Ozone, D. Reidel, 293-302, 1988.

A-15

Volz-Thomas, A. et al., Photo-oxidants and precursors at Schauinsland, Black Forest: results from continuous measurements of VOCs and organic nitrates, in: TOR Annual Report 1992, EUROTRAC ISS, Garmisch-Partenkirchen, 1993.

Warneck, P. Chemistry of the Natural Atmosphere, Academic Press, 1988.Watson, J.G.; Chow, J.C.; Richards, L.W.; Neff, W.D.; Andersen, S.R.; Dietrich, D.L.; Houck,

J.E.; Olmez, I. The 1987-88 Metro Denver Brown Cloud Study, Vol. 3, DRI Document No. 8810.1F3, 1988.

Weseley, M.L. et al., Daytime variations of ozone eddy fluxes to maize, Bound. Layer Meteor., 15, 361-373, 1978.

Whynot, J.; Quinn, G.; Perryman, P.; Votlucka, P. “Control of Fine Particulate (PM2.5) Emissions from Restaurant Operations,” in PM2.5: A Fine Particle Standard, Proceedings of the Specialty International Conference, 661-671, 1988.

Williams, J. et al., Regional ozone from biogenic hydrocarbons deduced from airborne measurements of PAN, PPN, and MPAN, Geophys. Res. Letters, 24, 1099-1102, 1997.

A-16

B. OZONE MODELING FOR THE COLUMBUS, GEORGIA, REGION: PRELIMINARY ANALYSIS OF THE IMPACT OF LOCAL AND REGIONAL EMISSIONS ON OZONE IN COLUMBUS AND THE SENSITIVITY TO VOC AND NOX EMISSIONS REDUCTIONS.

A.G. Russell, James G. Wilkinson, and Jim Boylan

Abstract

A preliminary air quality modeling study has been conducted to provide a information about the

extent and causes of elevated ozone levels in the Columbus area. Ozone production in Columbus

from local emissions appears to be small in comparison to what occurs in other, larger urban

areas. Results of the study suggest that transport of ozone from upwind areas plays a major, if

not dominating, role in leading to high ozone in Columbus. While periods where Columbus

appears to be significantly impacted from emissions from Atlanta and Birmingham were

identified, no one area is overtly and singularly responsible for exceedance levels of ozone in the

Columbus area. Indeed, based on this preliminary investigation, in some instances the elevated

ozone levels in the Columbus area appear to be due to more or less similar contributions of

transported emissions, transported ozone, and locally generated ozone on top of the regional

ozone background. The modeling further suggests that regional controls, and controls in non-

attainment areas in the region, will lead to significant ozone reductions in Columbus and result in

ozone levels in the city below the remanded 8-hr standard.

The study does have limitations, and further studies, e.g., similar to what is being conducted as

part of the Fall Line Air Quality Study, should provide an even stronger foundation for

determining and better quantifying the sources of exceedance levels of ozone in Columbus.

Such studies will also allow more detailed identification of how the current air quality controls

will help lower ozone in Columbus.

Introduction

In the last three years, Columbus, GA has experienced episodes of elevated ozone which exceed

the 0.08 ppm, eight-hour average National Ambient Air Quality Standard (NAAQS). Of note,

1

implementation of the eight hour ozone NAAQS is on hold pending the Supreme Court review

of a lower court ruling suggesting that the US EPA did not have the authority to enforce such a

standard (USDOJ, 2000). At present, the highest three-year average of the fourth highest eight-

hour average ozone levels at any monitor in Columbus is 0.089 ppm which is slightly above the

eight-hour average ozone NAAQS. Columbus has two ozone monitors: Columbus Airport

which has a three-year average of the fourth highest eight-hour average ozone levels of 0.087

ppm; and the Crime Lab which has the higher 0.089 ppm average (Table 1).

Table 1. Fourth highest eight-hour average maximum ozone levels for 1997-1999 in Georgia. As seen, the years 1998 and 1999 are responsible for Columbus having a running three-year average over 0.085 ppm.

FIPS County ID

AIRS Site ID Station Name Fourth Highest Eight-Hour Average Ozone

1997 1998 1999021 0012 Macon SE 0.095 0.105 0.113051 0021 Sav E Pres 0.071 0.077 0.083085 0001 Dawsonville 0.079 0.096 0.089089 0002 South DeKalb 0.092 0.111 0.112089 3001 Tucker 0.085 0.111 0.111097 9994 Douglasville 0.090 0.112 0.105111 0094 Cohutta 0.075 0.081 0.081113 0001 Fayetteville - 0.111 0.113121 0044 Confederate 0.104 0.125 0.124127 0006 Brunswick 0.079 0.082 0.077135 0002 Gwinnett 0.086 0.111 0.103215 0008 Columbus Airport 0.080 0.091 0.089215 1003 Columbus Crime Lab 0.081 0.090 0.097223 0003 Yorkville 0.086 0.104 0.103245 0091 Augusta Bungalow 0.087 0.099 0.090247 0001 Conyers 0.110 0.112 0.123261 1001 Sumter 0.081 0.085 0.082

Two questions arise: what is leading to the high ozone levels and what might be done to mitigate

ozone formation on days when ozone is elevated? These are not easy questions to answer given

the nature of ozone formation and the various processes involved in causing high ozone. Ozone

is not directly emitted, but is formed from non-linear photochemical reactions of other

compounds, most notably oxides of nitrogen (NOX) and volatile organic compounds (VOCs).

Together, in the presence of sunlight, these two sets of compounds can react to form ozone and

other photochemical oxidants. The relationship is not straightforward and reducing one of the

two precursors (i.e., VOCs or NOX) does not necessarily result in a similar reduction in ozone

levels (e.g. a ten percent reduction in VOCs does not necessarily result in a ten percent reduction

2

in ozone). Instead, the relationship between VOCs, NOX, and ozone involves a large number of

chemical reactions that, as a set, are highly non-linear. Indeed, reducing one of the precursors

may lead to very little change, or may even increase ozone, in some cases, and in others may

lead to very substantial decreases. In addition, other processes, most notably transport, can be

very important. Ozone and its precursors can be transported hundreds of kilometers causing high

ozone levels downwind of the location where the majority of the emissions occur.

In designing an effective control program, it is necessary to understand how local ozone levels

will respond to emissions changes, both locally and regionally. The most direct method for

developing such relationships is through the use of air quality models that include the applicable

physics and chemistry. One can exercise these models to determine how the air quality will

likely respond to emissions changes. Here, one such model, the Urban-to-Regional Multiscale

(URM) Model was used to assess the likely effectiveness of local control programs and to

suggest the factors responsible for higher levels of ozone in the Columbus area. The intent of the

study is to provide guidance on what type of programs may be useful for mitigating high ozone

levels in the Columbus area. Given the limited resources that were available to conduct the

preliminary study, the results are useful only for developing an initial understanding of the ozone

problem in the Columbus area. Use of the results from the preliminary study in any regulatory

framework is highly discouraged. Future studies of the ozone problem in the Columbus area

must address the following issues:

Lack of detailed air quality data for pollutants other than ozone in the Columbus area;

Detailed quality assurance of emissions data and the emissions inventory;

Selection of episodes which are representative of conditions conducive to ozone formation in

Columbus; and

Development of a finer resolution meteorological, emissions, and air quality modeling grid

for the Columbus domain.

3

Method

URM, an advanced photochemical oxidant model that has the ability to use multiple grid sizes,

was applied to two historical periods, May 9-13, 1993 and July 11-15, 1995. The observed peak

during the July period was moderately high, having a 113 ppb peak one-hour average and 92 ppb

peak eight-hour average. The observed peak during the May period was low, having a 65 ppb

peak one-hour average and 61 ppb peak eight-hour average. Using data sets adapted from the

Southern Appalachian Mountains Initiative (SAMI) meteorological, emissions and air quality

modeling data bases (Russell et al., 1998), URM was run for the following cases:

the May 1993 base case;

the July 1995 base case;

a zero anthropogenic emissions sensitivity run for the May 1993 episode; and

a zero anthropogenic emissions sensitivity run for the July 1995 episode.

Further, URM/DDM-3D was configured for the following experiments:

Birmingham anthropogenic and biogenic VOC emissions sensitivity;

Atlanta anthropogenic and biogenic VOC emissions sensitivity;

Columbus anthropogenic and biogenic VOC emissions sensitivity;

Birmingham anthropogenic NOX emissions sensitivity;

Atlanta anthropogenic NOX emissions sensitivity; and

Columbus anthropogenic NOX emissions sensitivity.

4

Shown in Figure 1 is the entire URM air quality modeling domain and the Columbus sub-

domain. The grid size in the Columbus area was twenty-four kilometers per side which is larger

than is desirable for detailed urban-scale air quality model analyses, but due to project resource

constraints, recasting the emissions, meteorological, and air quality modeling data for a finer grid

resolution was not performed. However, the twenty-four kilometer grid size can be used to

assess sensitivities to local and regional controls though with limitations. One such limitation is

that finer scale responses are lost. That is, those physicochemical processes that function at

scales finer than twenty-four kilometers (e.g. rapid, fine-scale photochemical production and

accumulation of ozone) are masked in the air quality model due to the use of a coarse grid.

However, the longer-range transport to the region will be captured, as well as the regional nature

of the ozone in the Southeast.

5

Figure 1. URM air quality modeling domain and Columbus sub-domain. Note that Atlanta and Birmingham are within the twelve kilometer domain while Columbus is in the twenty-four kilometer domain.

Results

July 1995 Base Case. The first step in analyzing the results was to compare the observed

Columbus ozone to the simulated ozone. Results for the July episode are shown in Figure 2.

While the performance is generally good, the peak ozone levels on the days with the highest

ozone are underpredicted, and the simulated ozone generally does not go down as far at night.

The predicted peak during the episode was about 80 ppb, somewhat less than the observed peak

of 92 ppb on July 14th. The underestimate of the very highest levels can be due to inaccuracies in

the meteorological fields or the lack of fine scale resolution, or both. For example, a peak

resulting from the nearby point source in Alabama, will be lost, as would an intense local peak

downwind from mobile source emissions in a more heavily-trafficked area. Also, the observed

ozone sometimes goes to zero during the nights and early morning, but the simulated ozone does

not. Both phenomena (i.e. day time underprediction of high ozone and nighttime overprediction

of low ozone) are consistent with the emissions being underestimated, which is not surprising

given recent evidence that in particular both mobile source VOC and NOX emissions are

6

Figure 2. Comparison of observed and simulated ozone at Columbus, GA. While the peak ozone levels on most days is well reproduced, the day with the highest ozone finds somewhat lower simulated ozone. At night, the simulated ozone does not go to zero, but the observed ozone does on about half of the nights. The model is correctly reproducing nighttime ozone on those nights it does not go to zero, and suggests that sub-grid scale emissions are impacting the observations (e.g. excess sub-grid scale NO emissions which are carried over from the daytime hours are scavenging ozone during the nighttime hours). Note that the first two days are used to “spin-up” the model and remove the impacts of initial conditions.

underestimated by the MOBILE model used to generate those estimates, but the extent of this

bias is not known (e.g. Brzezinski and Newell, 1998). Further evidence of the emissions being

underestimated can be inferred from the results of the sensitivity runs which are described later.

Also, the two Columbus sites, particularly the airport site, are located near roadways where NOX

emissions from mobile sources scavenge ozone. However, since the proximity of the monitor to

the road is below the resolution of the model as used here, the conclusion that underestimated

mobile source NOX emissions are responsible for the lack of agreement in the predicted

nighttime ozone to the observed nighttime ozone is only speculative. The underprediction of the

daytime ozone peaks may also be, in part, an artifact of the larger grid sizes used here, but to a

lesser extent than the overprediction at night. Air quality models tend to underpredict the peak

ozone levels if their overall predictions are unbiased. The simulations also do well during the

weekend period, which has proven difficult in the past. In summary, the good model results

would suggest that the major processes impacting ozone in Columbus are being reproduced,

though the relatively coarse grid, likely biased mobile source emissions estimates and

underprediction of the peak ozone on some of the days suggest that there are some limitations in

the use of the results.

The high levels of ozone in Columbus should be not be viewed as a purely local phenomena.

One finding of the Southern Oxidants Study is the regional nature of ozone in the southeast

United States (Chameides and Cowling, 1995). That is, periods of elevated ozone tend to

coincide throughout the southeast. This is shown in Figure 3, which shows that much of the

Southeast experienced elevated levels of ozone during the simulated period, particularly around,

and immediately downwind of, the most densely populated areas. Also, examination of the

observed ozone levels in Georgia confirms that high levels of ozone are not isolated at a few

sites, but generally are experienced at many sites throughout the state. This suggests that ozone

transport, as well as local generation of ozone from emissions, is important.

