Ambient Air Quality Monitoring and Health Research: Summary of April 16-17, 2008 Workshop to Discuss Key Issues U.S. ENVIRONMENTAL PROTECTION AGENCY Office of Research and Development Office of Air and Radiation/Office of Air Quality Planning and Standards DECEMBER 2008 ------- [This page intentionally left blank.] December 2008 ------- EPA452/S-08-001 December 2008 Ambient Air Quality Monitoring and Health Research: Summary of April 16-17, 2008 Workshop to Discuss Key Issues U.S. Environmental Protection Agency Office of Research and Development and Air Quality and Assessment Division Health and Environmental Impacts Division Office of Air Quality Planning and Standards Office of Air and Radiation Research Triangle Park, North Carolina December 2008 ------- DISCLAIMER This workshop summary report serves to document recommendations presented at a recent workshop held by the U.S. Environmental Protection Agency's Office of Research and Development (ORD) and the Office of Air Quality Planning and Standards (OAQPS) within the Office of Air and Radiation (OAR). The recommendations described in this report have been modified to reflect information developed during the review of this document and may be modified in subsequent discussions with internal and external experts. ORD and OAR will use the recommendations presented in this document as a tool to consider and prioritize both short- and long-term actions for EPA and others to undertake in the development and implementation of ambient air monitoring and health research strategies. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. December 2008 ------- TABLE OF CONTENTS LIST OF ACRONYMS Hi AUTHORS AND CONTRIBUTORS v INTRODUCTION. 1 Background 2 Workshop Structure, Objectives, and Outline of Session Summaries 2 Initial Successes and Next Steps 3 SESSION I: ELEMENTAL AND ORGANIC CARBON MEASUREMENTS 5 Background/Objectives 5 Session Overview 5 Major Points Raised by Participants 6 Recommendations/Actions for Consideration 9 SESSION II: ACCESSING AMBIENT AIR MONITORING DATA 11 Background/Objectives 11 Session Overview 11 Major Points Raised by Participants 12 Recommendations/Action Items for Consideration 14 SESSION III: AMBIENT AIR MONITORING FOR HEALTH RESEARCH 17 Background/Objectives 17 Session Overview 17 Major Points Raised by Participants 17 Recommendations 19 Action Items for Consideration 20 SESSION IV: THORACIC COARSE PARTICLE COMPONENTS AND POTENTIAL HEALTH IMPACTS 22 Background/Objectives 22 Session Overview 22 Major Points Raised by Participants 24 Recommendations/Actions for Consideration 26 SESSION V: AMBIENT AIR MONITORING REALITIES - EPA/STATE/LOCAL PERSPECTIVES-SUMMARY AND RECOMMENDATIONS. 30 Background/Objectives 30 Session Overview 30 Major Points Raised by Participants 30 December 2008 i ------- Recommendations/Actions for Consideration 35 BACKGROUND MATERIALS 39 Appendix A: Workshop Agenda and Participant List A-l Appendix B: Session I: Elemental and Organic Carbon Measurements - Chemical Speciation Network (CSN) Carbon Issues B-l Appendix C: Session II: Accessing Ambient Air Monitoring Data - Access to EPA's Air Quality Data for Health Researchers C-l Appendix C.I - Other Data Access Mechanisms C-10 Appendix D: Session III: Ambient Air Monitoring for Health Research - Air Quality Sampling: Benefits and Costs of Daily Health Targeted Monitors for Fine Particle Components D-l Appendix E: Session V: Ambient Air Monitoring Realities - EPA/State/Local Perspectives - Ambient Air Monitoring Network: Network Design and Site Selection Approval E-l Appendix F: Session V: Ambient Air Monitoring Realities - EPA/State/Local Perspectives - Ambient Air Monitoring Method Implementation F-l Appendix G: Preliminary Survey of Ambient Air Monitoring Sites Currently Being Considered in EPA-funded Epidemiology Studies Feb 2008 G-l December 2008 ii ------- LIST OF ACRONYMS The following acronyms have been used for the sake of brevity in this document: AMTIC AQS ARMs AQI BC BOSC CAA CAAAC CASAC CMAQ CSN DASH DEARS DRI DRUM EC EPA FDMS FEM FRM GEO GIS HEI IMPROVE KML MESA MYP NAAQS NACAA NARSTO NCER NCore OAQPS OC ORD PM PM2.5 PM10 PMiQ-2.5 PQAO QA SEM SLAMS Ambient Monitoring Technology Information Center Air Quality System Approved Regional Methods Air Quality Index Black carbon Board of Scientific Counselors Clean Air Act Clean Air Act Advisory Committee Clean Air Scientific Advisory Committee Community Multiscale Air Quality Model Chemical Speciation Network Denver Aerosol Sources and Health Detroit Exposure and Aerosol Research Study Desert Research Institute Davis Rotating Uniform size-cut Monitor Elemental carbon Environmental Protection Agency Filter Dynamic Measurement System Federal Equivalent Method Federal Reference Method Group on Earth Observations Geographic information systems Health Effects Institute Interagency Monitoring of Protected Visual Environment Keyhole markup language Multi-Ethnic Study of Atherosclerosis Multi-year Plan National ambient air quality standards National Association of Clean Air Agencies North American Research Strategy for Tropospheric Ozone National Center for Environmental Research National Core Monitoring Network Office of Air Quality Planning and Standards Organic carbon Office of Research and Development Particulate matter particles generally less than or equal to 2.5 jim in diameter particles generally less than or equal to 10 micrometers (|im) in diameter particles generally larger than 2.5 and up to 10 jim in diameter Primary Quality Assurance Organization Quality assurance Scanning electron microscopy State and local air monitoring stations December 2008 in ------- STAR Science to Achieve Results STN Speciation Trends Network TC Total carbon TEOM Tapered Element Oscillating Microbalance TTN Technology Transfer Network TOR Thermal-optical reflectance TOT Thermal-optical transmittance XRF X-ray fluorescence December 2008 iv ------- AUTHORS AND CONTRIBUTORS AUTHORS Dr. Dan Costa, National Program Director for Air Research, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC Mr. Neil Frank, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Dr. Barbara Glenn, National Center for Environmental Research, Office of Research and Development, US Environmental Protection Agency, Washington, DC Mr. Tim Hanley, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Ms. Beth Hassett-Sipple, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Dr. Bryan Hubbell, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Mr. Phil Lorang, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Mr. Nick Mangus, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Dr. Lucas Neas, National Health and Environmental Effects Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Chapel Hill, NC Dr. Venkatesh Rao, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Ms. Joann Rice, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Mr. Geoffrey Sunshine, Health Effects Institute, Boston , MA Mr. Tim Watkins, National Exposure Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC Mr. Lewis Weinstock, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC December 2008 v ------- CONTRIBUTORS Dr. Sherri Hunt, National Center for Environmental Research, Office of Research and Development, US Environmental Protection Agency, Washington, DC Ms. Sascha Lodge, National Center for Environmental Research, Office of Research and Development, US Environmental Protection Agency, Washington, DC Ms. Stacey Katz, National Center for Environmental Research, Office of Research and Development, US Environmental Protection Agency, Washington, DC Dr. Marc Pitchford, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Ms. Gail Robarge. National Center for Environmental Research, Office of Research and Development, US Environmental Protection Agency, Washington, DC Mr. Rich Scheffe, Office of Air Quality Planning and Standards, Office of Air and Radiation, US Environmental Protection Agency, Research Triangle Park, NC Ms. Laurel Schultz, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC Dr. William Wilson, National Center for Environmental Assessment, US Environmental Protection Agency, Research Triangle Park, NC Dr. Darrell Winner, National Center for Environmental Research, Office of Research and Development, US Environmental Protection Agency, Washington, DC December 2008 vi ------- INTRODUCTION Air pollution continues to have adverse impacts on the human and environmental health of the United States, despite clear evidence that overall air quality has improved. To understand the relationships between air pollutants and adverse health and welfare effects, researchers utilize ambient air measurement data collected through monitoring networks operated almost exclusively by State, local and Tribal air monitoring programs. These networks provide data for characterizing ambient air concentrations of the criteria air pollutants (particulate matter (PM), ozone, nitrogen dioxide, sulfur dioxide, lead, and carbon monoxide) as well as toxic air pollutants. However, the ambient air monitoring networks do not capture data everywhere or every day. Thus, in April 2008, the Environmental Protection Agency's (EPA's) Office of Research and Development (ORD) and Office of Air Quality Planning and Standards (OAQPS) within the Office of Air and Radiation (OAR) co-sponsored a workshop to discuss which modifications to the current ambient air quality monitoring networks would advance our understanding of the impacts of criteria air pollutant exposures on public health/welfare in the most meaningful way. In particular, EPA sought advice on concrete steps that could be taken to prioritize monitoring sites and/or specific fine particle components for more frequent monitoring in order to improve our understanding of the impact of these components on public health1. The purpose of this document is to present a summary of the April 2008 workshop and the recommendations emanating from the workshop discussions. This document is not intended to be a commitment to actually implement the recommendations but rather it will serve as a tool to consider and prioritize near- and long-term actions for EPA and others to undertake. To the extent that these recommendations can be incorporated into routine monitoring networks, they will help expedite research, better inform future reviews of the national ambient air quality standards (NAAQS), and, ultimately, reduce air pollutant exposures that are associated with adverse health and welfare effects.2 1 Research specific to the protection of public health remains a top priority and EPA has targeted PM as a high-risk pollutant. In the last PM NAAQS review, EPA focused on particle mass and primarily distinguished between two categories of particle pollution based on size (i.e., fine- and coarse-fraction particles), and conducted parallel evaluations of the available scientific evidence relating to each category. The importance of specific PM components and sources was evaluated within the context of this basic size differentiation. In the current PM NAAQS review, EPA is considering the extent to which new information has become available to assess and determine how particle pollution is defined. Specific characteristics to consider will include particle size/mass, composition, and sources/environments (e.g., urban and rural areas). This information will inform decisions related to whether sufficient evidence exists to warrant consideration of alternative indicators for PM, and, if appropriate, the development of new NAAQS. See http://www.epa.gov/ttn/naaqs/standards/pm/data/2008 03 final integrated review_plan.pdf for more information. 2 In a related effort, EPA recently issued the Clean Air Research Multi-Year Plan 2008-2012 which describes the objectives of leading-edge research to support regulatory decision-making. This plan outlines research that will provide critical information to add to the existing scientific foundation to inform the reviews of the NAAQS; develop regulations and advanced tools and models to implement the NAAQS; and improve methods to track progress in achieving health and environmental improvements. (See http://www.epa.gov/ord/npd/pdfs/Air-MYP-narrative-final.pdf) The multi-year plan builds upon recommendations from EPA's OAR as well as several scientific advisory boards. See also National Academy of Sciences (NAS) National Research Council (NRC): Research Priorities for Airborne Paniculate Matter at http://books.nap.edu/catalog.php?record_id=10957; Board of Scientific Counselors (BOSC) Report on the PM-Ozone Program Review: April 2005 at http://www.epa.gov/osp/bosc/pdf/pm0508rpt.pdf; and Clean Air Act Advisory Committee (CAAAC) Report on Air Quality Management in the United States at http://www.nap.edu/catalog/10728.html. December 2008 1 ------- Background The EPA is interested in having an open and continuing dialogue with representative experts regarding health research priorities for ambient air quality monitoring data that would best advance our understanding of the impacts of air pollutant exposures on public health. This dialogue was significantly advanced at an initial meeting co-sponsored by the Health Effects Institute (HEI) and EPA in late 2006. At that meeting, the primary focus was to discuss how the use of the accumulating data derived from nationwide monitoring of fine particulate matter (PM2.5) components could facilitate current and future health effects studies and improve comparisons of risk estimates across studies. The workshop illuminated issues associated with accessing and analyzing monitoring data and identified needs of the health effects research community regarding monitoring of fine particle components.3 In April 2008, EPA's ORD and OAQPS co-sponsored a follow-up workshop bringing together approximately 80 EPA and external air quality, monitoring, exposure, and health scientists4. As briefly described above, the major goal of the workshop was to discuss recommendations for modifications to the current ambient air quality monitoring that would advance our understanding of the impacts of criteria air pollutant exposures on public health/welfare in the most meaningful way, specifically for understanding the impact of PM and PM2.5 components. In addition, the workshop was designed to continue to facilitate communication and scientific interactions across disciplines (e.g., epidemiology, toxicology, atmospheric science, monitoring, risk/exposure assessment) to improve the availability and interpretation of air quality monitoring data for air pollution health studies dependent upon the national networks. The impetus for these two meetings was the growing recognition that current and future changes to the air quality monitoring system could significantly affect ongoing and future epidemiology research. This research serves as a foundation for EPA's reviews of the NAAQS. Yet resources at the Federal and State/1 ocal/Tribal levels for air pollution monitoring continue to diminish, while increasing demands (such as potential expansion of the lead ambient air monitoring network) tied to various aspects of NAAQS compliance continue to grow. Prominent health researchers have increasingly questioned EPA's commitment to health research in planning its monitoring programs, while the State and local monitoring experts who design and operate the monitoring programs wonder why some current data resources are not fully or properly accessed to explore opportunities to address some of the questions that are important to the health researchers with existing data sets. Clearly, improving the understanding of the objectives/mandates of each of these communities and fostering collaborative efforts between these communities is critical to moving forward in a positive manner. Workshop Structure, Objectives, and Outline of Session Summaries The April 2008 workshop began with introductory remarks by EPA's National Program Director for Air Research, Dr. Dan Costa, and by senior OAQPS managers, Ms. Lydia Wegman and Mr. Richard (Chet) Wayland, expressing their support for this effort. Dr. Morton Lippmann from New York University highlighted successful research that has benefited from collaboration among air quality experts and health researchers. Thoughtful discussions centered around five specific panel sessions, each facilitated by two co-chairs (one EPA and one non-EPA representative), addressing the following topic areas: 3 See http://www.healtheffects.org/AQDNov06/AQDWorkshop.html for more information. 4 The workshop agenda and list of participants is included in Appendix A. December 2008 ------- • Session I: Elemental and Organic Carbon Measurements • Session II: Accessing Ambient Air Monitoring Data • Session III: Ambient Air Monitoring for Health Research • Session IV: Thoracic Coarse Particle Components and Potential Public Health Impacts • Session V: Ambient Air Monitoring Realities - EPA/State/Local Perspectives The primary workshop objectives included: • To reexamine and assess progress to date on key issues identified at the earlier HEI/ EPA-sponsored workshop and in follow-up conference calls with the PM Research Center Directors, HEI National Particle Component Toxicity (NPACT) Principal Investigators, and other researchers. • To discuss specific recommendations for concrete steps that EPA and other organizations could take in the ambient air monitoring program to advance health research for the criteria air pollutants. • To seek constructive feedback on five draft "white papers"5 developed to aid in a common understanding of the issues under discussion. These draft white papers are included in Appendices B through F of this draft document. The summaries of the five workshop sessions presented below include: • Session overview, • Panel members, • Major points identified by the workshop participants, and • Recommendations o Presented at the workshop and/or o Developed by EPA staff based upon the workshop discussions Initial Successes and Next Steps This workshop was a major step in a series of interactions to foster improved long-term communication between air quality experts and health researchers. The air program has continually emphasized integration across disciplines, labs and EPA programs. Although this requires significant investment of time and effort for all involved, we believe such an investment is necessary to ensure that the ambient air monitoring program offers, and health researchers use, the best and most appropriate data possible to support the health research that serves as a foundation for EPA's NAAQS reviews. While follow-up from this workshop continues, there are already demonstrable outcomes from the efforts involved in planning, as well as holding the April workshop. A few examples of initial successes include: • Monitoring staff, in EPA and State/local agencies, are becoming aware of the need for daily measurements of PM components and have expressed willingness to save filters used for daily measurements of PM2.s mass so that they can be used to analyze PM components for health researchers. 5 The draft white papers presented relevant background information and critical issues, opportunities for improvement including draft EPA recommendations for possible short- and long-term activities, and, as appropriate, charge questions to stimulate discussion at the meeting. December 2008 1 ------- • Science to Achieve Results (STAR) grantees voluntarily analyzed data sets to help determine whether changes to the measurement method for organic carbon would affect the findings of ongoing epidemiological studies. • Based on information they received from air quality experts at the workshop, STAR grantees have acknowledged making changes to the designs of their health studies that will improve validity of results as a result of better information on spatial variability of PM components. • OAQPS monitoring experts and ORD epidemiologists and exposure scientists are working together to examine the covariance of specific PM components (e.g. metals) across various cities to develop a network of sites for refined epidemiological study. • ORD's National Center for Environmental Research has conducted a preliminary survey of EPA-funded health researchers to identify specific ambient monitoring sites that are being used in current epidemiological studies6. This information will be useful to State and local air agencies as they consider any future changes to their monitoring networks. A draft of this document was distributed to the workshop participants for review and comment. This final summary document incorporates the comments that were received. In addition, EPA has identified additional next steps that will be taken, including: • Stratifying the recommendations in this document as to their feasibility and prioritizing actions to be taken. These include providing important input into the strategy for the revised monitoring network (NCore7). EPA's ORD and OAQPS will work together to facilitate the incorporation of these recommendations into both short- and long-term monitoring strategies and leveraged program plans. This will include EPA staff briefing and receiving direction from senior management during planning cycles for both monitoring network design and health research planning. • Implementing data access improvements and communication tools as soon as possible to prevent disruption of data streams, loss of important monitoring sites, and developing a clearinghouse for other datasets not readily formatted to the Air Quality System (AQS) data system. • Requesting periodic consultations on the enhanced monitoring program with the Clean Air Scientific Advisory Committee (CASAC) and its Ambient Air Monitoring and Methods (AAMM) Subcommittee. 6 See Appendix G. 7 See http://www.epa.gov/ttn/amtic/ncore/index.html for more information on the NCore Multipollutant Monitoring Network. December 2008 ------- SESSION I: ELEMENTAL AND ORGANIC CARBON MEASUREMENTS Background/Objectives EPA has made changes in the urban Chemical Speciation Network (CSN) carbon sampling and analytical protocols for measurement of particulate carbon in order to address inconsistencies between the procedures previously used in that program and the rural Interagency Monitoring of Protected Visual Environments (IMPROVE) program. Health researchers have expressed concerns about this methodological change and potential impacts on monitoring data that are used for long- term time-series analyses. The session objectives were to: • find ways to minimize disruption to epidemiological studies, both current and future, • determine the extent of measurement change/error that is problematic, • determine which of the CSN changes are of most concern to epidemiological studies, • assess whether past and planned measurement comparisons are adequate, and • consider how blank filters and sampling artifacts should be handled. Session Overview This session contained two components: (1) a series of factual presentations on the measurement methods, the changes and/or errors that may occur, and the importance of carbonaceous aerosol sampling to the health research community and (2) a panel discussion to address approaches that can be used to assess potential types of measurement error, needed measurement comparisons, impacts on epidemiological studies, and additional steps necessary to identify and address information gaps. The presentations included: • Overview and Introduction to Key Issues—Venkatesh Rao (EPA) and Barbara Turpin (Rutgers University) • A Health Researcher's Perspective: What's so Special About Carbon?—Ed Avol (University of Southern California) • CSN Carbon Monitoring Changes and Issues—Joann Rice (EPA) • Carbonaceous Aerosol Sampling Artifacts in the National Monitoring Networks—John Watson (Desert Research Institute) • Transitions: Relating "Old" to "New" Methods—Warren White (University of California-Davis) • Predicting Carbonaceous Species Concentrations with Partial Least Squares—Philip Hopke (Clarkson University) • Impact of Method Transitions to Health Research—Michael Hannigan (University of Col orado-B oul der) • Air Quality Monitoring: Perspectives from East and West—Dirk Felton (NY Dept. of Environmental Conservation) Members of the discussion panel included: • Venkatesh Rao, co-chair, EPA, OAQPS • Barbara Turpin, co-chair, Rutgers University • Ed Avol, University of Southern California December 2008 f ------- • Michelle Bell, Yale University • Judith Chow, Desert Research Institute • Neil Frank, EPA, OAQPS • Philip Hopke, Clarkson University • Michael Kleeman, University of California-Davis • Allen Robinson, Carnegie Mellon University • Warren White, University of California-Davis This session provided a good forum for the exchange of information among diverse interests in monitoring, analysis, and health impacts. The panelists and those who spoke up from the floor shared the view that the impact of measurement changes on health research is unknown and complex, as is how much detail is needed to better understand the impact on epidemiological studies and resulting correlations with health effects. There was also agreement among those speaking about a need to: (1) collect more information from collocated measurements for an extended period of time at multiple sites; in order to (2) better document uncertainty, and associated differences between elemental carbon (EC) and organic carbon (OC) measurements for the new and old methods. Spatial location and source dependent differences in monitoring sites are important considerations, as are potential differences by season. Specific action items were not identified at this session, but there was a general consensus amongst the speakers about the proposal (see below) for more extensive co-located monitoring, and several researchers called for such analyses in a wide range of locations and seasons. Little discussion took place of other related issues that had previously been identified in the agenda (this included how to handle artifacts, how to handle field blanks vs. back-up filter measurements, and which measurement errors are most problematic for epidemiological studies). Additional background information is provided in Appendix B. Major Points Raised by Participants General • Health endpoints are affected by the physical, chemical and toxicological properties and attributes of carbon-containing particles as well as by the emission sources represented by the air quality measurements. These characteristics are not routinely measured, and currently can only be inferred—very indirectly—from the relative proportions of EC and OC. The reported proportions of EC and OC are likely to change as a result of the changes in the measurement protocol; this could be more important in assessing the impact on health than individual uncertainties in the OC and EC measurements. • Changes made in the CSN carbon measurements can be documented, but sampling artifacts (organic vapors adsorbed within the filter) are still being explored. Charring of these vapors within the filter is the main cause of differences between transmittance and reflectance corrections for OC charring (Chow et al., 2004 Relating the results of "old" and "new" methods is ongoing. The co-location of old/new monitors can provide information important to understanding the nature of changes in measurements. • In the past, EPA used several samplers in their chemical speciation network, which resulted in good agreement for EC, but poor agreement for OC and total carbon (TC) - which was a function of the differing sampling flow rates and having no good way to address the sampling artifacts. In the future, one sampler that is IMPROVE-like will be used for CSN with a higher flow rate, smaller filter, back-up quartz filters, and better field blanks (this will December 2008 6 ------- likely ensure uniformity of the OC measurements, but more work will likely be needed to understand what is actually being measured). • The change in the analytical protocol will change the OC-EC split in some cases, yielding more reported EC and less reported OC. The analytical change is not expected to change reported TC. The changes in sampling protocol will affect the sampling artifact, and thereby reduce reported TC and OC. The sampling change is not expected to change reported EC. • Hannigan et al. [2008] suggest that in the DASH (Denver, CO) study, EC would remain significant as an indicator for health endpoints. Of course, Denver is only one location in space, and generalizations cannot be made about the health effects of EC until more studies are completed. • For considering impacts of PM on health, measures of bias and of uncertainty are both important in determining health outcomes. • There is a need for semi-continuous OC/EC measurements as well as other PM chemical components. As important as the increased temporal resolution that will result from use of semi-continuous methods is the affordability of these measurements (daily (and shorter interval) sampling at multiple sites becomes more feasible). This can be important for better understanding sources of carbon since more detailed spatial and temporal information can aid in source attribution studies. The shorter sampling times associated with semi-continuous instruments also produce samples with much smaller variation in sampling conditions (good in terms of minimizing the potential for redistribution between the gas and particle phases). • In addition, there currently are sites with Sunset semi-continuous OC/EC analyzers, McGee aethalometers, and Thermo-Scientific MAAPs operating through the transition period and beyond. Some researchers suggested that the locations of these analyzers at sites important for epidemiologic studies (cities where health effects have been shown) and at other important sites should be identified so that the usefulness of these data for epidemiological studies and for "harmonizing" efforts can be identified. However, the Sunset Labs instrument collects material at a somewhat higher than ambient temperature, like the TEOM does. Thus, Sunset Labs' OC values may be somewhat smaller than those made at ambient temperature. Others noted that health research should not be limited to the locations of on- going studies due to the many studies that are national in design and the need to understand regional variation in health effects. • One presenter indicated that the sub-fraction measurements of OC and EC made in the IMPROVE and new CSN protocols can also add valuable information to source attribution studies. However, low temperature (i.e., volatile) OC1 and OC2 fractions are most sensitive to temperature changes during the analysis and during sample handling (Chow et al, 2007, Dillner et al., 2008). • The type of measurement error that is most important to a specific epidemiological study will depend on the study design, and on the scientific questions it aims to address (e.g., acute versus chronic effects). In other words, there is no single factor or set of factors that is the critical need for epidemiological researchers. The impact of measurement changes will December 2008 ------- affect various epidemiological studies differently, depending on their design. This issue needs to be sorted out and discussed further. What types of measurement error are problematic for epidemiologic studies? • All types of measurement errors are of potential concern. There is a need for a qualitative understanding of what the measurement error is comprised of, and how this might vary spatially. The analytical measurement error of the monitor is usually much smaller than the measurement error associated with using a centrally located monitor to represent an entire community, particularly when evaluating pollutants that are measured once every three or six days and in assessing pollutants that are heterogeneously distributed in the ambient air. Measurement error clearly matters when evaluating epidemiological evidence, but for time- series analyses, most of measurement error is Berkson, and hence does not include bias, just reduced power. The way to solve that problem is with more data, which again points to the need for daily measurements. • Changes in carbon components (bulk OC and EC) are important, as are changes in the percent measurement error of the components. Are there differential errors in EC and OC? We need to define "error" better. • Representativeness of sites is affected by spatial variability and by source types. • Balance can be attained by a central PM2.5 speciation monitor plus satellite sites (with lower costs) that indicate how representative the information collected at the central site is (and address how spatially divergent some EC/OC and other components are). Satellite sites must be based on lower-cost methodologies, for example, should optical measurements from Teflon filters be considered? Note: panelists did not address how we can be certain that relationships between "reference" and "cheap" methods observed at central sites will be the same at satellite sites. Are past and planned measurement comparisons adequate? • In evaluating the transition between carbon monitoring methods, workshop participants voiced concerns that the current process to conduct two months of measurement comparisons is too limiting, since seasonal variability is important; there is a need for more data to compare the different methods and inform the epidemiological research. There are 6 co- located CSN/IMPROVE sites ongoing (as identified in the draft White Paper for Session I, see Appendix B) and more discussion of how these could be used would be helpful. It is unlikely that we will be able to afford one year's worth of co-location at every site converted. So, it's important to specifically know where longer-term comparisons are required for health research or to know how to generalize from a small set of locations to the larger network. (see next bullet as well). • It might be useful to consider a whole year of data with less frequent measurements; also consideration of a limited number of samples at more individual sites would be helpful. As such, there is a need to know what kinds of locations are of most importance for health research studies. Some suggested that a wide range of locations representing different particle mixtures and sources would be useful. December 2008 ------- • There is a need to better understand uncertainty information. Proportional changes in EC & OC may create proportional changes in epidemiological studies; if changes are not proportional for different components, then there may be a problem. • In selecting sites for possible increased co-location of old and new measurements, the influence of variations in sources needs to be considered. These could include sites influenced by wood-smoke, mobile sources, large stationary sources, and a mix of all sources. Recommendations/Actions for Consideration Suggested by Workshop Participants: • There was general consensus on a proposal to run co-located "old" and "new" CSN measurements for a whole year; with consideration of potentially measuring every 6th (or 12th) day at 6 - 9 sites. Many researchers felt it was important to ensure that the specific sites include those currently being used for epidemiological studies where associations or relationships with OC or EC are being studied, although many other researchers noted the need for selection of sites that represent a wide array of locations and particle mixtures, as noted below. Inclusion of locations where carbon has been found to be associated with health effects may allow us to more easily determine if the measurement change has an impact on their health impact findings. Potential sites include Seattle (Beacon Hill), Rubidoux (CA), Bronx (New York City), Atlanta, Detroit, Cleveland, Denver, and an additional southern location. • Several individuals supported selecting sites based on source distribution/mix, not just on- going or planned epidemiological studies. Using this approach, consideration could be given to variations in soils, industries, and mobile sources in the source distribution. A site selection process could incorporate both this concept and that listed above, to include locations of ongoing studies. • There was also general consensus that this proposal be reassessed after a year of data collection in order to evaluate whether a longer study period or more sites are needed. • A limited number of individuals also supported an investigation of the availability of semi- continuous OC/EC data and its potential role in "harmonizing" old vs. new carbon measurements. • A limited number or participants also supported achieving better spatial resolution by implementing a main site and satellite sites for species with large spatial variability, to better characterize population exposures. Developed by EPA Staff Based on Workshop Discussions: • Continue UC-Davis (January 2008) "Carbon Summit" process: o Selected archived samples from IMPROVE and collocated IMPROVE/CSN network sites will be analyzed by the new IMPROVE carbon analysis methodology to develop December 2008 9 ------- additional data, which will be used along with the routine data collected by the two networks to better characterize the relationships between the various old carbon analysis methods and the new IMPROVE analytical method (i.e. the approach that is used now for both networks). o Changes will be made in the operations of both IMPROVE and CSN networks to increase the number and utility of carbon blanks and backup filters collected in order to further investigate and if possible develop more credible approaches to adjust for sampling artifacts. o Continue to conduct periodic assessments of the quality and comparability of PM carbon data from both networks and over time (i.e. as methods have changed) paying particular attention to data from continued operations of collocated monitoring sites. The results of these efforts to be the subject of possible future joint network workshops, publications and web postings. Further evaluate the performance of the Sunset EC/OC and other continuous analyzer and determine the role of continuous measurements to support daily monitoring and harmonize old and new CSN. Filter-based measurements will still be needed, but can be collected on a much reduced frequency. This idea can then be extended to other semi-continuous instruments, like the Aethelometer, MAAPs, and photoacoustic spectrometer which measure light absorption, strongly correlated with EC. Make a more targeted request (with much more specificity than was done for this workshop) to health researchers to help us better understand the sensitivity of their results to changes in carbon measurements. Further discussion of the topic "optimal OC artifact correction for large networks" is warranted. December 2008 10 ------- SESSION II: ACCESSING AMBIENT AIR MONITORING DATA Background/Objectives EPA's Air Quality System (AQS) is designed to collect and store ambient air monitoring information. EPA recently introduced the AQS Data Mart to facilitate access to this monitoring information. The AQS Data Mart is a generic "retrieval" tool that provides the ability to query any information, but it does not provide significant data exploration or analytic capabilities. These capabilities are left to the "analytical" tools. Various analytical tools, or interfaces, are available including the Health Effects Institute's (HEI's) Air Quality Database, which focuses on ambient air measurements of PM2.5 components and gaseous pollutants at and near STN and SLAMS sites. This session focused on data access issues and how to help health researchers obtain monitoring data for fine particle components and other critical pollutants more easily. Session Overview In general, access to ambient air monitoring data to support health research/ assessments falls into four general categories: • epidemiological studies, • exposure/risk assessments, • public health surveillance, and • health impact assessments Keeping these broad categories in mind, and understanding that the goal is to provide a framework for delivering consistent, well-documented