6
Measuring the Progress and Assessing the Benefits of AQM

INTRODUCTION

Implementation of the Clean Air Act (CAA) via the procedures and methods described in Chapters 2 through 5 has resulted in significant improvements in air quality in the United States, according to the U.S. Environmental Protection Agency (EPA), and these improvements have had a positive impact upon public health and welfare. EPA estimated that emissions of the six criteria pollutants have decreased by about 30% over the past three decades, despite sizable increases in population, energy use, and gross domestic product (see Figure 1-4 in Chapter 1). The estimated benefits of these reductions are substantial; they include an estimated 100,000 to 300,000 fewer premature deaths per year and 30,000 to 60,000 fewer children each year with IQs below 70 (EPA 1997). In economic terms, these benefits have been estimated to amount to trillions of dollars. In this chapter, the committee discusses how such estimates of the progress and benefits of AQM in the United States are made, the uncertainty of the estimates, and what can be done to reduce the uncertainty.

MONITORING POLLUTANT EMISSIONS

Direct Measurement

The most direct way to confirm that specific emission-control technologies are working effectively is to measure the rate at which pollutants are released from relevant sources. However, with a few exceptions, pollutant



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 216
Air Quality Management in the United States 6 Measuring the Progress and Assessing the Benefits of AQM INTRODUCTION Implementation of the Clean Air Act (CAA) via the procedures and methods described in Chapters 2 through 5 has resulted in significant improvements in air quality in the United States, according to the U.S. Environmental Protection Agency (EPA), and these improvements have had a positive impact upon public health and welfare. EPA estimated that emissions of the six criteria pollutants have decreased by about 30% over the past three decades, despite sizable increases in population, energy use, and gross domestic product (see Figure 1-4 in Chapter 1). The estimated benefits of these reductions are substantial; they include an estimated 100,000 to 300,000 fewer premature deaths per year and 30,000 to 60,000 fewer children each year with IQs below 70 (EPA 1997). In economic terms, these benefits have been estimated to amount to trillions of dollars. In this chapter, the committee discusses how such estimates of the progress and benefits of AQM in the United States are made, the uncertainty of the estimates, and what can be done to reduce the uncertainty. MONITORING POLLUTANT EMISSIONS Direct Measurement The most direct way to confirm that specific emission-control technologies are working effectively is to measure the rate at which pollutants are released from relevant sources. However, with a few exceptions, pollutant

OCR for page 216
Air Quality Management in the United States emissions are not routinely monitored in the United States. One notable exception is the congressional requirement for continuous emissions monitoring (CEM) of sulfur dioxide (SO2), nitrogen oxides (NOx), and particulate matter (PM) from any source regulated under the acid rain provisions of the 1990 Amendments to CAA. As discussed in Chapter 5, the inclusion of such monitoring is viewed as being essential to ensuring the success of the cap-and-trade mechanism incorporated into the legislation. Moreover, the application of CEM has provided direct evidence of substantial reductions in SO2 emissions from utilities since the implementation of the acid rain controls (see Figure 5-1 in Chapter 5). Inspection and maintenance (I/M) programs for motor vehicles, which were mandated in the CAA, could serve, in principle, as a check on the effectiveness of mobile-source emission controls. However, as discussed in Chapter 4, the effectiveness of I/M programs has been limited because of shortcomings in program design and effectiveness, and public resistance to such programs in some areas of the country. There are a number of reasons why emissions are not routinely monitored. There are myriad stationary and area sources that contribute to pollution, and technologies are not available to monitor their emissions routinely and reliably. Given the resources and measurement technologies available to the AQM system, a program that attempted to monitor emissions comprehensively through direct measurement would be unrealistically expensive and complex. In addition, efforts by the government to monitor certain types of emissions on a continuous basis (for example, mobile emissions) might be viewed by some as an unacceptable invasion of privacy. On the other hand, the application of new technologies and creative measurement strategies could help to make the task more tractable and less invasive. For example, a number of emerging technologies and methods could be deployed to augment I/M for mobile emissions. Remote sensors have been used to identify high-emitting vehicles without inconveniencing motorists or interfering with traffic (Stedman et al. 1997; Bishop et al. 2000); on-board diagnostic systems are being developed to automatically monitor and document problems that lead to increased emissions from individual motor vehicles; and standard air quality monitors could be deployed inside tunnels and along roadways to help characterize in-use emissions from fleets of vehicles (Kean et al. 2001). As discussed in Chapter 5, CEM technologies are very valuable in tracking stationary-source emissions and could be used more widely, but the development of a broader range of CEM systems has been slow. Using Ambient Concentrations to Confirm Emission Trends EPA estimates that nationwide emissions of volatile organic compounds (VOCs), SO2, PM, carbon monoxide (CO), and lead (Pb) have decreased