7

July 1995, zero anthropogenic emissions in the Columbus area. The URM was run to show how

predicted ozone levels in Columbus respond to the removal of the anthropogenic emissions in the

Columbus area. This experiment was performed by removing the anthropogenic emissions for

the Columbus area and running the URM with the modified emissions inventory. The effect of

removing the anthropogenic emissions from the Columbus area is shown in Figure 4. As shown,

the impact of removing the anthropogenic emissions was to reduce the predicted peak ozone in

the Columbus area by 1 to 2 ppb.

8

Figure 3. Modeled ozone levels in the southern Tennessee, Alabama, and Georgia region on July 12, 1995 at 2:00 pm CDT (i.e. the time of peak ozone in the area). The ozone is not limited to the major cities, but ozone is found throughout the region. This is corroborated by observations from remote sites, for example, Great Smoky Mountains, north Georgia mountains, and Leslie, Georgia.

A 1 to 2 ppb reduction in ozone is small given that all of the anthropogenic emissions associated

with the Columbus area were removed. However, as stated previously, there are known or

suspected limitations which impact the interpretation of the study results. Firstly, the mobile

source emissions, if not other sources as well, are likely underestimated. An increase in the on-

road mobile source emissions estimates will likely have the effect of increasing predicted ozone

during the daytime, and subsequent removal of those emissions estimates will likely result in a

larger (i.e. greater than 1 to 2 ppb) change in predicted ozone. Secondly, the period modeled is

not as stagnant as some less frequent but very severe episodes that can lead to higher ozone

levels and show greater local influence Finally, the coarse grid size over the Columbus area that

was used in this study (i.e. twenty-four kilometers) may not capture very local scale high ozone

levels. Examination of the observed ozone record indicates significant ozone concentration

gradients can exist in the Columbus area (i.e. high ozone levels in and around Columbus which

rapidly transition to much lower ozone concentrations). However, the results do strongly suggest

that a significant fraction of the ozone in Columbus is from the regional background.

9

Figure 4. Deficit-enhancement plot of the simulated change in ozone from removing anthropogenic emissions in the Columbus area. The graphic was created by subtracting the URM ozone predictions which were based on the zero anthropogenic emissions from the base case URM ozone predictions. Areas with negative ozone values indicate that ozone is predicted to decrease if anthropogenic emissions are removed from the Columbus area, and positive ozone values indicate that ozone is predicted to increase if anthropogenic emissions are removed from the Columbus area.

Birmingham, Atlanta, and Columbus URM/DDM-3D Sensitivities. A unique feature of URM

not present in other air quality models is the capability to determine the impact of emissions (e.g.

VOC, NOX, SO2, or speciated VOC such as benzene) from a source area (e.g. Atlanta) on a target

area (e.g. Columbus). This feature is based on a technique known as the Decoupled Direct

Method in Three Dimensions (DDM-3D) (Yang et al., 1999). The implementation of theDDM-

3D in the URM allows for the efficient computation of sensitivity coefficients (e.g. change in

ozone concentration per change in NOX emissions) at the same time that the state field (e.g.

ozone concentration) is estimated. That is, in only one run of the URM, not only is the predicted

ozone field computed, but also one or more sensitivity coefficients are computed. The sensitivity

coefficients are used to predict the impact from an emissions source on a target area.

The URM was used to assess the sensitivity of ozone in Columbus to reductions of ozone

precursors locally and in areas that, at times, may contribute to higher levels in the Columbus

area. The precursors tested were anthropogenic VOCs and NOx, and biogenic VOCs.

Birmingham and Atlanta were chosen as the source areas since their emissions are high and they

experience periods of elevated ozone (i.e. greater than 90 ppb eight-hour average). Thus, if

transport from the source area (e.g. Atlanta) to the target area (e.g. Columbus) is important, the

impact on the concentrations due to the effects of transporting pollution from the source to the

target will be exhibited in larger sensitivity coefficients along the axis of transport. Note that

when transport is important, it is not only transport from a single source location, but also

transport from the regionally high ozone levels. Hence, if the model results show that the plume

from Atlanta impacts Columbus, the regional ozone and ozone precursors impacting Atlanta also

contribute to the impact in Columbus, as do the sources along the axis of transport between

Atlanta and Columbus.

In these tests, the sensitivities of emissions from the source areas (i.e. Birmingham, Atlanta, and

Columbus) on ozone in the target area (i.e. Columbus) are presented as the change in predicted

ozone in the target area due to a twenty-five percent reduction in the emissions from the source

area. Thus, the size and magnitude of the plume will vary from city to city. For example, given

the magnitude of emissions in Atlanta, one would expect to see a much larger region of influence

10

than for emissions coming from Columbus. Also, the Atlanta region is larger, so there are more

biogenic emissions as well.

As shown in Figure 5, the peak impact of anthropogenic VOCs on ozone in Columbus is small,

less than 1 ppb, regardless of the source area for the period modeled. This is particularly true for

the anthropogenic emissions in Columbus, where almost no impact on ozone in Columbus is

found from a reduction of anthropogenic VOC emissions emitted from the Columbus area for the

period modeled. .

However, there is a somewhat larger peak impact on ozone, still less than 1 pbb but greater than

that predicted from anthropogenic VOCs, in Columbus from the Atlanta biogenic VOC

emissions, as seen in Figure 6. This is likely due to the highly reactive isoprene emissions,

which are ubiquitous to biogenic sources especially southeast deciduous forests, being

transported from the Atlanta area which is well known to be a region rich in biogenic VOCs (e.g.

Chang et al., 1996; Geron et al., 1995).

11

Figure 4. Sensitivity of ozone to a 25% reduction in anthropogenic VOCs in Birmingham, Atlanta and Columbus.

Figure 5. The predicted peak sensitivity of ozone in the Columbus area due to a twenty-five percent reduction in anthropogenic VOC emissions from Birmingham, Atlanta, and Columbus (12 July 1995, 10:00 am CDT.

A slightly different result is found when considering NOX emissions. As seen in Figures 7 and 8,

NOX emissions have a greater overall impact on predicted ozone both in magnitude and spatial

extent than do VOC emissions. At 1:00 p.m. on July 10th, Figure 7, NOX emissions from

Birmingham are predicted to impact ozone in Columbus to an extent about equal to that of the

locally generated NOX emissions (i.e. on the order of about 1 ppb for a twenty-five percent

change in NOX emissions. Similarly though of larger magnitude and spatial extent, NOX

emissions from Atlanta at 10:00 am on July 12, Figure 8, are predicted to have a peak impact on

Columbus area ozone greater than 1 ppb for a twenty-five percent reduction in NOX emissions

which is a larger impact on Columbus area ozone than that due to local NOX emissions.

On the other hand, the May 1993 episode exhibited different results for the NOX emissions

sensitivities, Figure 9. For the May 1993 episode, NOX emissions from Birmingham and Atlanta

impact ozone generally to the north of those cities indicating that transport is generally from the

south to the north. However, as can be observed in Figure 9, Columbus area NOX emissions

continue to impact only local ozone though in this case, removal of locally generated NOX

emissions is predicted to slightly increase the Columbus area ozone.

12

Figure 5. Sensitivity of ozone to a 25% reduction in biogenic VOCs in Birmingham, Atlanta and Columbus.

Figure 6. The predicted peak sensitivity of ozone in the Columbus area due to a twenty-five percent reduction in biogenic VOC emissions from Birmingham, Atlanta, and Columbus (12 July 1995, 10:00 am CDT).

These sensitivity results suggest a number of important features. Peak ozone in the Columbus

region appears to be NOX limited. That is, decreasing NOX emissions locally, and from upwind

areas such as Atlanta and Birmingham when Columbus is downwind, will generally decrease

Columbus area peak ozone levels, and reducing VOC emissions, whether from upwind or local

sources, will not impact Columbus area peak ozone levels significantly. However, reducing NOX

emissions can lead to increases in ozone at night. In the cases when Atlanta and Birmingham

NOX emissions impact Columbus area ozone, the impacts of NOX emissions reductions in those

cities are predicted to be similar to the impact from reductions in locally generated NOX

emissions. However, Columbus is not always downwind of those cities, so while local

reductions always impact air pollution to some extent, reductions in the other cities do not

13

Figure 6. Ozone sensitivity to a 25% reduction in NOx in Birmingham, Atlanta and Columbus on July 10 at 1:00 p.m.

Figure 7. The predicted peak sensitivity of ozone in the Columbus area due to a twenty-five percent reduction in NOX emissions from Birmingham, Atlanta, and Columbus (10 July 1995, 1:00 pm CDT). Note that the predicted peak impact on Columbus area ozone due to Birmingham NOX emissions is about equal to that predicted for locally generated NOX emissions.

Figure 8. The predicted peak sensitivity of ozone in the Columbus area due to a twenty-five percent reduction in NOX emissions from Birmingham, Atlanta, and Columbus (12 July 1995, 10:00 am CDT). Note that the predicted peak impact on Columbus area ozone due to Atlanta NOX emissions is greater than that predicted for locally generated NOX emissions.

necessarily effect ozone in Columbus. That is, the impact on Columbus area ozone due to NOX

emissions from Atlanta and Birmingham will depend on the wind direction and speed.

The regional nature of the ozone is important in the context of Columbus and the results found

here. Regional emissions controls will be effective in improving the air quality in Columbus, but

without regional controls, significantly reducing ozone levels in Columbus, at least on days like

those modeled here, will be difficult. Also, while the impact of local emissions on Columbus

area ozone is small, emissions from Columbus contribute to the regional emissions load which in

turn impacts the background and regional ozone levels. Further, though not exhibited in the

episodes modeled in this study, Columbus emissions likely impact ozone levels in downwind

areas, but the magnitude of this impact is unknown. Also, in areas that have a significant

transport of ozone in to the city, local NOX emissions will often lead to local decreases in ozone

(i.e. local NOX emissions scavenge the ozone when insufficient VOCs are present), but

downwind increases in ozone (i.e. local NOX emissions are transported downwind to areas where

sufficient VOCs are present and photochemical production and accumulation of ozone can

occur). This complicates identifying the most effective strategies to reduce ozone.

14

Figure 8. Ozone sensitivity to a 25% reduction in NOx in Birmingham, Atlanta and Columbus on May 11 at 4:00 p.m.

Figure 9. The predicted peak sensitivity of ozone in the Columbus area due to a twenty-five percent reduction in NOX emissions from Birmingham, Atlanta, and Columbus (11 May 1993, 2:00 pm CDT). Note the slight disbenefit on Columbus area ozone due to a twenty-five percent reduction in locally generated NOX emissions (i.e. ozone goes up when NOX emissions go down) attesting to the non-linear nature of ozone production.

Discussion

The recent decision to designate areas as non-attainment has brought added importance to the

results found here. Thus, it is important to consider the limitations mentioned in the report. The

conclusion that Columbus is subjected to very significant transport of ozone and ozone

precursors, and that transport significantly contributes to the high ozone readings found at the

two Columbus monitors is supported not only by the modeling but also consideration of the

observations which shows a regional ozone cloud. The very highest readings are likely due to

subgrid-scale variations in the ozone due to “local” emissions, such as the moderate point source

in Alabama and strong traffic emissions, on top of this regional cloud. There is some evidence

of this by examining the ozone observations at the two monitors in Columbus that sometimes

show moderate differences. A more extensive study can address some of the issues that remain.

Such a study would include additional episodes to look at the range of meteorologies leading to

high ozone in the area and would use a finer grid resolution over the Columbus area. It would

also be preferable to use more recent episodes, for example those occurring in August of 1999

that led to high ozone levels in Columbus. It would also be good to get a more detailed

inventory of emissions in the nearby area, particularly of mobile sources which are suspected of

being underestimated. A second type of study that could be useful is to diagnose the high ozone

levels and link them to particular meteorological patterns, e.g., wind direction and speed, to

demonstrate the importance of transport. Such a study would include analysis of ozone in the

surrounding areas. The definitiveness of such an analysis is hindered by the lack of ozone data

in the upwind counties in Georgia and Alabama.

15

Conclusion

An air quality modeling study has been conducted to provide a preliminary evaluation of the

extent and causes of elevated ozone levels in the Columbus area. Local ozone production from

local emissions appears to be small in comparison to what occurs in other, larger urban areas,

though some caveats are significant. Firstly, there is a very good chance that mobile source

emissions are underestimated. The same may be true for emissions from other sources as well.

Secondly, only two episodes were considered in this study. More extensive episode selection

work is warranted for the work to be considered more definitive. In particular, episodes chosen

for their representativeness to those conditions that most typically lead to ozone exceedences in

the Columbus area must be elucidated. Thirdly, Columbus has only two ozone monitors, and no

NOX or VOC measurements. Such measurements are desirable for air quality model evaluation.

Finally, the coarse scale grid that was used over the Columbus area (i.e. twenty-four kilometers)

is likely to have masked some sub-grid scale (i.e. less than twenty-four kilometers) phenomena

(e.g. proper capture of the nighttime scavenging of ozone by NO, and the impact of plumes from

nearby sources causing locally high levels). Use of grid scales on the order of five or six

kilometers should be considered in future studies in the Columbus area.