monitoring data to users including the health research community, the broad topics discussed in this session included: • To what key data do health researchers need access? • What formats are most useful? • What kind of access is most appropriate for health research uses? • How can the overall data context be improved and preserved in delivering the data to users? Members of the discussion panel included: • Michelle Bell, co-chair, Yale University • Bryan Hubbell, co-chair, EPA, OAQPS • Sara Adar, University of Washington • Kaz Ito, New York University • John Langstaff, EPA, OAQPS, Health and Environmental Impacts Division • Nick Mangus, EPA, OAQPS, Outreach and Information Division • Richard Poirot, Vermont Department of Environmental Conservation • Betty Pun, AER • Rich Scheffe, EPA, OAQPS The panel members were asked to be specific in their data needs, e.g., to clearly identify what specific documentation is needed, rather than just providing general recommendations to provide "more documentation." A special focus on versioning of the data was also recommended for the discussion. December 2008 11 ------- To lead off the session, two introductory presentations were provided by Nick Mangus and Rich Scheffe of EPA. Nick Mangus provided an overview of the draft white paper entitled "Access to EPA's Air Quality Data for Health Researchers (see Appendix C)," while Rich Scheffe provided a summary of the outcomes from a recent EPA-sponsored Air Quality Data Summit, held in March 2008.8 Major Points Raised by Participants Several persons mentioned the high value of EPA air pollution datasets, and noted that access has greatly improved over the years. Several broad themes emerged from the discussion: Versioning The most important theme seemed to be versioning of the EPA datasets. Several people expressed concerns over the current system, where researchers can download different datasets without realizing they differ. Currently, there is no way of knowing whether data files have changed, or merely been updated. This causes problems with reproducible research and with researchers not knowing whether to get new data due to corrections being made. It also is an issue when researchers update an epidemiological analysis with a new year of data; they need to know if there have been changes in the air quality database for previous years as well. It was noted that changes to data include not only updates of new data, but corrections to old data, even 10 years previous. Multiple options to address this issue were discussed: • time/date stamp every observation • version numbering (e.g., Version 10Jan08.7, etc.) • maintenance of snapshot datasets that are static • notice of changes to datasets • maintenance of datasets used by researchers Concerns were raised regarding the cost and storage of some of these options. Data Availability The draft white paper included in Appendix C provided background information for this session. This paper described and contrasted the data available through the AQS Data Mart and the HEI Air Quality Data Base, and other sources of information. Specific issues discussed related to data availability included: • EPA vs. other organizations' roles in providing access to data. Several participants expressed support for HEI or some other organization outside the EPA continuing to support analytical tools, or interfaces, to disseminate air quality data "packaged" with other relevant data for health researchers. • Information previously available through AQS that is no longer available. There were some concerns expressed that certain types of data, e.g. the TEOM data had been removed from the AQS archive website. EPA clarified that TEOM and FRM data were different and that the TEOM data were removed so unknowing users would not incorrectly compare it to FRM data. EPA has mitigated this issue by changing how we identify PM2.s data and will 8 See http://wiki.esipfed.org/index.php/Air Quality Data Summit#Documents. December 2008 12 ------- reconsider putting the TEOM data back on the website. It was noted that some local TEOM data are being adjusted to match FRM data as part of the joint CDC/EPA PHASE project. • Additional information for monitor locations. It was suggested that photographs of the monitoring station be easily accessed and keyhole markup language (KML) files be created. While the land-use around monitors could change, this would provide researchers with some guidance on the monitor location. • Information from other sources. The concept of a data clearinghouse providing links to non- EPA data sets was suggested, for example, access to additional State, local, and tribal air quality data that are not entered into AQS (e.g., AirNowTech data), special studies data, Supersites data, etc. EPA clarified that the Supersites data are available on the NARSTO website. However, workshop participants made it clear that the data on the NARSTO site is not in a format that is comparable to the FRM data available on AQS. Workshop participants noted that it would be helpful to have all the air quality data in one location, and provided in a similar format or have scripts for reading data from different formats into common software programs like SAS, R, or S-Plus. The need to leverage existing IT capabilities to gain efficiencies in creating a data clearinghouse was highlighted. Please see discussion of this issue below in the section Actions Items Under Consideration. • Requesting data from State, local tribal air agencies directly. EPA maintains a listing of regional, State, and local level air quality data contacts available on its website. Workshop participants noted that when going to State contacts for data, some provide data faster than others, and often in very different formats (sometimes still providing paper copies only). It was suggested that it would help State/local air agencies if health researchers would share final products (e.g., health studies) with these agencies so they could be better informed about how the air quality data are used and future research needs. Data Quality Some mention was made of data quality, and the need for researchers to be aware of data quality problems, even in EPA-vetted datasets. Examples provided included: • Quality and accurateness of existing metadata are varied, metadata can be wrong, and that there are often not even flags to note suspicious outlier values that might result from a misplaced decimal. For example, if ozone data at low levels are a decimal place off, this could play a major role in studies of thresholds and effects at low concentrations. • Incorrect identification of latitude/longitude has been noted for a subset of monitor sites as well as a high frequency of missing data in some of the critical fields (e.g. no monitoring objective listed for some sites). EPA notes that all data in AQS passes the EPA's QA checks for the tests we need for regulatory purposes. This may not be sufficient for some scientific applications. All metadata (and data) are ultimately controlled by the submitting State/local/tribal agency and EPA cannot change it. So improving quality is a community effort. This is another reason why researchers might want to engage the agencies whose monitors they are using for studies. • It was recommended that the data be stored at the finest spatial and temporal granularity possible, and then aggregated up as needed by the user. December 2008 13 ------- User Friendliness and Documentation It was noted that some data users are novices, some are expert, and each needs different levels of data access and, therefore, would have different needs. Many participants voiced the opinion that epidemiologists tend to be "very opportunistic," making use of whatever air quality data are provided. As such, epidemiologists will utilize available data, but would like more guidance as to what uses of the data are valid and/or appropriate (e.g. provision of data on the percent valid observations, monitoring objectives, monitor scale, use of flagged data). The need to increase user friendliness was discussed, such as with the ability to download specific States' data rather than data for the whole US for some data sources. Within the discussion of user friendliness, there was a general consensus that documentation needs to be improved. In addition, documentation is perhaps a bit scattered throughout various EPA sites. The point was made that "user-friendliness" is dependent on the needs and skill level of the user, so that no standard form will be perfect for all users. The issue is that it takes more effort to find and interpret documentation of the data than it does to obtain the data; and missing descriptive information can be devastating to an analyst. There are many attributes of the data that are not well understood by users (metadata is not owned by EPA and therefore can be outdated or data can be updated by the owner at any time; EPA data storage labels are in flux with regulation and policy changes; etc.). Users having a single site for documentation will greatly reduce the likelihood of incorrect interpretation. Secondary Data The panel discussed whether non-air pollution data would be useful to include in the air pollution datasets. Examples were weather, land-use, and census data. Some people thought this would be useful. Others thought this was less useful, given that such data are available elsewhere. This seemed to be far less of an issue than the versioning. Thus, resources might better be spent on the versioning issue than on incorporating secondary data. Recommendations/Action Items for Consideration The recommendations listed below include items discussed at the workshop as well as additional recommendations developed by EPA staff based upon the workshop discussions. Data Versioning • Explore options for adding the date (and time) of last modification to all data measurements in AQS (and the AQS Data Mart). Notes: Data in AQS can change at any time. Generally, EPA limits the ability of data submitters to change data for only the last few (3-5) years. However, we do open time windows where older data can be changed. There are implications to the submitter for changing data and the volume of older data changes is low, but they do happen. Researchers need to know if these data changes affect their analyses. Adding a date-of-change to each measurement would allow a data user to query the space and time domain of their initial query to see if any data had changed. EPA would not keep a record of the change, but new or changed values would be indicated with their date of creation/update. It would still be incumbent on the user to maintain the original data set and compare changed data to assess possible ramifications. Any change to a measured value or metadata (like the monitoring method or uncertainty) would trigger the date stamp to be updated. December 2008 14 ------- Data Availability • Continue to encourage HEI and other organizations support of analytical tools and interfaces to make the AQS data more useable for the health research community. Follow-up with HEI regarding the plans for continuing to support the HEI Air Quality Data Base. • Explore options for creating and maintaining a Clearinghouse of non-EPA datasets There are a lot of air quality data which are not in AQS. Some of these data have been collected by State and local air pollution control agencies and have not been reported to AQS for one reason or another (e.g., continuous data are collected on a minute basis, and EPA only requires that hourly averages be submitted; the information does not pass the regulatory quality assurance requirements, even though the quality may be high enough to fill a gap in a research time series; etc.) Another source of data that EPA does not have is data collected from special research studies run by academic or State/local public health organizations. This is generally high density data (in space and/or time) used to better understand the variability of measurements over smaller scales than regulatory monitoring requires. Even the other Federal air quality storehouse, AirNow Gateway will have State/local/tribal data that does not meet the policy requirements of AQS. Having a single location as the starting point for a dataset clearinghouse will dramatically help improve the inventory of available air quality information. • Location of Monitoring Sites. Make keyhole markup language (KML) files with AQS air monitoring site locations should be made available. KML files describe locations and can be used by most modern map-drawing applications. Including links within the KML file will allow users to download the actual measurements. Making these available would better allow people to visualize the monitoring network they would also confirm the monitor location when the listed latitude and longitude were suspect. • Develop new content and format for data on AQS archive page. Request feedback from key users on how to improve the AQS "Data Archive" (data download) page on EPA's website and update accordingly. Users complain about the limited number of parameters available, the frequency of updates, the format, and the size of the files. If agreement can be reached on how to improve these problems, EPA could make the appropriate changes. • Continue Data Summit follow-up to provide a system-of-systems for data integration and display. OAQPS will provide base data from Federal monitoring networks via the AQS Data Mart. The purpose of the data summit work is to make use of "interoperability" frameworks like GEO (Group on Earth Observations) for system developers to identify what part of the data value- chain they belong in and successfully connect with those up- and down-stream from them. The value-chain has roughly these divisions: base data provision, metadata provision, data integration, data processing (aggregation), data visualization, and communication. It is expected that the architecture standards board convened as a follow up to the Air Quality Data Summit will adopt specific recommendations on web services (machine queries) of data that all participating systems should support. If all systems agree to the web services, data integrators and interface builders would have a much easier time obtaining data from multiple systems. For example, the HEI Air December 2008 15 ------- Quality Database could be updated more frequently, and include more information about the progeny of the data. • Explore opportunities to encourage discussions between health researchers and State/local monitoring experts. For example, encourage health researchers to share published results with State/local monitoring community, specifically studies that use monitoring data collected by the State/local air agencies and encourage participation of a variety of experts in regional and national meetings (e.g., health researcher participation in monitoring meetings). User Friendliness and Documentation • Explore ways to highlight if a significant change in AQS has occurred. For example, EPA could provide an explanation on the AQS website so that regular AQS data users would be aware and know where to go to access the data. • Use the IMPROVE metadata as a model for developing AQS metadata, including the IMPROVE "Data Advisories" noting any changes or issues with the data. • Request that supplemental data be presented in the same format as primary data. December 2008 16 ------- SESSION III: AMBIENT AIR MONITORING FOR HEALTH RESEARCH Background/Objectives The purpose of Session III was to provide a discussion among a panel of health and exposure researchers who use ambient air monitoring data in their studies and to highlight the value of these data for the continued progress of research linking particulate matter (PM) sources, exposure and health effects. This session was intended to stimulate a discussion of the issues and identify creative solutions for the provision of daily speciation data in key locations while working within current resource constraints faced by local, State and Federal air quality agencies. For example, resolving key policy questions about health effects of specific size fractions, components and gaseous co- pollutants related to PM, requires more intensive temporal and spatial air quality data than are currently available. Background information for this session is included in Appendix D. Session Overview The panelists discussed the major sources of uncertainty that must be considered when designing and interpreting the results of studies of ambient PM mass, components and health. A number of important research questions were raised in this context that could be addressed if some changes were made to monitoring networks in some locations. If daily measurements of PM species were available in informative locations, studies would be better designed to detect the relative importance/toxicity of PM species and size classes. Studies designed to detect health effects of short- term exposure would develop more precise and valid exposure estimates and would be better able to generate hypotheses that address possible mechanisms for the observed health responses to PM. Although the panelists primarily focused on issues as they relate to epidemiology studies, the point was made that toxicological assessment of ambient particles must be generalizable to human exposure and be interpretable in relation to epidemiological results. Members of the discussion panel included: • Barbara Glenn, co-chair, National Center for Environmental Research, EPA • Joel Schwartz, co-chair, Harvard School of Public Health • John Godleski, Harvard School of Public Health • Patrick Kinney, Columbia University • Lucas Neas, National Health and Environmental Effects Research Lab, EPA • Roger Peng, Johns Hopkins University • George Thurston, New York University • Jay Turner, Washington University in St. Louis Major Points Raised by Participants Uncertainties Epidemiology studies must deal with several sources of error in estimating personal exposure to ambient PM. These include: • Instrument measurement error. Is the monitoring technique adequately measuring pollutants at the site of the monitor? The observation was made that while exposure measurement error December 2008 17 ------- at the monitor introduces uncertainty in estimates of ambient concentrations, this error is smaller than other measurement errors in epidemiology studies. • Spatial variability. How geographically homogenous are the particle concentrations in the cities where we are conducting epidemiology studies? Is the concentration of PM species measured at the monitor representative of the ambient concentrations experienced by human populations living and working in the city? • Temporal variability. Monitoring schedules of every third or sixth day results in data gaps that severely limit the ability to explore variations in the time lag of response for different PM components. Different PM components are believed to target different biologic pathways in generating health responses and the flexibility to explore a variety of time lags between exposure and outcome is necessary to reveal these effects. In addition, health events relevant to days with missing air quality data must be excluded, thus reducing sample size. Alternatively, air quality concentrations may be interpolated in some manner which increases uncertainty in exposure estimates. Interpolation results in a reduction of daily variability in PM concentration data with resulting loss of statistical power, and increases the correlation between concentration estimates for PM components in datasets. This inhibits efforts to differentiate species-specific toxicity. • Other errors in estimation of personal exposure based on ambient measurements. Infiltration of PM components indoors, time-location during the day etc. The comment was made that use of air conditioning during the summer months has a large effect on health models for certain geographic locations. Time scale Research has shown that the health effects of air pollution exposure on one particular day are spread out over several subsequent days. Therefore, the effects (e.g., death) observed on one day are the result of air pollution that occurred during a period of days on and before the deaths were recorded. Epidemiologists need to use pollution concentrations on a defined number of days before the date of an effect to study hypotheses regarding the relevant time between exposure and effect. These time scales likely are different for different PM components and different health outcomes (e.g., heart attack, asthma etc). The relevant time lags for a particular outcome also may vary by season and location as well. The study of time lags informs mode of action. In the absence of daily speciation data, PM components with a very acute or immediate effect (short-lags) on health would have less measurement error and thus an apparently stronger association with health outcomes than PM components with more delayed effects spread over several days (long-lags). This differential error could lead to a misattribution of PM effects to specific sources. Monitoring on a daily basis is needed in a reasonable number of cities to evaluate this variation. Statistical Power to differentiate between components The panel listed several key factors that influence the statistical power of an epidemiology study. These include: • Number of daily health events - need large population size. • Variability in estimated exposure (indicated by variation in pollutant concentration across space for long-term cohort studies or change in daily concentration for time-series studies). We want to maximize variance and minimize covariance. December 2008 18 ------- • Maximize measurement of concentration variance. Daily speciation measurements in a location will provide a dataset that documents the day-to-day variation in concentration that occurs at the monitoring site. The point was made that concentrations vary a great deal within a 24-hour period and, therefore, temporal variation in PM species within an area could be better quantified using continuous monitors. Continuous methods also would minimize concerns about artifacts. Local sources are another determinant of variation in PM species concentrations. Better spatial resolution in the data will better account for local sources. • Minimize covariance between PM components in datasets. Panelists emphasized that studies are not able to differentiate the relative importance of PM components in associations with health effects if concentrations are highly correlated in a study location. To avoid this problem they recommended that study sites be selected where covariance between PM components is minimized or different. The point was made that meteorological factors are a key determinant of correlation in daily concentration change for PM components, which poses analytical difficulties. Toxicology studies will help to sort out relative toxicities among PM components. Spatial scales For time-series studies, if daily change in pollutant concentration is homogenous across the population, then one "central" monitor will adequately characterize daily change in ambient concentration for that city. This assumption needs to be explored at multiple study sites. For long- term cohort studies, better within-city spatial resolution provides better estimates of annual average concentrations for individual study members, and increases the variation in exposure data. This results in an increase in statistical power. Some panelists emphasized that exposure estimates based on concentration data from one PM speciation monitor in the middle of the city will not add much information for epidemiology studies because exposure measurement error for the overall population is too large. Spatial variability in PM component concentrations is an important issue. Participants also were urged to integrate models that take into account meteorology and source information over space, such as CMAQ, into exposure estimation. At the same time, toxicology studies will have to play a large role in understanding PM component influences on health. Participants were cautioned that toxicology studies also are complex, and the use of pure components in toxicology studies has been disappointing. It is very difficult to generalize results from these studies to draw conclusions about responses to the ambient mixture. The influence of local sources in the vicinity of a monitor needs to be understood before those data are assumed to represent exposure for the population in a city. One panelist emphasized that during certain periods of time, local sources have been observed to highly influence ambient concentrations of a PM component measured by a speciation monitor. The point was made that the CSN monitors are sited to implement the NAAQS and assure compliance, not to support health effects studies. Recommendations Desirable Attributes of Locations Selected for Daily Speciation Measurements The following location characteristics were suggested as criteria for selecting a set of metropolitan areas to support daily speciation monitoring: December 2008 19 ------- • Population size. This is the most important attribute since the number of daily deaths in a city is the major determinant of statistical power of a time-series or case-crossover study. Large metropolitan areas, including New York City, Los Angeles, and Chicago were proposed. However, the comment was made that very large urban centers, such as New York City, have characteristics, such as street canyons, that increase spatial variability. Other less large cities with more homogenous geography may be better candidates. Once the population size is relatively large, resulting in daily deaths of around 20, then other selection factors should play a much greater role in the selection of urban areas for additional sampling. • Different covariance structure between PM species in different cities. • A variety of source contributions. • Location of speciation monitor(s) in the city that minimizes the influence of local sources of PM. • Topography and city attributes that increase the likelihood that one or a few monitors will characterize daily change appropriately. • Expertise and support of State and local air monitoring personnel to collect daily filters and maintain semi-continuous monitors. • Existence of additional monitoring data in that location to supplement information on spatial and temporal distribution of PM components and other pollutants. • Special State/local or academic studies using multiple monitors may have been conducted in the location to better characterize spatial and temporal variability and could be made available for analysis. FRM filters may have been archived by State/local air quality agencies and may be available for speciation analyses. The existence and availability of these studies and filters needs to be explored. A panelist commented that these filters are a national resource and need to be inventoried and kept, not thrown away. • Consider the impact of weather and seasons, differences in behavior that affect exposure. The Children's Health Study in Los Angeles was referred to during the discussion of site selection for additional monitoring. Twelve communities were initially chosen based on the hypotheses of interest for PM2.5 mass, the need to obtain a 3 - 5 fold difference in ambient concentrations, and an emphasis on long-term health effects. Hypotheses concerning different PM size fractions would result in selection of different communities. The design of the CHS was considered innovative because a monitoring approach of 26 two-week sampling periods was conducted which minimized the number of samples taken while capturing seasonal differences in ambient concentration. Action Items for Consideration • Develop a recommendation on the best locations to conduct additional sampling. More information must be collected to inform selection of proposed study sites and the minimum number of locations to conduct additional sampling. • Agree and prioritize important criteria for site selection. December 2008 20 ------- o Analyses and presentation of the correlation structure between PM components and between PM components and other pollutants at the CSN speciation monitors. Kaz Ito, New York University, volunteered to assist with this effort. Analyses are currently being conducted by EPA's OAQPS and ORD. o Identify locations with appropriate data from special studies on daily PM species concentrations or spatial variability. o Evaluate the feasibility of analyzing archived FRM filters from specific locations (for certain components as appropriate). o Develop a table summarizing relevant information. Information characterizing additional criteria will be obtained and the table filled in (ORD, OAQPS). * Kunzli, N.; Avol, E.; Wu, I; Gauderman, W.J.; Rappaport, E.; Millstein, I; Bennion, I; McConnell, R.; Gilliland, F.D.; Berhane, K.; Lurmann, F.; Winer, A.; and Peters, J.M. (2006). Health effects of the 2003 Southern California wildfires on children. Am. J. Respir. Crit. Care Med., 174(11):1221-1228. Wu, J.; Lurmann, F.; Winer, A.; Lu, R.; Turco, R.; and Funk, T. (2005). Development of an individual exposure model for application to the Southern California children's health study. Atmos. Environ., 39(2):259-273. December 2008 21 ------- SESSION IV: THORACIC COARSE PARTICLE COMPONENTS AND POTENTIAL HEALTH IMPACTS Background/Objectives In September 2006, the EPA revised the NAAQS for PM and amended the associated national air quality monitoring requirements.9 As part of the amended monitoring requirements, EPA finalized a Federal Reference Method (FRM) for thoracic coarse particles (i.e., PMio-2.s), even though a NAAQS for PMi0-2.5 was not adopted. This was done to facilitate consistent research on PMio-2.5 air quality and health effects and in promoting the commercial development of Federal Equivalent Methods (FEMs) (71 FR 61212). The amended monitoring requirements require the addition of PMio-2.5 measurements at 75 multi-pollutant monitoring sites (National Core or NCore sites) starting on January 1, 2011. A subset of these monitoring sites will include speciated coarse particle measurements. The purpose of this session was to discuss issues related to and relative priorities for EPA to consider as speciation of thoracic coarse particles is added to the monitoring networks to support future exposure and health studies. When discussing issues and priorities of thoracic coarse particle measurements to support health studies, it is important to acknowledge that we are starting from a different place than with fine particle measurements. This presents both challenges and opportunities. First, while there is an extensive network to monitor PMio and PM2.5, there exists no national network with the specific intent to consistently and accurately measure PMio-2.5-10 As a result, the amount of PMio-2.5 air quality data available and associated analyses are available at fewer locations than PMio or PM2.5 measurements. Second, there have been fewer health studies conducted to investigate relationships between thoracic coarse particle concentrations and health endpoints, which is due in part to the first challenge - limited available air quality data. Some health studies have been conducted using PMio measurements in areas where the PM concentrations are dominated by thoracic coarse particles. Other thoracic coarse particle air quality and health studies have relied upon data from locations where co-located PMio and PM2.s monitors exist, but there are uncertainties in the consistency of these data because the protocol for the PMio and PM2.s measurements is not usually identical.11 Despite these challenges, significant opportunities exist to inform the design of future thoracic coarse particle monitoring programs and to better harmonize thoracic coarse particle measurements with the needs of health effect researchers. Session Overview To investigate health effects associated with exposures to thoracic coarse particles will require an improved understanding of the intra-urban, inter-urban, and urban-rural variability of ambient thoracic coarse particle concentrations. Key uncertainties associated with intra-urban ambient thoracic coarse particles concentrations include spatial, temporal, and compositional variability, while key uncertainties with inter-city and urban-rural comparisons include variability in 9 See http://www.epa.gov/oar/particlepollution/actions.html for more information on amendments to EPA's National Air Quality Monitoring Requirements. 10 U.S. EPA. (2005) Review of the National Ambient Air Quality Standards for Paniculate Matter: Policy Assessment of Scientific and Technical Information, OAQPS Staff Paper (June 2005). U.S. Environmental Protection Agency, Washington, DC, EPA-452/R-05-005. 11 U.S. EPA. (2005) Review of the National Ambient Air Quality Standards for Particulate Matter: Policy Assessment of Scientific and Technical Information, OAQPS Staff Paper (June 2005). U.S. Environmental Protection Agency, Washington, DC, EPA-452/R-05-005. December 2008 22 ------- composition and temporal differences (e.g., seasonal). Ambient air monitoring networks can provide insights to address these uncertainties. As a result, general questions that were posed to the panelists in this session were as follows: • What is the relative value of thoracic coarse particle speciation at planned monitoring locations versus additional mass measurements? • What is the relative value of understanding intra-urban versus inter-urban/rural variability in thoracic coarse particle mass and composition? • What recommendations can be made to inform the design of a thoracic coarse particle monitoring network? Members of the discussion panel included: • Timothy Larson, co-chair, University of Washington • Tim Watkins, co-chair, EPA, ORD • David Diaz-Sanchez, EPA, ORD • Richard Flagan, CalTech • Terry Gordon, New York University • Michael Hannigan, University of Colorado • Thomas Peters, University of Iowa • Joann Rice, EPA, OAQPS • Jamie Schauer, University of Wisconsin There was a general consensus that thoracic coarse particles are very complex and pose significant challenges for both air quality and health scientists. Thoracic coarse particles differ from fine particles and, in many cases, present greater challenges. For example, thoracic coarse particles may exhibit larger spatial and temporal variability than fine particles and the composition of thoracic coarse particles can vary in rural versus urban areas. The composition of thoracic coarse particles can also differ from fine particles with introduction of biological materials and the increased importance of metals. In addition, thoracic coarse particles also present new challenges related to measurement technologies. The current FRM is based on the difference between measurements taken with co-located PMio and PM2.5 integrated filter samplers. While this method provides a good measurement of thoracic coarse particle mass, there is no direct thoracic coarse particle sample collected. Other methods exist, such as dichotomous or continuous mass samplers, which can provide valuable information about particle composition and temporal variability, but these methods need to be evaluated further. In addition, the use of emerging passive monitoring approaches may also provide opportunities for obtaining thoracic coarse particle data. The thoracic coarse particle monitoring program is still evolving, and EPA's OAQPS is seeking feedback and input on thoracic coarse particle network design issues. EPA should apply lessons learned from the fine particle monitoring program and closely follow ongoing research efforts, while moving forward with implementation of a thoracic coarse particle monitoring program. The emerging thoracic coarse particle monitoring program should include continuous monitoring technologies, where possible, and any speciation efforts should use consistent methodologies. Ongoing research may also yield important insights for designing a thoracic coarse particle monitoring program. The EPA's ORD recently awarded a series of grants to investigate the sources, composition, and health effects of thoracic coarse particulate matter. These grants integrate atmospheric measurements with toxicological and epidemiological investigations. ORD's in-house research program will also produce relevant research results to consider when designing a thoracic coarse particle monitoring program. Data from completed field efforts, such as the Detroit Exposure December 2008 23 ------- and Aerosol Research Study (DEARS)12 and FRM evaluations, should be analyzed to provide additional information about the intra- and inter-urban variability of thoracic coarse particles. Also, ongoing and planned studies, such as the Birmingham Saturation Sampling study and Near Roadway studies13, will also provide valuable information pertaining to thoracic coarse particles. Finally, ORD's in-house health research program includes thoracic coarse particle toxicological studies integrated with exposure and source apportionment analyses. While these research programs may produce results to inform decisions regarding thoracic coarse particle measurements (e.g., methods, monitoring locations, and components measured), the timing of the results will be critical since the initial thoracic coarse particle measurements will be required at NCore sites on or before January 1, 2011. Major Points Raised by Participants The following questions and the accompanying responses summarize the major points of this session: What lessons can be learned from the PMi.g chemical speciation network and applied to designing a thoracic coarse particle speciation network? • Monitoring Methods o Use consistent monitoring methods o Use continuous methods, where possible o Evaluate speciation methods now and conduct hypothesis driven pilots • Frequency of measurements o Consider conducting daily measurements at some thoracic coarse particle monitoring locations. • Archiving filters for future analyses What is the relative value of thoracic coarse particle mass versus speciation measurements? Initial efforts to monitor thoracic coarse particles should focus on mass measurements to inform our understanding of spatial and temporal variability. We need to learn more about the components of thoracic coarse particles and how composition varies across urban and rural areas, as well as, speciation measurement techniques before making significant investments in speciation monitoring. There is a significant amount of variation in thoracic coarse particle toxicity and we do not know enough to invest heavily in speciated thoracic coarse particle measurements at this time. What, potentially, are the most important thoracic coarse particle components to measure? Based on toxicology studies, metals, especially soluble metals, may play an important role in effects associated with thoracic coarse particle exposures. In addition, the role of biological materials in thoracic coarse particle health effects is unclear but it is also potentially very important. Analytical challenges limit our health-based hypotheses. Organics are quite different in thoracic coarse particles and difficult to measure. Furthermore, a significant mass fraction of many thoracic coarse particle samples is unknown/unidentified. 12 See www.epa.gov/dears for more information. 