OCR for page 216
Air Quality Management in the United States substantially since the early 1980s; the decrease in NOx emissions is estimated to be more modest (see Table 6-1, part A). These emission decreases can be reasonably ascribed to the promulgation of emission controls related to the nation’s implementation of the CAA requirements. Because the importance of NOx for ozone (O3) control was not recognized until the late 1980s or early 1990s, the slower pace of NOx emission decreases, compared with other pollutants, is probably due to the later implementation of NOx controls. The decline of pollutant emissions during a period of substantial growth in population, energy consumption, and gross domestic product in the United States is cited by EPA and others as evidence of the substantial progress of the AQM system. However, the trends listed in Table 6-1, part A, have been developed from inherently uncertain emission inventories (see Chapter 3), so significant uncertainties must also be attached to the emission trends portrayed in EPA’s reports. Because of such uncertainties, a technically robust AQM system should have mechanisms in place that could TABLE 6-1 Summary of EPA’s Trends in Estimated Nationwide Pollutant Emissions and Average Measured Concentrations Pollutant 1983–2002 1993–2002 A. Changes in Estimated Pollutant Emissions, %a NOx –15 –12 VOC –40 –25 SO2 –33 –31 PM10b –34c –22 PM2.5b No trend data available –17 CO –41 –21 Pb –93 –5 B. Changes in Measured Ambient Pollutant Concentrations, % NO2d –21 –11 O3 1-hr –22 –2 O3 8-hr –14 +4 SO2 –54 –39 PM10 No trend data available –13 PM2.5 No trend data available –8e CO –65 –42 Pb –94 –57 aNegative numbers indicate reductions in emissions and improvements in air quality. bIncludes only directly emitted particles. cBased on percentage change from 1985. dThe trends in NO2 should be viewed cautiously because of potential artifacts from the instrumentation. Also, because NO is readily converted to NO2 in the atmosphere, ambient monitoring data are reported as NO2. eBased on percentage change from 1999. SOURCE: Adapted from EPA 2003.

OCR for page 216
Air Quality Management in the United States independently establish the validity and accuracy of emission trends derived from emission inventories. Given the current inability to monitor all emissions comprehensively, trends cannot be verified directly. However, in principle, verification can be done indirectly by using long-term measurements of primary pollutant concentrations in the ambient air. The underlying assumption of this approach is that, all things being equal, there should be an approximate 1:1 relationship between a change in the total emissions of a primary pollutant (for example, CO and SO2) and a change in the pollutant’s average atmospheric concentration. Although this approach is straightforward in concept, it can be difficult to implement without an appropriately designed network of pollutant monitors. Since the 1980s, the United States has had an extensive air quality monitoring network that routinely measures the concentrations of selected air pollutants in some locations. (The objectives and features of this network are discussed in some detail in the next section.) Trends in the concentrations of relevant air pollutants derived by EPA from data obtained from this network are listed in Table 6-1, part B. Initial inspection of these trends indicates qualitative consistency with the estimated pollutant emission trends discussed above—that is, the trends of both emissions and concentrations are downward. However, a more detailed quantitative comparison of the trends indicates significant inconsistencies. As illustrated in Figure 6-1, the downward trends in the average pollutant concentrations tend to be significantly greater than those of the emissions.1 That result could be interpreted to mean that pollutant emissions have decreased more than estimated from emission inventories. However, there are other viable explanations. Significant uncertainties can exist in the concentration trends derived from the nation’s air quality monitoring network (see later discussion on trends analysis). More important, the nation’s air quality monitoring network was not designed to track nationwide emission trends or evaluate emission inventories; instead, it was largely designed to monitor urban pollution levels and compliance with National Ambient Air Quality Standards (NAAQS). As a result, most of the sites in the air quality monitoring network are urban; thus, the trends derived from them are more representative of urban pollution trends than national trends. Because many emission controls on stationary sources between 1970 and 1990 were aimed at urban emissions, urban areas might be expected to have larger decreases in pollutant concentrations than those seen overall. In any event, it would appear that air quality monitoring data provide qualitative but not quantitative confirmation that pollutant emission trends are downward (especially in urban areas) in the United States. 1   PM10 is an exception; however, note that the emissions change plotted in Figure 6-1 for PM10 is based on the estimated change from 1985 to 2002 and not 1983 to 2002.