Further, the results of the study suggest that transport of ozone from upwind areas plays a major,

if not dominating, role in leading to high ozone in Columbus. While periods where Columbus

appears to be significantly impacted from emissions from Atlanta and Birmingham were

identified, no one area is overtly responsible for exceedance level ozone in the Columbus area.

Indeed, based on this preliminary investigation, in some instances the elevated ozone levels in

the Columbus area appear to be due to more or less equal contributions of transported emissions,

transported ozone, and locally generated ozone on top of the regional ozone background.

16

Future Studies

A number of implied recommendations have been made throughout this report as to what should

be considered in future air quality modeling studies of ozone in the Columbus area. These

recommendations include:

Additional episode selection work needs to be performed to choose episodes that are

representative of the range of conditions conducive to high ozone levels in the Columbus

area

A finer scale grid mesh, on the order of five or six kilometers, should be used in future air

quality modeling so that fine scale phenomena are better captured; and

The emissions inventory should be further developed for the Columbus area. For one,

recent studies suggest that mobile source emissions are underestimated.

Such actions will provide greater confidence to those reviewing such work that our results to

date are applicable to the range of conditions found in the Columbus area.

17

References

Brzezinski, D. J. and T. P. Newell, 1998. “MOBILE6: A Revised Model for Estimation of Highway Vehicle

Emissions,” presented at the Eighth Annual Conference on Emissions Inventories, Emissions Inventory: Living in a

Global Environment, Air and Waste Management Association, New Orleans, LA 9-12 December.

Chameides, W. L. and E. B. Cowling, 1995. “The State Of The Southern Oxidants Study: Research

Accomplishments & Future Plans,” http://www2.ncsu.edu/ncsu/CIL/southern_oxidants/docs/state.html

Chang, M. E., D. E. Hartley, C. Cardelino, and W. Chang, 1996. “Inverse modeling of biogenic isoprene

emissions,” Geophysical Research Letters, Volume 23, Number 21, 3007-3010, 15 October.

Geron, C. D., T. E. Pierce, and A. B. Guenther, 1995. “Reassessment Of Biogenic Volatile Organic Compound

Emissions In The Atlanta Area,” Atmospheric Environment, Volume 29, Number 13, 1569-1578.

Russell, A. G., M. T. Odman, and J. G. Wilkinson, 1998. “Air Quality Modeling Protocol: Meteorological,

Emissions, and Air Quality Modeling for an Integrated Assessment Framework in Support of the Southern

Appalachian Mountains Initiative,” prepared for the Southern Appalachian Mountains Initiative (SAMI) Technical

Oversight Committee.

USDOJ, 2000. Petition For A Writ Of Certiorari To The United States Court Of Appeals For The District Of

Columbia Circut, No. 99-1257. In the Supreme Court of the United States, Carol M. Browner, Adminstrator Of The

Environmental Protection Agency, et al., Petitioners, versus American Trucking Association, Inc., et al.

http://www.usdoj.gov/osg/briefs/1999/2pet/7pet/99-1257.pet.aa.html

Yang, Y. J., J. G. Wilkinson, and A. G. Russell, 1999. “Direct Sensitivity Analysis of Multidimensional

Photochemical Models,” Environmental Science & Technology, Volume 33, Number 7, 1116-1126.

18

http://www2.ncsu.edu/ncsu/CIL/southern_oxidants/docs/state.html

C. 5 OCTOBER 1999 FALL LINE AIR QUALITY STUDY PROPOSAL

Fall line Air Quality Study (FAQS)

An outline for assessing urban and regional air pollution, identifying the sources of pollutants and pollutant precursors, and recommending solutions to realized and potential poor air quality in the Augusta, Macon, and Columbus, Georgia metropolitan areas.

Prepared by:The Fall line Air Quality Study Alliance

List MOU signatories

Total Project Cost: $3 millionTerm: January 1, 2000 – December 31, 2002

October 5, 1999

A. OVERVIEWThe Atlanta metropolitan area has failed to attain state and federal air quality standards since 1979. A 20-year effort to improve air quality in the metro area has not been successful. Now at crisis proportions, Governor Barnes and the General Assembly created the Georgia Regional Transportation Authority (GRTA) in response. GRTA has been granted broad powers to effect change in “nonattainment” areas, and through the Governor’s Development Council, other areas of the state.

It now appears that other metropolitan areas in Georgia may be experiencing poor air quality. On some days, concentrations of ground-level ozone in Augusta, Columbus, and Macon approach or exceed the threshold for clean air as defined by state and federal standards*. Rather than wait for a “nonattainment” designation before acting to improve air quality, these proactive metropolitan areas desire solutions to their air quality problems before the onset of crises†. Unfortunately, information is lacking in these regions regarding the composition of the air, the source of air pollutants and pollutant precursors, and the feasibility, effectiveness, and efficiency of any potential controls.

Ozone, the primary pollutant in the type of smog that most often afflicts Georgia’s communities, is produced in the atmosphere via a complex photochemical process involving two types of chemical compounds: volatile organic compounds (VOCs and also often referred to as “hydrocarbons”) and nitrogen oxides (NOx). The intense summer sun provides the energy needed to drive these reactions, and if the winds are light, ozone can accumulate rather than disperse.

VOCs + NOx + sunlight O3

Nitrogen oxides are a byproduct of the combustion process. As gasoline, diesel fuel, jet fuel, coal, natural gas, wood, or other combustibles are burned, nitrogen oxides are likely produced. Volatile organic compounds may be emitted into the atmosphere as paints, printing agents, fuels, or solvents evaporate or are otherwise released into the air. In the Southeast US, there is also an abundance of VOCs that are emitted from natural sources such as many types of trees, crops, and other vegetation.

* The US EPA’s National Ambient Air Quality Standard (NAAQS) for ground-level ozone is 0.12 ppmv averaged over 1-hour. Within Georgia but outside of Atlanta, only the Macon area meets the criteria for nonattainment of the 1-hour standard (i.e. four or more days over a three-year period on which 1-hour average ozone concentrations are greater than 0.12 ppmv). Macon reached this milestone on July 27, 1999 when the ozone monitor there registered its fourth violation of the 1-hour standard since 1997. In July of 1997, the US EPA promulgated a more stringent standard of 0.08 ppm averaged over 8-hours and remanded the 1-hour standard for all areas currently meeting it. On May 14, 1999 however, the United States Court of Appeals for the D.C. Circuit ruled that this new standard could not be enforced (American Trucking Associations, Inc., et al. v. United States Environmental Protection Agency; Case No. 97-1440). The court ruling aside, Augusta, Columbus, and Macon have all experienced sufficiently frequent violations of the 8-hour standard that, prior to the Court of Appeals ruling, could have warranted a nonattainment designation. If the US EPA successfully appeals the court ruling or corrects the deficiencies identified by the court, these cities could again face nonattainment.† Attached is a signed memorandum of understanding attesting to each community’s commitment to proactively address their air quality concerns, as well as a pledge to work with each other and several technical partners to resolve these problems autonomously.

2

During the 1980s, Atlanta pursued an ozone control strategy focused on reducing the emissions of man-made VOCs. By the early 1990s, continued exceedance of air quality standards indicated that this strategy was ineffective and concurrent research at the Georgia Institute of Technology suggested that the “VOC only” strategy failed because it did not account for the abundance of VOCs from natural sources. Even if all man-made VOC emissions were eliminated, sufficient natural VOC emissions remained, and combined with the prevalent man-made NOx emissions, could generate ozone concentrations greater than the health-based standards allow. Unfortunately, this became evident only after approximately $1 billion was expended on reducing man-made VOC emissions in the Atlanta area. Since then, strategies for improving Atlanta’s air quality promulgated by the Georgia Environmental Protection Division (EPD) have focused almost exclusively on controlling emissions of NOx. Still the Atlanta area fails to attain air quality standards. This is because 1) controlling emissions of NOx is both difficult and expensive, and 2) it is widely believed that regionally elevated concentrations of ozone exasperate Atlanta’s attempts to mitigate ozone locally.

In Atlanta, recent and ongoing research conducted or funded by the EPD, US EPA, and others seeks to discover the root causes of excessive ozone concentrations and the most effective and efficient methods to decrease this air pollution. As of the summer of 1999, the EPD has established a network of twelve ozone monitors, and four Photochemical Assessment Monitoring Stations (PAMS) in the Atlanta area. The PAMS sites collect data on the concentrations of VOCs, NOx, and meteorology – the key components controlling ozone formation and accumulation. Atlanta has also been the site of several research monitoring campaigns during which many additional monitors were deployed for a short period to gather intensive and more comprehensive data relevant to ozone air quality (Atlanta Ozone Precursor Study, US EPA, 1990; Atlanta Intensive, Southern Oxidants Study, 1992; PM Supersite, Southern Oxidants Study / Southern Center for the Integrated Study of Secondary Air Pollutants, 1999). All these endeavors have been used to analyze and assess Atlanta’s air pollution problem and have contributed to a better understanding of the causes and consequences of poor air quality in the area.

In preparing and revising the State Implementation Plan (SIP) for the Atlanta Ozone Nonattainment Area (1979, 1985, 1987, 1994, 1998, 1999), the EPD has also developed detailed VOC and NOx emission inventories that identify the source and type of ozone precursors and the location and time they are released into the atmosphere. Extensive computer simulations in support of the SIP and other research activities (e.g. the Ozone Transport Assessment Group, 1995-1997, and the Southern Appalachian Mountain Initiative, 1997-1999), have helped regulators and the public better understand the effects of past, present, or proposed changes in emissions resulting from change, growth, technological innovation, and regulation. These studies have also provided regulators a better understanding of Atlanta’s ozone problem in the context of the air quality conditions prevalent across the whole Southeastern US.

Taken collectively, these activities in Atlanta have significantly contributed to the development of strategies to improve air quality in the area. Of primary importance, they directly led to an initial strategy for decreasing ozone concentrations in the late 1970s and 1980s*; a major revision

* State Implementation Plans developed by the Georgia Environmental Protection Division in 1979, 1985, and 1987 pursued an ozone reduction strategy based on decreasing VOC emissions in the metropolitan Atlanta area.

3

of that strategy in the early 1990s as regulators shifted their efforts from controlling VOCs to controlling NOx*; and, by 1997, to a recognition of the regional contributions to poor air quality in Atlanta and the development of a strategy to address them†. Now as Augusta, Macon, and Columbus face the possibility of being designated nonattainment, it is evident that even the most fundamental of information needed to improve air quality in these communities is lacking. Whereas the single ozone monitors in Augusta and Macon, and two ozone monitors in Columbus are sufficient to identify the presence of excessive pollution, they do not provide sufficient information to diagnose the causes of poor air quality or to guide decision-making that will effectively and efficiently correct the problem‡.

Herein is an outline describing a study for assessing urban and regional air pollution, identifying the sources of pollutants and pollutant precursors, and recommending solutions to realized and potential poor air quality in the Augusta, Macon, and Columbus metropolitan areas. As the three cities lie along Georgia’s “fall line” – the line dividing the piedmont region from the coastal plain – this study is hereafter described as the Fall line Air Quality Study (FAQS). The FAQS will primarily address ground-level ozone but ancillary results will also provide better understanding of the mechanisms contributing to other pollutants such as fine particulate matter.

Included in this outline are the approximate costs associated with the FAQS. These costs were estimated under the assumption that the study leverages the capital resources (hardware, software, laboratories, field equipment, and expert personnel) afforded by the Georgia Institute of Technology (Georgia Tech) and more specifically the Center for Urban and Regional Ecology (CURE), the School of Earth and Atmospheric Sciences (EAS), and the School of Civil and Environmental Engineering (CEE). It may be possible as well as practicable that some components of the study draw on in-kind resources, or compensated services provided by others outside of Georgia Tech. In particular, universities, metropolitan planning organizations, and private environmental consulting firms local or external to Augusta, Macon, and Columbus may be able to provide equivalent or superior services at reduced costs. Inclusion of some groups may also be desirable from the perspective of educating, developing, or transferring capabilities to the local communities. Where such benefits are possible, these opportunities should be employed. Nevertheless, for the purpose of identifying costs only, this outline assumes Georgia Tech is the sole executor.