13.add reference/link December 2008 24 ------- What methods are available to measure thoracic coarse particles? There are several methods available for thoracic coarse particle measurements and we need to continue to evaluate these methods for measuring thoracic coarse particles. The current FRM for thoracic coarse particle is based upon the difference between PMi0 and PM2 5 measurements (using identical protocols and flow rates). Dichotomous instruments provide a separate measurement for thoracic coarse and fine particles. EPA studies show good agreement between the difference method and dichotomous measurements for both thoracic coarse particle mass and speciation, although additional analyses of speciation results from each method would be valuable. There are also continuous methods (e.g., FDMS Dichotomous TEOM) available for thoracic coarse particle measurement. In addition, other thoracic coarse particle sampling techniques may present opportunities for collecting more data for potentially lower costs. For example, the DRUM sampler may provide speciated size segregated measurements with a reduced operational burden. However, additional research is needed to evaluate the DRUM (Davis Rotating Uniform size-cut Monitor) sampler and compare results with other methods. Another promising monitoring approach uses passive techniques in combination with scanning electron microscopy (SEM) to provide thoracic coarse particle mass and speciation (elemental and morphological). These passive techniques can be deployed at relative low cost in a variety of locations which provides opportunities for improved spatial analyses of thoracic coarse particles. In what locations should thoracic coarse particle measurements be made? Consider monitoring in areas that are not in attainment for PMi0, but in attainment for PM2.s (see Table 1 below) or at least areas with higher PMio and lower PM2.5 concentrations. Analyses of data collected in these areas will provide insights into the thoracic coarse particle components or sources which are driving non-attainment (or high levels of thoracic coarse PM). Subsequent health studies can then provide information regarding the potential for health effects associated with exposure to these thoracic coarse particle components and sources. Rural locations should be included as well to improve understanding of differences between urban and rural thoracic coarse particle concentrations. At what height should thoracic coarse particles be monitored? Additional research is needed to understand the vertical profile of thoracic coarse particle concentrations and to identify the effect of monitoring height on thoracic coarse particle measurements (current requirements for thoracic coarse particle monitoring height range from 2 to 15 meters). How can the sources of thoracic coarse particles be identified? Research is needed to identify tracers or marker compounds for thoracic coarse particles for source apportionment analyses. Thoracic coarse particle sources likely fall into one of the following categories: direct emissions from mechanical processes that crush or grind larger particles (e.g., from industrial operations, construction and demolition activities, and agricultural and mining operations), resuspension of dusts (e.g., traffic-related emissions from tire and brake wear), biological materials, and secondarily formed aerosols. The influence of local sources will likely be relatively more significant for thoracic coarse particles than for fine particles. December 2008 25 ------- Why should State/Local/Tribal agencies be interested in thoracic coarse particle monitoring? Thoracic coarse particle monitoring in PMi0 non-attainment areas (particularly areas that are in attainment for PM2.5) could provide valuable information for State and local agencies, while also providing valuable information for air quality and health researchers that can be used in the design of thoracic coarse particle monitoring programs. What other tools are available for analyses of ambient thoracic coarse particle concentrations? Tools such as land use regression models, GIS, satellite data, and atmospheric dispersion models can supplement thoracic coarse particle monitoring data to provide enhanced information regarding spatial and temporal distributions of ambient thoracic coarse particles. However, it is important to note that there are uncertainties with outputs from these tools and while these tools were acknowledged at the workshop, detailed discussions regarding their potential application were beyond the scope of this workshop. The application of these tools may be included in future workshop discussions, as appropriate. Are there unique issues associated with thoracic coarse particle health studies? There are some unique issues related to thoracic coarse particle health studies. First, in vivo toxicology studies are more difficult because efficient animal models are not available, in part because inhalation toxicological studies are not possible in rodents. Differences in thoracic coarse particles typically found in rural versus urban areas also present challenges. Epidemiological studies in rural areas may not have enough statistical power. Finally, the role of exposures to biological components in health outcomes presents additional challenges. Are there existing data available to analyze? Various states, including Californian, New York, and Washington, have thoracic coarse particle data available for mass and composition analyses. For example, the state of California has collected and analyzed thoracic coarse particle measurements for spatial, temporal, and compositional patterns.14 In addition, data and results from previous research studies that included coarse particle measurements exist. One such study was conducted around major industries in southern Chicago.15 EPA also has thoracic coarse particle data available from previous field studies (e.g., FRM evaluations, source apportionment studies). Key science questions need to be developed for thoracic coarse particle monitoring and then the existing data sets need to be identified that could potentially be analyzed to address some of these questions. Recommendations/Actions for Consideration Building upon the general summary and major points above, the following is a list of recommendations expressed at the workshop or developed by EPA staff based on the workshop discussions. 14 Croes, B.E. (2003). Paniculate matter in California: Part 2 - Spatial, temporal, and compositional patterns of PM2.5, PM10-2.5, and PM10. J. Air Waste Manage. Assoc., 53(12): 1517-1530. 15 Watson, J.G., Chow, J.C.; Kohl, S.D.; Kuhns, H.D.; Robinson, N.F.; Frazier, C.A.; andEtyemezian, V. (2000). Annual report for the Robbins Paniculate Study - October 1997 through September 1998. Prepared for Versar Inc., Lombard, IL, by Desert Research Institute, Reno, NV. December 2008 26 ------- • When making decisions regarding investments in speciated thoracic coarse particle measurements, consider the following: o Analyze existing thoracic coarse particle data sets first. Such analyses could begin to address issues such as: • How does thoracic coarse particle composition vary across cities and in urban and rural areas? • To what extent do speciation analyses using samples collected via the difference methods differ from speciation of dichotomous samples? o Conduct targeted thoracic coarse particle speciation monitoring which is hypothesis- driven to decide what components to monitor and how. o Consider evidence generated from toxicological studies to inform decisions on which components to monitor. • When evaluating potential locations for thoracic coarse particle monitors, consider areas that are in attainment for PM2.5 but not for PMio. Speciated thoracic coarse particle monitoring in these locations may provide insights regarding sources that may be contributing to non-attainment. • Consider a thoracic coarse particle network design that includes a central site monitor collecting mass and speciation measurements with satellite locations that could potentially use alternative lower cost methods (e.g., passive methods). • Consider collecting daily thoracic coarse particle measurements at some subset of locations. • Encourage the use of thoracic coarse particle continuous methods, where possible. • Continue to evaluate thoracic coarse particle sampling and analytical methods. o Sampling methods to be evaluated include FDMS Dichotomous TEOM, DRUM, and passive sampling approaches. o Identify and evaluate potential methods to analyze biological components in thoracic coarse particles. • Incorporate thoracic coarse particle monitoring objectives in planned research field work. Examples objectives include: o Identification of marker compounds for thoracic coarse particle source categories. o Characterization of vertical distribution of ambient thoracic coarse particles and the associated implications for monitoring height. o Identification of composition and sources of unidentified thoracic coarse particle mass. • Consider application of land use regression models, GIS, satellite images, and atmospheric dispersion models in conjunction with ambient thoracic coarse particle measurements to conduct spatial and temporal variability analyses and to inform monitor locations. • Closely follow ongoing research for potential insights for thoracic coarse particle monitoring programs. The following research efforts may be particularly valuable: o EPA STAR Grants16 16 See http://cfpub.epa.gov/ncer abstracts/index.cfm/fuseaction/displav.rfa/rfa id/450 for more information December 2008 27 ------- Sources, composition, variability and toxicological characteristics of coarse (PMio-2.s) particles in Southern California (University of Southern California) Sources, composition, and health effects of thoracic coarse particulate matter (University of Colorado at Boulder) Cardiovascular effects of urban and rural thoracic coarse parti culate matter in African American and white adults (University of Michigan) Spatial investigation of sources, composition and long-term health effects of thoracic coarse particulate matter (PMio-2.s) in the Multi-Ethnic Study of Atherosclerosis (MESA) cohort (University of Washington) Comparative Toxicity of Thoracic coarse Particles (New York University) o EPA In-House Research17 Detroit Exposure and Aerosol Research Study (DEARS) Birmingham Coarse Particle Study Near Roadway Research Program Cleveland Source Apportionment Study Final County PMc Characterization Study (Region 9 RARE Project) 17 add a link to descriptions of these studies December 2008 28 ------- Table 1. PMio Non-Attainment Areas with PM2.s Designations (Source: EPA TTN Website) Area Phoenix, AZ Ckrk Co, NV Sacramento Co, CA Salt Lake Co, UT El Paso Co, TX Utah Co, UT Washoe Co, NV Eagle River, AK Coachella Valley, CA Eugene-Springfield, OR Imperial Valley, CA Mun. of Guaynabo, PR Yuma,AZ Ogden, UT Missoula, MT Bonner Co (Sandpoint), ID Butte, MT Nogales, AZ Sheridan, WY Paul Spur/Douglas (Cochise Comity), AZ Kalispell, MT Miami, AZ Juneaii, AK Shoshone Co, ID Ajo (Pima Comity), AZ Coso Junction, CA Owens Valley, CA Mammoth Lake, CA Hayden AZ Flathead County; Whitefish and vicinity, MT Poison, MT Columbia Falls, MT Trona,CA Lane Co, OR Anthony, NM Ronan, MT Pinehurst, ID Sanders County (part) ^Thompson Falls & vicinity,MT Fort Hall Reservation, ID Lame Deer, MT Rillito, AZ Mono Basin, CA Los Angeles South Coast Air Basin, CA San Joaquin Valley, CA New York Co, NY San Bernardino Co, CA Libby, MT Approximate Population 3,110,000 1,380,000 1,220,000 898,000 564,000 369,000 339,000 195,000 182,000 179,000 120,000 92,400 82,300 77,200 52,400 36,800 34,600 24,600 15,800 15,700 15,100 14,600 13,800 10,500 7,590 7,000 7,000 6,460 6,050 5,030 3,780 3,780 3,500 3,420 2,590 2,520 1,700 1,180 553 536 506 258 14,600,000 3,080,000 1,540,000 199,000 3,230 EPA Region 9 9 9 8 6 8 9 10 9 10 9 2 9 8 8 10 8 9 8 9 8 9 10 10 9 9 9 9 9 8 8 8 9 10 6 8 10 8 10 8 9 9 9 9 2 9 8 PM10 Non- Attainment Classification Serious Serious Moderate Moderate Moderate Moderate Senous Moderate Serious Moderate Serious Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Serious Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Moderate Serious Serious Moderate Moderate Moderate PM2.5 Attainment? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No No December 2008 29 ------- SESSION V: AMBIENT AIR MONITORING REALITIES - EPA/STATE/LOCAL PERSPECTIVES - SUMMARY AND RECOMMENDATIONS Background/Objectives This session was designed for EPA and State/local staff who manage monitoring programs to share their reactions (i.e., a reality check) to topics discussed in earlier sessions of the ambient air quality monitoring and health workshop. This included providing recommendations for addressing "low hanging fruit" as well as identifying significant challenges to making progress in ambient air monitoring to advance health research for the criteria and related (e.g., chemical speciation) air pollutants. Session Overview For this portion of the workshop a panel of experts in ambient air monitoring was assembled to provide their reaction to the papers, presentations, and discussions at the workshop. The panel included several staff and managers from EPA and State and local air agencies who are responsible for implementing and overseeing the operation of routine ambient air monitoring programs for air toxics, criteria, and other related air pollutants. Members of the panel included: • Dirk Felton, co-chair, New York State Department of Environmental Conservation • Tim Hanley, co-chair, EPA, OAQPS • Mike Gilroy, Puget Sound Clean Air Agency (Seattle, WA area local air agency) • Richard Payton, EPA Region 8 (Lead Monitoring Region) • Scott Reynolds, South Carolina Department of Health and Environmental Control • Eric Stevenson, Bay Area Air Quality Management District (San Francisco, CA area local agency) • Susan Zimmer-Dauphinee, Georgia Department of Natural Resources Major Points Raised by Participants The panel provided input and reactions on a range of topics from the workshop. Common themes across the panel members can be grouped into four categories: daily speciation sampling, PM2.5 continuous mass data, data management, and communications. Daily Speciation Sampling One of the major needs identified by health researchers is to have daily speciation in up to the largest 20 urban areas in the country. EPA currently works with State and local agencies to provide a Speciation Trends Network (STN) operating at midnight to midnight every third day at 53 locations around the country, plus an additional 120 locations identified as "supplemental speciation" that mostly operate on a sample schedule of every sixth day. Together the STN and supplemental stations comprise the Chemical Speciation Network (CSN). The CSN data together with collocated criteria pollutant gas measurements reported to AQS form the largest single source of ambient air pollution data used by the health community in researching the health effects of air pollution. December 2008 30 ------- Concerns The panel overwhelmingly supported the value of speciation data for use in health studies as well as to directly support State and local data uses such as source apportionment of fine particles and tracking control programs. However, the panel shared their concerns with the ability of State and local agencies to take on the workload and cost associated with daily sampling via filter-based speciation samplers. Even if funding were available to support the cost of the laboratory analysis, monitoring agencies would have a difficult time supporting the field operations due to frequent site visits and other logistical concerns. The panel identified that monitoring agencies were in some cases already proposing to cut back on existing monitoring systems due to diminishing resources via EPA grants and their own agency budgets. At the same time EPA's national contract laboratory has a pre-negotiated price increase each year. Therefore, even under the best of scenarios, which is usually flat funding, agencies are receiving less direct awards to support field activities every year. While the panel expressed their doubts that anything substantive could be accomplished through daily filter-based speciation, they did share thoughts on possible opportunities utilizing semi- continuous speciation methods. The panel also discussed the comparative uncertainty of health effect studies versus uncertainty in the ambient air monitoring data. Panel members suggested, and no one disagreed, that the uncertainty in the ambient air monitoring data was very small compared to the uncertainty in health studies. Health researchers pointed out that it was not just an issue of not having daily data to reduce uncertainty, but also an issue with interpreting the potential lag of health effects if daily characterization of particle species is not available. After the meeting, panel members suggested that through use of available PM2.5 filter-based mass, continuous mass, and filter-based speciation, reasonably good estimates of daily speciation could be derived (e.g., statistically interpolating chemical speciation on days 2 and 3 that the CSN sampler did not operate) for most major metropolitan areas. EPA could work with a group such as HEI to make such estimates available. Another option that could potentially support daily speciation is the rotating drum sampler, discussed in the white paper found in Appendix D entitled "Air Quality Sampling: Benefits and Costs of Daily Health Targeted Monitors for Fine Particle Components." The drum sampler was identified as a possible alternative measurement technology to support high time resolution (every six hours) of chemical speciation. The drum sampler measures various size classes of PM mass and PM components which would allow for a more comprehensive characterization of the sizes of PM by species. However, there is limited use of the drum sampler in applications other than special studies with anecdotal stories that it does not perform well. The panel expressed concern that the technology was not ready for routine application in State and local air monitoring networks, but were open to reconsidering the technology farther down the road if additional development and testing demonstrated that it would be easier to use than the current equipment and agencies would have at least the same confidence in the data as is available from the existing speciation sampling platform (i.e., the Met One SASS/SuperSASS for elements and ions). Even if the drum sampler advanced to the point of having comparable sampling performance to the existing CSN samplers, there remain two important issues that will need to be addressed. The first is the request made by health researchers to have a consistent method over the entire study period that they are researching and the second is ensuring data reporting to the AQS data system for integration with other ambient air monitoring data. December 2008 31 ------- Opportunities Acknowledging the value of highly time-resolved speciation data, the panel did express an interest in characterizing speciation on a daily basis utilizing some combination of semi-continuous and filter-based methods. Such an approach could be beneficial to both the health research community as well as supporting data needs of State and local agencies; which is very important in order to gain the support of senior management across monitoring agencies. The panel discussed how several agencies are reporting good results with semi-continuous methods such as the Sunset carbon analyzer, Aethelometer, and Thermo sulfate analyzer. Other semi-continuous speciation samplers exist, but have not been demonstrated to be comparable to filter-based methods or are so complex that data completeness and quality suffer due the instrument needing constant attention in the field. Utilizing some combination of semi-continuous carbon and sulfate characterization plus providing for elemental analysis via XRF analysis on daily Teflon filters may provide a reasonably complete characterization of the chemical speciation of most interest. The major species missing from such a protocol would be semi-volatile nitrate and organic carbon. A possible surrogate for the missing semi-volatiles could be the volatile channel of the Filter Dynamic Measurement System (FDMS) monitor, which provides highly time resolved characterization of stable and volatile PM. A protocol of semi-continuous and limited filter-based speciation for elements would be of interest to a number of State and local agencies to provide a more complete characterization of the source apportionment of their networks, especially on days above the NAAQS. Researchers are believed to value such a data set on all days to improve our understanding of potential health/welfare effects associated with ambient concentrations at or below the current standards. Review of existing collocated continuous speciation and filter-based method data (CSN or IMPROVE) available in AQS or from Regional or State monitoring organizations could provide an opportunity for health researchers to evaluate the potential value of semi-continuous speciated data and the need for filter based, co-located sampling. PM2.s Continuous Mass Data Several of the panel members commented on the availability of PM2.5 continuous mass data and potential value to the health research community. Some panel members were surprised to learn that these data were not being widely used in health studies. Across the country there are over 600 PM2.5 continuous mass monitors reporting hourly data with every major city covered by one or more stations. These data are stored and reported in near real-time through State and local agencies web sites and nationally through the AirNow program; see www.airnow.gov18. Long-term archiving of PM2.5 continuous data is provided for in the AQS data base. Acknowledging the variety of PM2.5 continuous methods used and their performance according to climate and measured aerosol components across the country, EPA set-up new AQS data storage protocols in 2006 so agencies could store their data in a way that allowed data users to better utilize the PM2.5 continuous mass data. Principally among the data storage protocols, data users can pull hourly data for the parameter "Acceptable PM2.5 AQI & Speciation Mass" (parameter code 88502). This parameter code is intended to represent the PM2.5 continuous mass data where the method meets performance criteria 18 www.airnow.gov provides illustrative maps of near-real time air pollution data and forecasts according to the EPA's Air Quality Index (AQI). These maps are intended for the general public and as such are color coded according to the AQI. Detailed near real-time air pollution and meteorology data from across the United States and Canada are available through www.airnowtech.org. This site is intended for technical users of the data. Health researchers or any other data user with a valid use for the information can request access to the site through the log-in screen on the web site. December 2008 32 ------- suitable for reporting the AQI19. In March of 2008, EPA-ORD approved the first Federal Equivalent Method (FEM) for PM2.5 (Met One BAM 1020). Data from this method will be stored under the parameter code "PM2 5 at local Conditions" (88101) as are all currently reporting PM25 FRM's, since these methods are approved for comparison to the NAAQS. All the available parameter codes (sometimes referred to as pollutant codes) for storing PM2 5 continuous data in AQS are provided below. Parameter Name PM2 5 LOCAL CONDITIONS PM2 5 TOTAL ATMOSPHERIC PM2 5 RAW DATA ACCEPTABLE PM2 5 AQI & SPECIATION MASS PM2 5 VOLATILE CHANNEL Parameter Code 88101 88500 88501 88502 88503 Purpose Code for all FRM's, FEM, and ARM's. Continuous FEM's will be stored with this parameter code. Data reported to this parameter code are generally eligible for comparison to the NAAQS. Valid data from methods measuring total PM2 5 aerosols in the atmosphere. FDMS is the method currently stored here Valid uncorrected data that does not meet DQO's for reporting at least the AQI Valid data that does meet the DQO's for AQI reporting with or without a correction or the mass data from the CSN network. Data reported to this parameter code are not eligible for comparison to the NAAQS. Store important related data such as the FDMS reference channel. ~Active Samplers/ Monitors 940 FRM's (-150 operate daily) 100 300 400 PM2 5 continuous mass; 200 CSN 5020 In total, there are -600 operating PM25 continuous monitors; some monitors dual report With the availability of data from a large number of PM2 5 continuous mass monitors, health researchers could utilize highly time-resolved data that can be used as a surrogate for the types of exposures in an urban area and combined with FRM data as a tool to better understand PM2 5 continuous mass measurements with a positive bias relative to the FRM. PM2 5 continuous mass data can be used to characterize different types of exposures by the time of day. For instance, weekday morning rush hour would be an indicator of automotive emissions, while early mornings in the winter might be an indicator of home heating (e.g., oil or wood smoke, depending on the neighborhood). In some cases, the measurement principle and time resolution of a PM2 5 continuous method results in data that have a positive bias relative to a 24-hour measurement on the filter-based PM2.s FRM. The positive bias with PM2 5 continuous methods are most likely associated with semi- volatile organics and nitrate that are not fully captured on the FRM due to evaporative losses that are exacerbated in warmer months when the sample filter is exposed at ambient conditions. These differences should be explored and included in analysis to health research data as the data are 19 Where bias is controlled to within +/-10% and correlation is at least 0.9 (R2 of 0.81) compared to collocated filter- based FRM's 20 Every FDMS monitor provides outputs of the volatile channel; however, some monitoring agencies have data management system limitations at their sample station; therefore only half the FDMS units (50 out of 100) in operation are reporting this channel. EPA will be working with monitoring agencies to improve reporting of this channel. December 2008 33 ------- already available and would help focus future research efforts on the most important particle species (e.g., to what extent do the semi-volatile organics have a stronger or weaker association with health effects?). Data Management The panel discussed several aspects of data management. One issue uncovered was the discovery that at least two major metropolitan areas were providing for daily sampling and analyses of the major fine particle species - elements, ions, and carbon (i.e., Los Angeles and Denver). The panel suggested that the applicable State and local agencies would highly value this additional speciation data being utilized by health researchers. To ensure data are utilized, it was suggested that, where appropriate, these important measurements be reported to the AQS data system. Specifically, in cases where health researchers are utilizing monitoring agency filters, plans should be made to load the data to AQS. In cases where health researchers are performing their own sampling, there would still be value in making the data available to a wider audience of users; however, monitoring agencies and health researchers would need to ensure data comparability with the existing network so other data users understand the usefulness of the data. Panel members offered that many agencies use EPA's AQS data system as the sole long-term repository of their data, but there were exceptions. While all routine State and local agency data make their way to AQS, in some cases agencies have their own long-term data record with unique or even routine measurements that are not in AQS. Health researchers should ask monitoring contacts about any such data when discussing availability of ambient air monitoring data with State and local agencies Panel members also offered that they can usually assist health researchers in the retrieval of data from their network, even when data are located on the AQS data system. Communications Each major stakeholder group (EPA health and monitoring programs, external health researchers, State and local air programs) has dozens of other groups that they communicate with. One of the goals of this workshop is to establish better communication between decision makers in routine ambient air monitoring programs and researchers that perform health effects studies that are used to inform NAAQS reviews. EPA, the routine ambient air monitoring programs, and the external health research are all motivated to invest time to improve communication that can lead to better use of ambient air monitoring data; however, all groups also have responsibilities to other stakeholders as well. So while monitoring programs and health researcher are making important strides to improve the use of data and therefore provide better products that can inform NAAQS reviews, each group also has a responsibility to other data users and clients of the health effects research studies. The panel suggested that there are already examples of good communications between health researchers and State and local air monitoring programs; however, communications could be improved. Developing a collaborative relationship is critical to maximize the benefit of both kinds of organizations. Health researchers will be able to better influence network decisions, and monitoring agencies will be able to better utilize health research results produced from their own ambient air monitoring system. Panel members cited examples where they had a good line of communication between health researchers and cases where they did not, even with their own sister December 2008 34 ------- health agencies in State governments. Panel members strongly support having health researchers review their annual monitoring network plans and to develop a line of communication as early as possible for identifying the most important monitoring stations being used in a health study. Where available, health and research community participation in plan review and development could be advantageous to all parties. Panel members suggested that they do usually attempt to accommodate the health research needs in cases where data are already being collected, especially if the researchers can identify the specific need and product in advance of a planned change in monitoring at a site. A couple of panel members commented on State/local review of grant proposals and how early involvement of air programs may help strengthen this review. For example, if EPA, the external health community, and State and local air monitoring programs already had a good line of communication, then research plans for a specific research study could be strengthened by making use of existing State and local monitors and data. Research resources could then be focused on augmenting the State/local data by targeting measurements and activities not planned for in the State air monitoring network. While this workshop has had a good deal of focus on what data are missing, there are actually some redundancies between monitoring data collected for specific health studies and State/local air monitoring programs. Future research work should focus on maximizing the use of the State/local monitoring networks and identifying the most critical "missing" data to collect. Some panel members commented on the wide variety of requests being made by the health community (e.g., lots of daily speciation, need for multiple speciation sites across a city, need for other measurements not currently being conducted such as ultrafmes, etc.) and suggested that it would be useful to have a prioritized strategic plan of health research needs for ambient air monitoring data. Such a plan should include commentary on the usefulness of co-located measurements such as pollutant gases and meteorological measurements. Recommendations/Actions for Consideration Daily Speciation Sampling Develop a protocol that utilizes a combination of semi-continuous and filter-based methods to characterize daily speciation. Ensure such a pilot has identified data users that can comment on the expected data quality so that if successful, these methods can be applied in other areas. Initially pilot this protocol in two or three major cities. Specific actions to develop hybrid semi-continuous and filter-based daily speciation protocol: • Inventory semi-continuous speciation methods operating across the country. Encourage entry of data into AQS where possible (OAQPS) • Perform data analysis and determine data quality on available co-located semi-continuous and CSN data. (OAQPS) • Evaluate intercomparability of Sunset carbon and Aethelometer data to CSN carbon data. Provide recommendations for relative value of each method to provide carbon data on days with no filter-based carbon sampling (ORD) • Develop analytical protocol for XRF analysis on PM2.5 FRM filters (ORD) o Test protocol by retrieving filters from a small number of monitoring agencies where there are co-located FRM and CSN data. Perhaps prioritize stations with collocated CSN sampler. o Analyze comparability of CSN and PM2.5 FRM Teflon filters December 2008 35 ------- o Provide recommendation on usefulness of XRF analysis on PM2 5 FRM filter on days when the CSN Teflon channel has operated and already has elemental assay • Develop laboratory protocol for additional carbon assay - Thermal Optical Reflectance - on PM2.5 Teflon filter (ORD) • Identify 2-3 test locations to pilot hybrid semi-continuous and filter-based daily speciation protocol. Prioritize based on both: o State/local partner(s) that will value daily speciation characterization for their program o Health research studies that are already underway, or soon to be underway, utilizing speciation data from an existing station Other related actions • Develop research action plan to address research questions from PMi0-2.5 chemical speciation Whitepaper (ORD) • Analyze data from co-located hourly PMio-2.5 by difference • Analyze data from FDMS Dichotomous monitor; focus on volatile channel • Evaluate potential for a urban PM-10 Sunset co-located with a PM-2.5 Sunset • Demonstrate long-term (1-2 years) successful operation of the rotating drum sampler and Synchrotron XRF analyses, including loading elemental data to AQS and comparability to CSN data (ORD). PM2 s Continuous mass data The availability of over 600 PM2.5 continuous monitors provides an opportunity to help health researchers utilize an important data set that might provide insights on the most important exposures of PM2.5 by time of day and averaging period. EPA has already set up data storage protocols in AQS so that data can be retrieved according to the performance of the PM2 5 continuous monitors and/or methods being utilized. EPA-OAQPS is actively working to store already reported PM2.5 continuous data going back to 200421 under these new storage protocols. Specific actions that may help facilitate better use of these data in health studies: • Provide recommendations on storage of PM2.s continuous mass data from years 2004 through 2006 under the appropriate parameter code so that data users retrieve data as expected (OAQPS). • Execute storage of PM2 5 continuous mass data under new parameter codes in AQS from years 2004 through 2006 (OAQPS and State/local agencies). • Develop long-term plan for including PM2.5 continuous mass data in next iteration of national ambient air monitoring dataset used by health researchers (OAQPS and F£EI). Data Management Availability of ambient air monitoring data to facilitate a two way communication of data can be improved by taking on the following actions: • EPA should include a discussion of data reporting expectations in the solicitation of any new long-term health studies where speciation data are involved. (ORD) 21 2004 was the first full year of national PM2 5 data reporting on AIRNow; the official launch date was October 1 of 2003. Prior to this date many agencies were just bringing their PM25 continuous data on-line, including developing statistical adjustments to have their data more closely resemble the PM2 5 FRM data. December 2008 36 ------- o In cases where the grant recipient will be utilizing methods consistent with the CSN network, the data will likely be highly valued, and provisions should be made to quality assure and report the data to AQS. • Localized health effects studies should establish a contact with the State or local monitoring agency to facilitate access to data that may not necessarily be reported to AQS. (Health Researchers) • Data leads on health effects research teams should be encouraged to contact State and local agencies for access to local data, even when the data are reported to AQS. (Health Researchers) Communications Many of the possible improvements to communication are already underway. For emphasis, the communication recommendations from the panel are provided here even if redundant or already underway: • EPA will facilitate a continued dialogue between health effects researchers and routine ambient air monitoring programs (OAQPS and ORD) o Invite health effects researchers to share their work at monitoring conferences o Invite leaders in the ambient air monitoring community to attend health effects research meetings o Engage the National Association of Clean Air Agencies (NACAA) at the Air Director level to emphasize the importance of this work (OAQPS and ORD) o Continue dedicated meetings between health effects researchers and routine ambient air monitoring programs. (All) o Encourage (State or PQAO) Air Monitoring Staff representation on science advisory boards for health effects research programs that have significant monitoring data needs. (Health Researchers, State/local agencies) • Facilitate input of health effects researchers into annual monitoring network plans22 (EPA Regions, OAQPS, ORD, Health effects researchers) o Maintain AMTIC web site with links to each State/local agencies' annual monitoring network plan o In early 2008, EPA ORD's National Center for Environmental Research (NCER) conducted a preliminary survey of EPA-funded epidemiology studies to develop an initial list of monitoring sites that are being used in current or planned health studies23. ORD will consider options for expanding/updating this preliminary list and ways to make it available to State/local air agencies when they are considering changes to their monitoring networks. o Health effects researchers should engage monitoring program early in the process when they have specific needs for continued data availability. Annual Monitoring Network Plans are due to the applicable EPA Regional Office by July 1 of each year. These documents represent the plans for monitoring in the subsequent calendar year and are subject to EPA approval. Each agency is required to make their plan available for public inspection for at least 30 days prior to submittal to EPA. See 40 CFR §58.10. An internet link to plans is available at: http://www.epa.gov/ttn/amtic/plans.html 23 See Appendix G. December 2008 37 ------- Develop strategic plan for health effects research ambient air monitoring data needs (ORD and HEI with input from OAQPS) o EPA and State/local agencies need to know the most valued data to help protect monitoring resources and guide future monitoring investments. Such a plan would be a guide for all monitoring investments including those by EPA-ORD, health effects researchers themselves and routine monitoring programs. Example topics to address include: • are Aethelometer BC data adequate? • is the UV channel useful? • are nitrate data necessary at all, or only in certain geographical areas? December 2008 38 ------- BACKGROUND MATERIALS December 2008 39 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Appendix A: Workshop Agenda and Participant List Ambient Air Quality Monitoring and Health Research: Workshop to Discuss Key Issues April 16 and 17, 2008 US EPA Main Campus, Building C-Auditorium, Research Triangle Park, NC EPA is interested in having an open dialogue with a small group of representative experts regarding health research priorities for ambient air quality monitoring data that could best advance our understanding of the impacts of air pollutant exposures on public health. In particular, we are interested in hearing comments and recommendations from experts on steps that could be taken to improve our understanding of the impact of fine particle components and other key air pollutants. These steps might include prioritization of