Collectively, these three plans sought a 68% decrease in VOC emissions from projected, uncontrolled emissions in 1987. * Findings from the Southern Oxidants Study (The State of the Southern Oxidants Study: Policy Relevant Findings in Ozone Pollution Research 1988-1994; April 1994) suggested that ozone concentrations in the South are more sensitive to changes in anthropogenic NOx emissions than anthropogenic VOC emissions. The 1994 SIP for the Atlanta Ozone Nonattainment Area demonstrated that “no amount of VOC reduction will be enough to show attainment with the ozone air standard.”† With the failure of many states to develop a plan for attainment of the 1-hour ozone air quality standard, the US EPA required all states east of the Mississippi River to participate in a two year study called the Ozone Transport Assessment Group (OTAG). Beginning in 1995, “OTAG was charged with assessing the significance of pollutant transport and with recommending control strategies for reducing that transport.” The policy outcome of OTAG was the Regional NOx SIP Call. Although currently delayed by pending litigation, the Regional NOx SIP Call mandates that states significantly decrease emissions of NOx by 2007 in and outside of current nonattainment areas.‡ The Georgia EPD began monitoring ozone in the Augusta area (Richmond county) in 1989. The sole Macon (Bibb County) monitor was deployed in 1997. Outside of Atlanta, the longest continuous ozone monitoring has been in the Columbus area. The EPD has been operating two monitors in Muscogee county since the early 1980s.

4

B. THE FALL LINE AIR QUALITY STUDYThe study shall span three years beginning January 2000 and terminating December 2002. It shall consist of four primary components: 1) enhanced monitoring; 2) emission inventory development; 3) scenario modeling; and 4) analysis, assessment, and recommendation. These four activities will be performed concurrently, but staged over four periods beginning with a six-month period of preliminary assessment and study definition. This will be followed by a one-year intensive period for emission inventory development, model validation, pilot field studies in the three metropolitan areas, and analysis. A second year-long intensive will use a more comprehensive field study, scenario modeling, and associated analyses to provide the primary information needed to properly assess the study areas and develop appropriate strategies. A final six-month period will be used to transfer the technologies implemented in the three areas to local or state authorities and develop comprehensive recommendations for improving air quality in the short and long term. Progress reports will be provided at six-month intervals with the final report, complete with recommendations, provided at the conclusion of the fourth period. Each of these four periods is described in more detail below.

Period 1 (January 2000 through June 2000)The FAQS will kickoff with a six-month concentrated effort to prepare for the first field campaign as well as initiate database development in support of the emissions inventory and air quality modeling activities. During this period, retrospective air quality analyses using existing information from the EPD chemical and National Weather Service meteorological monitoring networks will be completed to aid in designing the field study. The direction of winds coupled with corresponding ozone peaks will help identify the prevailing regions upwind and downwind of the metropolitan areas that will be targeted during the enhanced monitoring initiatives that will begin near the end of this period. This preliminary assessment also may be utilized to develop tentative action plans that may be implemented immediately to help mitigate the potential for future violations of air quality standards.

The preliminary assessment will also be used during this first period to determine the appropriate study domain (area of impact), and to identify relevant episodes for evaluation of potential strategies. Once the study domain is defined, it will be possible to begin creating a database of spatial and temporal information relevant to the study areas in general, and the emission inventories in particular. Among others, the database will house information on land use, demographics, vehicle travel, and industrial activity. Beginning in this period and extending into the next, man-made and natural emissions of VOCs and NOx will be derived from this information. These emissions will be disaggregated in time, space, and chemical speciation, but collectively, they constitute a full account or inventory of all emissions relevant to ozone air quality in the study domain.

At this time, training and other technology transfer for the benefit of the study as well as the local communities or study partners will also commence. Depending on need, training will be provided in any of the technical areas related to the field study, emissions inventory development, or air quality modeling.

Period 2 (July 2000 through June 2001)The first field study initiated in Period 1 will be completed in this term. The goal of this initial field study is to assess the relative contributions of local sources compared to the contributions of

5

sources beyond the metropolitan areas. VOC, NOx, and ozone data collected at locations upwind and downwind of the metropolitan areas should provide sufficient evidence to determine local contributions, with any residual pollutant loads due to regional background or external transport. Due to constraints on time, it is not possible to establish static upwind and downwind monitoring stations for VOCs, NOx, and ozone during the first ozone season (May 2000 – September 2000) in each of the three metropolitan areas. The Southern Center for the Integrated Study of Secondary Air Pollutants (SCISSAP) based at Georgia Tech however, has a fully equipped and functional mobile air quality laboratory for measuring VOCs, NOx, ozone, meteorology, and other pollution related variables. This mobile laboratory will be transported to each of the three metropolitan areas and deployed for 1 to 1.5 months during the ozone season.

Focused on the pollution events observed during the field study, meteorological modeling will commence to recreate the weather conditions, in a computational framework, that occurred during these episodes. Concurrently, the work to create baseline emission inventories for the study domains also will continue from Period 1. Once completed, emissions and meteorology will then be input into an advanced, three-dimensional photochemical transport air quality model to emulate the chemical conditions that were observed during the field study. NOx, VOC, and ozone data collected during the initial field study will be used to validate or otherwise ensure that the model is performing satisfactory (as determined by US EPA guidance for model performance). These simulations may also be used to identify key local and distant contributors to local and regional air quality.

Based on analyses from the first field study, the baseline emission inventory, and baseline modeling activities, a second field study will be designed for implementation the following year.

Period 3 (July 2001 through June 2002)A second ozone season-long field study will be conducted to augment the data collected the previous year. As meteorological conditions vary from day to day and year to year, it is important to capture the variability in these conditions to determine how they affect pollutant concentrations. In addition and with knowledge gained from the initial study, permanent stations will be established to begin monitoring long-term trends in NOx, VOC, and ozone concentrations, and to provide continuous assessment as the regions change. At the conclusion of the study, all capital and operations related to these permanent stations will be transferred to local partners or the EPD. Georgia Tech will provide training sufficient to operate these stations. The mobile air quality laboratory used exclusively during the first field study will again be deployed in each of the three metropolitan areas to supplement data from the permanent monitors.

Working with local stakeholders, several future year scenarios will be developed and evaluated using the emissions and air quality modeling system developed and validated in Period 2. In this process, potential changes in either local or distant sources can be examined relative to their impact on local ozone concentrations. Through this process, it may be possible to identify the optimal or least cost strategy that must be implemented in order for the area to meet or maintain the air quality standard. Options related to growth, voluntary or regulatory controls, or technology development and implementation may be explored. These results will be used in the final period to develop recommendations or action plans for the community.

6

Period 4 (July 2002 through December 2002)In this final six months, training will be completed, and all monitoring stations, emissions inventories, and air quality models will be transferred to local partners or the EPD. With guidance by the local stakeholders, final unique recommendations for meeting or maintaining air quality standards will be developed for Augusta, Macon, and Columbus and presented to each community for possible implementation.

C. MANAGEMENTAs the results and recommendations of the Fall line Air Quality Study will have far reaching impacts, it is important that all stakeholders are informed, represented, and accepting of the study. To ensure this, project oversight shall be provided by a Coordinating Council consisting of representatives from the Augusta, Macon, and Columbus metropolitan areas, the Georgia Environmental Protection Division, the Georgia Regional Transportation Authority, the United States Environmental Protection Agency Region IV, and other stakeholders such as representatives from business and industry, environmental advocacy groups, and concerned citizens.

The Coordinating Council shall also seek advice from an independent Scientific Advisory Committee. This committee shall consist of university, government, and industry scientists and engineers with expertise in monitoring, emission inventories, and air quality modeling.

Implementation of the FAQS itself shall be guided by a Director that will coordinate the activities of the Chemical and Meteorological Measurement group and the Emissions, Models, and Effects group.* The Director shall also be responsible for organizing and coordinating activities within and between the Coordinating Council, the Scientific Advisory Committee, and the technical work groups. Finally the Director shall work with all FAQS participants to interpret and communicate all technical results and develop comprehensive air quality policy recommendations for the Augusta, Macon, and Columbus metropolitan areas.

This management structure is illustrated in Figure 1. It is patterned after the management structure utilized in the highly successful Southern Oxidants Study. The important features of this broad-based approach include:

Stakeholders select the policy-relevant research themes and priorities;

* At Georgia Tech, Dr. Michael Chang of the Center for Urban and Regional Ecology (CURE) will manage at the project level. Professor C.S. Kiang will serve as senior advisor. CURE will also direct policy development (e.g. recommendations for controls) and implementation guidance (e.g. forecasting / smog alert programs).

The Air Resources Engineering Center (AREC), coordinated by Professor Ted Russell, will develop the emission inventories and relevant air quality models, and execute all simulations exploring the consequences of different scenarios (working in collaboration with each metro area to develop the scenarios), and analyze the results and associated data.

The Southern Center for the Integrated Study of Secondary Air Pollutants (SCISSAP), coordinated by Professor William Chameides, will execute the pollutant monitoring and diagnostic analysis component of this project.

Cooperation among these three centers is a natural outcome of their organization within Georgia Tech in the Institute for Sustainable Technology and Development.

7

The project utilizes the existing advanced instrumentation and expertise residing in the government, university, and private communities in general, and in Georgia Tech in particular; and The FAQS provides opportunities for education and technology transfer within the university setting as well as between Georgia Tech and local, state, and regional constituents

Figure 1. Organization of the FAQS.

8

Coordinating Council

Scientific Advisory Committee

Director

Chemical & Meteorological Measurement Work Group

Emissions, Models, & Effects Work Group

D. DELIVERABLESBiannual progress reports will be provided. Meetings for the Coordinating Council and Scientific Advisory Committee will be scheduled at their request, but no less than biannually. Other milestones and deliverables are described as follows. Jan 2000 Project Kickoff / Organizational Meeting

Selection of Coordinating Council and Scientific Advisory Committees

Apr 2000 Initial Field Study Design Completed

May 2000 Initial Field Study Begins

Jun 2000 Six-month Progress ReportPreliminary Assessment Completed

Sep 2000 Initial Field Study Completed

Dec 2000 Six-month Progress ReportBaseline Emission Inventories Completed

Apr 2001 Field Study #2 Design Completed

May 2001 Field Study #2 BeginsPermanent Monitoring Stations Functional

Jun 2001 Six-month Progress ReportAnalysis of Initial Field Study Completed

Sep 2001 Field Study #2 Completed

Dec 2001 Six-month Progress ReportBaseline Modeling and Validation CompletedScenario Emissions Inventories Completed

Jun 2002 Six-month Progress ReportAnalysis of Field Study #2 CompletedScenario Modeling Completed

Dec 2002 Final ReportStudy Results and Policy RecommendationsTransfer of Monitoring Station OperationsTransfer of Emission Inventories and ModelsProject Completed

9

E. COSTS OF THE STUDYCosts associated with the Fall line Air Quality Study assume Georgia Tech as executor. These costs do not reflect matching real or in-kind funds that may be provided by local stakeholders, or services that may be reasonably provided by sources external to Georgia Tech. Costs described below do however, reflect the contribution from Georgia Tech to the project through the use of the Southern Center for the Study of Secondary Air Pollutants (SCISSAP) mobile air quality lab. This contribution is valued in excess of $1 million.

Period 1 (January 2000 through June 2000)

Database DevelopmentRetrospective air quality analysisLand useRegional statistics

$75,000

Preliminary Emissions Inventory Development $90,000Air Quality Model SetupEpisode electionDomain selection

$90,000

Field Study Design $100,000Equipment (includes SCISSAP mobile laboratory) $1,125,000Training $20,000Total Period 1 $1,500,000

Period 2 (July 2000 through June 2001)

Model Development and Validation $300,000Initial Field Study $500,000Analysis & Field Study #2 Design $200,000Total Period 2 $1,000,000

Period 3 (July 2001 through June 2002)

Scenario Development and Simulation $300,000Permanent Monitoring Network Design, Training and Technology Transfer

$200,000

Field Study #2 $400,000Analysis $100,000Total Period 3 $1,000,000

Period 4 (July 2002 through December 2002)

Technology Transfer $150,000Recommendations and Action Plan Development $325,000Final Report $25,000Total Period 4 $500,000

10

Total (January 2000 through December 2002)

Period 1 $1,500,000Period 2 $1,000,000Period 3 $1,000,000Period 4 $500,000Subtotal Periods 1 – 4 $4,000,000Georgia Tech Contribution (SCISSAP mobile laboratory) ($1,000,000)Total Project Cost $3,000,000

F. LEGACY OF THE FAQSAt the conclusion of the Fall line Air Quality Study, Augusta, Macon, and Columbus will be provided with sufficient information to begin making the difficult decisions necessary to meet or maintain state and federal air quality standards. Where poor air quality is identified by the FAQS to result from factors beyond the political jurisdiction of these metropolitan areas, EPD or the US EPA will be provided with information necessary to address these externalities. Further, as political, environmental, or economic conditions change in the future, each community will have sufficient skills and computational tools to evaluate and modify alternate action plans. Finally each community will possess a basic monitoring system capable of assessing regional and local air quality conditions, and tracking progress in meeting and maintaining air quality standards.