monitoring sites and/or the designation of sites for more frequent monitoring. This meeting is another step in a series of interactions to foster improved long-term communication between air quality experts and health researchers. This communication is critical for ensuring that the ambient air monitoring program offers, and health researchers use, the best and most appropriate data possible to support the health research that serves as a foundation for EPA's reviews of the national ambient air quality standards (NAAQS). Primary Meeting Objectives • To discuss specific recommendations for concrete steps that EPA and other organizations could take in the ambient air monitoring program to advance health research for the criteria air pollutants. • To reexamine and assess progress to date on key issues identified at an earlier workshop sponsored by the Health Effects Institute (HEI) and EPA24 and in follow-up discussions with the EPA-PM Center Directors, HEI National Particle Component Toxicity (NPACT) Directors, and other researchers. • To provide constructive feedback on the following draft "white papers" developed to aid in a common understanding of the issues under discussion: • Chemical Speciation Network (CSN) - Carbon Issues • Access to EPA's Air Quality Data for Health Researchers • Air Quality Sampling: Benefits and Costs of Daily Health Targeted Monitors for Fine Particle Components 24 HEI and EPA co-sponsored a meeting in late 2006 to discuss how the use of the accumulating data derived from nationwide monitoring of fine paniculate matter (PM) components can facilitate current and future health effects studies and improve comparisons of risk estimates across studies. The workshop illuminated issues associated with accessing and analyzing monitoring data and identified needs of the health effects research community regarding monitoring of fine particle components. See http://www.healtheffects.org/AODNov06/AQDWorkshop.html for more information. A-l ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite • Long-term communication strategies for improving interactions between health researchers and air quality staff on changes in ambient air monitoring specifically: o Network Design and Site Selection Approval o Methods Implementation Expectations for Meeting Participants This workshop is designed to inform EPA staff plans for the criteria pollutant ambient air monitoring program to ensure that the most effective approaches for providing monitoring data to support health research studies are included. A large portion of the workshop agenda is devoted to discussion - the goal of which is to talk about recommendations for concrete steps that could be made to move the alliance between health and monitoring objectives forward. Thus, to maximize the effectiveness of the meeting, workshop participants will be expected to be familiar with background information distributed prior to the workshop, including draft white papers. The workshop discussions will need to be forward looking - to identify specific near- and long-term steps that EPA's health and monitoring staff, as well as external organizations and science communities, can take to improve the ambient air monitoring program to appropriately advance our understanding of the health impacts of criteria air pollutants. This workshop is designed to be an honest and objective endeavor to address health research needs, however, participants must also understand that EPA resources are, and will most likely continue to be, limited. Therefore, providing prioritization of recommendations for EPA and other organizations to consider is essential if we are to make some clear steps forward and, hopefully, build from anticipated initial successes. Wednesday, April 16,2008 8:30-9:45 Welcome/Introductory Remarks 8:30 - 8:45 Purpose of the Meeting/Overview of Key Issues/Summary of Progress Dr. Daniel Costa, EPA National Program Director for Air Research/ORD 8:45-9:00 EPA Program Office Perspective Ms. Lydia Wegman, Director, Health and Environmental Impacts Division/OAQPS Mr. Richard Way land, Director, Air Quality and Assessment Division/OAQPS 9:00 - 9:45 Air Quality Experts and Health Researchers Working Together: Why Communication is Critical - Stories of Success Dr. Morton Lippmann, New York University 9:45-10:00 BREAK A-2 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite 10:00-2:45 Session I: Elemental and Organic Carbon Measurements Background In 2007, EPA made changes in the monitoring network to address inconsistencies in carbon sampling and analysis procedures used in urban CSN (STN/SLAMS) and rural IMPROVE programs. Health researchers have repeatedly raised concerns to EPA regarding this methodology change and potential interruptions in monitoring data used for time-series analyses. This session will include: (1) a brief overview of what measurements are currently being made; (2) highlights from a recent CSN/IMPROVE: Carbon PM monitoring workshop with emphasis on issues of most interest to the health research community25; (3) approaches that are being evaluated for relating different data sets and the potential impacts for on-going epidemiological studies, and (4) opportunities to discuss steps that are being taken or could be taken to identify and address information gaps, including continuous carbon measurements. Background Information - Draft White Paper: "Chemical Speciation Network (CSN) - Carbon Issues" 10:00 - 10:20 Overview and Introduction to Key Issues Dr. Venkatesh Rao, EPA/OAQPS Dr. Barbara Turpin, Rutgers University 10:20 - 10:40 A Health Researcher's Perspective: What's So Special About Carbon? Dr. EdAvol, University of Southern California 10:40 - 11:00 CSN Carbon Monitoring Changes and Issues Ms. Joann Rice, EPA/OAQPS 11:00 - 11:20 Carbonaceous Aerosol Sampling Artifacts in the National Monitoring Networks Dr. John Watson, Desert Research Institute 11:20 -11:40 Transitions: Relating "Old" to "New" Methods Dr. Warren White, University of California-Davis 11:40 - 11:50 Predicting Carbonaceous Species Concentrations with Partial Least Squares Dr. Philip Hopke, Clarkson University 11:50 - 12:10 Impact of Method Transitions to Health Research Dr. Michael Hannigan, University of Colorado-Boulder 12:10-12:30 Air Quality Monitoring: Perspectives from East and West Mr. DirkFelton, NY Department of Environmental Conservation 25 See http://vista.cira.colostate.edu/improve/Publications/Workshops/Carbon Jan2008/CarbonMeeting2008.htm for more information. A-3 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite 12:30-1:30 LUNCH Session I: Elemental and Organic Carbon Measurements (cont.) 1:30-2:45 Panel Discussion Dr. EdAvol, Univ. of Southern California Dr. Michelle Bell, Yale University Dr. Judith Chow, Desert Research Institute Mr. Neil Frank, EPA/OAQPS Dr. Philip Hopke, Clarkson University Dr. Michael Kleeman, University ofCA-Davis Dr. Allen Robinson, Carnegie Mellon Univ. Dr. Warren White, University ofCA-Davis Suggested Issues for Discussion: • What types of measurement error are problematic for epidemiology? • Knowing this, what is of most concern (to epidemiology) given the CSN changes? Step changes in detection limits/precision? • Step changes in OC and EC (but not TC)? • Bias due to sampling artifacts? • Other? • Are past and planned measurement comparisons adequate: • to aid epidemiology study analyses? • to "harmonize" results from old and new methods? • How should blanks and sampling artifacts be handled? • What other types of carbon measurements are good candidates for examination in large epidemiology studies? 2:45-3:00 BREAK 3:00-4:30 Session II: Accessing Ambient Air Monitoring Data Background EPA's Air Quality System (AQS) is designed to collect and store ambient air monitoring information. EPA recently introduced the AQS Data Mart to facilitate access to this monitoring information. The AQS Data Mart is a generic "retrieval" tool that provides the ability to query any information, but it does not provide significant data exploration or analytic capabilities. These capabilities are left to the "analytical" tools. Various analytical tools, or interfaces, are available including HEFs Air Quality Database, which focuses on levels of PM2.5 components and gaseous pollutants at and near STN and SLAMS sites. This discussion will focus on data A-4 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite access issues and how to help health researchers obtain monitoring data for fine particle components and other critical pollutants more easily. Background Information - Draft White Paper: "Access to EPA's Air Quality Data for Health Researchers" 3:00 - 3:10 Overview and Introduction to Key Issues Dr. Bryan Hubbell, EPA/OAQPS Dr. Michelle Bell, Yale University 3:10 - 3:20 Overview of Draft White Paper: "Access to EPA's Air Quality Data for Health Researchers" Mr. NickMangus, EPA/OAQPS 3:20 - 3:30 Summary of Recent Data Summit Mr. Rich Scheffe, EPA/OAQPS 3:30 - 4:30 Panel Discussion Dr. Sara Dubowsky Adar, Univ. of Washington Dr. Kaz Ito, New York University Mr. John Langstaff, EPA/OAQPS Mr. NickMangus, EPA/OAQPS Mr. Richard Poirot, Vermont DEC Dr. Betty Pun, AER Mr. Rich Scheffe, EPA/OAQPS Suggested Issues for Discussion: In general, we see that access to ambient air monitoring data needs to support health research/assessments falls into four general categories: . epidemiological studies . exposure/risk assessments . public health surveillance . health impact assessments Keeping these broad categories in mind and understanding that the goal is to provide a framework for delivering consistent, well-documented monitoring data to users including the health research community, the issues discussed in this session will focus on data type/format, access, and context including: . What key data and formats do health researchers need access to? . How user-friendly are the data bases currently available to health researchers? What are the similarities/differences between the various data bases and how are they communicated? A-5 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite . What potential changes could be made to improve access to ambient air monitoring data? . What mechanisms are currently being used to communicate the limitations associated with the ambient air monitoring data? Are there specific recommendations for improving how the limitations are characterized? . How can we improve the understanding of the limitations, quirks, and context of the ambient air monitoring data and assist end users in determining the subsequent appropriate use of the raw data? 4:30 ADJOURN Thursday, April 17, 2008 8:00-8:30 Summaries of Day 1 Discussions and Comments on Draft White Papers/Next Steps 8:00 - 8:15 Session I: Elemental and Organic Carbon Measurements Dr. Barbara Turpin, Rutgers University 8:15-8:30 Session II: Accessing Ambient Air Monitoring Data Dr. Bryan Hubbell, EPA/OAQPS Dr. Michelle Bell, Yale University 8:30 - 10:15 Session III: Ambient Air Monitoring for Health Research Background EPA has been measuring fine particle components in urban areas since 2001. The network consists of 54 sites intended to capture long-term trends (Speciation Trends Network or STN) and approximately 150 other State and local air monitoring stations (SLAMS). Collectively the urban locations are part of the EPA Chemical Speciation Network (CSN). Due to cost considerations the CSN was reduced in 2006 from its original size of approximately 240 stations to its present size. Currently, fine particle components are measured at each location every third or sixth day. The Interagency Monitoring of Protected Visual Environments (IMPROVE) network, covering background sites in national parks and wilderness sites in addition to Washington, DC and the South Bronx in New York City, provide additional data of great value to researchers. Health researchers have requested EPA implement daily fine particle speciation measurements, however resource constraints have impeded any real plans to this end. This session will explore opportunities to obtain these types of data in critical locations to make true inroads in improving our understanding of the temporal variability of fine particle components in ambient air. The monitoring issue presently is fine particles (and components), but looming ahead is the issue of thoracic coarse particles - what steps can we make to prepare for this new data-source? A-6 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Background Information - Draft White Paper: "Air Quality Sampling: Benefits and Costs of Daily Health Targeted Monitors for Fine Particle Components" 8:30-8:45 Overview and Introduction to Key Issues Dr. Barbara Glenn, EPA/ORD/NCER Dr. Joel Schwartz, Harvard School of Public Health 8:45-10:15 Panel Discussion Dr. Robert Devlin, EPA/ORD/NHEERL Dr. Patrick Kinney, Columbia University Dr. Lucas Neas, EPA/ORD/NHEERL Dr. Roger Peng, Johns Hopkins University Dr. George Thurston, New York University Dr. Jay Turner, Washington Univ. in St. Louis Suggested Issues for Discussion: . Sources of Error - which are the most limiting? o For time-series studies that rely on air monitoring data collected every third or sixth day from a single (or a few) central site monitors, which major sources of error are the most important? Why? Which is the most important? • Uncertainties in exposure assessment associated with: > Missing days. > Spatial variation. > Monitor location. > Instruments measurement error or analytic methods. o Could exposure modeling to "fill-in" missing days adequately address uncertainties associated with every third or sixth day monitoring data? . If daily monitoring was going to commence in a few cities in the U.S., what is the best monitoring plan to study the relative health importance of PM components in the ambient mixofPM? o What are the best sites? Why? What site criteria are the most important? o What minimum number of locations for daily sampling is adequate to address a particular research area? o What components would you evaluate first? . Integrating previous or ongoing data collection to obtain retrospective data. o In some locations, FRM filters may have been archived and could be analyzed to learn more about daily variation of fine particles and components. o In some areas, data from continuous monitors for previous years is available. o In previous years, special studies or grant-funded studies have collected data on PM species. o How could these data be integrated to provide daily data for time-series studies? If data from different instruments or methods were combined to obtain a set of daily ambient concentration data for a city, would this introduce a significant source of uncertainty/error? A-7 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite o What components should be measured and what are the issues associated with their measurement and use in analyses? 10:15-10:30 BREAK 10:30-12:00 Session IV: Thoracic Coarse Particle Components and Potential Public Health Impacts Background On September 27, 2006, the U.S. Environmental Protection Agency (EPA) amended its national air quality monitoring requirements.26 As part of these changes, EPA and the states will add measurements of "thoracic coarse particles" (i.e. PMio-2.s) at 75 multi-pollutant monitoring sites (National Core or NCore sites). Some monitors will provide at least hourly measurements in near real-time (continuous mass concentration monitoring); while other monitors will sample the air over a 24-hour period and require laboratory processing of the sample (filter based sampling). Filter-based monitoring will enable development of PMi0-2.5 methods for chemical speciation of thoracic coarse particles. This session will explore criteria to consider as EPA adds speciation of thoracic coarse particles to the ambient air monitoring network. 10:30-11:00 Overview and Introduction to Key Issues Dr. Timothy Larson, University of Washington Mr. Timothy Watkins, EPA/ORD/NERL 11:00-12:00 Panel Discussion Dr. David Diaz-Sanchez, EPA/ORD/NHEERL Dr. Michael Hannigan, Univ. of CO - Boulder Dr. Philippine, SCAQMD Dr. Thomas Peters, University of Iowa Dr. Richard Flagon, CalTech Ms. Joann Rice, EPA/OAQPS Dr. Terry Gordon, New York University Dr. Jamie Schauer, Univ. of WI-Madison Suggested Issues for Discussion : . What is the relative value of coarse particle speciation at planned monitoring locations versus additional mass measurements? . What is the relative value of understanding Intra-urban versus Inter-urban/rural variability in coarse particle composition and spatial and temporal distributions? . What are your recommendations for coarse particle network design? 12:00-1:00 LUNCH 26 See http://www.epa.gov/oar/particlepollution/actions.html for more information on amendments to EPA's National Air Quality Monitoring Requirements. A-8 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite 1:00-2:30 Session V: Ambient Air Monitoring Realities - EPA/State/Local Perspectives Background EPA works with State, local, and tribal air agencies to design and implement ambient air monitoring networks to meet several monitoring objectives including: . Determining compliance with standards (i.e., the NAAQS) . Providing air pollution data to the general public on a timely basis . Supporting the development and tracking of the effectiveness of emission control programs . Providing input data for health and welfare effects and exposure research studies . Providing input data for health and welfare risk/exposure assessments conducted for NAAQS reviews . Measuring overall progress of air pollution control programs Opportunities are available for interested parties to provide comments on monitoring network plans to ensure input from health researchers and other interested users is considered in the design of these plans. Two draft white papers review the current processes for public comments on the monitoring network plans and changes to monitoring methodologies, respectively, as well as options for future efforts to improve communications with the health research community regarding ambient air monitoring networks. This session is designed for EPA and State/local staff who manage monitoring programs to share their reactions (i.e., a reality check) to topics discussed in earlier sessions. This may include providing recommendations for addressing "low hanging fruit" as well as significant challenges that may need to be addressed in order to make considerable progress in the ambient air monitoring program to advance health research for the criteria air pollutants. Background Information - Draft White Papers: "Ambient Air Monitoring Networks: Network Design and Site Selection Approval" and "Ambient Air Monitoring Method Implementation" 1:00 - 1:15 Overview and Introduction to Key Issues Mr. DirkFelton, NY DEC Mr. Timothy Hanky, EPA/OAQPS 1:15-2:30 Panel Discussion Dr. Philippine, SCAQMD Mr. Michael Gilroy, Puget Sound CAA Mr. Richard Pay ton, EPA/Region 8 Mr. Scott Reynolds, SCDHEC Mr. Eric Stevenson, BAAQMD Ms. Susan Zimmer-Dauphinee, GA DNR A-9 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite 2:30-2:45 BREAK 2:45-3:30 Summaries of Day 2 Discussions and Comments on Draft White Papers/Next Steps 2:45-3:00 Session III: Ambient Air Monitoring for Health Research Dr. Barbara Glenn, EPA/ORD/NCER Dr. Joel Schwartz, Harvard School of Public Health 3:00 - 3:15 Session IV: Thoracic Coarse Particle Components and Potential Public Health Impacts Dr. Timothy Larson, University of Washington Mr. Timothy Watkins, EPA/ORD/NERL 3:15-4:00 Concluding Remarks/Emerging Issues/Next Steps Dr. Morton Lippmann, New York University Ms. Lydia Wegman, EPA/OAQPS Mr. Richard Wayland, EPA/OAQPS Dr. Daniel Costa, National Program Director for Air Research, EPA/ORD 4:00 ADJOURN A-10 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Ambient Air Quality Monitoring and Health Research: Workshop to Discuss Key Issues - Participants April 16-17, 2008 Last Name Arnold Avol Baldauf Baxter Bell Brook Bucky Chow Costa Devlin Diaz- Sanchez Dubowsky Adar F el ton Fine Flagan Foley Frank Garbe Garcia Gilliland Gilroy Glenn Godleski Gordon Hall Hanley Hannigan Hansen Hassett-Sipple Holland Hopke Hubbell Ito Jenkins Katz Kim First Name Jeff Ed Rich Lisa Michelle Jeffrey Barbra Judy Dan Robert David Sara Dirk Phillip Richard Kristen Neil Paul Val Alice Mike Barbara John Terry EricS Tim Michael Craig Beth David Phil Bryan Kaz Scott Stacey Jee- Young Affiliation EPA National Center for Environmental Assessment University of Southern California EPANRMRL EPA National Exposure Research Laboratory Yale University Environment Canada EPA National Center for Environmental Assessment Desert Research Institute EPA Office of Research & Development EPA National Health & Environmental Effects Research Laboratory EPA National Health & Environmental Effects Research Laboratory University of Washington New York State Department of Environmental Conservation South Coast Air Quality Management District California Institute of Technology EPA National Exposure Research Laboratory EPA Office of Air Quality Planning & Standards Centers for Disease Control & Prevention EPA National Exposure Research Laboratory EPA National Exposure Research Laboratory Puget Sound Clean Air Agency EPA National Center for Environmental Research Harvard School of Public Health New York University EPA National Exposure Research Laboratory EPA Office of Air Quality Planning & Standards University of Colorado at Boulder EPA National Center for Environmental Assessment EPA Office of Air Quality Planning & Standards EPA National Exposure Research Laboratory Clarkson University EPA Office of Air Quality Planning & Standards New York University EPA Office of Air Quality Planning & Standards EPA National Center for Environmental Research EPA National Center for Environmental Assessment A-ll ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Last Name Kinney Kirrane Kleeman Kryak Lamason Langstaff Larson Lippman Long Lorang Luben Mangus Martin Mikel Mintz Mukerjee Neas Ozkaynak Payton Peltier Peng Peters Pierce Pinto Poirot Pun Rao Reynolds Rice Richmond Robarge Robinson Ross Sacks Schauer Scheffe Schultz Schwartz Sheldon First Name Patrick Ellen Mike David Bill John Tim Morton Tom Phil Tom Nick Karen Dennis David Shailbal Lucas Haluk Richard Richard Roger Thomas Tom Joe Rich Betty Venkatesh Scott Joann Harvey Gail Allen Zev Jason Jamie Rich Laurel Joel Linda Affiliation Columbia University EPA National Center for Environmental Assessment University of California-Davis EPA National Exposure Research Laboratory EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards University of Washington New York University EPA National Center for Environmental Assessment EPA Office of Air Quality Planning & Standards EPA National Center for Environmental Assessment EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards EPA National Exposure Research Laboratory EPA National Health & Environmental Effects Research Laboratory EPA National Exposure Research Laboratory EPA Region 8 New York University Johns Hopkins University University of Iowa EPA National Center for Environmental Assessment EPA National Center for Environmental Assessment Vermont Department of Environmental Conservation Atmospheric and Environmental Research, Inc. EPA Office of Air Quality Planning & Standards South Carolina Department of Health and Environmental Control EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards EPA National Center for Environmental Research Carnegie Mellon University ZevRoss Spatial Analysis EPA National Center for Environmental Assessment University of Wisconsin EPA Office of Air Quality Planning & Standards EPA Office of Research & Development Harvard University EPA National Exposure Research Laboratory A-12 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Last Name Stanek Stevenson Stewart Stone Sunshine Thurston Tikvart Turner Turpin Vandenberg Vanderpool Vette Watkins Watson Way land Wegman Weinstock White Williams Willis Wilson Winner Wyzga Zimmer- Dauphinee First Name Lindsay Eric Michael Susan Geoffrey George Joe Jay Barbara John Robert Alan Tim John Richard (Chet) Lydia Lewis Warren Ron Robert William Darrell Ronald Susan Affiliation EPA National Center for Environmental Assessment Bay Area Air Quality Management District EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards Health Effects Institute New York University EPA Office of Air Quality Planning & Standards Washington University in St. Louis Rutgers University EPA National Center for Environmental Assessment EPA National Exposure Research Laboratory EPA National Exposure Research Laboratory EPA National Exposure Research Laboratory Desert Research Institute EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards EPA Office of Air Quality Planning & Standards University of California-Davis EPA National Exposure Research Laboratory EPA National Exposure Research Laboratory EPA National Center for Environmental Assessment EPA National Center for Environmental Research Electric Power Research Institute Georgia Department of Natural Resources A-13 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Appendix B: Session I: Elemental and Organic Carbon Measurements - Chemical Speciation Network (CSN) Carbon Issues Questions on this draft white paper should be directed to Neil Frank, EPA/OAQPS, frank.neil(g)epa.gov, (919) 541-5560. Introduction The purpose of this draft white paper is to provide an overview of urban and rural carbon measurement protocols and to identify issues associated with data reporting and usage, particularly with respect to the CSN transition to IMPROVE-protocol for carbon measurements. This document will also serve as a discussion piece to gather input from the health research community on related issues and next steps. Background State and local air agencies, under EPA grants, have been measuring organic carbon (OC) and elemental carbon (EC) in urban areas since 2001. The network consists of 54 sites intended to capture long-term trends (speciation trends network or STN) and approximately 150 other State and local air monitoring stations (SLAMS). Collectively, the urban locations are part of the EPA Chemical Speciation Network (CSN). To support the regional haze and PM2.5 programs, EPA also funds a largely rural network called IMPROVE. Together with support from NFS and other Federal agencies, the IMPROVE network provides carbon measurements at approximately 160 national park, wilderness, and other rural locations nationwide. Two different thermal-optical analysis methods are currently used by the CSN and IMPROVE networks for the analysis of carbon. The IMPROVE method is based on the Desert Research Institute/Oregon Graduate Center (DRI/OGC) thermal-optical reflectance (TOR) method27. The CSN method has historically used a modified version of the National Institute for Occupational Safety and Health (NIOSH) 5040 thermal optical transmittance (TOT) method28. The latter CSN method is different from the NIOSH method in that it has a different thermal temperature profile. The CSN is transitioning to the IMPROVE sampling and analysis protocols for carbon. Currently 56 sites have changed. Additional sites will be changed in the future. Research has shown that differences in the thermal profile, optical correction (transmittance versus reflectance), and specific analyzer used will result in differences in the OC and EC values obtained (Schmid et al, 2001; Currie, et al., 2002). In addition, sampling and sample handling differences also have an impact, especially for OC. Other carbon monitoring networks and measurement studies (e.g. SEARCH, Supersites) use variations of the CSN and IMPROVE protocols and are not discussed here. For the first 6 years of CSN operation, urban and rural carbon have been collected with different samplers and analyzed by different thermal optical methods. For chemical analysis, CSN has used the NIOSH-type thermal optical transmittance (TOT) method for measures of OC 27 The current IMPROVE_A method for Organic and Elemental carbon is described by the Standard Operating Procedure (SOP) provided at: http://vista.cira.colostate.edu/improve/Publications/SOPs/drisop2005.asp 28 The CSN method for Organic and Elemental Carbon is described in the SOP provided at: http://epa.gov/ttn/amtic/files/ambient/pm25/spec/ocecsop.pdf B-l ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite and EC. IMPROVE measurements are based on thermal optical reflectance (TOR) with a different thermal evolution temperature profile. Both measurement protocols provide operationally-defined measures of OC and EC. The IMPROVE protocol generally results in a lower OC/EC ratio and also provides 4 sub-fractions of OC (OC1, OC2, OC3, OC4), pyrolized organic carbon (OP), and 3 fractions of EC (EC1, EC2, ECS). These sub-fractions have been used in source attribution studies (Kim and Hopke, 2006). Starting in 2005, however, IMPROVE switched to an upgraded TOR analyzer with more accurate temperature settings. This change to "IMPROVE_A" results in approximately the same total OC and total EC but relatively different amounts of the sub-fractions (Chow et al., 2007). The new IMPROVE analyzer also provides TOT measurements which may have value in relating the two networks' data. In addition, the two networks have their different approaches to address various sampling artifacts and public reporting of its carbon data. While CSN collects field and trip blanks (but with limited ambient exposure) at all sites as a measure of passively collected organic vapors (positive artifact), the IMPROVE program uses longer duration field blanks at all its sites, and additionally has deployed secondary carbon (sc), i.e. backup quartz filters, at 6 of their sites to provide a network-wide measure of sampling artifacts. IMPROVE data shows that sc is greater than field blank carbon (fbc) and IMPROVE uses the monthly median sc value (by carbon sub- fraction) to correct the entire network's organic and elemental carbon values. The adequacy of using 6 sites to represent the entire network is currently under review by IMRPOVE. IMPROVE reports publicly only the artifact adjusted data. EPA has been publicly reporting in AQS carbon data produced by the primary collection filters, and separately the carbon values for the field and trip blanks.29 All field and trip blank data since 1999 are now available in AQS. Until CSN sites are transit!oned to the new IMPROVE protocol, CSN sites are not measuring carbon on backup filters. To correct for urban sampling artifacts, CSN data users have used the CSN fbc data together with sampler specific flow rates to "blank correct" the reported CSN data and have also used material balance or statistical approaches (Frank, Solomon, Kirn). These sampling artifact procedures have also considered the need to differentiate particulate carbon collected on quartz vs. Teflon filters. Some users have not made any adjustments in some work (Pun). The correction for sampling artifact can be as much as 30% of the organic carbon, as reported at the 2006 HEI meeting in Boston (www.healtheffects.org/AQDNov06/AQDWorkshop.htmn. The new CSN samplers whose quartz filters are analyzed with "IMPROVE_A" are deploying sc and 24-hr duration fbc filters. The potential use of those filters to adjust for sampling artifact is currently under study. EPA is also exploring what adjustments can/should be applied to old- protocol-CSN data to best correct for sampling artifacts and whether a single universally acceptable approach or multiple approaches for artifact corrected data exists. A question to pursue is: What is the impact of using CSN data that have not been corrected for sampling artifacts in epi studies (i.e., inclusion of a large and variable positive bias which may possibly have a seasonal component)? To help understand the differences between CSN and IMPROVE carbon-protocol measurements, EPA has collocated CSN samplers with IMPROVE samplers in various urban and rural environments over a 1-3 year period (See Table 1). Because of the many separate influences on carbon measurements (e.g., sampler, specific analytical method, and artifact 29 Arifact corrected CSN OC data, using network average fbc values, are available on http://www.epa.gov/airexplorer B-2 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite correction), the only definitive data to show comparisons between CSN and IMPROVE are these data generated by the CSN and IMPROVE networks. From data analyzed to date, the results show that the IMPROVE-protocol EC is generally higher (+10 to +30%), except at 3 locations (Phoenix and Tonto, AZ and Rubidoux, CA) where the IMPROVE EC is lower (-2 to -8%). The average differences appear to vary by location and the difference may therefore be related to the type or composition of the carbon aerosol. On the other hand, OC concentration is greatly affected by sampling artifact, sampler flow rate and filter size, and therefore the inter-network differences are more difficult to characterize. Application of simple adjustments, say using field blanks, may not be sufficient to adjust CSN data to look like IMPROVE-protocol concentrations (Flanagan). Chow, Watson (at DRI) and White (at UC Davis) are also examining this issue for EPA and recommendations will be forthcoming. National consistency in carbon measurements for source attribution, model evaluation and urban-rural comparisons is very important. Starting in calendar year 2007, EPA began transitioning the urban CSN to the IMPROVE analytical protocol, with an IMPROVE-like sampler (i.e. URG3000N sampler, with identical PM2.s particle size separator, filter size and flow rate, but with mass flow control) and will be employing secondary filters and 24-hr duration field blanks to help estimate carbon sampling artifacts. Fifty-six sites have been established and produced two months of collocated data during May-June 2007. Preliminary analysis of these collocated data show similar IMPROVE-CSN relationships as discussed above. The transition of CSN will continue in two additional phases. Phase 2 will begin early 2008 with the conversion of about 65 sites and Phase 3 (the last phase of about 65 sites) will begin late 2008-early 2009. See http://www.epa.gov/ttn/amtic/files/ambient/pm25/spec/faqcarbon.pdffor more information on the conversion. EPA has also reorganized the parameter codes and data field definitions in AQS to better differentiate current and future carbon measurement data according to collection sampler, analytical protocol and adjustments if any for sampling artifacts. Closing the Gap in EC Monitoring to Support PM Health Effects Research The association between ambient concentrations of EC in PM2.5 and human health effects is a subject of considerable interest. This section describes a number of possible steps that could be taken towards minimizing the affects of CSN protocol changes on the ability of epidemiology projects to report useful results for consideration in planned periodic reviews of the PM2.5 NAAQS. The purpose of this section is to facilitate communication about next steps along the lines of these steps or alternatives that are more promising. In addition, the OAQPS Air Quality Assessment Division (AQAD) convened an in-person workshop of CSN and IMPROVE monitoring program experts and selected atmospheric scientists in January 2008 to discuss outstanding issues related to EC and OC measurement in the two networks. The participants in this workshop are continuing the discussion by e-mail and conference calls to develop a 1-2 year research plan, possibly leading to changes in the operation of one or both monitoring networks and/or the post-processing of their monitoring data. In order to improve the data usability of EC for epidemiological studies and subsequent PM NAAQS reviews, the following steps may be explored: Possible Data Analysis Steps 1. Relating Old CSN TOT and New CSN TOR Data: Available data should be rigorously analyzed to determine if there is a reasonably good method for predicting the former from B-3 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite the latter (including using sub-fraction information and possibly using site/day variables) or vice versa. Data that includes the effect of the sampler difference should also be analyzed. The purpose would be to determine whether double laboratory analysis of the sort described in Items 2 and 3 below is actually needed to close the time series discontinuity, versus relying on a mathematical conversion or algorithm. a. There are some data sets available now that can be used to address this question. Discussion Question: If new and old protocol CSN carbon data cannot be quantitatively related or adjusted so that a consistent time series is available for OC and EC, can epi studies use the unadjusted time series as long as it recognizes or accounts for the changes or intervention(s) in the measurement process? 2. Epidemiological Sensitivity Analysis: After some number (TBD) of months of doubly- analyzed samples are available from Items 5 and 6 below, epidemiologists should test whether the two physical measures of EC are similarly associated with health effects of interest. This would help determine the length of time and scope needed for items 4, 5 and 6 below. Discussion Question: Can sensitivity testing of epi models be done to explore whether the change in carbon measurements has a significant impact on results? Possible Pilot/Investigative Studies 3. Sampling Effects on EC measurements: Test the hypothesis that sampler model, artifact correction (subtraction of field blank or backup filter blank values), and quartz filter brand (Whatman QMA vs. Pallflex) have a small enough effect on measured EC, such that epidemiology studies can span discontinuities in these aspects provided there is consistency in the EC thermal analysis, by taking second punches from relevant filters already in cold storage. There are 517 filter pairs available from 56 sites in May and June 2007, each pair consisting of an old-CSN filter and a new-CSN filter. The two kinds of filters have already been tested for EC once each, with the old and new laboratory method respectively. Of these, 53 sites used the old CSN method with the MetOne sampler, which has the flow rate most different from the new URG3000N sampler (-6.7 vs. -22.7 L/min). The sensitivity of EC to sampler type (independent of lab analysis) can be tested by taking a second punch from the old CSN filters and analyzing them with IMPROVE_A, and comparing the results to the IMPROVE_A result on the new CSN filter. Alternatively or in addition, the comparison can be done the other way by taking the extra punch from the new CSN filter. This comparison may not be indicative of sensitivities during other seasons. 4. There are about 250 site-months of collocation data between old CSN TOT and IMPROVE TOR (up until January 2005) and IMPROVE_A TOR (after January 2005), spanning all seasons, in selected urban areas.30 Some of these filters could be analyzed a second time, as described immediately above to evaluate sampling effects on EC measurements. Discussion Questions: What are the most important data assessment attributes or metrics for the comparison to satisfy the needs for health studies? Can sensitivity testing of 30 Sites are in Atlanta, Birmingham, Allen Park, MI, Fresno, New York City, and Pittsburgh. B-4 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite epidemiologic models be done to explore whether the change in carbon measurements has a significant impact on results? Possible Gap-Filling Data Collection (Short and Long-term) to Extend the Continuity of the OC/EC time series 5. At some number (TBD) of converted CSN sites of most importance to ongoing epidemiology studies, analyze some number (TBD) of quartz filters from the new URG3000N sampler with the old CSN-TOT method, in addition to the IMPROVE_A measurements. Double analysis is possible because a single filter can usually allow three separate analyses using three separate punches from the filter. This would give a continuous time series using the old CSN-TOT lab method. There would be a discontinuity in sampling method. a. Same-time, double analysis would be implemented for newly collected filters as they are received.31 b. Filters collected since conversion that have completed analysis would be retrieved from cold storage and re-analyzed also. 6. (Additionally or Alternatively to Item 2 above) At some number (TBD) of converted CSN sites of most importance to ongoing epidemiology studies, retrieve pre-conversion filters from cold storage and perform a second analysis for EC using IMPROVE_A. This would give a continuous time series using the new IMPROVE_A lab method. There would be a discontinuity in sampling method. 