11

D. 29 OCTOBER 1999 REVIEW OF FALL LINE AIR QUALITY STUDY PROPOSAL BY HAROLD REHEIS, DIRECTOR, GEORGIA ENVIRONMENTAL PROTECTION DIVISION

1

E. 22 DECEMBER 1999 REVIEW OF FALL LINE AIR QUALITY STUDY PROPOSAL BY CATHERINE ROSS, EXECUTIVE DIRECTOR, GEORGIA REGIONAL TRANSPORTATION AUTHORITY

1

F. POLICY STATEMENTS GOVERNING THE CREATION, ROLE, AND PROCEDURES OF THE FAQS COORDINATING COUNCIL AND SCIENTIFIC ADVISORY PANEL.

POLICY STATEMENT

FAQS COORDINATING COUNCIL

The Environmental Protection Division (EPD) will contract with the Georgia Tech Research Corporation (GTRC) to do the work of the Fall Line Air Quality Study. Although GTRC, in matters of this study, will be under the direction of EPD, the contract will describe the guidelines and schedules under which the study will be conducted. In most cases, these guidelines will be in the form of U.S. Environmental Protection Agency guidance documents. A Fall Air Quality Study Coordinating Council, consisting of study stakeholders, will be created to advise EPD, which will in turn direct GTRC, in the event that a situation arises where the guidance is either not explicit or is ambiguous, This situation is expected to be a rare event. The Council will also serve as a forum and a medium for information dissemination for FAQS-related concerns such as non-attainment designations, implications of non-attainment status, the State Implementation Plan (SIP) and attainment demonstration timelines and process.

The Coordinating Council shall consist of one representative each from the Environmental Protection Division (who will also serve as Chair), the Georgia Regional Transportation Authority, the Georgia Department of Transportation and the Department of Defense; and two representatives each from the cities of Augusta, Columbus, and Macon. Representatives from the U.S. Environmental Protection Agency and the States of Alabama and South Carolina will serve as ex-officio [non-voting] members. The Council will meet on an as-needed basis as the Council determines. Meeting times and places will be determined by Council consensus.

The Council shall strive for and make a reasonable and determined effort to reach all decisions by consensus. Should this reasonable and determined effort fail to provide consensus, the majority of voting members shall prevail.

SCIENTIFIC ADVISORY PANEL

The Scientific Advisory Panel will serve two purposes: 1) Provide peer review of the study process; and 2) Upon the request of the FAQS Coordinating Council, provide technical advice on monitoring, inventory, modeling and control strategy issues. The Scientific Advisory Panel shall consist of appointees made by GTRC and members of the Coordinating Council and shall be experts in the areas of ambient monitoring, emissions inventories, air quality modeling and air pollution control equipment, strategies and/or cost effectiveness.

The Panel will meet initially upon the request of the Coordinating Council and thereafter on an as-needed basis as the Panel determines. GTRC shall appoint the Panel Chair. The Panel shall determine its meeting times and places.

The Panel shall strive for and make a reasonable and determined effort to reach all decisions by consensus. Should this reasonable and determined effort fail to provide consensus, the majority of voting members shall prevail. A report of the decision shall be provided in writing to the Coordinating Council within 3 days of the decision. In cases where the decision was reached by majority vote, a dissenting minority opinion may be provided.

1

G. 30 JUNE 2000 LETTER FROM GOVERNOR ROY BARNES TO US EPA REGION IV ADMINISTRATOR JOHN HANKINSON RECOMMENDING NONATTAINMENT AREAS IN GEORGIA FOR THE 8-HOUR OZONE NATIONAL AMBIENT AIR QUALITY STANDARD.

1

H. GENERAL CRITERIA FOR IDENTIFYING POTENTIAL AMBIENT AIR QUALITY MONITORING SITES AND SPECIFIC CRITERIA FOR DEPLOYING THE MOBILE AIR QUALITY RESEARCH LAB.

Set-up criteria specific to the Mobile Air Quality Research lab.

1. Access [semi-truck can maneuver it into a relatively level position, we can fine-adjust].

2. Electrical Power

Standard American style "single phase", ie two phase service; hardwired into it. Each phase is 120 volts from neutral, and the two phases are 240 volts from each other. We run 150 amps, though our main breaker is 200 amps. Fuses or a main breaker on the host's service panel is recommended in order to prevent an overcurrent condition. We have a service disconnect switch box mounted on the front-outside of our trailer, where the wiring will finally be connected.

At previous sites, 3 individual copper 00 gauge wires for phases I + II, and neutral were provided to us by the local electrical service contractor who also connected us to their main power panel [cable type or insulation type unknown]. The 00 gauge requirement is assuming that the cable won't be running across pavement or in a similar condition that heats it up excessively; otherwise we might need to derate to the next heavier wire gauge. Local code may require (even for just 2 weeks) that the wiring be run in PVC conduit. Running 200-300’ cable should not be a problem!

In the past, we drove a ground rod at our trailer, and could provide it again, but ground could also be provided from a service panel, requiring a 4th wire to carry it. In addition to the AC ground, we require a separate ground for our data acquisition system, and a 3rd ground for our tower as lightning protection. We may require a special tool if the ground rods need to be driven into asphalt or concrete!

3. Space [to lower/raise tower, to span guy-wires]

The 10m tower attached at the rear end to the long side of the trailer is hinged and 'folds out'; therefore requiring an additional 10m [33'] clearance.

General criteria for selecting appropriate ambient air quality monitoring sites.

4. 'Fetch' [distance to nearest bldg/obstacle should be at least 20 times its height]

5. 'Foot-print' or Representativeness

Street canyons or close-by point sources are not ideal for the scope of this study. Wide open fields in some residential or park area without lots of through-traffic would be most suitable. Locations near sports fields of local communities or colleges would also work.

6. Safety [minimize the risk of vandalism].

With all the criteria that need to be met, the electric power one requires the most lead time!! It is advised to get local, certified electricians involved asap for preparations of el power service.

1

I. DESCRIPTION OF SITES PROPOSED FOR USE AS LOCATIONS TO DEPLOY THE MOBILE AIR QUALITY RESEARCH LAB OR TO ESTABLISH THE STATIONARY SATELLITE MONITORING SITES IN COLUMBUS, MACON, AND AUGUSTA, GA.

Potential FAQS Monitoring Sites in Muscogee County, GA (visited 4/6/00)

Site 1: Columbus State University, Parking lot (near College of Sciences).Site description: Located ~north of urban core and about 1.5 miles south of existing EPD O3 monitor at Columbus Airport. Elevated location with easy access. Located at edge of campus andnearby (CSU owned) residential housing (detached homes).Power requirement: Must drop new line but source is nearby (50 to 100 ft).Fetch: Open to the west. Some trees and nearby single story homes on other sides.Security: Subject to availability of campus police patrol.Potential interference: roof vented chemical hoods from labs in College of Sciences.Intangibles: Possible utilization of laboratories or other facilities afforded by proximity to CSU College of Sciences.PRO'S: CLOSE TO LAB FACILITY, CLEAN WATER (DDW), EDUCATIONAL OUTREACHCON'S: CLOSE TO AUTOMOTIVE [COLD STARTS] AND BIOGENIC [TREES] SOURCES, LIMITED SECURITYSUITABILITY: LESS LIKELY.

Site 2: GBI Crime Lab.Site description: Current site of EPD O3 monitor. Located ~northeast of urban core and about 3 miles east of EPD O3 monitor at Columbus Airport. Elevated location with moderate access (may be difficult for semi to maneuver and fence is open only during business hours).Power requirement: Probably can be met relatively easily.Fetch: Single story buildings in front of monitor (~20 yards) and to either side (200 yards). Area behind has a large pasture and golf course. (note: pasture might be ideal, but no road to access it).Security: Locked fenced area. Major nearby police presence. Low visitor traffic site.Potential interference: 4 lane road ~100 yards in front of monitoring site (being able to push the monitor back ~300 yards into the pasture would correct for this); Georgia Driver's License test facility located directly adjacent to the right could generate nearby auto related emissions (but activity is low); and new building construction located directly adjacent to the left (diesel powered earth moving equipment and construction related architectural coatings, roofing materials, or paving processes could confound if not completed prior to study).Intangibles: existing EPD site with long monitoring history. Only location shown that significantly varies from North-South corridor.PRO'S: GREAT FETCH TO N AND W DIRECTIONSCON'S: DIFFICULT ACCESS, FENCE, CONSTRUCTION, EPD SITESUITABILITY: LESS LIKELY FOR ONLY 10 DAYS

Site 3: Columbus AirportSite Description: Current site of EPD O3 monitor. Located ~north of urban core overlooking Columbus airport and adjacent to aircraft radar tower. Elevated location with easy access.

Power requirement: Probably can be met relatively easily.Fetch: Other than open beam ~50m radar tower, excellent in all directions.Security: Open area, but highly visible. EPD's shelter has resided there for many years with no incidents. Low visitor traffic site.Potential interference: 4 lane road ~100 yards down a small hill; surrounding commercial "strip mall" type development along 4 lane road (relatively heavy vehicle traffic); adjacent airport and fuel tanks; idle auxiliary generator for aircraft radar immediately adjacent to site (possible concern over generator testing).Intangibles: existing EPD site with long monitoring history.PROS: OPEN FIELD, SUPPORTING MSMTS FROM AIRPORT OPERATIONCONS: LESS SECURE, NEARBY HC SOURCE [FUEL TANKS], CO SOURCE [GENERATOR], EPD SITESUITABILITY: LESS LIKELY FOR ONLY 10 DAYS

Site 4: Columbus WaterworksSite Description: Located far north of urban core along Chattahoochee River. Elevated location with easy access.Power requirement: Must drop new line with source moderately nearby (100 to 200 yards).Fetch: With exception of large nearby water storage tank (~30 ft tall, by 150 ft diameter), excellent in all directions.Security: Excellent - fenced area with access via code.Potential interference: Some light chemical application at water works (e.g. chlorine and alum) but otherwise none noted.Intangibles: Columbus task force recommends this site as viable "upwind" monitoring station.PROS: ACCESS, SECURITY, FETCHCONS: STATE ROAD ~1200' ARCHING 1/4 AROUND THE HILL, CL EMISSIONSSUITABILITY: YES

Site 5: Benning Hills Water TowersSite Description: Located far south of urban core near Chattahoochee River in residential neighborhood. Last elevated area before terrain becomes more flat towards south. Elevatedlocation, but with relatively poor access.Power requirement: Must drop new line, but nearby source not readily evident (unless drawing from residential power poles located around other side of large water tanks).Fetch: Very Poor (mature trees on three sides with two very large water tanks on fourth side).Security: No fence. Isolated but could be subject to neighborhood curiosities.Potential interference: none noted.Intangibles: Columbus task force offered site as possible "downwind" monitoring station.WOULD NOT CONSIDER AS OPTION!!

Site 6: Oxbow Environmental Learning CenterSite Description: Located far south of urban core on Chattahoochee River along southern end of "Riverwalk" and adjacent to Fort Benning. Area not elevated, but surrounding areais relatively flat. Access fairly easy, except main entrance closed after business hours (we gained access via employee entrance). Medium size parking lot (~100 spaces) adjacent.Power requirement: Must drop new line with nearby sources located 100 to 300 yards away.

2

Fetch: Excellent in all directions.Security: Uncertain -- "Riverwalk" remains accessible 24 hours a day (primarily used by walkers/joggers, rollerbladers, and bicyclists).Potential interference: Site located near capped and reclaimed former landfill (potential for CH4 emissions); some land application of sewer treatment sludge by waterworks department also done ~1/2 mile away.Intangibles: Environmental learning center (run by CSU) is highly conducive to affiliated education and outreach. Site located directly adjacent to Fort Benning.PROS: FARTHEST AWAY FROM EPD SITES, ACCESS, FETCHCONS: EL POWER PROBLEMATIC, CLOSE TO AL BLACK-C PLANT [POLITICALLYSENSITIVE?]SUITABILITY: YES

3

Potential FAQS Monitoring Sites in Bibb County, GA (visited 4/13/00)

Site 1: Georgia Forestry Service.Site description: Current site of EPD O3 monitor (est. 1997). Located ~east/southeast of urban core and about 1.5 miles away from largest industry in Bibb County (Brown and Williams – cigarette production). Site is adjacent to forestry vehicle “junkyard” and forestry nursery.Power requirement: Probably can be met relatively easily.Fetch: Fair – proximity to industrial park concerns locals. Trailer-style offices and single story nursery located within 200 yards.Security: Isolated site located within controlled access area.Potential interference: Possible interference from nearby industrial parks but difficult to determine. Possible use of ag related chemicals in nursery.Intangibles: Existing O3 site. With NW’ly flow, site probably captures Atlanta+Macon AQ.