7. Depending on the outcomes of steps 3 and 4, the number of sites subject to double analysis could be reduced (because no important differences are discerned) or increased to include more sites of interest (because it becomes clear that only consistent physical measurements are useful.) 8. Presently, EPA has no plan for long term operation of any sites at which the old CSN method (using the dominant old sampler type and the old TOT analysis protocol) and the new CSN method (URG300N and EVIPROVE_A) are collocated. A possible step is to establish some such sites and commit to their operation until these EC (and related OC) issues are well settled. Depending on logistics and monitoring agency agreement, these could be the same 6 sites where IMPROVE and one CSN samplers already operated on a collocation basis.32 Discussion Questions: What are the number and location of sites that are of most interest or importance to ongoing epidemiological studies? If long-term comparisons of old and new CSN are needed, where should collocated measurements be obtained, and at what frequency, and for how long? What are the most important data attributes or metrics for the comparison of new vs. old CSN protocol measurements to satisfy the needs for epi studies (e.g. sufficiently high correlation; consistent day-day and seasonal variability)? What is judged to be sufficiently high correlation; what is "consistent" temporal behavior? 31 It is not urgent to begin this same-time double analysis because any filters tested only with IMPROVE_A can be retrieved from cold storage later if necessary. 32 Sites are in Atlanta, Birmingham, Allen Park, MI, Fresno, New York City, and Pittsburgh. B-5 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite 9. As presented by Frank at the 2006 meeting in Boston, carbon by material balance between non-carbon species and FRM mass ("SANDWICH" technique) may have value in providing a consistent and independently derived time series of carbonaceous mass as measured on Teflon filters. This alternative indicator can minimally assist with quality control of newly derived procedures. To help isolate the OC and EC portions of the mass balance estimates, new measurements from archived Teflon filters may be needed to compensate for network changes in the thermal optical procedures used on collocated CSN measurements (e.g., optical measures of black carbon in combination with statistical procedures to establish site specific correction factor for "EC"). What About Daily EC? EPA ORD is investigating the feasibility of limited speciation on daily FRM collected Teflon filters, where available. This may involve performing XRF analyses and possibly an additional optical measure of black carbon in combination with statistical procedures to establish a site-specific correction factor for "EC". This work has not started and is not expected to be available by April 2008. ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Table 1. List of EPA CSN/IMPROVE Collocated Study Sites Site Name Atlanta (Decatur), GA Birmingham, AL Detroit (Allen Park), MI Fresno, CA New York (IS52), NY Pittsburgh, PA Houston, TX Chicago, IL Rubidoux (Riverside), CA Phoenix, AZ Tonto NP, AZ Seattle, WA Mt. Ranier, WA Washington DC Dolly Sods, WV AQS Site ID 13-089-0002 01-073-0023 26-163-0001 06-019-0008 36-005-0110 42-003-0008 48-201-1039 17-01-0076 06-065-8001 04-013-9997 04-007-0010 53-033-0080 53-053-0014 11-001-0042 54-093-9000 Urban/ Rural Urban Urban Urban Urban Urban Urban Urban Urban Urban Urban Rural Urban Rural Urban Rural CSN Sampler Andersen RAAS until 1/2006 then MetOne SASS MetOne SASS (URG3000N for carbon May 2007) MetOne SASS MetOne SASS R&P2300 until 1/2006 then MetOne SASS (URG3000N for carbon May 2007) MetOne SASS URG MASS URG MASS MetOne SASS MetOne SASS MetOne SASS URG MASS URG MASS Andersen RAAS Andersen RAAS Start Date 4/2004 4/2004 1 1/2003 9/2004 8/2004 4/2004 5/2004 1 1/2003 9/2004 10/2001 10/2001 10/2001 10/2001 10/2001 10/2001 End Date Ongoing * Ongoing Ongoing * Ongoing Ongoing Ongoing * 9/2005 9/2005 9/2005 12/2003 12/2003 12/2003 12/2002 12/2003 12/2003 * Continuing with IMPROVE carbon aerosol measurements starting July 2005. Full IMPROVE speciation at other sites. B-7 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite References Chow, J. C., Watson, J.G., Chen, L-W., Chang, M.C.O., Robinson, N.F., Trimble, D., Kohl, S., (2007) The IMPROVE_A Temperature Protocol for Thermal/Optical Carbon Analysis: Maintaining Consistency with a Long-Term Database, J. Air & Waste Management Association; 57: 1014-1023. Currie, L. A., B.A. Benner, Jr., J.D. Kessler, D.B. Klinedinst, G.A. Klouda, J.V. Maroif, J.F. Slater, S.A. Wise, H. Cachier, R. Cary, J.C. Chow, J. Watson, E.R.M. Druffel, C.A. Masiello, T.I. Eglinton, A. Pearson, C.M. Reddy, O. Gustafsson, J.G. Quinn, P.C. Hartmann, J.I. Hedges, K.M. Prentice, T.W. Kirchstetter, T. Novakov, H. Puxbaum, H. Schmid. (2002). A critical evaluation of interlaboratory data on total, elemental, and isotopic carbon in the carbonaceous particle reference material, NIST SRM 1649a. J. Res. Natl. Inst. Stand. TechnoL, 107: 279-298. Flanagan, James B., Max R. Peterson, R.K.M. Jayanty, and Ed E. Rickman. Analysis of PM2.5 Speciation Network Carbon Blank Data, Research Triangle Institute, Research Triangle Park, North Carolina 27709 (2003). http://www.rti.org/pubs/OCEC_flanagan_2003.pdf Frank, N. H., Retained Nitrate, Hydrated Sulfates, and Carbonaceous Mass in Federal Reference Method Fine Particulate Matter for Six Eastern U.S. Cities, J. Air & Waste Manage. Assoc. 2006, 56,500-511. Frank, N. H., Carbon Measurements and Adjustments. Presented at HEI/EPA Workshop on Air Quality Data in Health Effects Research , Newton Marriott Hotel, Newton, MA, November 30- December 1, 2006 http://www.healtheffects.org/AQDNov06/AQD_Frank.pdf Kim, E. , P.K. Hopke and Y. Qin (2005). Estimation of Organic Carbon Blank Values and Error Structures of the Speciation Trends Network Data for Source Apportionment. J. Air & Waste Manage. Assoc. 55:1190-1199. Kim, E. and P.K. Hopke (2006). Characterization of fine particle sources in the Great Smoky Mountains area; Science of the Total Environment: 368: 781-794. Pun, B. HEI Air Quality Database. http://hei.aer.com/aboutDatabase.php Schmid, Heidrun, Lothar Laskus, Hans Jiirgen Abraham, Urs Baltensperger, Vincent Lavanchy, Mirko Bizjak, Peter Burba, Helene Cachier, Dale Crow, Judith Chow, et al. (2001). Results of the "carbon conference" international aerosol carbon round robin test stage I. Atmos. Environ. 35:2111-2121. Solomon, Paul. Organic Carbon Artifacts. Presented at Mid-Atlantic Regional Air Management Assoc., June 29, 2004. http://oaqpswww.epa.gov/tom/wiki/pmteam/files/MARAMA PC Artifacts 6-29- 04 Conf Call.ppt ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Appendix C: Session II: Accessing Ambient Air Monitoring Data - Access to EPA's Air Quality Data for Health Researchers Questions on this draft white paper should be directed to Nick Mangus, EPA/OAQPS, mangus.nick@epa.gov, (919) 541-5549. Introduction A common refrain from policymakers, analysts, and scientists is that obtaining the air quality data which they need is a challenge. This paper outlines the current collection and dissemination framework for air quality data and poses "charge questions" to the health-research/ epidemiology community. The answers to these questions will help us at the EPA improve our offerings. To frame the charge questions, this document describes a relatively new EPA system, the AQS Data Mart, and contrasts it with the HEI Air Quality Database, which was put in place to provide access to PM components and other data for health researchers. Finally, the charge questions are presented. Background The collection, storage, and dissemination of air quality data is a complex process achieved by a series of separate groups of hardware, software, and people. As technology has advanced and the number of distinct sets of user groups (those with different data or analytical needs) have proliferated, the problem for any individual finding precisely what they need has only gotten more complex. Adding to this complexity are intermediate "value-added" providers who may integrate, visualize, or otherwise post-process data from various sources. Thus, users can invest in their own data gathering and processing or they can rely on an array of intermediary providers. We also have data from special studies. The quality is (probably) high, but the data may not be readily available to others. So, EPA will always be the provider of certain base data, but we may not have it in the desired form, integrated with other desirable data (emissions or population), or presented in the desired manner. There will always be the possibility for a value- added provider to enhance the EPA data or integrate it with other data. The following diagram is a simplified view of the components that accomplish the collection and dissemination tasks at the EPA. It will be used to explain how data are collected, stored, and provided by EPA and how the HEI acts as a value-added post-processor. C-l ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite AQS Data Flow Diagram Monitors / Analyzers (in the field) Data acquisition, LI aid analysis, and QA collection (State, tribal, and local agencies plus labs) facilities Inward & outward facing databases Rest of EPA AQS Application TIN Data Page AirData AQS Query Trends ! NAA Designations FOIAs & Requests j AQSDM Direct Interface Air Compare Software Application Process EPA partners Others The main part of the diagram shows the major components of the EPA's Air Quality System (AQS). Beginning from the left hand side, samples are collected in the field by monitors. Some of these samples are analyzed in situ, others are collected by the State, tribal, or local agency responsible for the monitor and analyzed at laboratories. Either way, the agency responsible for the monitor is also responsible for ensuring the measurements are reported to AQS. It should be noted that only monitors within the EPA national ambient air quality monitoring network must have their data reported to AQS, for other monitoring networks or special studies (e.g., The Texas PM2.5 Sampling and Analysis Study) it is optional and the information may be stored in another system (e.g., NARSTO). AQS is the EPA system designed to collect and store the monitored information. When users are allowed unlimited access to download information from such collection systems, the demands put on the system by voluminous requests can compromise the ability of the system to fulfill its collection function. To alleviate this problem, software engineers developed the AQS Data Mart which stores a copy of the information from the AQS and allows users to download data. It is a generic "retrieval" tool that provides the ability to query any information, but it does not provide significant data exploration or analysis capabilities. These capabilities are left to downstream "value-added" tools. EPA is in the process of transit!oning our user applications designed for downloading information from the AQS database to the AQS Data Mart database. The right hand side of the diagram represents the several places to query or download air quality information that EPA C-2 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite provides. Each has been targeted to a specific audience: the general public, data analysts, or researchers. The diagram indicates which ones are still connected to AQS and the ones that have been transit! oned to the AQS Data Mart. Note that the small cylinders by three of the systems still getting their data from AQS indicate that they must copy data and store it separately so as not to impose large loads on AQS. One of the advantages of using a data mart is to alleviate the need to store these data again. As an example, raw PM2.5 data collected by EPA is available to external users in three of these EPA "front-ends". Large text files can be downloaded from our website (The TTN Data Page at http://www.epa.gov/ttn/airs/airsaqs/detaildata/downloadaqsdata.htm). The AirExplorer site can be used to query, plot, and map these data. Finally, the Data Mart Direct Interface can be used to query the data. Each of these tools has advantages and disadvantages depending on the needs of the user. For more information about all of the front-ends listed in the diagram, please see Appendix A. Beyond AQS and the related EPA systems, there are many other stakeholders involved in the collection and dissemination of air quality data, each with their own activities and possibly systems. AQS is likely the largest repository, but there may be additional information of interest to health researchers stored in other places. These additional stakeholders are represented by the other "layers" in the diagram. Elsewhere in EPA there are data collection and dissemination systems (CASTNET and AirNow in the Office of Air and Radiation; RSIG and PHASE in the Office of Research and Development; and Environmental Geoweb in the Office of Environmental Information). Additionally, EPA has other systems that present public and management views of air quality data. The next layer out represents EPA partners, those who operate in cooperation with EPA, like the Health Effects Institute, Colorado State University, etc. who maintain data dissemination systems (many that integrate data from outside of AQS). Also in this layer are special studies (DEARS, NMMAPS, etc.) that manage the full lifecycle of air quality data management from collection to dissemination. Generally these non-governmental partners and EPA communicate with each other and the action that one takes may influence the other. Considering again the PM2.5 example, the HEI Air Quality Database uses the EPA provided data for PM and the nearest gas phase monitors, and integrates EPA emissions and non-EPA population and meteorological information. This is a value-added service to provide a custom-tailored solution to a specific community. Finally, there is the layer entitled "Others," which represents those stakeholders who operate independently. These are the "unknown unknowns" in terms of additional data that may be collected or made available. Each of these groups brings with them a different list of what they can do easily, what they can do with difficulty, and what they cannot do. That is, each provides a degree of flexibility or constancy that makes them the best at providing a particular product or service. Collaboration, building on the strengths of each organization, is critical and one organization may have to take up the role of integrator and communicator so the research community knows where to get vital information. That is, if a clearinghouse listing all available databases, datasets, and access systems is needed, someone will have to manage its creation and maintenance. C-3 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite The remainder of this paper discusses only one EPA access mechanism, the AQS Data Mart Direct Interface, which was designed specifically to address the needs of the research community. EPA perceived these needs as primarily the ability to locate and extract large sets of data. The Data Mart was made available for internal EPA use in mid-2006 and for external use, along with the Direct Interface, in early 2007. Use has been growing steadily since then. Overall, it has been well received by most of those who have accessed and used it. Initially a pilot project, the reaction from users has been positive enough that EPA management has committed to ongoing support for the system. Most of the negative reaction falls into two categories: the user friendliness of the system and the documentation of the data. To address the first, we continue to add features and improve usability to make the Data Mart as friendly as possible to the research community. Documentation of the data is not a problem inherent to the Data Mart, but we realize it is much needed, so we are also addressing this as we can. The remainder of this paper will introduce the Data Mart Direct Interface, compare it to the HEI Air Quality Database, and place "charge" questions to the research user community to help us continue to improve these systems to meet your needs. Contents of the Data Mart The Data Mart contains every measured ("raw") and aggregated ("daily and annual summary") value reported to AQS from January 01, 1980 to the present. It also contains all of the same site and monitor descriptive data and measurement metadata in AQS. We have converted most data-entry codes to plain English words to help with the interpretation of downloaded data. There are no additional quality assurance steps performed on the data in the Data Mart, as the data in AQS are generally considered to be of the highest quality. Data must undergo many quality control steps as part of the loading process before it is saved in the AQS database. Likewise, submitters are required to assure that the monitor is operating properly and has passed precision and bias checks before loading the data. Finally, each year, EPA and the submitter review the data for completeness and correctness before the data are "certified" for regulatory use. It should be noted that IMPROVE (visibility network) and SANDWICH (modeled PM2.5 species) data are not generally reported to AQS. However, EPA staff has recently loaded the IMPROVE data for 1988-2005 into AQS and the loading of SANDWICH data is planned. As of January 14, 2008, there were 1.67 billion raw measurements for 885 different parameters in the database (there is a profiling spreadsheet under the documentation section of the web page). The Data Mart is refreshed from AQS each weekday night, so it always has the latest available information. However, since data up to 4 years old can be submitted to AQS at any time, and there are special windows for "historical" data updates, any of the contents can change at any time. That is, there is no freezing or snapshotting of data into a static version in the database. C-4 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Accessing the AQS Data Mart The AQS Data Mart can be accessed by visiting the webpage, http://www.epa.gov/ttn/airs/aqsdatamart, and following the "Access" link. Registration is required, and a user ID and password needed for access. You may sign up for your own account or use a guest account with user = aqsdatamart@epa.gov and password = AQSdatamartl (case sensitive). Access is provided by an application that you can either run in your web browser or download and run on a PC. The application is used to submit a query. A query lets the user select the geography, substance (parameter), time, metric, and optional data to return. The Data Mart currently has five queries, summarized below. Query Values Monitor Annual Summary Raw Data Sites by Threshold Description Recommended, returns any single raw, daily, or annual variable with metadata and is very efficient Returns descriptions of the monitoring site and equipment Returns all annual summary aggregate statistics for the monitors selected Returns raw data in the AQS transaction format - recommended only for AQS users Returns a list of sites that meet a specific data-related threshold that you specify When the query is complete, results can be downloaded using the application or by following a link in an email message sent to the user. All output is in XML format, but with embedded links to stylesheets for user-friendly display. The Data Mart is intended as an extraction system only and EPA does not plan to provide analytic or graphical capabilities with the Data Mart. However, some of the other tools that EPA provides do have these capabilities (see Appendix A for details). Contents of the HEI Air Quality Database In September 2005, a group funded by the Health Effects Institute (HEI) and led by Christian Seigneur and Betty Pun at Atmospheric and Environmental Research (AER) launched a website/database to facilitate health effects studies that require detailed knowledge of air pollutant levels and other relevant information at selected sites across the US. The HEI Air Quality Database combines information on PM2.5 components collected at monitoring sites in the Chemical Speciation Network (CSN); meteorological variables; and levels of gaseous pollutants (SC>2, Os, NOx, and CO) from monitoring sites at or near each CSN site. Metadata are provided for each monitoring site, such as its geographic coordinates, state, as well as county, city location information, population, and emissions data for nearby point, area, and mobile sources. AER updates information in the HEI Database every few months and is currently funded to do this through 2008. Accessing the HEI Air Quality Database The HEI Air Quality Database can be accessed by visiting the webpage, http://hei.aer.com. Once you obtain an account by following the instructions on this page, you can access the site browser and list building, database queries, and users' guides. The general C-5 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite data retrieval process consists of four steps: browsing sites, defining and saving a list of sites, extracting data for the sites in a saved site list, and, downloading the extracted air quality data. Comparison of AQS Data Mart and the HEI Air Quality Database The HEI Air Quality Database represents a value-added service over what EPA provides for a scientist looking for specific speciated PM2.5 data to evaluate in health research studies. So, a natural starting point for such a user would be the more tailored FIEI system. If, however, that system does not have some particular information that the user needs, they can revert to using the EPA system. The EPA system is broader, but less refined; the closer the user gets to the source, the more raw material they must process to get a finished product. The following table compares some of the features of the HEI Air Quality Database and the AQS Data Mart to illustrate some of these trade-offs. Feature Site browser Site finder Query from saved list Query by any geography Air quality data for PM2 5, O3, CO, NOx, NO2, & SO2 Air quality data for all other parameters AQS met data Integrated non-EPA met data Emissions data Census data On-line help Off-line help File format Data returned in one file Update frequency (versions) Build your own query HEI Air Quality Database Yes, with maps to help Yes, with multiple-variable filter Yes. Station lists may be saved and re-used Yes Yes No Yes Yes Yes Yes Yes Yes CSV No Quarterly Yes AQS Data Mart No Yes, via a single-variable "sites by threshold" query No. Query based on geography and parameter or single site Yes Yes Yes Yes No No No No Yes XML (CSV planned) Yes Daily No To summarize the key differences: . The HEI interface is more tailored to the PM2.5 analyst. . The HEI interface contains emissions, census, or NCDC meteorology data, the Data Mart does not. . The Data Mart contains all ambient data reported to AQS (not just PM, meteorological, and NAAQS gases). C-6 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite . The Data Mart only contains special studies data (e.g., supersites) if it has been loaded into AQS. Interpreting the Data Between data element names, report headings, and data transfer formats, there are almost 2000 named data elements relating to air quality that EPA makes available. In addition, some of the values in those fields need individual documentation to properly describe them (for example, what is the difference between a SLAMS and a NAMS monitor type). To help the user identify and perhaps understand the data they have, the EPA created an annotated, cross-referenced index called the "Field Guide to Air Quality Data". It is available in the documentation section of the Data Mart web page. There is also a list server that can be used to ask questions or monitored for system status. Charge Questions - Introduction To help prioritize and define future activities so that we can better meet the needs of the members of the research community, EPA has compiled a list of "charge questions" for invitees to this conference to consider. The overarching issue is connecting the data users to the data providers. For EPA and our partners to improve on this, we need to fully understand the data needs of the health research community. The more specifically the needs can be elucidated, the more concrete actions that can be taken to improve the situation. We are interested in feedback from users and potential users of air quality data and retrieval tools. This paper is concerned only with access to existing data; possible new data collection activities are covered elsewhere. Standout Charge Questions In previous interactions with data users and the health research community, three questions repeatedly come to the forefront as seemingly ubiquitous and critical. These issues are also at a high level and decisions on them will potentially impact decisions on the other charge questions. To complicate matters, there is not a single unifying idea that all agree is progress in the right direction on these issues. Thus, these questions are presented in more detail and with possible solutions to initiate discussions. 1. Data versioning/snapshotting: How often should EPA release data and how should we indicate that it has changed? The EPA, HEI, and others currently provide data via many applications. The data in those applications are generally updated on a schedule or as new data become available. For example, the AQS Data Mart is updated every day with new submissions and changes to AQS. However, new data or changes coming into AQS may be 10 years old. So a value in the AQS Data Mart representing a sample taken in the late 1990s may change today. Likewise, the HEI Air Quality Database is generally updated as the EPA makes new AQS "flat file" data extracts available on our web sites. This is usually done quarterly and without notice, thus the HEI database changes about quarterly; and the same 10 year rule applies. The key difference is that if you get data from the AQS Data Mart and your colleague gets the "same" data the next day, the data may have changed. If you are using the HEI database, the data may also have changed in one day, but the odds are less and the data vintage is clear in the "about" pages of the website. The stability of data for verifying and comparing research is essential, so the charge question is this: How often should EPA release data and how should we indicate that it has changed? One solution to this issue is to only make new data C-7 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite available outside EPA once per year. These data would be released on Independence Day and would be up-to-date through the prior year. This option provides greater stability to the data but may not be timely enough for particular studies or NAAQS revisions. A second solution is for EPA to continue to release data as it is received. Each value would be date-stamped with the date it last changed along with the date it represents. This allows for comparisons of data sets but requires more data to be downloaded and analyzed by the user. There are many intermediate options that could be implemented. 2. Topic-focused portals: Are topic focused portals needed for air quality data? If so, what should those portals be and what should they contain? A strength of the HEI Air Quality Database is that it is geared towards health researchers evaluating speciated PM2.5 data and the user interface provides tools and information specifically targeted to this user. The AQS Data Mart, on the other hand, is generic and targeted at anyone wishing to download air quality data. An annotated map of the PM2.5 speciation sites on the HEI page helps the user understand and find the data they need. An analogous map of all 5,000 sites represented in the AQS Data Mart would only overwhelm and confuse users. Custom tailored "portals" into data, like HEFs, are very helpful to the user, especially when they have an interest limited to less than everything available. The EPA is reasonably good at providing data but is often constrained in the technology we can use to provide descriptive and analytical tools. Likewise, we are sometimes not able to quickly secure funding to add tools to respond to developing areas of interest. This may be a place where the flexibility of external organizations can be used to provide a more custom, and therefore useful, experience. Are topic-focused portals needed for air quality data? If so, what should those portals be and what should they contain? For example, there could be portals specific to PM2 5 speciation, ozone and precursors, toxics, organic compounds, etc. Given the new technologies, a portal that resides outside of the EPA can have live access to a single, consistent, stable database within EPA. 3. Accessibility of non-AQS data: The AQS Data Mart stores data from the national ambient air quality monitoring network(s) and, as previously mentioned, has recently begun to add some data from other networks and "special studies." Is it important to have access to data from local, short-term, air quality special studies? Examples include MESA-Air, DEARS, Supersites, and ultrafme particle projects. If these data should be included, how should it be done? For example, to be loaded into the AQS Data Mart the data must match the monitor paradigm (no remote sensing or mobile monitors), it must meet format and quality requirements, and it must have associated descriptive data (e.g., method used, sampling schedule). Getting new data to match EPA's data standards are often labor-intensive activities - are they worth it? Would EPA have to correct and load these data into the AQS database (or a "research" copy of the AQS database)? Would EPA be able to commit the resources to doing this? As an alternative, EPA can provide information to managers of new studies about the data format and content standards we have so that the data can be collected in a way that could be more easily shared and compared with AQS data or other new data collected using the standards. If this special study data remains outside of EPA systems, is there a role for a clearinghouse? The clearinghouse could keep an up-to-date list of monitoring efforts, ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite databases, contents, and appropriate uses. Issues to be considered include: how resource intensive would this effort be and who would develop and maintain this clearinghouse? Other Charge Questions The remaining charge questions are more straight-forward than the standout charge questions. They are related to how individuals gather and use data rather than community-wide concerns. 4. What are the key data that you need? Is any of this currently collected but not available? 5. Is there a particular way that you need data organized, grouped, or formatted? 6. What data elements other than measurements do you need? 7. What is the typical domain of the data you need (time, space, and parameter selections; for example, 3 years, several cities, and 4 parameters; or 1 year, national, 44 parameters)? 8. Are there "profiling" reports - descriptions of which sites collect which data, how complete the data are, etc. - that you need? 9. Would you rather query a database or have a large list of files that you can select from to download (like http://www.epa.gov/ttn/airs/airsaqs/detaildata/downloadaqsdata.htm but with more geographic resolution)? 10. What would your ideal query builder/interface look like? 11. Are there pieces of data that we provide or questions that we ask that confuse you? C-9 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Appendix C.I- Other Data Access Mechanisms EPA has many places to access air quality data. Each of these websites or applications was designed for a specific target audience, for example, the general public concerned with acute health issues, the general public concerned with long-term air quality where they live, the general public interested in air quality comparisons between multiple locations (for living, vacationing, etc.), data analysts concerned with regulatory compliances, data analysts contributing to policy decisions, and health researchers. We consider a researcher to be someone who is looking to download raw data; either in large volume or in small, discrete sets that are difficult to tease out of large published datasets. Each of these websites or applications presents a unique front end for queries, charts, or maps that are geared toward their target audience. EPA is developing a "portal" to list all of the sources of air quality (and emissions) data that are available and link directly to their access pages. This portal is at the following web address: http://www.epa.gov/oar/airpolldata.html. Below is a table comparing key information about each of the available EPA-maintained systems for air quality data (including AirNow and CASTNET which contain data not in AQS). The systems are described at the link above. (Key: a filled circle means "yes" and an empty circle means "some".) System AQS AirNow (Tech) AirData AirExplorer AirCompare AQS Data Page NATA (modeled) Air Trends AQS Data Mart CASTNET Level of Detail ^.^ C/5 o* •^ z ^^ 4J s. s £ sS • • • • • t» k w Q^ B s sS Q • • • • • > * QQ % S g "3 s ^ • • • • • • • • S . Time •a 0) "c4 •a S o & a 2 S5 .5 -5 H -a 1994 - Present 1999 - Present 1996 - 2006 1996 - 2006 2000 - Present 1994 - 2006 1996 & 1999 1990-2005 1980 - Present 1987-2005 Substances • fS S PH + 0) o (S . . . . . . . . 0 CS 0) -*^ U •_ 0) J 6 • 0 . . . . . . 'B • o • • | t/3 ••? "1 PH • ^ 8. "7 sS a >r> PH • • • • 0) j 6 • • • Outputs j. sS S -2 K 0) •c 0) c- •a 0) a sS U • • • • • • • • • • s. CS S 0) •c 0) c- •a 0) e sS U • • • • • • u o •a • Audience ;3 PH • • % X >: 13 ^ H -J !