Site 2: Jones Road Firestation.Site description: Located due south of urban core at county firestation. 7 meter tower adjacent to firehouse could be utilized. Access is relatively easy. If used, preference is to use the tower and housing the monitor and data acquisition instruments inside the firehouse. Not likely a good site for trailer.Power requirement: If using firehouse, power requirement can be met easily.Fetch: Firestation itself is an obstacle. Trees located within 10 yards of any potential site not in the firestation. Use of tower could overcome some fetch issues.Security: 24-hour presence of firefighting crew. Nearby residence seems harmless.Potential interference: Located at intersection of two 2-lane roads. One of the roads is in the process of being widened to 4-lanes.Intangibles: Only site shown between Macon and Robins AFB.

Site 3: Arrowhead ParkSite Description: Bibb County park located due west of urban core and on southern shore of Lake Tobeskoffee (directly across lake from Sandy Beach potential site). Isolated and elevated location, but site would require some preparation by County engineers (removal of trees and grading of “island”). Concern over high-voltage lines running nearby.Power requirement: Probably can be met relatively easily.Fetch: Elevation is good, but mature pine and hardwood trees obscure in almost every direction (within 100 yards).Security: Recreational area, but closes at sundown. With beach nearby, could see significant visitor traffic.Potential interference: High-voltage line, but none otherwise noted.Intangibles: none noted.

4

Site 4: Springdale Elementary SchoolSite Description: Located far northwest of urban core (between Atlanta and Macon) at edge of parking lot fronting the elementary school playground. Easy access when school is not in session.Power requirement: Closest source probably located in the school about 50 yards away.Fetch: Single story elementary school ~50 yards away, but other wise open in all directions.Security: Nothing special. Subject to school and local police protection.Potential interference: None noted.Intangibles: Viable "upwind" monitoring station.

Site 5: Sandy Beach ParkSite Description: Located due west of urban core and on the northern shore of Lake Tobeskoffee (directly across lake from Arrowhead Park potential site). Location is a ballfield. Access is good.Power requirement: Must drop new line (probably from nearby tennis courts).Fetch: Excellent on all sides.Security: Park is restricted to daytime use only and security is maintained by family with residence near gate. Potential site is near actively used ballfield with nothing separating.Potential interference: none noted.Intangibles: Probably only site that could accommodate the trailer.

Additional Site(s) (not visited) Robins AFB (Houston County)Time did not permit us to visit the base, but the folks at Robins have offered to host a visit May 4 if desired. Robins AFB is located due south of Macon in the next county.Intangibles: Robins is a probably a big “source” in the Macon area. Houston will likely be included in the Macon nonattainment area. Representative Larry Walker of Houston County is the Majority Leader in the Georgia House.

5

Potential FAQS Monitoring Sites in Richmond and Columbia Counties, GA (visited 4/18/00)

Site 1: Bungalow Road Elementary School.Site description: Current site of EPD O3 monitor (est. 1989). Located ~south of urban core on edge of playground. Access relatively easy.Power requirement: Probably can be met relatively easily.Fetch: Fair – proximity to two-story residential apartments about 25 yards one way, and to single story school about 25 yards the other way. Third direction brings in flow across school parking lot. Only fourth direction is substantially open across playground. Site is located within 4 miles of some of Augusta area’s heaviest industries. Security: Area is questionable. Recommend that site is fenced.Potential interference: Possible nearby industry and concerns with fetch.Intangibles: Existing O3 site.

Site 2: Fort GordonSite description: Located southwest of urban core on plot adjacent to the Parade Grounds on the Fort Gordon base. Access is relatively easy.Power requirement: Adjacent pole with transformer = power requirement can be met easily.Fetch: Parade field is huge with single row of trees lining the edges. Site is located just on other side of row of trees. Otherwise, fetch is excellent.Security: Fort Gordon requires permission to enter base, but with several thousand employees and residents, access to the base is relatively open. Site is highly visible with Inspector General’s office located nearby. Potential interference: Power station and “peak-shaver” generators located within 2 miles.Intangibles: Fort Gordon is the area’s 2nd largest employer (SRS is #1) and geographically is comparable in size to Augusta proper. Gordon also offered more remote site (nature center) further southwest.

Site 3: Columbia County FairgroundsSite Description: Located due west of main Columbia County population center and Augusta urban core at the Fairgrounds. Fair is held in March with little other activity the rest of the year. Access is excellent. Site is located between Atlanta and Augusta. Only possible concern is that it is fairly remote from Augusta itself.Power requirement: Excellent. Site is already wired for 200 amps to a panel (used for carnival rides).Fetch: Excellent in all directions.Security: We got in, but I believe area is usually gated. Site is particularly visible. Potential Interference: Softball complex located nearby may generate some traffic, but probably not worth noting.Intangibles: Site is between Atlanta and Augusta, but with E’ly flow prevailing, could actually be downwind site.

6

Site 4: Lakeside High School and Middle SchoolSite Description: Located in residential area of Columbia County and west of Augusta urban core (between Atlanta and Augusta proper) at edge of parking lot fronting the HS and MS baseball fields. Easy access when school is not in session.Power requirement: Underground utilities – “green box” about 20 yards from potential site.Fetch: Area slopes down toward ballfields. High School is located at highest point, which then slopes down to potential site, which then slopes down to the baseball fields. School buildings are all single story.Security: Nothing special. Subject to school and local police protection.Potential interference: None noted.Intangibles: Near population center of Columbia County – a bedroom community of Augusta. Uncertain if Columbia County will be included in Augusta nonattainment area.

Sites 5 & 6: Columbia County Park (name not noted) and Georgia Welcome CenterNeither site appeared suitable due to topography and proximity to I-20 respectively.

7

J. MATERIALS USED IN THE 1999 GEORGIA EMISSION INVENTORY POINT SOURCE SURVEY

October 9, 2000

To the Plant Manager(s):

The Georgia Environmental Protection Division (GaEPD) requests your assistance in the development of a 1999 Emission Inventory for the Augusta, Columbus and Macon and surrounding areas. Per Georgia Rules for Air Quality Chapter 391-3-1-.02(6)(b)1(i), please complete the questionnaire and mail it to Drs. Ted Russell and Michael Baker at the address shown below by November 30, 2000. Information reported should represent 1999 operational and emissions data. You may refer to record keeping or other monitoring data for process information.

The EPD will hold three workshops for facilities needing assistance in completing the questionnaire. The workshops will be held sometime during the week of October 30-November 3. A notification of the exact date, time, and location of the workshops for each city will be included on the website http://www.cure.gatech.edu as soon as this is determined. You can also call the following contacts, if needed: for Augusta, Eliot Price, (706)-737-1415; for Columbus, John Mills, (706)-649-1661; for Macon, George Lee, (912)-751-6190. At these workshops, staff will provide instruction on each form in the questionnaire and will answer any questions. Please R.S.V.P. to Scott Southwick (contact listed below) if you are able to make it to one of the workshops so we can gauge the number of people who may attend.

The information collected will be used in a joint research project between the Georgia Institute of Technology school of Civil and Environmental Engineering and the GaEPD. This study, the Fall Line Air Quality Study (FAQS), focuses on emissions from three cities in Georgia: Columbus, Macon, and Augusta, and their effects on the entire state. The emissions information that you provide will be catalogued by the GaEPD and the Federal Environmental Protection Agency (EPA) and become part of their standard inventory database. In addition, it will provide vital input to our study in order to assess the impact of these emissions on the state of Georgia and to find cost-effective means by which to improve air quality.

Contained in the following package are: (1) Brief instructions on filling out the survey, (2) a Plant General Information page, (3) a page for the inclusion of additional information you may have about chlorine, ammonia, and particulate matter, and (4) a spreadsheet containing the survey itself. Within the survey are 6 sheets with information for fuel burning processes, evaporative emissions, miscellaneous processes, stack information, control device information, and finally an emission factor calculation worksheet. Please fill these forms out as completely as possible. A zip file containing these forms can also be downloaded from http://www.cure.gatech.edu (zipped file is in WinZip format; unzipped files are in Microsoft Word 97 and Excel 97 format). If you have any questions feel free to email [email protected] , call the contacts listed above for Augusta, Columbus, or Macon, or contact the following people directly:


http://www.cure.gatech.edu/

http://www.cure.gatech.edu/

Scott SouthwickEnvironmental Engineer, Planning and Support Program

Georgia Environmental Protection Division, Air Protection Branch4244 International Pkwy., Suite 120

Atlanta, GA 30354Phone: (404)-362-4569FAX: (404)-363-7100

Internet Address: [email protected]

Dr. Ted Russell and Dr. Michael BakerSchool of Civil and Environmental Engineering

The Georgia Institute of TechnologyAtlanta, GA 30332

Phone: (404)-894-3760FAX: (404)-894-8266

Sincerely, Sincerely,Ted Russell, Ph.D. Scott SouthwickProfessor, Environmental EngineerGeorgia Institute of Technology GaEPD

2

Instructions for filling out inventory sheets (Feel free to make additional copies as needed)Please return by NOVEMBER 30, 2000

PLANT GENERAL INFORMATION:(1) Is there ANY information contained in this report confidential in accordance with the Freedom of Information Act and the Public Control Act? If answer is YES, please follow confidentiality procedures found at “other info & forms” at: http://uam.air.dnr.state.ga.us/sspp/titlev/permit.html(2) Self explanatory.(3) Self explanatory.(4) Provide information for the specific location of the plant.(5) Provide information for the Emissions Contact at your facility. We will call this person if there are questions about the contents of this report. Future Emission Inventory Questionnaires and other correspondence related to Emission Inventories will be sent to this person. If there is an extension, please provide that number along with the phone number. Provide your Internet Address if you prefer correspondence through the Internet.(6) SIC (Standard Industrial Classification) codes are descriptive codes for facilities. The primary SIC code should be found on your Title V application. If you do not know your SIC code, leave it blank and we will assign the appropriate code.(7) NAICS (North American Industrial Classification System) code. This is the new industry coding system that will replace the existing SIC coding system. It has been developed because of the North American Free Trade Agreement. It has been left blank and is optional. We intend to convert existing SICs to NAICSs in the future.(8) What is the principal product manufactured at your facility?(9) Please enter the latitude and longitude of your facility OR the UTM coordinates. Providing both is not necessary.(10) Self explanatory.(11) If actual plant emissions are below 25 tpy for all of VOC, NOx, CO, SO2, PM, Cl, and NH3, then the rest of the form does not have to be filled out. Simply check YES, sign the PLANT GENERAL INFORMATION form and return.(12) A signature from an authorized plant official is REQUIRED.

FUEL BURNING PROCESS INFORMATION:This page should list each significant fuel burning emission unit at your plant. The type of fuel and the size of the fuel burning equipment determine how significant a source is. Natural gas fired units smaller than 10 mmBTU/hr (per unit) have low emissions and their combined fuel use can be reported on one line.

Process ID Unit: The unit ID refers to the boiler, generator, dryer, etc. Create a number for each, starting with “F” (e.g., “F1”) to differentiate from the other process units.Process ID Fuel: Number each fuel at a given unit separately. Start fuel numbers over at 1 for each unit.Description: Unit descriptor, as identified by facility personnel. Fuel is self-explanatorySCC: List the 8 digit SCC code of the process in which the unit is running.% Sulfur: List the percent sulfur by weight; required for coal and fuel oil.% Ash: List the percent ash by weight; required for coal only.Operating rate: Maximum is the manufacturer’s rated maximum rate of operation. For Normal, enter the average operating rate during the year of inventory. Units should be million BTU per hour.Normal Operating Schedule: Average operating schedule per day, per week, per year.Seasonal Throughput: Percentage of production for each quarter of the year. The total MUST equal 100%. For winter throughput, use December of 1999, and January and February of 2000.Process Throughput: Annual and Ozone season daily fuel feed. Ozone season includes the months of May through September. Daily summer is typical ozone season day. List the appropriate units [note: typical units are Millions of Cubic feet for Natural Gas, 1000s of Gallons for Fuel Oils (including LPG), and Pounds for Solid Fuels (including coal)].

3

EVAPORATIVE EMISSIONS PROCESS INFORMATION:This page is for reporting VOC (volatile organic compounds) emissions resulting from evaporation. To report VOC use, it must be derived from material used and its VOC content. An MSDS (Material Safety Data Sheet) will provide some information to determine the pounds of VOC per gallon or % makeup. After the amount of VOC containing material is established, SUBTRACT any amounts Shipped Out in Product and Contained in Waste Stream. Any remaining material is assumed to have been emitted into the air unless a control device is in place. This is the Uncontrolled VOC. If a control device is in place, MULTIPLY the resulting Uncontrolled VOC result by (1-Control Eff’cy) and report this as the Uncontrolled VOC. Report in units of Pounds of VOC.Process ID: The Process ID refers to the coating booth, batch chemical process, etc. Create a number for each starting with “E” (e.g., “E1”) to differentiate from other process units.SCC: List the 8 digit SCC code of the process in which the unit is running.Raw Material: Name of material.Operating Rate: Maximum is the manufacturer’s rated maximum rate of operation. For Normal, enter the average operating rate during the year of inventory. Units should be POUNDS of material applied per hour.Normal Operating Schedule: Average operating schedule per day, per week, per year.Seasonal Throughput: Percentage of production for each quarter of the year. Total MUST equal 100%. For winter throughput, use January, February, and December of 1999.Material Balance:

Gross VOC Inputs: For each material, multiply pounds of chemical/coating/adhesive/solvent used by its VOC content by weight. Results are VOC use in pounds.Residual VOC Shipped in Product: For each input, determine the ratio of starting material left in the product and its VOC content by weight. For paint manufacturers, chemical manufacturing, or facilities that ship dye, adhesive, or VOC-containing materials, this blank is critical to determining emissions.VOCs Contained in Waste Stream: For solvent cleanup or other processes that dispose of VOC-containing materials, use this blank to list the pounds of VOC disposed through wastewater or shipped offsite for treatment.

MISCELLANEOUS PROCESS INFORMATION:This is for reporting any miscellaneous emissions (VOC, NOx, CO, PM, NH3, Cl, or SO2) that result from neither fuel burning or evaporative loss. THERE IS NO NEED TO FILL THIS PAGE OUT IF ALL EMISSIONS FROM A GIVEN EMISSIONS UNIT ARE EITHER FUEL OR EVAPORATIVE LOSS RELATED.

Process ID: The unit ID refers to that process which is generating miscellaneous emissions. Create a number for each, starting with “M” for Miscellaneous (e.g., “M1”) to differentiate from other processes.SCC: List the SCC of the process in which the unit is running.Description: Unit descriptor, as identified by facility personnel. Also, list each product produced by the process.Operating Rate: Maximum is the manufacturer’s rated maximum rate of operation. For Normal, enter the average operating rate during the year of inventory. List in terms of pounds of throughput per hour.Seasonal Throughput: Percentage of production from each quarter. The total MUST equal 100%. For winter throughput, use January, February, and December of 1999.Process Throughput: Annual and Ozone season daily fuel feed. Ozone season includes the months of May through September. Daily summer is typical ozone season day. Units should be Pounds of product.

STACK INFORMATION:Stack information should list each stack that exhausts a process listed on one of the process information pages.

Stack ID: Assign a unique number to each stack. This can be the facility designation for the stack, not to exceed 3 characters.Process ID(s) Exhausted: Each process from any of the Process Information pages exhausted by this stack should be listed here. If one process exhausts to multiple stacks, list each stack separately with the common process ID indicated, and show the percentage of exhaust flow routed to each stack in the “multiple stacks” column.Description: Facility designation or other description of the stack.Height: Stack height above grade, in feet.Diameter: Interior stack diameter, in feet.Velocity: Exit gas velocity, linear feet per second.

4

Temperature: Exit gas temperature, degrees Fahrenheit.Flow Rate: Flow rate in actual cubic feet per minute (ACFM) under maximum emissions conditions.

CONTROL DEVICE INFORMATION:Use this page to describe equipment that VOC, NOx, CO, PM, NH3, CL, or SO2 emissions. The device must be associated with the process unit it controls. If only one device is present, list it under Primary Control Device.

Process ID: List the process ID from one of the Process Information pages controlled by this device.Pollutant: List whichever pollutant is controlled by the device.Control Efficiency: Two factors contribute to the overall efficiency of a control device: Capture and Destruction. List separate efficiencies for primary and secondary devices, if present.Capture Efficiency: Describes the percentage of process exhaust that enters the control device. In a closed system, this would be 100% (1.00 as decimal percent). For a hood or other enclosure, less than 100% of the emissions of a process enter a control device.Destruction Efficiency: Equivalent to destruction efficiency in a thermal oxidizer, this factor describes the percent of pollutant entering the device that is removed by the device.

EMISSION FACTOR CALCULATION WORKSHEET:The Emission Factor Calculation Worksheet is provided for facilities that would like to supply their own emissions calculations. Facilities with source testing, continuous emissions monitoring (CEM), or other source-specific emissions data may want to provide this data to ensure the best emissions calculations. Duplicate the worksheet as needed to list all emissions calculations.

Process ID: List the process ID from the appropriate Process Information page.Source Classification Code: If available, list the EPA source classification code describing the process. This may have been reported in a permit application.Process Throughput: Taken from the appropriate Process Information page.Emission Factor: The pollutant-specific emission factor for this process. Each factor should include a method code as a reference for the factor. Method codes are: 1=Continuous Emission Monitor, 2=Material Balance, 3=Source Test, 4=AP-42 emission factor, 5=Other.Control efficiency: Control efficiency here is the overall efficiency, which is the product of the capture and destruction efficiencies listed on the control information page. Multiply the efficiencies as decimals to get the overall percentage (as a decimal).Total Emissions: Multiply annual throughput, emission factor, and (1-control efficiency) together to get total annual emission, i.e., (annual throughput) x (emission factor) x (1-control efficiency).

5

Directions: Please complete the information as needed.

PLANT GENERAL INFORMATION:

FIPS State/County/Plant ID:

(1) Is there ANY information contained in this report confidential in accordance with the Freedom of Information Act and the Public Control Act? YES: NO: If answer is YES, please follow confidentiality procedures found at “other info & forms” at: http://uam.air.dnr.state.ga.us/sspp/titlev/permit.html

(2) Plant Name:

(3) Permit Number(s):

(4) Plant Location:

City: Zip Code: County:

(5) Emission Inventory Contact Person:Contact Name: Phone # + ext.: Fax #:

Mailing Address: Mailing City: Mailing State: Mailing Zip Code:

Internet Address:

(6) Primary/Secondary SIC Codes:

(7) NAICS Code:

(8) Principal Product:

(9) Latitude/Longitude for Plant coordinates:

Latitude (DDMMSS):

Longitude (DDMMSS):

(10) Is the Plant a portable facility (e.g., asphalt plants, portable concrete plants, soil remediation units, or portable diesel generators)? YES: NO: If YES, in what county were the emissions generated:

(11) Are actual emissions below 25 tons per year (tpy) all of the following: VOC, NOx, CO, PM2.5, Chlorine, and Ammonia? YES: NO: If YES, then the rest of the form does not have to be filled out. Simply check YES, sign the PLANT GENERAL INFORMATION form and return.

(12)When the entire package has been completed, please sign and date below. By your signature you are declaring that the information is complete and accurate.

Name: ____________________________________ Title:___________________________ Date:____________

6

ADDITIONAL INFORMATION

This survey includes sections dealing with chlorine, ammonia, and particulate matter (PM) emissions. Chlorine may impact the ozone formation rate, while ammonia and particulate matter emissions are needed in order to better study aerosol and PM2.5 (particulate matter less than 2.5 microns). Particulate matter is of special importance to this study, and PM2.5 may become the new federal standard affecting our area. Any additional knowledge about the size and chemical composition of emitted particulate matter, as well as the amount of chlorine and/or ammonia emissions that you have, would be of help. Please describe these below. If you have any further specific questions as to what additional information would be useful, feel free to email [email protected].

7

K. THE FALL LINE AIR QUALITY STUDY IN THE NEWS”

Newspaper

8/11/99* “Columbus might do regional air study;” Niesse, M.; Columbus Ledger-Enquirer.

8/12/99* “Bibb’s air standard escape only temporary;” Editorial Board (Corson, Ed); Macon Telegraph.

8/12/99* “Political boundaries don’t halt pollution;” Editorial Board; Columbus Ledger-Enquirer.

8/18/99* “Cities seek air studies;” Pavey, R.; Augusta Chronicle.

8/20/99* “Clean air study needed;” Editorial Staff;" Augusta Chronicle.

8/27/99* "Augusta will be named a polluted-air city;" Pavey, R.; Augusta Chronicle.

10/8/99* "Pollution may limit Augusta;" Pavey, R.; Augusta Chronicle.

10/29/99* “Air quality a large legislative concern;” Houston, J.; Columbus Ledger-Enquirer.

2/10/00* “Tech could aid Macon’s air problem;” Badertscher, N.; Macon Telegraph.

2/15/00* “Bad air target of state study;” Houston, J.; Columbus Ledger-Enquirer.

4/12/00* “EPD official says Macon’s air not always safe;” Schwarzen, C.; Macon Telegraph.

4/14/00* “Monitors going up for ozone study;” Schwarzen, C.; Macon Telegraph.

4/26/00* “Georgia to Name Counties That Must Help Macon Clean Its Air;” Schwarzen, C.; Macon Telegraph.

4/27/00* “Macon enters prime ozone season, EPD to track levels;” Schwarzen, C.; Macon Telegraph.

5/5/00* “Atlanta smog rivals L.A.’s;” Seabrook, C.; Atlanta Constitution.

5/5/00* “Macon may fail air standard in addition to ozone;” Schwarzen, C.; Macon Telegraph.

5/5/00* “Atlanta, L.A. close on bad air Pollution figures show conditions worsening;” Williams, D.; Florida Times-Union (Jacksonville, FL).

5/6/00* “Georgia Tech Research: Atlanta’s smog could be deadly;” Seabrook, C.; Atlanta Journal & Constitution.

5/10/00* “Augusta pollution rivals Pasadena;” Williams, D; Augusta Chronicle.

6/15/00* “Bad air tag will cost Bibb, Houston;” Schwarzen, C.; Macon Telegraph.

6/16/00* “Houston vows to help tackle ozone problem;” Schwarzen, C.; Macon Telegraph.

6/22/00* “Experts to study local air;” Pavey, R.; Augusta Chronicle.

6/23/00* “Congress gagging EPA on air quality;” Editorial Board; Atlanta Journal-Constitution.

6/24/00* “Clearing the air;” Editorial Staff; Augusta Chronicle.

7/1/00* “Bibb on ozone failure list; Houston off;” Schwarzen, C.; Macon Telegraph.

7/4/00* “Suburbs dodge pollution stigma;” Pavey, R.; Augusta Chronicle.

7/8/00* “Bibb County dodges ozone list, for now;” Schwarzen, C.; Macon Telegraph.

7/9/00* “Taking on Toxins: Air, water pollution plague area;” Pavey, R.; Augusta Chronicle.

7/9/00* “Researchers analyze quality, contaminants of Augusta’s air;” Pavey, R.; Augusta Chronicle.

7/9/00* “Scientists looking at air pollution in Augusta, Macon, Columbus;” Associated Press.

7/13/00* “Bibb ozone level again above EPA standards;” Schwarzen, C.; Macon Telegraph.

7/22/00* “Big-city pollution migrating south;” Seabrook, C.; Atlanta Journal-Constitution.

7/22/00* “Augusta, Macon, Columbus wary of being classed as air polluters;” Associated Press.

7/23/00 "Air Quality puts Columbus on Hot Seat;" Wehunt, W. C.; Columbus Ledger-Enquirer.

7/24/00* “Manager of air quality study offers observations;” Wehunt, W. C.; Columbus Ledger-Enquirer.

8/18/00* "DOT board wants later start for schools;" Badertscher, N. and C. Schwarzen; Macon Telegraph.

8/20/00* "Wetlands offer answers;" Pavey, R.; Augusta Chronicle.

8/28/00* "Chamber group lists area needs;" Cline, D.; Augusta Chronicle.

8/30/00* "Tougher air controls not likely;" Williams, D.; Augusta Chronicle.

9/1/00* "Study tracks pollution patterns;" Pavey, R.; Augusta Chronicle.

9/1/00* "Smog invades the quality of Macon's air;" Schwarzen, C.; Macon Telegraph.

9/1/00* "Georgia Tech seeks more funds for air quality studies across state;" Schwarzen, C.; Macon Telegraph.

10/22/00* "Macon seems to be making its own smog;" Schwarzen, C.; Macon Telegraph.

2

Television

6/15/00 WMAZ (channel 13) in Macon.

6/15/00 WGXA (channel 24) in Macon.

6/15/00 WPGA (channel 58) in Macon.

6/26/00* WAGT (channel 26) in Augusta.

7/6/00* WAGT (channel 26) in Augusta.

7/6/00 WJBF (channel 6) in Augusta.

7/10/00 WTVM (channel 9) in Columbus.


7/20/00 WRBL (channel 3) in Columbus.

7/20/00 WXTX (channel 54) in Columbus.

7/25/00 WRBL (channel 3) in Columbus.


Radio

7/10/00 Lead local (GA) news story on Peach State Public Radio (NPR).