/5 • • • • • • C 0) A ^ S 1 . . C-10 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Appendix D: Session III: Ambient Air Monitoring for Health Research - Air Quality Sampling: Benefits and Costs of Daily Health Targeted Monitors for Fine Particle Components Questions on this draft white paper should be directed to Dr. Barbara Glenn, EPA/NCER, glenn.barbara@epa.gov; (202) 343-9721. INTRODUCTION EPA's air quality datasets are generally recognized and valued primarily for their use in ascertaining compliance with the National Ambient Air Quality Standards (NAAQS), developing State Implementation Plans for the improvement in air quality, and providing timely air quality data to the public. EPA's air quality datasets also are essential to extramural and intramural health research addressing scientific uncertainty related to the current NAAQS and to the assessment of the possible health benefits of any new air quality standard. These health-related uses of EPA's air quality datasets are an important consideration in the design and conduct of the national air quality monitoring network. As specialized monitoring networks have begun providing information on the composition of particulate matter, epidemiologic researchers are striving to address a major research priority defined by the National Academy of Sciences National Research Council (NRC)—assessing the health effects of PM components and sources. The NRC reports on "Research Priorities for Airborne Particulate Matter" repeatedly emphasize the importance of research to assess the relationships between particle composition and health responses. According to the fourth report, "Progress on assessment of hazardous PM components is central to the national research portfolio and to any refinement of the current mass-based NAAQS for PM... .A better understanding of characteristics that modulate toxicity could lead to targeted control strategies specifically addressing those sources having the most significant adverse effects on public health." (NRC 2004) On November 30, 2006, the Health Effects Institute (HEI) and EPA, in conjunction with the annual EPA PM Centers meeting, convened a meeting of the research and air quality management communities to discuss the use of EPA's air quality datasets for health research on particulate matter (PM). Participants raised several issues that complicate the design and interpretation of epidemiologic research on PM2.5 mass, components, and sources. Participants emphasized that the lack of daily concentration measurements for fine particle mass and components in key locations was severely affecting their ability to design and conduct epidemiologic studies that would address issues of scientific uncertainties highlighted in air quality standard setting at Federal and state levels. This paper summarizes these issues, proposing a range of options to address the need for daily data based on the November 2006 meeting and subsequent information exchange with EPA grantees, state/local air quality monitoring representatives, HEI and EPA staff. These challenges would not exist if resources to collect daily, speciated monitoring data were readily available. The pressing needs for these data are increasing at a time when resources are decreasing and monitoring costs are generally increasing. This draft white paper provides background information to facilitate a broad discussion on the benefits of obtaining daily fine particle speciation measurements and to D-l ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite encourage health researchers and air quality experts to work together to creatively identify solutions that address the need for daily data while understanding the resource constraints and competing needs for monitoring data. The goal of this discussion paper is to draw attention to the importance of specific monitoring data needs in planning health research studies, specifically related to evaluating potential public health impacts of fine particles. Epidemiologic studies relating daily variation in ambient air pollutant concentrations with disease-specific mortality or morbidity have been very important for providing the scientific basis for recent standard-setting for PM. Some of the key policy relevant issues considered in evaluating the PM NAAQS include: • What are the potential public health impacts associated with exposures to specific size fractions, chemical components, sources and/or environments (e.g., urban and non-urban areas) of PM? • What is the relationship between various health endpoints and different lag periods (e.g., less than one day, single day, and multi-day distributed lags)? • How does spatial and/or temporal heterogeneity of PM exposures vary with different size fractions and/or components? Providing daily ambient air monitoring data for fine particle components from several cities to health researchers would reduce exposure misclassification, allow the use of all health events in statistical analyses, and thereby increase the precision of risk estimates. In addition, the availability of these data would significantly decrease the length of time necessary to produce study results. BACKGROUND Epidemiologic studies of the adverse human health effects of short-term exposures to air pollutants have generally relied upon air quality monitoring systems established to ensure compliance with ambient air quality standards. These epidemiologic studies contributed to decisions in 1987 to change the indicator for the PM NAAQS from total suspended particles (TSP) to PMio and to decisions in 1997 to add new standards to consider fine and coarse fractions of PMio separately, using PM2.s as the indicator for fine particles and using PMio as the indicator for purposes of regulating thoracic coarse particles.. As regulatory efforts have increasingly focused on reducing the mass of fine particles from combustion sources, the air quality monitoring network has successfully responded, at considerable cost and human effort, to the monitoring challenges. Since promulgation of the fine particle NAAQS in 1997, subsequent epidemiologic and toxicologic research has confirmed the earlier scientific findings and validated the substantial investment in ambient air monitoring. The 2004 Air Quality Criteria Document (CD) for PM highlighted the importance of epidemiologic studies in its evaluation of the scientific evidence. In particular, the CD emphasized new multi-city studies that investigated the effects of short-term human exposures to PM on mortality and morbidity using data from multiple locations with varying climate and air pollution mixes. These epidemiologic studies were valued because they provided information about areas not previously studied, reported risk estimates for all study locations, and used the D-2 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite same analytical approach at each location allowing comparisons. In addition, multi-city studies contributed to an increased understanding of the role of various potential confounders, including gaseous co-pollutants, on observed associations. These studies, which combined risk estimates across all locations, provided more precise estimates of the magnitude of an effect of exposure to PM than most smaller-scale individual city studies because of their larger sample size. Because model results were reported for all study locations regardless of the magnitude of the observed risk estimate, these studies also avoided the potential for publication bias. The National Mortality and Morbidity Air Pollution Study (NMMAPS) was the first multi-city time-series study of air pollution and health and serves to illustrate the data- availability issues highlighted in this discussion paper. NMMAPS, funded by HEI, evaluated associations between daily mortality rates in 90 U.S. cities with the largest population and the daily level of PMio reported for that locality in the EPA Air Quality System (Samet et al, 2000, Dominici et al, 2003). Mortality data from 1987 to 1994 was obtained from CDC's National Center for Health Statistics. While location-specific risks were reported, the objective was to construct precise national and regional estimates of mortality risk from daily changes in ambient PM and other criteria pollutants, thus increasing confidence in the values of the disease-specific risk estimates, and that these estimates were representative of those experienced by the U.S. population as a whole. In 14 cities with daily monitoring on at least 50% of study days, NMMAPS also evaluated the association of hospital admissions with PMi0 (Samet et al., 2000; Schwartz et al., 2003). Despite its national scope, NMMAPS was limited by the amount of air quality data available for analysis. The 90-city mortality analyses were based on air quality data in the AQS primarily collected using l-in-6 day sampling schedules. The 8-year mortality dataset was necessarily restricted to only those days where PMio data were available between 1987 and 1994 in each county. County-specific mean PMio concentrations were calculated for each day with PMio measurements contributed by one or more monitors. Almost half (43) of the 90 cities had data from only one or two monitors and only 28 cities had the equivalent of two or more years (730 days) of monitoring days available. The dataset for the 20 city analysis of PMio adjusting for other pollutants was further restricted for multi-pollutant models because data on all pollutants had to be available on the same day. Consequently, these adjusted risk estimates were less precise. Even with these limitations, the risk estimates were determined to be robust in several sensitivity analyses to investigate residual confounding and exposure misclassification. An NMMAPS sub-study often cities with PMio monitoring on a daily schedule (New Haven, Birmingham, Pittsburgh, Detroit, Canton, Chicago, Minneapolis/St. Paul, Colorado Springs, Spokane, and Seattle) systematically evaluated the potential bias associated with the use of a single day PM concentration (Samet et al., 2000, Appendix B, pp. 54-61). Mortality data was fit using a generalized additive Poisson regression model and polynomial constrained and unconstrained lag models for ambient PM. Risk estimates from these models were compared to those obtained using 1 or 2 day means. Overall effects estimated using the distributed lag models were larger compared to effects estimated using the single or two-day mean lags. These analyses showed that the mortality effects of an increase in pollution levels on a single day are spread over several succeeding days, or conversely, that deaths on a single day are the result of pollution over several preceding days. For the 90-cities, the NMMAPS mortality analyses estimated D-3 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite increased mortality per each 10 |ig/m3 increase in daily PMio using a specified lag structure (0, 1 or 2 days prior to the day deaths occurred). The 10 city sub-study using distributed lags demonstrated that the larger multi-city study underestimated the risk associated with PMio. In addition, consistent with previous reports, the NMMAPS results for morbidity in 14 cities demonstrated that use of PMio concentrations on a single day (a one day lag) resulted in an underestimation of the cumulative PMio effect on hospitalizations. Therefore, the flexibility to analyze effects in relation to PM concentrations over several days is key to a complete understanding of the magnitude of risk and the relevant time period for exposure to PM components. Many of the other times-series studies reviewed in the CD also relied on l-in-6 day PM measurements including the Canadian eight cities study (Burnett et al., 2000; Burnett and Goldberg, 2003). The 2004 CD was able to cite very few studies that relied on daily ambient concentrations for PM indicators. As a consequence, the CD discussed in depth the trade-offs of increased representativeness and precision provided by the larger, multi-site studies with the increased uncertainties in the reported risk estimates due to exposure misclassification. MISSING AMBIENT CONCENTRATION DATA AND STATISTICAL POWER Data collection frequency is a key component of statistical power for time-series studies, and missing data results in increased uncertainty in study results (discussed in PM Staff Paper, Dec. 2005, p. 3-39 and CD, p. 9-41). The Staff Paper concluded that, "consistent with the CD's observation that uncertainty is increased in studies using infrequently collected PM data, staff judges that greater weight should be placed on those studies with daily or near-daily PM data collection in drawing quantitative conclusions." Daily PM measurements in locations where enough health events occur will support future health studies that reduce uncertainties and thereby improve our understanding of the public health impacts of PM. Such studies will provide important information on specific components within the ambient mix of particles to inform the review of the PM NAAQS and strategies to implement these standards. Statistical Power and Potential Bias The statistical power of any proposed study is the probability that the completed study will correctly reject the null hypothesis with a specified confidence level, usually 95% confidence or a p-value of less than 0.05. In the calculation of statistical power, one must specify the expected magnitude of the exposure-health association, the expected exposure gradient, the sample size, the variability of the health outcome measure, and the specified confidence level. The minimum desired statistical power is usually 80%, and the formula may be inverted to calculate the necessary minimum sample size for a specified statistical power. Generally, increased exposure variability is associated with an increased exposure gradient and with increased statistical power. However, exposure variability due to measurement error and unmeasured exposure variability does not increase statistical power. For many epidemiologic study designs, sample size is directly related to the person-time of observation. For time-series studies of air pollution in a large metropolitan area, the usable sample size is largely determined by the number of days with complete exposure information. The variability of the health outcome measure depends on the type of measure (e.g., mortality counts or continuous biological indicators) and on the precision with which the outcome is measured. For time-series studies of D-4 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite the daily variation in mortality for large urban areas, the variability of the outcome measure is largely determined by the average number of daily deaths. Everything else held constant, statistical power for a time series study increases relative to the square root of the average number of daily deaths. Hence, the weights in tables 1 and 2 are the square root of the estimated number of daily deaths, or 1% of the population (In the US, about 1% of the population dies each year) / 365. For air pollution time-series studies, when air quality measurements are missing, the deaths or heart attacks that would be studied for those days must be excluded. Therefore, the sample size (number of mortality-days) available to analyze in a locality is reduced because the exposure data may not exist for the desired lag days. For example, if one wants to evaluate deaths in relation to ambient pollution levels on the same day, the day before, and the day before that (lags 0, 1 and 2), then air pollution concentrations must be available at that location for three consecutive days. Statistical power has implications for the selection of cities in epidemiologic studies. The cities that can be analyzed become restricted to locations with a high number of daily events (e.g., deaths, hospital admissions, etc.) so that the required sample size will be obtained in a reasonable time period. Dr. Kazuhito Ito, NYU, presented a slide at the November 2006 meeting that showed statistical power curves for hypothetical time-series studies by the number of daily deaths in a location and study duration (see Figure 1). These curves indicate statistical power achieved to evaluate a hypothesized increase in daily total (non-accidental) deaths of 2.5% per 25 |ig/m3 increase in PM2.5. The hypothesized increase in mortality rate was derived from city-specific relative risks reported in the PM literature. With l-in-3 day monitors, six years of monitoring data would be required to achieve 80% statistical power to evaluate non-accidental mortality in cities with 100 mean daily deaths (6 years x 365 days/year x l/3 monitoring days = 730 days). For example, a study in New York City, with 180 to 200 deaths per day, would achieve 80% power to evaluate the effect of PM2.5 on nonaccidental mortality in about three years (3 years x 365 days/year x % monitoring days = 365 days). Conversely, a study conducted in a city the size of Seattle, with about 30 deaths per day would not achieve 50% power even if the study were extended beyond six years! Furthermore, study power decreases when the focus of study is cause-specific mortality or the identification of susceptible subgroups. For many fine particle components, temporal and spatial variation within and between localities may be different than the variation for PM2.5 mass. With greater temporal exposure variation and spatial variation between cities, the statistical power to study specific fine particle components is likely higher than for PM2.5 mass. Statistical power is also enhanced if the mortality risk associated with a specific component is higher than for PM2.5 mass. However, exposure error caused by uncertainties in a study's exposure assessment can result in an attenuation of risk estimates and an inability to reject the null hypothesis. Exposure error also may lead to biased risk estimates. Within a metropolitan area, some PM ambient air measurements, such as total PM2.5 mass or sulfates, show less spatial variability and a metropolitan area may be well characterized by a single central-site monitor. Other PM2.5 components, such as elemental carbon, show considerable spatial variability and a metropolitan area may not be well characterized by a single central-site monitor. However, this uncertainty in D-5 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite exposure assessment related to the unmeasured spatial variability will be differential with respect to PM2 5 component; and PM2 5 components with less spatial variability would have less exposure uncertainty. For a multi-city study focused on regional air pollution gradients, city-to- city differences in exposure uncertainty related to monitor location could affect city-to-city differences in the observed associations with health outcomes and hence could be misinterpreted as related to city-to-city differences in PM components. Within the PM2 5 ambient air monitoring network, there are approximately 900 Federal Reference Method (FRM) filter-based samplers that provide 24-hour PM2.5 mass concentration data and about 600 continuous PM2 5 mass monitors that provide hourly data on a near real-time basis. Due to the complex nature of fine particles, EPA implemented the Chemical Speciation Network (CSN) to better understand the components of fine particle mass at selected locations. Chemical speciation measurements are made at 54 "Speciation Trends Network (STN)" sites that are intended to remain in operation indefinitely and about 150 other, potentially less permanent sites used to support State Implementation Plan (SIP) development and other monitoring objectives.33 Specific components of fine particles also are measured through the Interagency Monitoring of Protected Visual Environments (IMPROVE) monitoring program which supports regional haze characterization and tracks changes in visibility in Class I areas as well as many other rural and some urban areas. Together, the CSN and IMPROVE data provide chemical species information for fine particles that are critical for use in health and epidemiologic studies to help inform reviews of the PM NAAQS. The cities where the CSN monitoring sites are located are very important for studying health effects associated with fine particle exposures. There are more than 200 sites in the CSN. Table 1 (see associated pdf file) lists the Primary Metropolitan Statistical Areas (PMSAs) and Metropolitan Statistical Areas (MSAs) ordered by population size and a weight determined by the contribution to precision that area would make to a statistical analysis of risk. The site locations for CSN monitors are listed within the relevant PMSA where they operate. It is encouraging that more than 50% of the U.S. population resides in census areas with at least one CSN monitor. There also are some large population centers where PM components are not measured such as Orange County and Oakland in California, northern New Jersey, and Long Island, New York (PMSAs and MSAs where no speciation monitors are located are highlighted in red on Table 2). Currently, the CSN sites measure fine particle mass and components every third day or every sixth day. A change to daily sampling would increase the statistical power for time-series studies. In evaluating criteria for identifying potential locations for increased monitoring, consideration could be given to CSN locations representing varied fine particle sources in the eastern, western, mid-western and southern parts of the U.S. Future epidemiologic studies that examine PM exposures at or below the current level of the PM2 5 NAAQS will contribute significantly to reducing scientific uncertainty concerning health effects. Most major US metropolitan areas are below or close to the current PM NAAQS and can contribute useful information on the public health impact of PM exposures. The collection of daily measurements for PM2.5 mass and key PM2.5 components in metropolitan areas with high numbers of deaths and 33 See http://www.epa.gov/ttn/amtic/speciepg.html for more information on the PM2 5 speciation monitoring program. D-6 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite in locations with variation in ambient concentrations near the current standard and in the mix of PM2 5 components would inform our understanding of the relative health significance of specific PM2.5 constituents and sources of PM2.5. MISSING DAYS: IMPACT ON VALUE OF RISK ESTIMATES Missing days of sampling data presents problems for epidemiology studies of health effects associated with PM exposures over and above a decrease in statistical power and the problems are compounded for studies of fine particle components. PM2.5 components are likely to present a large degree of variation involving associations with different health endpoints and time from exposure to response. This requires flexibility in constructing statistical models and lag structures. In addition, variation between localities presents complexities in the interpretation of the results of multi-site time series studies. Different components predominate in different regions of the U.S. and the correlation between PM components in each area will vary. Finally, the temporal and spatial variability of each PM component of interest within a city will vary and, if not adequately captured in sampling data, will result in exposure misclassification and an effect on the value of risk estimates. The NMMAPS study and other time-series studies of mortality and morbidity indicate that risk estimates may vary between metropolitan areas or regions. There are multiple explanations for these observations including, random variation (chance), residual confounding, exposure misclassification, and the existence of real source-specific differences in risk. The differences in the precision of city-specific estimates of mortality risk associated with daily change in PM concentration complicate the interpretation of heterogeneity in risk reported by multi-city time-series studies or when single-city estimates show differences between localities. Some authors have used interpolation approaches to fill in the missing days of ambient concentration data in order to avoid excluding cases for days with no air quality measurements. These imputation methods are often based on hourly or daily air quality measurements of PM mass or gaseous co-pollutants. Unfortunately, imputed values never carry as much information about population exposures as measured values and any evaluation of the improved health associations with speciated PM would be diluted by the high proportion of imputed values. Moreover, the error in imputation is not likely to be constant for each specific PM component; some components will be imputed with more error than other components. For example, filling in missing data with imputed values may be associated with more error in studies of coarse particle mass and fine particle components which are associated with larger spatial and day-to- day variation than fine PM mass. Along with the simple imprecision of the imputed values, any use of the gaseous co-pollutants in an imputation algorithm will necessarily increase the co- linearity between the measured values for the gaseous co-pollutants and the imputed values for various PM components. Thus, imputation methods will generally tend to bias any epidemiologic studies of the differential associations of PM components with human health outcomes. The 2004 CD discussed results from a study conducted in Chicago, IL, which illustrates the impact on risk estimates caused by the use of l-in-6 day ambient concentration data. In this study, a significant association was reported between daily change in PMio concentration and D-7 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite mortality between 1985 - 1990 using data from one monitor collected on a daily basis (Ito et al., 1995). However, when the data set was divided into 6 subsets representing a l-in-6 day monitoring frequency, the effect estimates for the PMio-mortality association were quite variable. Moreover, the confidence intervals were wider and analysis of only one of the subsets indicated a statistically significant association. This analysis indicates reduced precision due to the markedly decreased number of events available for analysis, but also indicates that risk estimates may be affected by exposure misclassification, chance, or selection bias. Selection bias could occur if the analyzed group associated with any particular sampling schedule were different from the unanalyzed group in a way that was systematically associated with exposure estimates. However, selection bias is not as prevalent a concern for time-series studies as is exposure misclassification due to inadequate characterization of the spatial variation of the PM exposure measure within a locality. Exposure Misclassification and Spatial Variability Exposure misclassification can occur when ambient concentration from one or only a few monitors in a geographic area is assigned to estimate the PM exposure of the individuals who died in that area. If the ambient concentration that is calculated for a particular day is higher than what some of those who died actually experienced, but lower than what others who died experienced, the resulting "noise" in the PM indicator makes it harder to distinguish a statistical association with mortality. If the calculated ambient concentration is not consistently higher or lower than the concentration experienced by those who died on that day (that is, nondifferential misclassification), the size of the relative risk may be attenuated. The importance of the attenuation depends on the degree of spatial variability characteristic of the pollutant under analysis and the resulting amount of exposure misclassification. While the impact of spatial variability on estimates of exposure is of less concern for studies of fine PM mass, a PM exposure with relatively homogenous local distribution, this is an important issue for epidemiology studies of PM2.5 components or thoracic coarse particle mass and components. The NCER STAR program is funding five studies beginning in early 2008 that will provide information about spatial variability in coarse particle mass and components and effects on health. Additional studies, to be awarded in 2008, will address strategies to incorporate data on spatial and temporal variability of PM components in atmospheric and exposure models. Evaluation of Cumulative Effects of Air Quality on Health Studies using distributed lag models indicate that risk estimates using zero or one-day lags may underestimate the magnitude of mortality associations with air quality. Distributed lag models allow the examination of the combined effect of air pollution across a range of prior days on mortality for one particular day, e.g. today's mortality with today's air quality (lag 0), yesterday's air quality (lag 1), and day before yesterday's air quality (lag 2). The lagged effects of air quality over multiple previous days are compared with each day's mortality throughout the study period. With l-in-3 or l-in-6 day monitors, the relationship with the health measurements is disordered; the lagged effects of air quality on a single day must be compared with mortality on different days. For example, today's air quality is compared to today's mortality at lag 0, with tomorrow's mortality at lag 1, and with the day after tomorrow's mortality at lag 2. Distributed lag models using daily monitors are advantageous because a specific lag structure for D-8 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite modeling the association of a pollutant with health does not have to be selected in advance. This modeling flexibility will be especially important for the study of PM components, which may have differing lags between exposure and health outcome. SUMMARY The purpose of the preceding discussion has been to highlight the importance of time- series studies using air quality data obtained from EPA's Air Quality System to identify health risks associated with ambient PM mass concentration and the limitations of the national monitoring networks for similar studies of fine particle components as indicators of PM sources. This issue was raised by the health research community at the Fffil/EPA workshop in Boston November, 2006 and in subsequent discussions. The lack of daily speciation monitoring for PM2.5 components is an important research need identified by the epidemiology community. This issue has been highlighted because the number of PM2 5 speciation monitors per location is much smaller and variability (temporal and spatial) for many fine particle components is much greater than for PM2 5 mass. Obtaining daily PM2 5 speciation monitoring in a set of key locations will enhance our understanding of the health effects associated with fine particles by: • providing improved statistical power for epidemiologic studies of PM components within a reasonable time period, • providing analytical flexibility to examine distributed lags, and • reducing exposure misclassification to improve the validity and precision of health effect estimates. In addition, targeted studies in some metropolitan areas will help to characterize the spatial variability of PM2.5 components and quantitative impact on risk estimates. POTENTIAL OPTIONS FOR OBTAINING DAILY PM2.5 SPECIATION MEASUREMENTS A: Retrospectively fill in the missing data Actions could be taken to construct a dataset containing daily values for PM2.5 mass, metals, elements, sulfate, nitrate, and carbon for previous years. Options could include analyzing archived daily PM2 5 mass (Teflon) filters collected at CSN or nearby sites, using data collected at nearby continuous (hourly) monitors. These efforts most likely could be done at most at a limited number of sites due to resource constraints and the limited historical use of the relevant samplers. 1. Analyze archived filters to obtain daily measurements of metals and elements (XRF) The 54 STN monitors operate on a l-in-3 day sampling schedule. Some non-STN monitors in the CSN network may also operate on this schedule. There are some things that could be done to achieve daily measures at some of these sites retrospectively (back to 1999) for PM2.s and some components that are key source indicators. XRF analysis of archived daily PM2.s mass (Teflon) filters collected from an adequate number of locations of daily FRM monitors (primarily ones co-located with some of the l-in-3 day CSN monitors) would provide critical D-9 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite information about daily concentrations of metals and elements. There are alternative methods to measure particle mass, EC, and XRF-elements that may address the data needs for epidemiology studies and be cost-effective. For example, Joel Schwartz and George Thurston recommended using reflectance, a measure of black carbon, at the CSN sites to obtain estimates of daily EC (where daily samples for PM2.5 mass are available). Reflectance or other optical analyses of the same filters may provide an indicator of EC. It must be noted that the existence of all historical, Teflon filters of interest remains to be determined and gaining access to those filters that have been archived will require a collaborative effort with State and local agencies. 2. Evaluate continuous measurement data to obtain daily organic carbon (OC)/elemental carbon (EC) at TRENDS sites. Continuous (hourly) data for sulfate, nitrate, EC, and OC are available in some cities including, Chicago, IL, Indianapolis, IN, Davenport, IA, Bar Harbor, ME, Cedar Rapids, IA, Raleigh, NC, New York City, Seven Oaks, SC, Greenville, SC, Rockwell, NC, Seattle, WA, and Detroit, MI. However, all components are not measured at all cities. Daily measures for sulfate, nitrate, EC, and OC could be obtained over multiple years for Chicago (2002-2007), Bar Harbor (2004-2007 for sulfate, 2004 -? for OC, EC, & TC), New York City (2001 - 2007 for sulfate & nitrate, 2005 - 2007 for OC & EC), and Raleigh (2003 - 2006 for OC, 2003 - ? for sulfate, 2003 - 2007 for nitrate, and 2003 - 2006 for total carbon). These data could be used to construct a data set containing daily concentrations. Of these sites, STN monitors are located in the vicinity of the continuous monitors at Chicago, New York City, and Raleigh. OAQPS notes that continuous speciation monitors have their own measurement uncertainties, which may include systematic biases that are not well characterized; data from them cannot simply be merged with CSN data to fill in missing days. However, the continuous data could be used on a site-by-site basis if a relationship between the continuous analyzer and the filter-based monitor was established. B: Expand current monitoring schedules at selected locations in order to conduct daily speciation measures in selected metropolitan areas. 1. Locations Population size and the number of health events that occur each day are location attributes that contribute the greatest amount to the power of a time-series study to detect an association with exposure to an air pollutant if one exists. The top 22 PMSAs or MSAs with the highest weight were selected from Table 1 and are listed in Table 2 along with any CSN monitors currently operating in that location. It should be noted that there are three PMSAs (highlighted in red) in this group where there is no CSN monitor located. All of the other locations, except for Los Angeles, have an STN monitoring site. In addition to population size and mortality and the opportunity to take advantage of an existing CSN monitor, there are a number of factors that are important to consider for site selection for daily speciation monitoring. Important information might include: • What are the major sources in an area? - Are components of interest present in measureable concentrations? D-10 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite • Regional representation of study locations under consideration • Area characteristics - topography, which influences spatial variability • Existing data collection at a location - • Are data available on a daily timescale for components of interest retrospectively? • Are multiple speciation monitors operating in the location? • Are special studies (government or grant-supported) being conducted in the location? Participants at the April 2008 workshop will be asked to provide feedback about the importance of daily monitoring for fine particle components for advancing our understanding of the impacts of air pollutant exposures on public health and criteria to consider for prioritizing locations for daily PM2.5 speciation monitoring. For discussion purposes, three columns have been added to Table 2 indicating significant area characteristics, notes on predominant PM components, and any special studies known to be conducted in the location that may contribute information on temporal or spatial variability. 2. Components For what PM species would it be beneficial to have daily ambient measurements? For discussion purposes, Ito (HEI/EPA workshop, 2006) suggested the following components: OC, EC, nitrate, sulfate, Se, As, Si, Fe, Mn, Cr, Zn, Pb, V, Ni based upon information from toxicology and source apportionment studies. Some of these components may have more homogenous distributions in certain regions but others are likely to have a high degree of spatial variability. CSN and IMPROVE currently analyze for these components and EPA plans are to continue to do so. 3. Costs The estimated annual cost for shipping and lab analysis to add daily PM2.5 mass and speciation monitoring at one CSN site that is currently operating on a l-in-3 day sampling schedule would be $100,000 based on current EPA contracts. The State/local monitoring agency would incur additional labor and equipment costs to operate the monitors as well. OTHER MONITORING ISSUES Two additional issues should be mentioned in this discussion of the use of air quality monitoring data in time-series studies. A: Spatial Variability: Set up additional monitoring sites within certain cities to increase understanding of spatial variability of specific components. Several components of research interest will be associated with a high degree of spatial variability across a location. A small number of monitors (4 - 6) distributed to capture concentrations throughout an area could give enough information to conclude whether or not a specified component has a uniform distribution in that area. If the distribution appeared uniform for a certain component, multi-site time-series analyses could be conducted using data from one or more centrally located monitors per location. If not, a more detailed exposure analysis would D-ll ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite be necessary in certain locations to obtain a finer spatial resolution and develop exposure models. Such a detailed, nonroutine study could potentially involve a research grant. Based on current EPA contracts, the estimated annual cost to establish a CSN-like site operating on a daily schedule in a new location but probably where other monitors are located is $150,000 for shipping and lab analysis. B. Alternative Measurement Methodologies There are alternative methods to measure particle mass, EC, and XRF-elements that may address the data needs for epidemiology studies and be cost-effective. For example, Joel Schwartz and George Thurston recommended using reflectance, a measure of black carbon, at the CSN sites to obtain estimates of daily EC. In addition, the use of a rotating drum sampler which measures various size classes of PM mass and PM components with a finer time resolution (six hours) could be considered. At a meeting with State/local monitoring managers, reservations were expressed about whether this sampler has demonstrated adequate repeatability. BENEFITS OF OBTAINING DAILY PM2.5 SPECIATION DATA If resources can be secured or re-programmed to support daily PM2.5 monitoring at a well prioritized set of monitoring sites of most value in health studies, the following benefits would be obtained: 1. Time-series studies will have enough statistical power to determine which particles are more toxic than others without having to wait ten or more years for results. 2. We will develop more accurate estimates of health effects that fully address lag issues due to the availability of daily health and monitoring data. Studies have shown that the use of distributed lag models evaluating several consecutive days prior to the occurrence of death result in a higher estimated relative risk. Studies of PM components need more flexibility in choice of lag models because not all components are predicted to have the same lag structure for effects. This has obvious implications for RIA, accountability studies, and basis for NAAQS decisions. 3. With multiple daily speciation monitoring sites in some of the larger cities, especially those with more complicated geologic features, it would become possible to improve our understanding of the impact of spatial variability on exposure estimates for PM components. Studies of within-city spatial variability will allow assessment of whether the "noise" in exposure estimates is so large for some components that no excess risk is observed. 4. Researchers could base their analyses on actual data, rather than using creative approaches to get around the fundamental issue of missing data. These methods are helpful, but introduce more uncertainty into the exposure estimates by increasing the co- linearity with co-pollutants. 5. The primary recommendation in the final NRC report was the need for EPA to systematically examine which PM components and sources are most important for public health. Since PM components may exert their adverse influences over different lags between exposure and outcome, the ability to correctly evaluate lagged effects may be D-12 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite crucial to the correct assessment of the relative toxicity of PM components. The availability of daily air quality information would enable the differential assessment of PM components and PM sources, and eliminate a potential bias in favor of the assessment of those components with very immediate effects. 6. EPA is investing heavily in studies of key components and sources of PM (e.g., Hopkins PM Center studies, HEI's NPACT study, recent STAR RFAs). The return on this investment would be increased significantly if daily PM2.5 data are available to increase confidence in the findings and reduce uncertainties in the estimates as explained above. 7. Accountability: Assessing the health improvements attributable to reduced air pollution is already a difficult challenge. Without daily data, such research will be even more difficult and take many years to demonstrate benefits. REFERENCES Burnett RT, Brook J, Dan T, Celocla C, Philips O, Cakmak S, Vincent R, Goldberg MS, Krewski D. 2000. Association between particulate- and gas-phase components of urban air pollution and daily mortality in eight Canadian cities. Inhalation Toxicol 12 (Suppl4): 15-39. Burnett RT and MS Goldberg. 2003. Size-fractionated parti culate mass and daily mortality in eight Canadian cities. In: Revised Analyses of Time-Series Studies of Air Pollution and Health. Special Report. Health Effects Institute, Boston, MA. Dominici F, McDermott A, Daniels M, Zeger SL, Samet JM. 2003. Mortality among residents of 90 cities. In: Revised Analyses of Time-Series Studies of Air Pollution and Health. Special Report. Health Effects Institute, Boston, MA. Ito K, Kinney PL, and Thurston GD. 1995. Variations in PM-10 concentrations within two metropolitan areas and their implications for health effects analyses. Inhalation Toxicol 7:735- 745. National Research Council (NRC). 2004. Research Priorities for Airbourne Particulate Matter IV: Continuing Research Progress. The National Academies Press, Washington D.C. Samet JM, Zeger SL, Dominici F, Curriero F, Coursac I, Dockery DW, Schwartz J, Zanobetti A. 2000. The National Morbidity, Mortality, and Air Pollution Study, Part II: Morbidity and Mortality from Air Pollution in the United States. Research Report 94. Health Effects Institute, Cambridge MA. Schwartz J, Zanobetti A, Bateson T. 2003. Morbidity and Mortality Among Elderly Residents of Cities with Daily PM Measurements. In: Revised Analyses of Time-Series Studies of Air Pollution and Health. Special Report. Health Effects Institute, Boston, MA. D-13 ------- DRAFT 4/02/08 - For Discussion Purposes Only - Do Not Quote or Cite Figure 1. Implication of every-6th-day and every-3rd-day data: Statistical power for time- series studies. Slide provided by Dr. Kazuhito Ito, presented at Fffil/EPA Workshop on Air Quality Data, Newton, MA, November 2006. Adapted slightly for this discussion paper. c/i • S ra a; un (a) non-accidental mortality 3 yrs of every- 3rd-day data 6 yrs of«very- 3rd-day data 200 400 600 800 number of monitoring days D-14 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE Table 2. Twenty-two MSAs with largest population and associated speciation monitors. # 2000 Pop Wgt34 MSA Local Site Name City Name State Site ID Sample Collection County Name 1 9,3 14,235 16.0 New York, NY PMSA; New York-Northern New Jersey-Long Island, NY-NJ-CT-PA CMSA IS 52 DIVISION STREET QUEENS COLLEGE 2 2 9,519,338 16.1 Los Angeles-Long Beach. CA PMSA: New York NY 0110 Met One SASS Teflon Bronx New York NY 0134 Met One SASS Teflon New York New York NY 0124 R&P MDL2300 PM2.5 Queens SEQ SPEC Los Angeles— Riverside— Orange County, CA CMSA STN? Area characteristics Large urban increment plus NE background STN Complex coastal and mountain topography, extreme traffic, ozone. and sun Components High sulfate and organics, low nitrates Extremely high nitrate and organics, low sulfate Special Studies NYDOC, MESA-AIR, Supersite GARB, PM Center, CHS, MESA-Air, Supersite NULL Los Angeles CA 1103 Met One SASS Teflon Los Angeles 3 8,272,768 15.1 Chicago. IL PMSA: Chicago-Garv-Kenosha. IL-IN-WI CMSA SPRINGFIELD PUMP STATION Chicago COM ED MAINTENANCE BLDG Chicago IL IL NORTHBROOK WATER PLANT Northbrook IL 0057 Met One SASS Teflon Cook 0076 Met One SASS Teflon Cook 4201 Met One SASS Teflon Cook CITY HALL Naperville IL 4002 Met One SASS Teflon DuPage 4 5,100,931 11.8 Philadelphia. PA-NJ PMSA: Philadelphia-Wilmington-Atlantic City. PA-NJ-DE-MD CMSA CAMDEN LAB AMS Laboratory Philadelphia PA ON AMTRAK RIGHT OF WAY - Philadelphia PA NEAR AIRPORT HI SPEED LINE (ELECTRIFIED) A420450002LAT/LON POINT IS Chester OF CORNER OF TRAILER Camden NJ 0003 Met One SASS Teflon Camden 0004 Met One SASS Teflon Philadelphia 0136 Met One SASS Teflon Philadelphia PA 0002 Met One SASS Teflon Delaware STN STN Industrial center with lake High sulfate and influences, windy organics, low nitrates Industry plus NE background High sulfate and organics, low nitrates MESA-Air 34 Weights (Wgt) are proportional to expected inverse-variance regression weights: sqrt((population * 0.01) / 365) D-15 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE # 2000 Pop Wgt MSA Local Site Name Site Special City Name State ID Sample Collection County Name STN? Area characteristics Components Studies CHESTER COUNTY Not in a city PA 0100 Met One SASS Teflon Chester TRANSPORT SITE INTO PHILADELPHIA CORNER OF MLK BLVD AND Wilmington DE 2004 Met One SASS Teflon New Castle JUSTISON ST, NO TRAFFIC DATA AVAILABLE 5 4,923,153 11.6 Washington. DC-MD-VA-WV PMSA: Washington-Baltimore. DC-MD-VA-WV CMSA MCMILLAN PAMS Washington DC 0043 Andersen RAAS Teflon District of Columbia STN HOWARD UNIVERSITY Beltsville MD 0030 Andersen RAAS Teflon Prince George's 6 4,441,551 11.0 Detroit. MI PMSA: Detroit-Ann Arbor-Flint. MI CMSA NULL Allen Park MI 0001 Met One SASS Teflon Wayne PROPERTY OWNED BY Dearborn MI 0033 Met One SASS Teflon Wayne DEARBORN PUBLIC SCHOOLS STN NE background plus High sulfate and Supersite traffic organics, low nitrates Industry and high traffic High sulfate and EPA organics, low nitrates DEAN ROAD DEAD-ENDS AT Luna Pier SITE, 200 FT WEST MI 0005 Met One SASS Teflon Monroe 7 4,177,646 10.7 Houston. TX PMSA: Houston-Galveston-Brazoria. TX CMSA SOUTH OF DETERMINED & Not in a city TX ALDINE MAIL RD INTERSECTION NW OF W. LAMBUTH & DURANT INTERSECTION 8 4,112,198 10.6 Atlanta. GA MSA 2390-B WILDCAT ROAD, DECATUR, GA Deer Park TX Decatur 9 3,519,176 9.8 Dallas. TX PMSA: Dallas-Fort Worth. TX CMSA 0024 R & P Model 2025 PM- Harris 2.5 Sequential Air Sampler w/VSCC 1039 URG MASS400 Teflon Harris WINS GA 0002 Met One SASS Teflon DeKalb STN STN Extreme chemical industry and ozone, coastal sunny Medium (if interesting) organics, sulfate? Extreme biogenics, high High sulfate and EPRI, traffic, sunny organics, low Supersite nitrates Traffic and cattle, sunny high ammonia? WESTOFS. AKARD& CANTON STREETS INTERSECTION Dallas TX 0050 R & P Model 2025 PM- Dallas 2.5 Sequential Air Sampler w/VSCC D-16 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE # 2000 Pop Wgt MSA Local Site Name City Name State Site ID Sample Collection County Name STN? Area characteristics Components Special Studies DALLAS HINTON Dallas TX THIS SITE WILL MONITOR Midlothian TX THE SAME AREA AS SITE 1390015 0069 URG MAS S400 Teflon Dallas WINS 0016 R & P Model 2025 PM- Ellis 2.5 Sequential Air Sampler w/VSCC 10 3,406,829 9.7 Boston. MA-NH PMSA: Boston-Worcester-Lawrence. MA-NH-ME-CT CMSA DUDLEY SQUARE ROXBURY Boston MA 0042 Met One SASS Teflon Suffolk 11 3,254,821 9.4 Riverside-San Bernardino. CA PMSA: Los Angeles-Riverside-Orange County. CA CMSA NULL 12 3,251,876 9.4 Phoenix-Mesa. AZ MSA Rubidoux (West Riverside) CA 8001 Met One SASS Teflon Riverside PHOENIX SUPERSITE Phoenix AZ 9997 Met One SASS Teflon Mancopa 13 2,968,806 9.0 Minneapolis-St. Paul. MN-WI MSA ANDERSON SCHOOL - Minneapolis MN 0963 Met One SASS Teflon Hennepm PHILLIPS NEIGHBORHOOD 14 2,846,289 8.8 Orange County. CA PMSA: Los Angeles-Riverside-Orange County. CA CMSA No speciation monitors? STN STN STN STN STN NE background plus traffic Complex coastal and mountain topography, extreme traffic, ozone. and sun Extreme traffic, ozone. and sun Continental urban Complex coastal and mountain topography, traffic, ozone, and sun High sulfate and organics, low nitrates Extremely high nitrate and organics, low sulfate High nitrate and organics, low sulfate Medium High nitrate and organics, low sulfate Harvard MESA-Air MESA-Air 15 2,813,833 8.8 San Diego. CA MSA NULL NULL El Cajon CA 0003 Met One SASS Teflon San Diego Escondido Ca 1002 Met One SASS Teflon San Diego 16 2,753,913 8.7 Nassau-Suffolk. NY PMSA: New York-Northern New Jersey-Long Island. NY-NJ-CT-PA CMSA No speciation monitors? 17 2,603,607 8.4 St. Louis. MO-IL MSA STN Complex coastal and mountain topography, traffic, ozone, and sun Medium nitrate and organics, low sulfate Large urban increment plus NE background Industry and traffic High sulfate and organics, low nitrates High sulfate and organics, low nitrates Super site D-17 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE # 2000 Pop Wgt MSA Local Site Name SIU DENTAL CLINIC NULL City Name Alton Not in a city State IL MO Site ID 2009 0012 Sample Collection Met One SASS Teflon Met One SASS Teflon County Name STN? Madison Jefferson Area characteristics Components Special Studies 18 BLAIR STREET CATEGORY A St. Louis CORESLAMPM2.5. 2,552,994 8.4 Baltimore. MD PMSA: Washington-Baltimore. DC-MD-VA-WV CMSA STN NE background plus High sulfate and MESA-Air, traffic organics, low nitrates Supersite ESSEX Essex MD 3001 Met One SASS Teflon 19 2,414,616 8.1 Seattle-Bellevue-Everett. WA PMSA: Seattle-Tacoma-Bremerton. WA CMSA SEATTLE DUWAMISH BEACON HILL OLIVE STREET Seattle Seattle Seattle 20 2,395,997 8.1 Tampa-St. Petersburg-Clearwater. FL MSA Baltimore WA 0057 Andersen RAAS Teflon King WA 0080 Met One SASS Teflon King WA 0048 Met One SASS Teflon King STN STN Coastal urban, high wood burning Medium organics?, nitrate?, low sulfate? Coastal with some SE Medium sulfate and and power plant organics influence SYDNEY NULL Plant City FL 3002 Met One SASS Teflon Hillsborough Pmellas Park FL 0026 Met One SASS Teflon Pmellas 21 2,392,557 8.1 Oakland. CA PMSA: San Francisco-Oakland-San Jose. CA CMSA No speciation monitors? 22 2,358,695 8.0 Pittsburgh. PA MSA STN Coastal urban, high wood burning Medium organics?, nitrate?, low sulfate? Industry & traffic plus High sulfate & eastern background organics, low nitrates NULL NULL S ALLEGHENY HS DOWN WIND FROM USS CLAIRTON COKE WORKS LAT/LON POINT IS TRAILER Pittsburgh PA 0008 Met One SASS Teflon Allegheny Not in a city PA 5001 Met One SASS Teflon Washington Liberty PA 0064 Met One SASS Teflon Allegheny Greensburg PA 0008 Met One SASS Teflon Westmoreland STN Supersite D-18 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE Appendix E: Session V: Ambient Air Monitoring Realities - EPA/State/Local Perspectives - Ambient Air Monitoring Network: Network Design and Site Selection Approval Questions on this draft white paper should be directed to Tim Hanley, EPA/OAQPS, hanley.tim@epa.gov; (919) 541-4417. Introduction The purpose of this white paper is to briefly describe the process for designing major ambient air monitoring networks including: • the roles of EPA and State, local, and Tribal monitoring agencies in selecting and approving monitoring stations; • the ways that health and other researchers can provide input currently to State/1 ocal/Tribal monitoring agencies and EPA on the usefulness and approval of monitoring stations; and • suggestions to improve facilitation of soliciting input on monitoring station selection from health and other researchers. Background The measurement of ambient air pollution in the United States is provided through a number of ambient air monitoring networks operated almost exclusively by State, local, and Tribal air monitoring programs. The EPA identifies key parameters to measure such as criteria pollutants35, pollutant precursors, chemical composition of particles, and air toxics. Ambient air monitoring networks are implemented through a combination of Federal requirements and voluntary programs.36 EPA provides required siting criteria and network deployment strategies for measurement of pollutants as one of several key components to implementing air monitoring networks. Monitors are categorized as State and local Air Monitoring Stations (SLAMS) when they are approved as part of the long-term operating network or Special Purpose Monitors (SPMs) when they are being used for short-term investigations (i.e., less than two years). Air toxic monitoring stations are not required by regulation, and do not carry the SLAMS distinction. Additional, Federally run networks provide monitoring coverage in primarily rural areas to meet specialized objectives. The Interagency Monitoring of Protected Visual Environments (IMPROVE) network, a cooperative measurement effort guided by a steering committee composed of representatives from Federal and regional-state organizations, provides important data for implementing both regional haze and PM2.5 attainment programs. The Clean Air Status and Trends Network (CASTNET), managed by EPA's Clean Air Markets Division, provides atmospheric data on the dry deposition component of total acid deposition, ground-level ozone and other forms of atmospheric pollution. The National Atmospheric Deposition Program (NADP), another cooperative program involving several governmental agencies, provides measurements of pollutants in precipitation, including sulfate, nitrate and ammonium. 35 The Clean Air Act requires EPA to set National Ambient Air Quality Standards (NAAQS) for six common air pollutants. They are particle pollution (often referred to as paniculate matter), ground-level ozone, carbon monoxide, sulfur oxides, nitrogen oxides, and lead. 36 All Tribal monitoring programs are provided for cooperatively as compliance with Federal rules cannot be required of Tribes. E-l ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE EPA designs and implements ambient air monitoring networks to meet several monitoring objectives: • Determining compliance with health- and welfare-based standards (i.e., the NAAQS); • Providing air pollution data to the general public on a timely basis; and • Supporting the development and tracking of emission control programs. In addition to the monitoring objectives above, EPA recognizes important additional objectives that need to be factored into designing ambient monitoring networks such as: • Supporting health and welfare effects and exposure research studies • Providing air pollution data for human health risk/exposure assessments and NAAQS reviews; • Providing air pollution data for welfare effects assessments; and • Supporting atmospheric research studies. EPA recently made changes to the NAAQS-related monitoring regulations. Specifically, the general monitoring network design requirements for the minimum number of ambient air monitors were modified to focus more on populated areas with air quality problems and to significantly reduce the requirements for criteria pollutant monitors that have measured ambient air concentrations well below the applicable NAAQS. A number of the changes related to the monitoring of PM2 5 include revisions to the requirements for reference and equivalent method determinations (including specifications and test procedures). These regulations also added a requirement for a new multi-pollutant monitoring network called National Core (NCore) and revised certain provisions regarding monitoring network descriptions and periodic assessments, quality assurance, and data certifications (71 FR 61236, October 17, 2006)37. Design criteria for required ambient air monitoring networks are provided in Appendix D to 40 CFR Part 58. Network design criteria include monitoring objectives, scale of representation, and specifications for locating monitors (e.g., a requirement to be in the area of expected maximum concentration). In many cases, there are multiple monitoring objectives for a site with the highest concentration of a pollutant. For instance, a neighborhood scale site in the area of maximum fine particle exposure could be thought of as a central monitoring station. These central monitoring stations might have several PM measurement samplers such as a PM2 5 FRM for comparison to the NAAQS, a PM2.5 continuous mass monitor for reporting the Air Quality Index (AQI), and a fine particle speciation sampler to develop and track emission control strategies. All of these data could be useful in health studies depending on the purpose and availability of health endpoint data. Requirements for the minimum number of monitors to operate are identified for PM, ozone, Photochemical Air Monitoring Stations (PAMS), and NCore (which include several measurements); however, monitoring agencies are encouraged to operate additional stations to adequately characterize pollutants. Siting criteria are provided in Appendix E to Part 58. Siting criteria include the specifications for probe and inlet height, distance from obstructions, and traffic. 37 See also http://www.epa.gov/ttn/amtic/ for more information on the Ambient Monitoring Technology Information Center (AMTIC) operated by EPA's Ambient Air Monitoring Group (AAMG). AMTIC contains information and files on ambient air quality monitoring programs, details on monitoring methods, relevant documents and articles, information on air quality trends and nonattainment areas, and Federal regulations related to ambient air quality monitoring. E-2 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE Roles of EPA and Monitoring Agencies in selecting and approving changes to a monitoring network: The EPA requires each State monitoring agency to develop and submit an Annual Monitoring Network Plan to the applicable EPA Regional Office by July 1 of each year. States may delegate portions of the Annual Monitoring Network Plan to applicable local agencies (e.g., in California there are several plans, while in New York there is one plan for the whole State). The annual monitoring network plan must be made available for public inspection for at least 30 days prior to submission to EPA. Any annual monitoring network plan that proposes SLAMS network modifications including new monitoring sites is subject to the approval of the EPA Regional Administrator. Air Toxic monitoring stations are encouraged to be included in annual monitoring network plans, but are not formally required. Modifications to PAMS, the Speciation Trends Network (STN), and the NCore network are to be approved by EPA's Office of Air Quality Planning and Standards. The EPA Regional Office will provide an opportunity for public comment and approve or disapprove the plan and schedule within 120 days of submission. If the State or local agency has already provided a public comment opportunity on its plan and has made no changes subsequent to that comment opportunity, the Regional Administrator is not required to provide a separate opportunity for comment. The annual monitoring network plan must contain the following information for each existing and proposed site: 1. The AQS site identification number. 2. The location, including street address and geographical coordinates. 3. The sampling and analysis method(s) for each measured parameter. 4. The operating schedules for each monitor. 5. Any proposals to remove or move a monitoring station within a period of 18 months following plan submittal. 6. The monitoring objective and spatial scale of representation for each monitor as defined in appendix D to this part. 7. The identification of any sites that are suitable and sites that are not suitable for comparison against the annual PM2.5 NAAQS as described in §58.30. 8. The metropolitan area (e.g., MSA, CBSA, CSA) or other area represented by the monitor. The annual monitoring network plan must document how State and local agencies provide for the review of changes to a PM2.5 monitoring network that impact the location of a violating PM2.5 monitor or the creation/change to a community monitoring zone. The affected State or local agency must document the process for obtaining public comment and include any comments received through the public notification process within their submitted plan. What factors are critical in decisions to change the location of a monitoring station? In most cases, monitoring stations are located for many years in the same location; however, from time to time a monitoring station is moved or shut down due to either planned or unforeseen reasons. The following list provides a summary of the most common reasons why monitoring stations are moved or shut down: Logistical reasons: • The lease for the land or building where the monitoring station is located cannot be extended due to redevelopment or other reasons. E-3 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE • Construction adjacent to a monitoring station renders the site inappropriate to use during construction and in some cases unable to meet siting criteria after construction is complete. Failure to meet probe and siting criteria: • Growth of trees around a monitoring station renders the site no longer able to meet siting criteria and the owner of the trees is unwilling to have them cut or trimmed. • Increases in motor vehicles traffic, including the addition of new traffic lanes lead to re- categorizing the scale of representation or failure to meet set-back requirements; which no longer meets the network design criteria. • Site inspections reveal that some aspect of the siting criteria is no longer acceptable (e.g., a new HVAC system or emission source is located too close to an inlet). Changes in responses to emission and/or ambient monitoring trends: • Changes to existing point, area, or mobile source emission inventories (e.g., establishment of new beltways, shut-down of manufacturing facilities) that alter the original premise for site placement. • Long-term trends analysis demonstrates monitor's objective has been fulfilled (taking into account future alterations of the NAAQS). For example, steep declines in ambient carbon monoxide levels at micro-scale compliance sites led to discontinuation of a large number of CO monitors. • Network assessment indicates that monitoring resources should be devoted to other issues (e.g., air toxics measurement) or conserved to preserve high priority objectives (e.g., ozone, PM2 5) What feeds into the decision-making process for moving or shutting down a monitoring station? With any number of reasons why a monitoring station may need to be moved or shut down, monitoring agencies must plan for network changes. For situations where it may be possible to stay at the existing site, if barriers can be overcome, an agency would likely make the necessary efforts to maintain the site if the monitoring objective were critical and no other suitable location were available. Knowing that data from a monitoring station were being used in an important health or epidemiological study would provide a persuasive argument to keep the station in the same place if the agency knew the data were being used. For example, an agency may be willing to petition a land owner to trim a tree or move an obstruction given the more compelling use of the data in a health research study. If a monitoring station has to be moved, how is a new location picked and approved? Although usually not possible, the best way to handle moving a monitoring station is to identify a new site location within the same general area such that: • the scale of representation and impacts from emission sources is the same as the original site (so long as this is what is intended to be measured at the monitor); • the old and new monitoring stations can be both operated simultaneously for one year or, at minimum, during the season(s) of maximum expected concentrations; and • the statistical analyses of the data from the old and new monitoring stations are deemed to be sufficiently comparable. E-4 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE In many cases, the decision to move a monitoring station does not allow enough time to provide for simultaneous operation, so careful selection of the new station and analysis of its data can only be compared to historical levels and other operating monitors. Use of emission inventories, traffic counts, pictures, and satellite imagery can help document site characteristics for comparability of old and new sites. Monitoring station removals or relocations that are anticipated for the next 18 months must be identified in the annual monitoring network plan that is required to be made available for public inspection and is due to the EPA Regional Office by July 1 of each year. Although not required, ideally the applicable EPA Regional Office will visit and perform a site inspection to assure the new station meets siting criteria and is acceptable. In recognition of uncontrollable circumstances (e.g., a natural disaster such as Hurricane Katrina), the EPA provides for moving an air monitoring station outside the window of an annual monitoring plan by review and approval of the applicable EPA Regional Office. In what ways can communication with the health research community be improved concerning possible changes in the ambient air monitoring networks? Improved outreach concerning currently available tools: • EPA has developed a web site that provides a link to each available State and local agency annual monitoring network plan (see: http://www.epa.gov/ttn/amtic/plans.html). • EPA will continue to update the website as plans are revised. EPA is seeking input regarding recommendations for how often a reminder should goes out when plans are updated. o State/1 ocal/Tribal monitoring agencies can be encouraged to summarize their anticipated monitoring network changes in one place within the annual monitoring network plans, or in a companion summary document, that could be easily scanned by interested parties without wading through an extensive plan. o Although not currently available, there is recognition that a mechanism allowing for quick review of all anticipated network changes across the nation in one place would be beneficial to the health research community. Note: this is not available as all network plans are currently summarized by the appropriate State, local, and/or Tribal agency. • EPA has developed a web site dedicated to documenting the site characteristics, including photos, and links to satellite imagery of candidate NCore monitoring stations (http://www.epa.gov/ttn/amtic/ncore/). At the bottom of each individual NCore station web page, there is an opportunity to provide comments on the candidate station. EPA encourages health researchers to offer comments on the usefulness of candidate NCore stations. • EPA maintains a relatively easy to use public web site that can be used to generate maps and lists of active ambient air monitors (http://www.epa.gov/air/data/). o An agency contact list is maintained as part of this web site so that data users can reach State and local contacts concerning monitors of interest (http ://www. epa. gov/air/data/contsl. html). o A contact list for EPA Regional Office monitoring staff is available at: http ://www. epa. gov/ttn/amtic/namscon.html. E-5 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE • EPA's Office of Air Quality Planning and Standards (OAQPS) has recently developed a list serve that is used as a communication tool for ambient air monitoring and health researchers can be added to our distribution list. Sign-up instructions are available at: http://www.epa.gov/ttn/amtic/airlist.html. Increased participation in health-focused gatherings: • Monitoring experts from OAQPS can participate in key, annual national health research conferences to present information on ambient air monitoring networks and plans for method improvements or changes. This would also improve monitoring experts' knowledge of health research needs, improve communication, and build a bridge between these two communities. o OAQPS can also work with key State and local agency monitoring and network leads by inviting them to participate in annual national health research conferences to present information on ambient air monitoring networks for which they are responsible. • EPA will continue to engage CASAC's Ambient Air Monitoring and Methods Subcommittee, and in doing so can specifically engage or address health research interests. Communication initiatives: • Health researchers are encouraged to communicate with the ambient air monitoring community on the key monitoring sites that provide data to their research. Communications should be at multiple levels to ensure an understanding of the importance of the work; however, the most important communication needs to be directly with the State and/or local air monitoring agency responsible for operating ambient air monitoring stations. • EPA will facilitate and encourage the participation of health researchers at national air monitoring conferences to provide presentations on how their research is using ambient data. This will serve to educate and sensitize monitoring staff to the importance of the ambient air monitoring program to health researchers especially if the issue of relocation or termination of long-term monitoring sites is being considered. • Health researchers and EPA should work collectively to establish the requirements for a website or other publicly available forum to serve as an inventory of all on-going and planned health studies utilizing ambient air monitoring data, the monitoring sites and key ambient monitoring data being used, and the time period of the study. This would be extremely beneficial so that monitoring agencies can make contacts with researchers who are using the information from their networks. E-6 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE Appendix F: Session V: Ambient Air Monitoring Realities - EPA/State/Local Perspectives - Ambient Air Monitoring Method Implementation Questions on this draft white paper should be directed to Joann Rice, EPA/OAQPS, rice.joann@epa.gov, (919) 541-3372. Introduction The purpose of this draft white paper is to describe the current process and communication strategy used by the EPA to implement monitoring methods and method improvements in support of National Ambient Air Quality Standards (NAAQS) criteria pollutants, criteria pollutant precursors, and air toxics and to encourage discussion on how to improve communications with the health research community. Background The Office of Air Quality Planning and Standards (OAQPS) is responsible for identifying ambient monitoring needs based on the NAAQS review process and other air quality data requirements. OAQPS implements the nation's ambient air monitoring networks to ensure that they meet critical air program needs by leading and collaborating on the development of data quality objectives (DQOs), monitoring methods, and a quality assurance (QA) program for achievement of monitoring objectives. The best approach is utilized to optimize the value of the monitoring networks to meet multiple program objectives and regularly assesses the network's effectiveness in continuing to meet those objectives. This is done in collaboration with other key partners, including EPA Headquarters and Regional Offices, the Office of Research and Development (ORD), other Federal agencies, the Ambient Air Monitoring Steering Committee (AAMSC), State/1 ocal/Tribal agencies, the National Association of Clean Air Agencies (NACAA), Multi-State Organizations, the National Academy of Sciences (NAS), the Clean Air Scientific Advisory Committee (CASAC), and private entities such as instrument manufacturers. The EPA requires approved methods for measuring criteria pollutants. The monitoring staff participates in the NAAQS review process to help identify monitoring network issues and new monitoring technology needs. Once these needs are identified and articulated, the staff works with ORD to develop new monitoring technologies and Federal Reference Methods (FRMs) to support these needs. EPA also engages the CASAC, and their subcommittee on ambient air monitoring and methods, in review of the methods developed. Once EPA develops and specifies the FRM requirements, the instrument manufacturers are involved to develop candidate FRM and Federal Equivalent Methods (FEMs). ORD is responsible for testing and approval of equivalent and reference methods. A method has several components: sample collection, analysis, handling, archival, and data processing and reporting, etc. Once FRM/FEMs are approved, they are implemented in the national ambient air monitoring network to support the NAAQS. As the NAAQS review cycle repeats, EPA reviews the monitoring networks and monitoring method needs in consultation with monitoring agencies at the State, Local and Tribal level. If adjustments to the FRM/FEMs are needed, the AAMG works with ORD to develop new methods, or make adjustments or improvements to methods to meet the data needs in support of the NAAQS. Then the method development, review, consultation, approval, and implementation cycles repeat as described above. G-l ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE In the case of criteria pollutant precursors, like PM2.5 chemical species or air toxics, there are no requirements for FRM development and approval. EPA rules or method plans may specify the species or components and methods needed. In this case, OAQPS works with ORD to identify the best methods and technologies available to meet the data use objectives. Once these methods/technologies are identified, OAQPS/ORD consults with Regional Offices, State/1 ocal/Tribal agencies, Multi-State Organizations, and CASAC to obtain feedback on the appropriateness of the methods chosen. Once recommendations are provided on the method/technological approach, the monitoring methods are implemented with the help of the Regions and State/local agencies. Method plans are documented in the monitoring agency's quality assurance project plan (QAPP). States and local agencies often adopt the methods employed in the national monitoring programs for additional monitoring in their networks. As EPA regularly reviews and assesses the monitoring networks to confirm that they are meeting the data quality objectives and data use needs, revisions to the monitoring methods may be recommended or warranted. What factors are critical in decisions to change - why make changes or improvements? Changes to the FRM/FEM are done in support of the NAAQS review process and any resulting changes in the form or level of the standards, as well as to address needed operational efficiencies. Changes for non-criteria pollutants or precursor species are largely made to improve consistency and data usability across our monitoring networks, and to support multiple monitoring objectives such as: • Supporting the development of modeling tools and the application of source apportionment modeling for control strategy development in support of the NAAQS; • Assessing the effectiveness of emission reductions strategies through the characterization of air quality trends; • Supporting health effects and exposure research studies; and • Supporting programs aimed at improving environmental welfare (e.g., the regional haze program). What feeds into the decision-making process? Some changes are intentionally made and others inadvertently or unknowingly happen as a result of changes at the sample collection or analysis stages (e.g., changes in field or laboratory instrument operation). In the case of intentional plans for change, EPA may invoke special field or monitoring studies and data analysis efforts to assess the need for, and the impact of change. Plans for change are then vetted within EPA, and the monitoring, expert, and academic community (disciplines covered include monitoring, modeling and data analysis researchers, as well as health scientists) in a variety of ways and forums to obtain feedback from key partners. These forums include participation in and presentation or communication of plans for change at conferences, meetings, workshops, and Regional/State/Local and NACAA conference calls. In addition, OAQPS holds a tri-annual monitoring conference specific to monitoring issues (the last one was held November 2006). OAQPS may also issue letters, memorandums, program Newsletters, and other forms of written communication through our list serve (link to sign up instructions provided below). The list serve sends an email notification to all parties on the distribution about posting of information on our Ambient Monitoring Technology Information Center (AMTIC) website. In addition, special consultation with the AAMSC and CASAC is held if appropriate. How do we communicate plans for change? G-2 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE Several opportunities exist along the way for interested parties and key partners to provide feedback. EPA begins to consider possible changes well in advance of implementation. It takes several months if not years to perform any special studies, analyze the data, consult with ORD, the academic and expert community and other groups before change can occur. The EPA has (and continually develops) a variety of mechanisms to communicate plans for method improvements and changes. These mechanisms have already been mentioned above (e.g., participation in conferences, meetings and conference calls, newsletters, consultations, etc.). In what new ways can we engage health researchers and improve communication? Communications between OAQPS and the health research community can be improved by the following: • OAQPS can participate in key, annual national health research conferences to present information on ambient air monitoring networks and plans for method improvements or changes. This would also improve OAQPS's knowledge of health research needs, improve communication, and build a bridge between these two communities. o Important conferences and dates need to be identified. • EPA will continue to engage CASAC, and in doing so can specifically engage or address health research interests. o If the monitoring subcommittee is restored, make sure "right" health person(s) involved • OAQPS has recently developed a list serve that is used as a communication tool and health researchers can be added to our distribution list. Sign-up instructions are available at: http://www.epa.gov/ttn/amtic/airlist.html. o NCER can help OAQPS focus what health researchers need to pay attention to or focus the distribution versus "mass mailing" (see below). • EPA can improve internal communications by instituting regular forms of communication between ORD and OAQPS' divisions. o Need regular process of communication across EPA on changes/plans, etc. o Need to "institutional" process to formalize communications between OAQPS and health researchers through NCER. o Build additional relationships and channels for communication. • OAQPS is involved in ORD's air research implementation planning process where OAQPS research needs are identified and conveyed across ORD laboratories. This forum can also be used to communicate plans for change across ORD. o OAQPS can communicate plans to ORD and ORD can help to convey messages and information across ORD labs and centers. • ORD can participate in the AAMSC to improve communication with health researchers regarding monitoring method issues and to monitoring agencies regarding ongoing and planned research efforts. G-3 ------- DRAFT 3/27/08 - FOR DISCUSSION PURPOSES ONLY - DO NOT QUOTE OR CITE Appendix G: Preliminary Survey of Ambient Air Monitoring Sites Currently Being Considered in EPA-funded Epidemiology Studies Feb 2008 G-l ------- Preliminary Survey of Ambient Air Monitoring Sites Currently Being Considered in EPA-funded Epidemiology Studies Feb 2008 for more information, contact Sascha Lodge PLEASE NOTE' There is no distir the Bakersfield (( only one researc Colors signify that a given monitor is being used by multi ction between blue and orange. For example, four resea DA) monitor, three researchers are using the El Cajon (C her is using the Phoenix (AZ) monitor. State STN Sites Alabama Arizona California California California California City Birmingham Phoenix Bakersfield Bakersfield Bakersfield Bakersfield County Jefferson Maricopa Kern Kern Kern Kern Site Name NULL PHOENIX SUPERSITE FLAT TERRAIN.OIL REFINERY 1.3 Ml NNW.TRAIN 1.4 Ml N.FREEWAY 1.3 Ml E FLAT TERRAIN.OIL REFINERY 1.3 Ml NNW.TRAIN 1.4 Ml N.FREEWAY 1.3 Ml E FLAT TERRAIN.OIL REFINERY 1.3 Ml NNW.TRAIN 1.4 Ml N.FREEWAY 1.3 Ml E FLAT TERRAIN.OIL REFINERY 1.3 Ml NNW.TRAIN 1.4 Ml N.FREEWAY 1.3 Ml E State Code 01 04 06 06 06 06 County Code 73 013 029 029 029 029 Site ID 0023 9997 0014 0014 0014 0014 lodge.sascha@epa.gov; (202) 343-9769 pie researchers. rchers are using A) monitor, and Priority Medium High Medium High High High Address 5558 California Ave; Bakersfield Mon Objectivel Population Exposure Mon Object ive2 Latitude 33.553056 33.503643 35.356111 35.356111 35.356111 35.356111 Longitude -86.815000 -112.095001 -119.040278 -119.040278 -119.040278 -119.040278 Parameters Measured if available N020ZPM10 PM25 PM25species Researcher Name Kaz Ito Kaz Ito Bart Ostro Antonella Zanobetti Kaz Ito Kaz Ito Organizatio n/Affiliation NYU NYU CAOEHHA Harvard University NYU NYU Duration of Study For time-series and case-crossover, the longer into the future. the better the analyses. 2000-2003 ------- California California California California California California California California California California California California El Cajon El Cajon El Cajon Riverside Rubidoux (West Riverside) Rubidoux (West Riverside) Rubidoux (West Riverside) Rubidoux (West Riverside) Rubidoux (West Riverside) Sacramento Sacramento Sacramento San Diego San Diego San Diego Riverside Riverside Riverside Riverside Riverside Riverside Sacrament 0 Sacrament 0 Sacrament 0 NULL NULL NULL Mira Loma NULL Riverside-Rubidoux NULL NULL NULL NULL NULL NULL 06 06 06 06 06 06 06 06 06 06 06 06 073 073 073 065 065 065 065 065 065 067 067 067 0003 0003 0003 8005 8001 8001 8001 8001 8001 0006 0006 0006 Medium High High High Medium High High High High Medium High High 1155 Redwood Ave.; El Cajon 5130 Poinsettia Place 5888 Mission Blvd.; Rubidoux 5888 Mission Blvd., Rubidoux Del Paso- 2701 Avalon Dr; Sacramento Population Exposure Other Population Exposure Population Exposure (Riverside- San Bernardino, CA) Population Exposure 32.791389 32.791389 32.791389 33.995638 33.999580 33.99958 33.999580 33.999580 33.999580 38.614167 38.614167 38.614167 -116.941667 -116.941667 -116.941667 -117.493304 -117.416010 -117.41601 -117.416010 -117.416010 -117.416010 -121.366944 -121.366944 -121.366944 if available N020ZPM10 PM25 PM25species PM2.5, PM10 if available N020ZPM10 PM25 PM25species PM2.5, PM10, S02, N02, Oz CO if available N020ZPM10 PM25 PM25species Bart Ostro Antonella Zanobetti Kazlto Joel Kaufman Bart Ostro Joel Kaufman Antonella Zanobetti Kaz Ito Kaz Ito Bart Ostro Antonella Zanobetti Kazlto CAOEHHA Harvard University NYU Univ. of Wash. MESA Air project CAOEHHA Univ. of Wash. MESA Air project Harvard University NYU NYU CAOEHHA Harvard University NYU For time-series and case-crossover, the longer into the future, the better the analyses. 2000-2003 8/1/2004-7/31/2014 For time-series and case-crossover, the longer into the future, the better the analyses. 8/1/2004-7/31/2014 2000-2003 For time-series and case-crossover, the longer into the future, the better the analyses. 