3

L. PEER REVIEW PANEL’S CRITICAL REVIEW OF FAQS PHASE I

Comments and Recommendations for the Future of the Fall Line Air Quality Study (FAQS)

This report was prepared by an External Review Panel for FAQS consisting of:Ellis Cowling, North Carolina State University

Rafael Ballagas, Georgia Environmental Protection DivisionJohn Jansen, Southern Company

This report was prepared after the FAQS Science and Policy Workshops which were held at

Georgia Tech's School of Earth and Atmospheric Sciences, and in theGeorgia Regional Transportation Authority Offices

on October 19-20, 2000

Prior to our arrival at the Science Workshop, the Panel received a series of documents including:

1) A detailed Agenda for the FAQS Science Workshop and the following FAQS Policy Workshop

2) An Overview of FAQS describing:

A) The general objectives of FAQS:"A scientific assessment of urban and regional air pollution, identifying the sources of pollutants and pollutant precursors, and recommending solutions to realized and potential poor air quality in the Augusta, Macon, and Columbus, Georgia metropolitan areas"

B) The total budget for FAQS -- $3 million provided by the Georgia General Assembly and $1 million provided by Georgia Tech (SCISSAP mobile

laboratory)

C) The time period of FAQS -- January 1, 2000 -- December 31, 2002with four distinct study periods:

Period 1 -- January 2000-June 2000 -- Organization and Initial Setup PhasePeriod 2 -- July 2000-June 2001 -- Initial Field Study, Model Development, Design of Field Study #2Period 3 -- July 2001-June 2002 -- Field Study #2, Monitoring Network Design, Training, Tech TransferPeriod 4 -- July 2002-December 2002 -- Technology Transfer, Recommendations for Action

D) The organizational structure of FAQS which is patterned after that of the Southern Oxidants Study.

E) The following statements of the expected Legacy of FAQS:"At the conclusion of FAQS, Augusta, Macon, and Columbus will be provided with sufficient information to begin making the difficult decisions necessary to meet or maintain state and federal air quality standards. Where poor air quality is identified by FAQS to result from [regional] factors beyond the political jurisdiction of these metropolitan areas, the EPD [Georgia Environmental Protection Division] or the USEPA will be provided with information necessary to address these externalities. Further, as political, environmental, or economic conditions change in the future, each community will have sufficient skills and computational tools to evaluate and modify alternate action plans. Finally, each community will possess a basic monitoring system capable of assessing regional and local air quality conditions , and tracking progress in meeting and maintaining air quality standards."

3) A printed "Summary of Year to Date Accomplishments and FY 01-FY02 Budget Justification"

Introduction

During the FAQS Science Workshop at Georgia Tech, the External Review Panel mainly listened to (but also asked pertinent questions about) reports by various members of FAQS Science Team. The topic covered included:

-- Overview of the Phase I Pilot Studies in and around Augusta, Macon, and Columbus-- Results of:

-- Gas phase and meteorology measurements-- VOC measurements-- PM mass and optical properties measurements-- PM composition measurements

-- Emissions inventory update-- Initial synthesis and integration of Phase I results including comparison of

measurements made in and around Augusta, Macon, and Columbus during the summer of 2000

-- FAQS Phase II planning

Following these FAQS Science and Policy Workshops, the three Panel Members prepared and exchanged our individual impressions, comments, and recommendations. On the basis of these written statements and verbal and E-mail communications, we offer the attached General Comments about Phase I Results and Recommendations for the Future of FAQS. These comments are offered in the hope that they will help ensure the success of the FAQS research program in fulfilling the expectations outlined in the FAQS Legacy (see above).

2

GENERAL COMMENTS ABOUT PHASE I RESULTS ANDRECOMMENDATIONS FOR THE FUTURE OF FAQS

General Comment 1) The FAQS Science Team includes a very high quality group of investigators.

The FAQS Chemical and Meteorological Measurements Work Group is led by Professor William Chameides of Georgia Tech's School of Earth and Atmospheric Sciences. He coordinates the pollutant monitoring and diagnostic analysis components of FAQS in connection with his leadership of the Southern Center for Integrated Study of Secondary Air Pollutants (SCISSAP). Professor Karsten Baumann provides leadership for the chemical and meteorological field measurements using the facilities of the Georgia Tech Mobile Laboratory and other chemical analytical facilities within the Georgia Tech School of Earth and Atmospheric Sciences.

The FAQS Emissions, Models, and Effects Work Group is led by Professor Armistead (Ted) Russell, of Georgia Tech's Air Resources Engineering Center (AREC). He and his graduate student colleagues are responsible for development of the FAQS emissions inventories and relevant air quality models which will be used to execute the simulations of various air quality scenarios in cooperation with personnel from Augusta, Macon, and Columbus.

Overall project management for FAQS is provided by Dr. Michael Chang of Georgia Tech's Center for Urban and Regional Ecology (CURE) with general guidance provided by professor C. S. Kiang who serves as Senior Advisor for FAQS.

General Comment 2) A vary good initial effort has been made by the FAQS Science Team to get a "good intellectual grip" on the air quality problems of all three cities with the limited financial and human resources that are available.

General Comment 3) The three sponsoring municipalities, the Georgia state legislature, and the FAQS Science team all are to be commended for "getting out in front" of the air quality challenges of these three Near-Non--Attainment Areas within the state of Georgia. Having even a limited amount of measurement data and information that is carefully assessed, is certainly better than no data and only continuing uninformed speculation about the air-quality status and future of the three FAQS cities.

We caution, however, that these same three groups of stakeholders should not have too high expectations for the general outcome of FAQS. The data so far accumulated are very limited -- only 2-3 weeks in each metropolitan area within one particular year -- and analysis and interpretation have only just begun. The high goals of FAQS will be achieved only after very careful analysis and interpretation of these initial observations in the light of what is known from previous research both within and outside of Georgia. These "initial observations" are a good foundation for development of a "climatology" of observations in and around these three cities. Such a climatology will need to be constructed before firm conclusions can be drawn with reliability. In essence, more observations in Field Season #2 and beyond will be needed in order to firm up, improve, and probably revise these initial insights over time.

3

General Comment 4) The most intensive field measurements for FAQS will begin in earnest during Field Season #2. Additional measurement sites to be operated under the guidance of the Georgia Tech Science Team are already in place. That is very good.

General Comment 5) The FAQS Website [http://www.cure.gatech.edu/faqs.asp] established by Director Michael Chang and maintained in cooperation with his several colleagues on the FAQS Science Team provides a publicly accessible record of:

1) The origins of FAQS, the original FAQS Proposal, the FAQS Memorandum of Understanding, 2) The qualifications of the FAQS Science Team, 3) Minutes of the several meetings of the FAQS Coordinating Council, 4) FAQS Outreach and Extension efforts, and 5) Progress achieved during the Phase I Science and Policy Workshop on October 19-20, 2000.

In this connection, we were pleased to see again, some of the 68 graphic displays on the FAQS Website which included much of the original data which were shared with us and among member of the FAQS Science Team during the FAQS Science Workshop meetings on October 19 and 20. These detailed graphic displays and verbal expositions included the following topics: 1) review of Georgia's Air Quality "Climate," 2) Regional Relationships, 3) Local Relationships, 4) Other Chemical Data, 5) FAQS Phase I Questions, 6) "Eyeball" Analysis of VOC Data, 7) Fine particulate Matter -- Concentrations, Composition, and Effects, 8) Relationships between Ozone and PM, and 9) Forming Hypotheses about "What it all means."

A preliminary assessment of results from the Pilot Study in Field Season #1 (2000) also was posted on the FAQS Website on October 24, 2000 -- a few days after the FAQS Science and Policy Workshops on October 19-20, 2000:

"The key findings from the pilot study were: 1) local areas may simultaneously contribute to and be affected by high concentrations of ozone across the southeastern region -- this is most evident in Augusta and Columbus; 2) in Macon, like Atlanta, the underlying regional effects sometimes may be dwarfed by local effects; 3) high ozone concentrations in each of the three Fall Line cities appear to be associated most frequently with light or stagnant winds; 4) total hydrocarbon concentrations and isoprene fractions observed in the three cities were similar in relative magnitude to those observed in previous studies of Atlanta suggesting controls on NOx emissions may be needed to reduce ozone concentrations; and 5) fine particulate matter was observed to be composed largely of organic carbon and sulfates."

In this connection, we believe it is important to repeat the caution offered in General Comment 3 (above). We believe that the verbal interpretations associated with the graphic display of data presented at the FAQS Science and Policy Workshops and on the FAQS Website constitute what we suggest be called "Initial Impressions" rather than "Key Findings." In our opinion, the limited observations so far made in the Pilot Study in Phase I of the FAQS Study (2000) should be regarded mainly as "Hypotheses which have not yet been fully tested." In essence we believe the term "Key Findings" should be reserved for analyses and interpretations made after Field Study #2 has been completed and that the term "Initial Impressions" is more appropriate to the present verbal interpretations of "What it all means."

4

We appreciate the completeness and timeliness of many of the publicly accessible FAQS Website records. The originating documents, qualifications of team members, minutes of meetings, etc. help satisfy the need for "public education" and are consistent with the idea of the "public's right to know." At the same time, we emphasize the need to ensure that high standards of quality are maintained both in the FAQS data (numbers) and the FAQS information (written statements) posted on the FAQS Website. As indicated in the immediately preceding paragraph, we believe special care should be taken, especially when posting:

-- FAQS field measurement data,-- Iinitial impressions from preliminary analyses of FAQS data and information, and

especially when-- Major conclusions and (later) policy implications of observations are formulated for

posting on the FAQS Website.

Recommendation 1) If FAQS is to be maximally productive, full funding for the remaining three phases of the FAQS study plan should be secured and provided to the FAQS Science Team in a timely way. Also, care should be taken to allow adequate time for thorough analyses and interpretations of the data collected in Field Season #1 before Field Season #2 is initiated.

Recommendation 2) All data collected at the present measurements sites and additional measurement sites as discussed in Recommendations 4 and 5 below, should follow carefully determined quality control and quality assurance procedures. These quality control and quality assurance procedures should be identified in a written QA plan in order to maintain data integrity to the highest degree possible.

Recommendation 3) We recommend that the Georgia EPD be invited to participate in these QA/QC processes to provide oversight of data collection processes. In this way, EPD will have the opportunity to review the measurement and QA/QC processes as they develop.

Recommendation 4) Regional monitoring should be considered as important as the current urban-focused (upwind-downwind) design for Field Season #2. The proposed study of ozone transport between the three Fall Line cities as well as transport from the Atlanta metropolitan area should be realized. Additional sampling stations should be established at carefully selected locations between Atlanta and Macon, between Macon and Columbus, and between Macon and Augusta. Ideally, these additional stations should include continuous monitoring of ozone, CO, SO2, NOx, NOy, PM2.5 mass, basic meteorology, and canister sampling for VOCs.

Recommendation 5) In addition to the regionally focused monitoring sites discussed in Recommendation 4 (above), we suggest that consideration should be given to outfitting the Georgia EPD sampling stations in most parts of Georgia, with the same type of continuous monitoring equipment (including instrumentation for measurements of ozone, CO, SO2, NOx, NOy, PM2.5 mass and chemical speciation, basic meteorology, and canister sampling for VOCs). This would provide information on the effectiveness of each existing and new sampling station in determining three important features: 1) The age of the air masses passing these established sampling stations,2) The effectiveness of each sampling station in explaining peak ozone concentrations, and 3) Identifying regions of Georgia that are tending toward "NOx sensitivity," "VOC sensitivity,"

or the "transition" between these two air-quality management regimes.

5

Recommendation 6) While FAQS should be seeking to understand what is happening at both peak and low (near background) ozone and PM concentrations, we also encourage the FAQS Science Team and its various stakeholders to recognize that the question of attainment vs. non-attainment in the three Fall Line cities is likely to have appreciably different dimensions and significance for air-quality management in the state or Georgia, if the new ozone standard is adopted (0.08 ppmv averaged over 8 hours) compared to the old ozone standard (0.120 ppmv averaged over 1 hour).

Recommendation 7) Additional funding should be sought for investment in FAQS for the long term. These additional resources will be especially useful for three interrelated purposes: 1) Additional field measurements, 2) Additional analysis, interpretation, and assessment activities, and 3) Additional training and technology-transfer functions as outlined in the FAQS study plan and

Legacy statements.

Recommendation 8) The External Review Panel is aware that a more detailed Progress Report for the Pilot Study completed in Field Season # 1 will soon be completed for consideration and response by the FAQS Coordinating Council. All of us on the External Review Panel would welcome the opportunity to review and comment on this Progress Report whenever it s completed. We suggest that this be done both before the Progress Report is submitted to the Coordinating Council and perhaps also after the Coordinating Council has offered their comments and reactions to the Progress Report.

In Conclusion:

We hope that the six General Comments and eight Recommendations offered above will be useful to the FAQS Science Team, the FAQS Coordinating Council, officials in the three Fall Line cities (Augusta, Macon, and Columbia), and other stakeholders in the management of air quality in the state of Georgia -- as they continue to work together to make further progress toward fulfilling the Legacy of FAQS.

If we can be of further help in these matters, please let us know. Please also let us know if we have misunderstood or misinterpreted any aspects of the purposes of FAQS and/or the significant progress that has been achieved by FAQS to date.

6

the fall line air quality study (faqs) - georgia institute...

Documents