2000-2003 ------- California California California Colorado Connecticu t District Of Columbia Florida Florida Georgia Idaho Illinois Illinois Illinois Indiana San Jose San Jose Simi Valley Commerce City New Haven Washington Davie Plant City Decatur Meridian Chicago Chicago Chicago Indianapolis Santa Clara Santa Clara Ventura Adams New Haven District of Columbia Broward Hillsboroug h DeKalb Ada Cook Cook Cook Marion SAN JOSE JACKSON ST SAN JOSE JACKSON ST NULL ALSUP ELEMENTARY SCHOOL- COMMERCE CITY NULL MCMILLAN PAMS NULL SYDNEY 2390-B WILDCAT ROAD, DECATUR, GA NULL Lawndale Comm-Ed COM ED MAINTENANCE BLDG COM ED MAINTENANCE BLDG IN PARKING LOT NEXT TO POLICE STATION 06 06 06 08 09 11 12 12 13 16 17 17 17 18 085 085 111 001 009 001 011 057 089 001 031 031 031 097 0005 0005 2002 0006 0027 0043 1002 3002 0002 0010 0076 0076 0076 0078 Medium High High High High High High High High Medium High High High High 156B Jackson Street; San Jose 7801 Lawndale Population Exposure Population Exposure (Chicago, IL Northwester n Indiana) 37.348500 37.348500 34.277500 39.825739 41.301111 38.918889 26.082778 27.965650 33.688007 43.607568 41.751369 41.751369 41.751369 39.811097 -121.895000 -121.895000 -118.684722 -104.936987 -72.902778 -77.012500 -80.237778 -82.230400 -84.290325 -116.348434 -87.713745 -87.713745 -87.713745 -86.114469 if available N020ZPM10 PM25 PM25species PM2.5, S02, N02, Oz Bart Ostro Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Joel Kaufman Kaz Ito Antonella Zanobetti Kaz Ito CAOEHHA NYU NYU NYU NYU NYU NYU NYU NYU NYU Univ. ol Wash. MESA Air project NYU Harvard University NYU For time-series and case-crossover, the longer into the future, the better the analyses. 8/1/2004-7/31/2014 2000-2003 ------- Kansas Kansas Louisiana Maryland Maryland Massachus etts Massachus etts Massachus etts Michigan Michigan Minnesota Minnesota Minnesota Kansas City Kansas City Baton Rouge Essex Essex Boston Boston Chicopee Allen Park Detroit Minneapolis Minneapolis Minneapolis Wyandotte Wyandotte East Baton Rouge Baltimore Baltimore Suffolk Suffolk Hampden Wayne Wayne Hennepin Hennepin Hennepin JFK JFK NULL Essex ESSEX DUDLEY SQUARE ROXBURY DUDLEY SQUARE ROXBURY NULL NULL Phillips ANDERSON SCHOOL -PHILLIPS NEIGHBORHOOD ANDERSON SCHOOL -PHILLIPS NEIGHBORHOOD 20 20 22 24 24 25 25 25 26 26 27 27 27 209 209 033 005 005 025 025 013 163 163 053 053 053 0021 0021 0009 3001 3001 0042 0042 0008 0001 0001 0963 0963 0963 Medium High Medium High High High High High High High High High High Woodward And Franklin Roads Essex 2727 10th St. Minneapolis Population Exposure (Baltimore, MD) Population Exposure (Minneapoli s-St. Paul, MN) 39.117500 39.117500 30.461111 39.310833 39.310833 42.329444 42.329444 42.194460 42.228611 42.228333 44.955396 44.955396 44.955396 -94.635556 -94.635556 -91.176944 -76.474444 -76.474444 -71.082778 -71.082778 -72.555711 -83.208333 -83.209167 -93.25827 -93.258270 -93.258270 PM2.5, S02, N02, Oz, CO PM2.5 Kaz Ito Antonella Zanobetti Kaz Ito Joel Kaufman Kaz Ito Kaz Ito Antonella Zanobetti Kaz Ito Kaz Ito Antonella Zanobetti Joel Kaufman Kaz Ito Antonella Zanobetti NYU Harvard University NYU Univ. ol Wash. MESA Air project NYU NYU Harvard University NYU NYU Harvard University Univ. of Wash. MESA Air project NYU Harvard University 2000-2003 8/1/2004-7/31/2014 2000-2003 2000-2003 8/1/2004-7/31/2014 2000-2003 ------- Mississippi Missouri Missouri Montana Nebraska Nevada New Jersey New Jersey New York Gulfport St. Louis St. Louis Missoula Omaha Reno Elizabeth North Brunswick (Township of) New York Harrison St Louis (City) St. Louis City Missoula Douglas Washoe Union Middlesex Bronx BEHIND HARRISON COUNTY YOUTH COURT BLAIR STREET CATEGORY A CORE SLAM PM2.5. BLAIR STREET CATEGORY A CORE SLAM PM2.5. NULL NULL NULL ELIZABETH LAB NEW BRUNSWICK I.S.52 New York North Carolina New York Charlotte Bronx Mecklenbur 9 IS 52 Garinger High School North [Charlotte |Mecklenbur|Garinger High School Carolina North Dakota Ohio Fargo Cleveland g Cass Cuyahoga FARGO NW GT CRAIG 28 29 29 30 31 32 34 34 36 36 37 37 38 39 047 510 510 063 055 031 039 023 005 005 119 119 017 035 0008 0085 0085 0031 0019 0016 0004 0006 0110 0110 0041 0041 1004 0060 High High High Medium High Medium High High High High High High High High E 156th St Bet Dawson And Kelly Population Exposure (New York, NY- Northeaster n New Jersey) 30.390139 38.656300 38.656300 46.874912 41.247222 39.525083 40.641440 40.472790 40.81616 -89.049722 -90.198100 -90.198100 -113.995253 -95.975556 -119.807717 -74.208360 -74.422510 -73.90207 PM2.5, S02, Kaz Ito Antonella Zanobetti Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Kaz Ito Joel NYU Harvard University NYU NYU NYU NYU NYU NYU Univ. ol 2000-2003 8/1/2004-7/31/2014 |N02,Oz JKaufman |Wash. | 40.816160 35.240278 -73.902070 -80.785556 35.240278 -80.785556 46.933754 41 .493955 -96.855350 -81.678542 N02 OZ PM10 CO Kaz Ito Adel Hanna Kaz Ito Kaz Ito Kaz Ito MESA Air project NYU University Of Noth Carolina NYU NYU NYU 01/01/06-12/21/2008 ------- Ohio Oklahoma Oregon Pennsylva nia Pennsylva nia Pennsylva nia Pennsylva nia Rhode Island South Carolina Tennessee Texas Texas Texas Texas Texas Texas Cleveland Tulsa Portland Philadelphia Philadelphia Pittsburgh Pittsburgh Providence Charleston Knoxville Dallas Dallas Deer Park Deer Park El Paso El Paso Cuyahoga Tulsa Multnomah Philadelphi a Philadelphi a Allegheny Allegheny Providence Charleston Knox Dallas Dallas Harris Harris El Paso El Paso GT CRAIG NORTH TULSA - FIRE STATION#24 AT 36TH AND PEORIANR NULL AMS Laboratory AMS Laboratory NULL NULL BUILDING ROOFTOP CHARLESTON PUBLIC WORKS NULL DALLAS HINTON DALLAS HINTON NW OF W. LAMBUTH & DURANT INTERSECTION NW OF W. LAMBUTH & DURANT INTERSECTION CHAMIZAL CHAMIZAL 39 40 41 42 42 42 42 44 45 47 48 48 48 48 48 48 035 143 051 101 101 003 003 007 019 093 113 113 201 201 141 141 0060 1127 0080 0004 0004 0008 0008 0022 0049 1020 0069 0069 1039 1039 0044 0044 High High High High High High High High High Medium High High High High High High 41 .493955 36.204902 45.496667 40.008889 40.008889 40.465556 40.465556 41.807949 32.790984 36.019440 32.819952 32.819952 29.670046 29.670046 31 .765673 31 .765673 -81 .678542 -95.976537 -122.602222 -75.097778 -75.097778 -79.961111 -79.961111 -71.415000 -79.958694 -83.873610 -96.860082 -96.860082 -95.128485 -95.128485 -106.455225 -106.455225 Antonella Zanobetti Kaz Ito Kazlto Kaz Ito Antonella Zanobetti Kazlto Antonella Zanobetti Kazlto Kazlto Kazlto Kazlto Antonella Zanobetti Kaz Ito Antonella Zanobetti Kaz Ito Antonella Zanobetti Harvard University NYU NYU NYU Harvard University NYU Harvard University NYU NYU NYU NYU Harvard University NYU Harvard University NYU Harvard University 2000-2003 2000-2003 2000-2003 2000-2003 2000-2003 2000-2003 ------- Utah Vermont Virginia Washingto n Washingto n Washingto Salt Lake City Burlington Not in a city Seattle Seattle Seattle n West Virginia Wisconsin SLAMS Not in a city Milwaukee Salt Lake Chittenden Henrico King King King Kanawha Milwaukee UTM COORDINATES = PROBE LOCATION ZAMPIERI STATE OFFICE BUILDING, CORNER OF CHERRY STREET NULL BEACON HILL BEACON HILL Beacon Hill NULL DNR SER HQRS SITE California Escondido San Diego NULL California California Fresno Fresno Fresno Fresno NULL 49 50 51 53 53 53 54 55 06 06 06 035 007 087 033 033 033 039 079 073 019 019 3006 0012 0014 0080 0080 0080 0011 0026 1002 0008 0008 High Medium High High High High Medium High Medium Medium High 4103 Beacon Ave. S. 600 E. Valley Pkwy.; Escondido 3425 N First St; Fresno Population Exposure (Seattle- Tacoma- Bellevue, WA) Population Exposure Population Exposure 40.736389 44.480278 37.558333 47.570273 47.570273 47.570273 -111.872222 -73.214444 -77.400278 -122.308596 -122.308596 -122.308596 PM2.5, S02, N02, Oz, CO 38.448611 43.061111 -81.683889 -87.912500 Kaz Ito Kaz Ito Kaz Ito Kaz Ito Antonella Zanobetti Tim Larson Kaz Ito Kaz Ito NYU NYU NYU NYU Harvard University Univ. d Wash. NYU NYU 2000-2003 7/1/2008-6/30/2009 33.127778 -117.074167 if available Bart Ostro CAOEHHA For time-series and 36.781389 36.781389 -119.772222 -119.772222 N020ZPM10 PM25 PM25species if available N020ZPM10 PM25 PM25species Bart Ostro Antonella Zanobetti CAOEHHA Harvard University case-crossover, the longer into the future, the better the analyses. For time-series and case-crossover, the longer into the future, the better the analyses. 2000-2003 ------- California California California California Illinois Illinois Missouri New York Los Angeles Los Angeles Los Angeles Sacramento Chicago Northbrook Not in a city Rochester Los Angeles Los Angeles Los Angeles Sacrament 0 Cook Cook Clay Monroe NULL Los Angeles-North Main Street NULL Springfield Pump Station Northbrook Water Plant Rochester 2 06 06 06 06 17 17 29 36 037 037 037 067 031 031 047 055 1103 1103 1103 0010 0057 4201 0005 1007 Medium High High Medium Medium High High High 1630 N Mainl Population St; Los Angeles 1630 N Main St, Los Angeles 1309 T St.; Sacramento 1745 N. Springfield 750 Dundee Road Yarmouth Re (RG&E Substation) Exposure Population Exposure (Los Angeles- Long Beach, CA MSA) Population Exposure Population Exposure (Chicago, IL Northwester n Indiana) Population Exposure (Chicago, IL Northwester n Indiana) Population exposure 34.066590 34.06659 34.06659 38.558333 41.914733 42.14 39.303056 43.146198 -118.226880 -118.22688 -118.22688 -121.491944 -87.722725 -87.799167 -94.376389 -77.54813 if available N020ZPM10 PM25 PM25species PM2.5, PM10, S02, N02, Oz CO if available N020ZPM10 PM25 PM25species PM2.5 PM2.5, PM10, S02, N02, Oz CO PM2.5, S02, Bart Ostro Antonella Zanobetti Joel Kaufman Bart Ostro Joel Kaufman Joel Kaufman Antonella Zanobetti Philip Hopke CAOEHHA Harvard University Univ. of Wash. MESA Air project CAOEHHA Univ. d Wash. MESA Air project Univ. of Wash. MESA Air project Harvard University Clarkson For time-series and case-crossover, the longer into the future, the better the analyses. 2000-2003 8/1/2004-7/31/2014 For time-series and case-crossover, the longer into the future, the better the analyses. 8/1/2004-7/31/2014 8/1/2004-7/31/2014 2000-2003 6/2006to12/2009 |CO, 03 [University ------- North Asheville Buncombe BOARD OF ED 37 021 Carolina BLDG NW CORNER PARKING LOT North Raleigh Wake NULL 37 183 Carolina North Winston- Forsyth Hattie Avenue 37 067 Carolina Salem Ohio Akron Summit 39 153 Ohio Columbus Franklin 39 049 Ohio Toledo Lucas 39 095 Pennsylva Erie Erie 42 049 nia Pennsylva Harrisburg Dauphin 42 043 nia Washingto Seattle King 53 033 n Other network sites CA Azusa Los Azusa 06 037 Angeles CA Los 06 037 Angeles 0034 Medium 35.609722 -82.350833 N02 OZ PM10 Adel Hanna CO 0014 High 35.856111 -78.574167 N02 OZ PM10 Adel Hanna CO 0022 High 1300 Blk. Population 36.110556 -80.226667 PM2.5, PM 10, Joel Hattie Exposure S02, N02, Oz Kaufman Avenue (Winston- Salem, NC) 0023 High 41.088056 -81.541667 Antonella Zanobetti 0081 High 40.087778 -82.959722 Antonella Zanobetti 0026 High 41.620556 -83.641389 Antonella Zanobetti 0003 High 42.14175 -80.038611 Antonella Zanobetti 0401 High 40.245 -76.844722 Antonella Zanobetti 0057 High 47.563333 -122.338333 Antonella Zanobetti 0002 Low 803 N. Loren HIGHEST OTHER 34.1365 -117.923 PM, NOx Joel Ave., Azusa CO Kaufman 0016 Low 34.1443 -117.85 NOx Joel Kaufman University of 01/01/06 - 12/21/2008 North Caroliana University of 01/01/06 - 12/21/2008 North Caroliana Univ. of 8/1/2004 -7/31/2014 Wash. MESA Air project Harvard 2000-2003 University Harvard 2000-2003 University Harvard 2000-2003 University Harvard 2000-2003 University Harvard 2000-2003 University Harvard 2000-2003 University Univ. of 8/1/2004 -7/31/201 4 Wash. MESA Air projGct Univ. of 8/1/2004 -7/31/2014 Wash. MESA Air project ------- CA CA CA CA CA CA CA CA CA Burbank Los Angeles Reseda Lynwood Pico Rivera Pico Rivera Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Burbank Reseda Lynwood Pico Rivera #1 Pico Rivera #2 06 06 06 06 06 06 06 06 06 037 037 037 037 037 037 037 037 037 0031 0113 1002 1103 1201 1301 1601 1602 1701 Low Low Low Low Low High Low High Low 228 W. Palm Ave., Burbank 18330 Gault St., Reseda 11220 Long Beach Blvd., Lynwood 4144 San Gabriel River Pkwy, Pico Rivera OTHER GENERAL7 BA HIGHEST CO OTHER Population Exposure (Los Angeles, CA) HIGHEST CO Population Exposure (Los Angeles, CA) POPULATIO N OTHER OTHER POPULATIO N MAX PRECUR 33.7861 34.0511 34.176 34.0665 34.1992 33.92899 34.014 34.01407 34.067 -118.246 -118.456 -118.317 -118.226 -118.532 -118.21071 -118.06 -118.06995 -117.751 NOx NOx PM, NOx PM, NOx PM, NOx PM2.5, N02, Oz, CO PM, NOx PM2.5, N02, Oz, CO NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- CA CA CA CA CA CA CA CA CA Pasadena Long Beach Long Beach Lancaster Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Orange Pasadena North Long Beach (Long Beach) South Long Beach 06 06 06 06 06 06 06 06 06 037 037 037 037 037 037 037 037 059 2005 4002 4004 5001 5005 6012 9002 9033 0001 High Low Low Low Low Low Low Low Low 752 S. Wilson Ave., Pasadena 3648 N. Long Beach Blvd., Long Beach 1305 E. Pacific Coast Hwy., Long Beach 43301 Division St., Lancaster, Ca Population Exposure (Los Angeles, CA) HIGHEST CO OTHER MAX OZONE UPWIND BAG OTHER POPULATI ON POPULATI ON OTHER OTHER POPULATIO N POPULATIO N POPULATIO N 34.1326 33.8237 33.7923 33.9228 33.9508 34.3834 34.69 34.6713 33.8306 -118.1272 -118.189 -118.175 -118.37 -118.43 -118.528 -118.131 -118.13 -117.938 PM2.5, N02, Oz, CO PM, NOx PM NOx NOx NOx NOx PM, NOx PM, NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- CA CA CA CA CA CA CA CA CA CA Anaheim Riverside Mission Viejo Palm Springs Riverside Indio Palm Springs Rubidoux (West Riverside) Orange Orange Orange Orange Riverside Riverside Riverside Riverside Riverside Riverside Anaheim-Loara School Riverside-Magnolia Mission Viejo Riverside-Magnolia Big Bear 06 06 06 06 06 06 06 06 06 06 059 059 059 059 065 065 065 065 065 065 0007 1003 2022 5001 0012 1003 2002 5001 8001 9001 Low Low Low Low Low Low Low Low Low Low 1630 W. Pampas Lane 7002 Magnolia Ave., Riverside 26081 Via Pera, Mission Viejo, Ca 92691 Fs-590 Racquet Club Ave, Palm Springs 7002 Magnolia Ave., Riverside 46-990 Jackson St., Indio Fs-590 Racquet Club Ave, Palm Springs 501 W. Valley Blvd., Big Bear City POPULATI ON OTHER POPULATI ON POPULATI ON HIGHEST CO OTHER OTHER HIGHEST CO POPULATI ON POPULATIO N UPWIND BAG OTHER POPULATIO N POPULATIO N OTHER 33.8306 33.6746 33.63 33.9251 33.9208 33.946 33.7085 33.8527 33.9995 33.6764 -117.938 -117.925 -117.675 -117.952 -116.858 -117.4 -116.215 -116.541 -117.416 -117.33 PM, NOx NOx PM NOx NOx PM PM PM, NOx PM, NOx NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- CA CA CA CA CA CA CA CA CA CA Ontario Victorville Fontana San Bernardino Thousand Oaks Piru Simi Valley San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino San Bernardino Ventura Ventura Ventura Ventura Ontario-Fire Station Fontana- Arrow Highway San Bernardino 06 06 06 06 06 06 06 06 06 06 071 071 071 071 071 071 111 111 111 111 0001 0025 0306 1004 2002 9004 0007 0009 1004 2002 Low Low Low Low Low Low Low Low Low Low 1408 Francis St. 14306 Park Ave., Victorville, Ca 14360 Arrow Blvd., Fontana 24302 4th St., San Bernardino, Ca 2323 Moorpark Road, Thousand Oaks 3301 Pacific Avenue, Piru, Ca 93040 5400 Cochran Street, Simi Valley POPULATI ON OTHER POPULATI ON GENERAL/ BA HIGHEST CO HIGHEST CO OTHER OTHER OTHER HIGHEST CO REGIONAL T POPULATIO N REGIONAL T HIGHEST CO OTHER OTHER POPULATIO N POPULATIO N POPULATIO N OTHER 34.895 34.0372 34.51 34.1037 34.1 34.1068 34.21 34.4046 34.4483 34.2775 -117.023 -117.69 -117.33 -117.629 -117.492 -117.274 -118.869 -118.81 -119.23 -118.684 NOx PM PM, NOx NOx PM, NOx PM, NOx PM, NOx PM NOx PM, NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- CA CA CT CT CT CT CT CT CT CT CT El Rio Ventura Ventura Fairfield Fairfield Fairfield Fairfield Fairfield Fairfield Hartford New Haven New Haven 06 06 09 09 09 09 09 09 09 09 09 111 111 001 001 001 001 001 001 003 009 009 2003 3001 0010 0113 1123 2124 3005 9003 1003 0018 0026 Low Low Low Low Low Low Low Low Low Low Low Rio Mesa School, El Rio GENERAL/ BA GENERAL/ BA GENERAL/ BA HIGHEST CO POPULATI ON HIGHEST CO HIGHEST CO GENERAL/ BA GENERAL/ BA HIGHEST CO POPULATI ON POPULATIO N HIGHEST CO HIGHEST CO OTHER POPULATIO N POPULATIO N POPULATIO N HIGHEST CO OTHER 34.2804 34.255 41.1708 41.1836 41 .3991 41 .063 41.1125 41.1183 41 .7847 41 .2938 41.2911 -119.313 -119.142 -73.1947 -73.1902 -73.443 -73.5288 -73.4072 -73.3366 -72.6316 -72.9013 -72.8941 NOx PM, NOx PM PM PM PM PM PM, NOx NOx PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- CT CT CT CT CT DE DE DE DE DE DE Dover Bellefonte Newark Wilmington New Haven New Haven New Haven New Haven New Haven Kent Kent New Castle New Castle New Castle New Castle 09 09 09 09 09 10 10 10 10 10 10 009 009 009 009 009 001 001 003 003 003 003 0027 1123 2008 2123 9005 0002 0003 1003 1007 1012 2004 Low Low Low Low Low Low Low Low Low Low Low State Roac 384, Killens Pond Rd Water St. Dover River Roac Park Lums Pone State Park Univ. Del. North Campus Mlk Blvd Anc GENERAL/ BA HIGHEST CO POPULATI ON HIGHEST CO HIGHEST CO GENERAL/ BA POPULATI ON HIGHEST CO OTHER OTHER HIGHEST HIGHEST CO OTHER OTHER MAX PRECUR OTHER POPULATIO N POPULATIO N POPULATIO N OTHER 41.3011 41.3108 41.3313 41 .5505 41.3411 38.9847 39.155 39.7611 39.5511 39.6919 39.7394 -72.9027 -72.9169 -72.9197 -73.0436 -72.9213 -75.5555 -75.518 -75.4919 -75.7308 -75.7616 -75.558 PM, NOx PM, NOx PM PM PM, NOx PM PM PM PM PM PM, NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- DE DC DC DC DC IL IL Seaford Chicago Chicago Sussex District of Columbia District of Columbia District of Columbia District of Columbia Cook Cook Farr Dormitory Washington 10 11 11 11 11 17 17 005 001 001 001 001 031 031 1002 0025 0041 0042 0043 0014 0022 Low Low Low Low Low Low Medium 350 Virginia Ave Seaford 3300 S Michigan Ave 3535 E. 114th St. OTHER POPULATI ON HIGHEST CO GENERAL/ BA HIGHEST CO POPULATI ON Population Exposure (Chicago, IL Northwester n Indiana) POPULATIO N POPULATIO N HIGHEST CO MAX OZONE 38.6444 38.9752 38.8972 38.8808 38.9188 41 .8342 41.689195 -75.613 -77.0227 -76.9527 -77.0325 -77.0125 -87.6238 -87.539318 PM NOx PM, NOx PM PM, NOx PM PM2.5 Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IL IL IL IL IL Chicago Chicago Chicago Cook Cook Cook Cook Cook S.E. Chicago Mayfair Pumping Stn. Springfield Pump Station 17 17 17 17 17 031 031 031 031 031 0050 0052 0057 0063 0075 Medium Medium Low Low Low 103rd Anc Luella 4850 Wilson Ave. 1745 N. Springfield Source Oriented (Chicago, IL Northwester n Indiana), Population Exposure (Chicago, IL Northwester n Indiana) Population Exposure (Chicago, IL Northwester n Indiana), Highest Concentrati on (Chicago, IL Northwester n Indiana) POPULATI ON HIGHEST CO POPULATI POPULATIO N 41 .709561 41 .967429 41.9147 41 .8772 41 .9641 -87.568576 -87.749819 -87.7227 -87.6344 -87.6586 PM2.5, S02 PM2.5, PM10 PM NOx NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IL IL IL IL IL IL IL IL IL Chicago Mccook Blue Island Schiller Park Summit Des Plaines Northbrook Cicero Cook Cook Cook Cook Cook Cook Cook Cook Cook Lawndale Comm-Ed Northbrook Water Plant 17 17 17 17 17 17 17 17 17 031 031 031 031 031 031 031 031 031 0076 1016 2001 3103 3301 4002 4007 4201 6005 Low Low Medium Low Low Low High Low Low 7801 Lawndale 50th St. Anc Glencoe 12700 Sacramento 4743 Mannheim Rd. 60th St. & 74th Ave. 9511 W. Harrison St 750 Dundee Road 13th St. & GENERAL7 BA HIGHEST CO Population Exposure (Chicago, IL Northwester n Indiana) HIGHEST CO POPULATI ON HIGHEST CO Population Exposure (Chicago, IL Northwester n Indiana) MAX OZONE POPULATI HIGHEST CO POPULATIO N POPULATIO N POPULATIO N POPULATIO N 41.7513 41.8011 41 .663997 41 .9652 41 .7827 41 .8552 42.060278 42.14 41 .8642 -87.7137 -87.8319 -87.696468 -87.8763 -87.8052 -87.7524 -87.863333 -87.7991 -87.7488 PM, NOx PM PM2.5, PM10 PM, NOx PM NOx PM2.5, Oz PM, NOx PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IL IL IL IL IL IL IL IN IN IN IN Naperville Elgin Pittsboro Cook DuPage Kane Lake McHenry Will Will Gibson Hendricks Hendricks Hendricks City Hall 17 17 17 17 17 17 17 18 18 18 18 031 043 089 097 111 197 197 051 063 063 063 8003 4002 0003 1007 0001 1002 1011 0010 0001 0002 0003 Low Low Low Low Low Low Low Low Low Low Low 400 S. Eagle St. 258 Lovell St. Illinois Beach State Park Cr 800 N And Cr 275 E 206 N. Meridian St. High School, POPULATI ON POPULATI ON POPULATI ON EXTREME DO POPULATI ON HIGHEST CO GENERAL/ BA HIGHEST CO HIGHEST CO HIGHEST HIGHEST CO POPULATIO N POPULATIO N OTHER OTHER OTHER 41.6313 41.7711 42.0502 42.4675 42.2214 41 .5266 41.2215 38.2762 39.8769 39.8633 39.8808 -87.568 -88.1525 -88.2802 -87.81 -88.242 -88.1163 -88.1909 -87.5529 -86.4738 -86.4707 -86.5421 NOx PM PM PM PM PM NOx NOx NOx NOx NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IN IN IN IN IN IN IN IN IN East Chicago Gary Gary Griffith Gary Hammond Hammond Indianapolis Lake Lake Lake Lake Lake Lake Lake Lake Marion 18 18 18 18 18 18 18 18 18 089 089 089 089 089 089 089 089 097 0006 0022 0026 0027 1003 1016 2004 2010 0073 High High High High High High High High Low Franklin School Alder &142ndSt 201 Mississippi St., litr Bunker 25th Anc Burr Street Ready Eldon School, 1345 N. Broad St. Ivanhoe School 15th & Gerry Sts Purdue Univ Calumet- Powers Building 6937 1921 Davis St., Robertsdale, Clark H.S. Naval Avionics Center, 6125 E. 16th St. HIGHEST CO HIGHEST CO HIGHEST CO POPULATI ON POPULATI ON HIGHEST CO POPULATI ON POPULATI ON HIGHEST CO POPULATIO N POPULATIO N POPULATIO N POPULATIO N OTHER 41 .6361 41 .6066 41 .573 41 .5466 41 .5888 41 .6002 41 .5852 41 .6783 39.7891 -87.4408 -87.3047 -87.4058 -87.4263 -87.4077 -87.3347 -87.4744 -87.5083 -86.0608 PM PM, NOx PM PM PM PM PM PM NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IN IN IN IN IA IA IA IA IA IA Ogden Dunes (Wickliffe) Evansville Waterloo Clinton Iowa City Cedar Rapids Porter St. Joseph Spencer Vanderburg h Black Hawk Cerro Gordo Clinton Emmet Johnson Linn 18 18 18 18 19 19 19 19 19 19 127 141 147 163 013 033 045 063 103 113 0024 1008 0008 0012 0008 0019 0021 0003 2001 0033 Low Low Low Low Low Low Low Low Low Low 84 Diana Rd/ Water Treatment Plant 425 West Mill Road/ Fire Station #17 Roosevelt St. 2200 East Court 408 E. Linn St. Coggon, Iowa HIGHEST CO OTHER HIGHEST CO HIGHEST CO OTHER OTHER GENERAL/ BA POPULATI ON POPULATI ON POPULATIO N POPULATIO N OTHER POPULATIO N POPULATIO N POPULATIO N POPULATIO N 41.6175 41 .6936 37.9811 38.0216 42.493 43.1616 41 .8749 43.3975 41 .6573 42.2805 -87.1991 -86.2366 -87.0325 -87.5694 -92.3438 -93.2083 -90.1774 -94.8172 -91 .5034 -91 .5269 PM NOx NOx NOx PM PM PM PM PM NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IA IA IA IA IA IA IA IA IA Cedar Rapids Muscatine Emmetsburg Des Moines Des Moines Clive Davenport Linn Muscatine Palo Alto Polk Polk Polk Polk Polk Scott 19 19 19 19 19 19 19 19 19 113 139 147 153 153 153 153 153 163 0037 0015 1002 0030 0058 0059 2510 2520 0014 Low Low Low Low Low Low Low Low Medium 1599 Wenig RdNe 1409 Wisconsin Iowa Lakes Community College 1907 Carpenter, Des Moines Iowa Se 18th Anc Scott, National By- products 9401 Indian Hills Drive, Clive 50325 Scott County Park POPULATI ON POPULATI ON GENERAL/ BA OTHER POPULATI ON HIGHEST CO OTHER OTHER GENERAL/ BA POPULATIO N POPULATIO N OTHER POPULATIO N POPULATIO N POPULATIO N 42.0083 41 .4008 43.1233 41 .603 41 .6077 41 .5833 41 .6027 41 .6647 41 .6991 -91 .6786 -91 .0677 -94.6933 -93.643 -93.5719 -93.5838 -93.7477 -93.6141 -90.5219 PM PM PM PM NOx PM PM PM NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- IA IA IA IA MD MD MD MD MD MD Davenport Davenport Clarion Fort Meade (U.S. Army) Glen Burnie Riviera Beach Cockeysville Essex Scott Scott Story Wright Anne Arundel Anne Arundel Anne Arundel Anne Arundel Baltimore Baltimore Davidsonville Family Recreation Center 19 19 19 19 24 24 24 24 24 24 163 163 169 197 003 003 003 003 005 005 0018 0019 2530 0004 0014 0019 1003 2002 1007 3001 Medium Medium Low Low High Low High Low High High 3029 N Division St. Davenport 300 Wellman St. Davenport 2446 Quincy Ave. Clarion Queen Anne And Wayson Roads 9001 YStreet.Ft.M eade 7409 Balto And Annapolis Blvd 8515 Jenkins Rd Riviera Beach Padonia E.S. 9834 Greenside Dr. Cockeysv 600 Dorsey Avenue, POPULATI ON POPULATI ON OTHER GENERAL/ BA POPULATI ON GENERAL/ BA POPULATI ON GENERAL/ BA GENERAL/ BA HIGHEST CO SOURCE ORI POPULATIO N POPULATIO N POPULATIO N POPULATIO N HIGHEST CO MAX PRECUR 41.55 41.5177 42.0413 42.6953 38.9025 39.1011 39.1695 39.1597 39.4608 39.3108 -90.6 -90.6186 -93.6138 -93.6559 -76.653 -76.7294 -76.6279 -76.5116 -76.6311 -76.4744 PM PM PM PM PM PM, NOx PM PM PM PM, NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. Air 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- MD MD MD MD MD MD MD MD MD Edgewood Rockville Beltsville Greater Upper Marlboro Hagerstown Baltimore Baltimore Baltimore Cecil Harford Montgomer y Prince George's Prince George's Washington Baltimore Baltimore Baltimore NEPS NWPS SEPS 24 24 24 24 24 24 24 24 24 015 025 031 033 033 043 510 510 510 0003 1001 3001 0030 8003 0009 0006 0007 0008 Low High Medium Medium Medium Low Medium Medium Medium 4600 Telegraph Road, Fairhill, Cecil Co. Edgewood Army Chem Center, Waehli Road Lathrop E. Smith Env.Ed Center 51 10 Me Howard University's Beltsville Laborator P.G. Co. Equestrian Cntr, 14900 Pennsylv 18530 Roxbury Road, Hagerstown N E Police Sta, 1900 Argonne Dr, Balto N W Police Station 5700 Reistertown Rd. S E Police POPULATI ON HIGHEST CO POPULATI ON GENERAL/ BA POPULATI ON POPULATI ON Population Exposure (Baltimore, MD) Population Exposure (Baltimore, MD) Population POPULATIO N HIGHEST CO 39.7011 39.41 39.1144 39.0552 38.8119 39.5655 39.340556 39.344444 39.28768 -75.86 -76.2966 -77.1069 -76.8783 -76.7441 -77.7219 -76.582222 -76.685278 -76.547616 PM PM PM NOx PM PM PM2.5 PM2.5 PM2.5 Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Air 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- MD MD MD MN MN MN MN MN MN Baltimore Baltimore Baltimore Blaine Cloquet Rosemount Apple Valley Baltimore (City) Baltimore Baltimore (City) Anoka Carlton Dakota Dakota Dakota Dakota Oldtown 24 24 24 27 27 27 27 27 27 510 510 510 003 017 037 037 037 037 0035 0040 0049 1002 7416 0020 0423 0470 6018 High Medium High High Low High High High Low Fmc Corp. 1701 E Patapsco Avenue Old Town Fire Station 1100 Hillen Street Anoka County Airport 2289 Co. Rd. J 175 University Rd 12821 Pine Bend Trail 2142 120th Street East 225 Garden View Drive GENERAL/ BA Population Exposure (Baltimore, MD) POPULATI ON GENERAL/ BA GENERAL/ BA POPULATI ON POPULATI ON HIGHEST CO POPULATI ON HIGHEST CO POPULATIO N POPULATIO N SOURCE ORI SOURCE ORI OTHER REGIONAL T 39.2327 39.298056 39.2616 45.1397 46.7052 44.7653 44.775 44.7407 44.75 -76.5797 -76.604722 -76.6375 -93.2076 -92.5236 -93.0324 -93.0627 -93.2372 -92.8877 PM PM2.5 PM NOx NOx NOx NOx PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- MN MN MN MN MN MN MN MN MN Minneapolis Minneapolis Minneapolis St. Louis Park Rochester St. Paul St. Paul Hennepin Hennepin Hennepin Hennepin Mille Lacs Olmsted Ramsey Ramsey Ramsey Richfield Phillips St. Louis Park Red Rock Road St Paul Health Centre 27 27 27 27 27 27 27 27 27 053 053 053 053 095 109 123 123 123 0961 0963 1007 2006 3051 5008 0864 0866 0868 High High High High Low Medium Low Medium Medium 7020 12th Ave S, Minneapolis, Mn 27271 Qh St. Mpls 4646 Humboldt Ave. N. 5005 Minnetonka Blvd. Her 67 Box 194 1801 9th Ave S. E. Rochester, Mn 55904 1450 Rec Rock Road, St. Paul, Mn 555 Cedar Street POPULATI ON POPULATI ON POPULATI ON POPULATI ON POPULATI ON GENERAL/ BA POPULATI ON Highest Concentrati on (Minneapoli s-St. Paul, MN) Population Exposure WELFARE RE SOURCE ORI REGIONAL T HIGHEST CO 44.8775 44.9553 45.0418 44.95 46.207 43.9969 44.9919 44.899379 44.952442 -93.2588 -93.2582 -93.2987 -93.3428 -93.7594 -92.4503 -93.183 -93.017155 -93.098475 PM PM PM PM PM PM NOx PM2.5, PM10 PM2.5, PM10 Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. Air 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- MN MN MN MN MN MN MN MO NJ St. Paul Virginia Duluth Duluth Shakopee St. Cloud Fort Lee Ramsey Ramsey Saint Louis Saint Louis Saint Louis Scott Stearns Mercer Bergen Harding High School Fort Lee Library 27 27 27 27 27 27 27 29 34 123 123 137 137 137 139 145 129 003 0871 0872 7001 7550 7551 0505 3052 0001 0003 Medium Low Low Low Low High Low Low High 1540 East 6th Street City Hall Roof 1202 East University Circle 2424 W 5th St 917 Dakota St., Shakopee, Mn 55379 1321 Michigan Ave, St. Cloud Mn 56304 Fort Lee Library.Cente r Avenue Population Exposure (Minneapoli s-St. Paul, MN) POPULATI ON POPULATI ON POPULATI ON HIGHEST CO POPULATI ON POPULATI ON OTHER POPULATI ON POPULATIO N SOURCE ORI 44.961451 44.9311 47.5233 46.8201 46.7666 44.7914 45.5498 40.56 40.8516 -93.035894 -93.156 -92.5363 -92.0894 -92.133 -93.5125 -94.1334 -93.4183 -73.9733 PM2.5 PM PM PM PM PM PM NOx PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NJ NJ NJ NJ NJ NJ NJ NJ NJ Pennsauken (Pensauken) Newark Jersey City Lawrence (Township Of) Trenton North Brunswick (Township Of) East Brunswick (Township Of) Morristown Camden Essex Hudson Mercer Mercer Mercer Middlesex Middlesex Morris 34 34 34 34 34 34 34 34 34 007 013 017 021 021 021 023 023 027 1007 0015 1003 0005 0008 8001 0006 0011 0004 Low High High Low Low Low Medium Medium High Pennsauken Twp; Morris- Delair Wtp Mary Willis Cultural Ctr, 18th Av, Newark 355 Newark Ave.Consolid ated Fire House Rider College;Lawr ence Township 120 Academy Street, Trenton Public Libr. Washington Crossing State Park Cook College, Log Cabin Road R.U. Veg Research Farm 3,Ryders Ln, Newb 16 Early St, Morristown HIGHEST CO POPULATI ON HIGHEST CO HIGHEST CO HIGHEST CO POPULATI ON HIGHEST CO GENERAL/ BA POPULATI ON POPULATIO N POPULATIO N MAX OZONE POPULATIO N REGIONAL T POPULATIO N POPULATIO N 39.9888 40.7319 40.7254 40.283 40.2222 40.3124 40.4727 40.4621 40.803 -75.0491 -74.2052 -74.0522 -74.7426 -74.7636 -74.8726 -74.4225 -74.4294 -74.4833 PM PM PM NOx PM PM PM NOx PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NJ NJ NJ NJ NJ NY NY Paterson Elizabeth Elizabeth Rahway Phillipsburg New York New York Passaic Union Union Union Warren Bronx Bronx Morrisania NY Botanical Gardens 34 34 34 34 34 36 36 031 039 039 039 041 005 005 0005 0004 0006 2003 0006 0080 0083 High High High High High Medium Medium Health Department 176 Broadway New Jersey Turnpike Interchange 13 Mitchell Building,600 North Broac Street Rahway Fire Dept, 1300 Main Street Pburg Municipal Bldg, 675 Corliss Ave Morrisania Center, 1225 57 Gerarc Ave. 200th Street And Southern Blvd POPULATI ON HIGHEST CO HIGHEST CO POPULATI ON POPULATI ON Population Exposure (New York, NY- Northeaster n New Jersey) Population Exposure (New York, NY- Northeaster n New Jersey) POPULATIO N POPULATIO N 40.9186 40.6414 40.673 40.606 40.6872 40.83608 40.86586 -74.1677 -74.2083 -74.2136 -74.2749 -75.1813 -73.92021 -73.88075 PM PM PM PM PM PM2.5 PM2.5, S02, N02, CO Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NY NY NY NY NY NY NY NY NY New York East Meadow Cedarhurst Dutchess Kings Kings Kings Nassau Nassau Nassau Nassau New York 36 36 36 36 36 36 36 36 36 027 047 047 047 059 059 059 059 061 1004 0052 0076 0122 0005 0008 0012 0013 0010 Low Low Low High Low Low Low Low High Jhs 126 424 Leonard St Eisenhower Park.Merrick Av&Old Country R Lawrence High School ,Arling ton Place OTHER HIGHEST CO OTHER POPULATI ON GENERAL/ BA POPULATI ON OTHER OTHER HIGHEST CO POPULATIO N OTHER POPULATIO N HIGHEST CO POPULATIO N POPULATIO N OTHER 41 .6948 40.6415 40.6718 40.7198 40.7432 40.631 40.789 40.7607 40.7394 -73.9144 -74.0183 -73.9782 -73.9478 -73.5854 -73.7347 -73.6364 -73.4906 -73.9861 PM PM PM PM NOx PM PM PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NY NY NY NY NY NY NY NY NY New York New York New York Newburgh New York New York New York New York Orange Queens Queens Queens Queens Queens JHS45 Queens College 2 36 36 36 36 36 36 36 36 36 061 061 061 071 081 081 081 081 081 0062 0079 0128 0002 0094 0096 0097 0098 0124 High Medium High Low Low Low Low Low High Post Office,350 Canal Street School Is 45, 2351 1st Avenue Ps 19 185 1st Avenue 55 Broadway 14439 Gravett Road GENERAL/ BA Population Exposure (New York, NY- Northeaster n New Jersey) POPULATI ON GENERAL/ BA OTHER OTHER GENERAL/ BA GENERAL/ BA GENERAL/ BA HIGHEST CO POPULATIO N POPULATIO N POPULATIO N OTHER SOURCE ORI OTHER 40.7205 40.79937 40.73 41 .4994 40.7779 40.7703 40.7552 40.7842 40.7362 -74.004 -73.93334 -73.9844 -74.0097 -73.8431 -73.8284 -73.7586 -73.8475 -73.8231 PM PM2.5 PM PM PM PM NOx NOx PM, NOx Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air Air 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NY NY NY NC NC NC NC NC NC New York Mamaronec k Burlington Black Mountain Fayetteville Durham Richmond Suffolk Westcheste r Alamance Buncombe Caswell Chatham Cumberlan d Durham 36 36 36 37 37 37 37 37 37 085 103 119 001 021 033 037 051 063 0067 0001 1002 0002 0034 0001 0004 0009 0001 High Low High High Medium Low Medium Medium Medium Susan Wagner Hs, Brielle Ave.S Manor Rd, 5th Avenue & Madison, Thruway Exit 9 827 S Graham & Hopedale Rd 175 Bingham Road Asheville Nc 7074 Cherry Grove Rd, Reidsville Rt 4 Box 62 4533 Raeford Rd Health Dept, 300 E Main GENERAL/ BA POPULATI ON POPULATI ON EXTREME DO POPULATI ON GENERAL/ BA GENERAL/ BA POPULATI ON HIGHEST CO POPULATIO N POPULATIO N POPULATIO N POPULATIO N POPULATIO N 40.5973 40.7458 40.93 36.089 35.6097 36.307 35.7572 35.0414 35.9919 -74.1261 -73.4202 -73.7692 -79.4078 -82.3508 -79.4674 -79.1597 -78.9531 -78.8963 PM PM PM PM PM PM PM PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NC NC NC NC NC NC NC NC Rocky Mount Winston- Salem Winston- Salem Greensboro Charlotte Spruce Pine Candor Edgecomb e Forsyth Forsyth Guilford Guilford Mecklenbur 9 Mitchell Montgomer y North Forsyth H.S. Clemmons Mendenhall 37 37 37 37 37 37 37 37 065 067 067 081 081 119 121 123 0004 0024 0030 0009 0013 0041 0001 0001 Medium Low High Low Low Medium Low Low 900 Springfield Road North Forsyth High School Fraternity Church Road 205 Wiloughby Blvd 1130 Eastway Drive City Hall Summit St 112 Perry Drive, GENERAL/ BA POPULATI ON Population Exposure (Winston- Salem, NC) POPULATI ON Population Exposure (Greensbor o, NC), General/Ba ckground (Greensbor o, NC) OTHER GENERAL/ BA GENERAL/ BA POPULATIO N POPULATIO N POPULATIO N POPULATIO N 35.9335 36.1713 36.026 36.0758 36.109167 35.2402 35.9152 35.26 -77.75 -80.2819 -80.342 -79.7944 -79.801111 -80.7855 -82.0733 -79.84 PM {shut down Jan. '06} PM2.5, Oz PM PM2.5, PM10, Oz NOx PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- NC NC NC NC NC Ohio PA PA PA PA Lumberton Raleigh Boone Goldsboro Dayton Robeson Wake Wake Watauga Wayne Montgomer y Adams Berks Bucks Chester 37 37 37 37 37 39 42 42 42 42 155 183 183 189 191 113 001 011 017 029 0005 0014 0015 0003 0005 0031 0001 0009 0012 0100 Low Medium Low Low Medium High Low Low Low Low 1170 Linkhaw Road 3801 Spring Forest Rd. 361 Jefferson Road, Boone Dillard Middle School, Devereau St GENERAL/ BA GENERAL/ BA POPULATI ON EXTREME DO POPULATI ON EXTREME DO HIGHEST CO OTHER POPULATI ON POPULATIO N MAX OZONE GENERAL/B A POPULATIO N OTHER POPULATIO N REGIONAL T 34.6425 35.8561 35.79 36.2219 35.3692 39.759444 39.92 40.3202 40.1072 39.8344 -78.9902 -78.5741 -78.6197 -81 .663 -77.9938 -84.144444 -77.31 -75.9266 -74.8822 -75.7686 PM PM PM PM PM PM, NOx NOx PM, NOx PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Antonella Zanobetti Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Harvard University Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 2000-2003 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- PA PA PA PA PA PA PA PA PA PA Dauphin Delaware Lackawann a Lancaster Lehigh Luzerne Montgomer y Northampto n Perry Philadelphi a 42 42 42 42 42 42 42 42 42 42 043 045 069 071 077 079 091 095 099 101 0401 0002 2006 0007 0004 1101 0013 0025 0301 0004 Low Low Low Low Low Low Low Low Low Low HIGHEST CO HIGHEST CO HIGHEST CO HIGHEST CO OTHER OTHER OTHER OTHER GENERAL/ BA HIGHEST CO OTHER OTHER OTHER OTHER POPULATIO N POPULATIO N POPULATIO N POPULATIO N POPULATIO N POPULATIO N 40.245 39.8355 41 .4427 40.0466 40.6119 41 .2655 40.1122 40.628 40.4569 40.0088 -76.8447 -75.3725 -75.623 -76.2833 -75.4325 -75.8463 -75.3091 -75.3411 -77.1655 -75.0977 PM, NOx NOx NOx PM, NOx PM, NOx NOx PM, NOx PM, NOx PM, NOx PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- PA PA PA PA SC SC SC SC SC SC Philadelphi a Philadelphi a Philadelphi a York Chesterfiel d Florence Greenville Greenville Greenwood Lexington 42 42 42 42 45 45 45 45 45 45 101 101 101 133 025 041 045 045 047 063 0024 0047 0136 0008 0001 0002 0008 0009 0003 0008 Low Low Low Low Low Low Low Low Low Low POPULATI ON HIGHEST CO HIGHEST CO HIGHEST CO GENERAL/ BA OTHER OTHER GENERAL/ BA OTHER GENERAL/ POPULATIO N POPULATIO N OTHER POPULATIO N POPULATIO N POPULATIO N OTHER POPULATIO N OTHER 40.0763 39.9447 39.9275 39.9652 34.6171 34.1676 34.8404 34.901 34.2145 34.0528 -75.0119 -75.1661 -75.2227 -76.6994 -80.1987 -79.8504 -82.4029 -82.313 -82.1731 -81.1549 PM PM PM PM, NOx PM PM PM PM PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- sc sc Texas VA VA VA VA VA Wl Wl Port Arthur Madison Richland Spartanbur 9 Jefferson Arlington Fairfax Fairfax Fairfax Loudoun Dane Dane 45 45 48 51 51 51 51 51 55 55 079 083 245 013 059 059 059 107 025 025 0019 0010 0022 0020 0030 1005 5001 1005 0025 0047 Low Low High Low Low Low Low Low Low Low City Well #6, 2557 OTHER OTHER POPULATI ON GENERAL/ BA POPULATI ON POPULATI ON POPULATI ON HIGHEST CO HIGHEST CO POPULATIO N POPULATIO N HIGHEST CO POPULATIO N POPULATIO N 33.9932 34.9268 29.863889 38.8575 38.7727 38.8375 38.9319 39.0244 43.0819 43.0733 -81 .0241 -82.0052 -94.317778 -77.0591 -77.1055 -77.1632 -77.1988 -77.49 -89.3766 -89.4358 PM PM PM PM PM PM PM PM PM Joel Kaufman Joel Kaufman Antonella Zanobetti Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Harvard University Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. Air 8/1/2004-7/31/2014 8/1/2004-7/31/2014 2000-2003 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- Wl Wl Wl Wl Wl Wl Wl Wl Wl Pleasant Prairie Milwaukee Milwaukee Milwaukee Milwaukee Douglas Grant Jefferson Kenosha Milwaukee Milwaukee Milwaukee Milwaukee Milwaukee 55 55 55 55 55 55 55 55 55 031 043 055 059 079 079 079 079 079 0025 0009 0008 0019 0010 0026 0041 0043 0050 Low Low Low High Medium Medium Medium Medium Low 128 Hwy 61, Potosi Township Chiwaukee Prairie, 11838 First Court Health Center, 1337 So16thSt Dnr Ser Hdqrts, 2300 N M. L. King JrDr Uwm North Campus, 2114 E Kenwood Blvd Virginia Fire Station, 100 W Virginia St HIGHEST CO POPULATI ON HIGHEST CO GENERAL/ BA OTHER HIGHEST CO HIGHEST CO POPULATI ON POPULATI REGIONAL T HIGHEST CO POPULATIO N MAX PRECUR MAX PRECUR SOURCE ORI 46.7302 42.6921 43.1838 42.5047 43.0166 43.0611 43.0752 43.0264 43.0977 -92.0797 -90.6863 -88.9941 -87.8093 -87.9333 -87.9125 -87.8844 -87.9111 -88.0077 PM PM PM PM PM PM, NOx NOx PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. ol Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- Wl Wl Wl Wl Wl Wl Wl Wl Wl Milwaukee Milwaukee Grafton Somerset Waukesha Milwaukee Milwaukee Ozaukee Ozaukee St. Croix Sauk Taylor Waukesha Waukesha 55 55 55 55 55 55 55 55 55 079 079 089 089 109 111 119 133 133 0059 0099 0008 0009 1002 0007 8001 0027 0034 Medium Medium Medium Medium High Low Low Medium Low Federal Aviation Adm, 4942 S 16thSt Milw Fire DeptHq, 711 W Wells St Grafton, Hwy32 And I43 Harrington Beach State Park, 531 HwyD Hwy 64, Somerset Town Hall Devils Lake State Park, E12886 Tower Rd 1 Mi E. Perkinstown OnSr.M 1310 Cleveland Ave GENERAL/ BA HIGHEST CO HIGHEST CO GENERAL/ BA POPULATI ON GENERAL/ BA GENERAL/ BA HIGHEST CO POPULATI ON HIGHEST CO POPULATIO N POPULATIO N HIGHEST CO REGIONAL T OTHER POPULATIO N OTHER 42.955 43.0397 43.343 43.498 45.1244 43.4355 45.2038 43.0202 43.0072 -87.9341 -87.9205 -87.9208 -87.81 -92.6625 -89.6802 -90.6 -88.215 -88.2297 PM PM PM PM PM PM PM PM PM Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Joel Kaufman Univ. of Wash. MESA Air E°M___ Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air project Univ. of Wash. MESA Air E2Ei___ Univ. of Wash. MESA Air project Univ. of Wash. MESA Air E°M___ Univ. of Wash. MESA Air project Univ. of Wash. project 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 8/1/2004-7/31/2014 ------- United States Office of Air Quality Planning and Standards Publication No. EPA 452/S-08-001 Environmental Protection Health and Environmental Impacts Division December 2008 Agency Research Triangle Park, NC ------- |