OCR for page 216
Air Quality Management in the United States FIGURE 6-1 Scatterplot of estimated trends in pollutant emissions derived from emission inventories and changes in average pollutant concentrations derived from air quality monitoring networks. The squares indicate the trends for the period 1983–2002 (except for PM10 emissions, which are for the trend period 1985–2002) and the circles for 1993–2002. SOURCE: Data from EPA 2003. MONITORING AIR QUALITY The 1977 Amendments of the CAA state that the Administrator shall promulgate regulations establishing an air quality monitoring system throughout the United States which (1) utilizes uniform air quality monitoring criteria and methodology and measures such air quality according to a uniform air quality index, (2) provides for air quaity monitoring stations in major urban areas and other appropriate areas throughout the United States to provide monitoring such as will supplement (but not duplicate) air quality monitoring carried out by the States required under any applicable implementation plan, (3) provides for daily analysis and reporting of air quality based upon such uniform air quality index, and (4) provides for record keeping with respect to such monitoring data and for periodic analysis and reporting to the general public by the Administrator with respect to air quality based upon such data.

OCR for page 216
Air Quality Management in the United States In response to this and subsequent congressional mandates, EPA has overseen the development and operation of several national monitoring networks. These networks generally fall into one of two categories: ambient air quality monitoring networks that measure the atmospheric concentrations of pollutants at various locations, and deposition networks that measure the rate at which pollutants are deposited on the earth’s surface. Collectively, these networks provide the best and most detailed (although by no means comprehensive) data available today for assessing the progress of the AQM system. Because air quality and atmospheric deposition can exhibit large daily, seasonal, and interannual variations independent of any changes in pollutant emission rates, long-term monitoring is needed to detect and interpret air quality trends and thereby determine the effectiveness of regulations. The requirement for complete, precise, and accurate long-term monitoring of air quality presents significant technical challenges that require substantial investments in financial and human resources. In the United States, more than $200 million is spent annually to support air monitoring (EPA 2002p). In addition to providing an essential performance measure of the effectiveness of air quality regulations, air quality monitoring networks provide critically important information to scientists attempting to advance the understanding of the causes and remedies of air pollution. Thus, these networks represent a valuable and significant national resource. The major federally supported monitoring networks for atmospheric composition and deposition operating in the United States are summarized below and in Table 6-2. An example of the spatial distribution of one of these networks (used to monitor O3) is presented in Figure 6-2. Atmospheric Composition Monitoring Networks National, State, and Local Air Monitoring Stations The CAA requires every state to establish a network of air monitoring stations for the criteria air pollutants. These networks are called the state and local air monitoring stations (SLAMS). SLAMS currently consist of approximately 4,000 monitoring stations whose size and distribution is largely determined by the needs of state and local air pollution control agencies to meet their respective state implementation plan (SIP) requirements (for example, to assess their NAAQS attainment status). A subset of the SLAMS monitoring sites (1,080 stations) comprises the national air monitoring stations (NAMS). They are located in urban and multisource areas to provide air quality data in areas where the pollutant concentrations and population exposures are expected to be the highest.

OCR for page 216
Air Quality Management in the United States TABLE 6-2 Summary of Major U.S. Monitoring Networks Network Start Year Lead Agency No. of Sites Measurements National air monitoring stations and state and local air monitoring stations (NAMS/SLAMS) 1980 EPA ~4,000 Continuous O3, NO2, CO, and SO2 measurements; PM10 and PM2.5 and total suspended particulates (TSP) at least once every sixth day; other measurements taken at selected sites Photochemical assessment monitoring stations (PAMS) 1994 EPA 85 O3, NO, NO2, surface meteorological conditions (NOy at some sites); 56 VOCs; some sites include 3 carbonyl compounds. PM2.5 networks 1999 EPA 1,100 1,100 federal reference method: mass (every third day, 5% daily); 300 continuous mass (hourly); 200 speciation (every third day) Interagency Monitoring of Protected Visual Environments (IMPROVE) 1987 NPS EPA Increasing with time, ~160 All sites: every third day elemental and organic C, SO42–, NO3–, Cl–, elements between Na and Pb, and PM2.5 and PM10 mass Selected sites: hourly light-scattering and/or extinction coefficient, humidity, temperature; photographic scene monitoring Clean Air Status and Trends Network (CASTNet) 1990 EPA 84 (additional sites scheduled to begin operation) Weekly particulate SO42–, NO3–, NH4+, gaseous HNO3 and SO2; and continuous O3; meteorological conditions for calculating dry deposition rates

OCR for page 216
Air Quality Management in the United States National Atmospheric Deposition Program and National Trends Network (NADP/NTN) 1978 Cooperatorsa ~250 Weekly measurements of wet deposition: pH, SO42–, NO3–, NH4+, Cl–, Ca2+, Mg2+, Na+, and K+, precipitation amounts NADP/Mercury Deposition Network (NADP/MDN) 1996 — ~80 Total mercury and sampling period precipitation; some sites report methylmercury Atmospheric Integrated Research Monitoring Network (AIRMoN) 1984 NOAA Wet: 9 Dry: 5 Research and measurements of wet deposition, including SO42–, NO3–, NH4+, PO43–, Cl–, Ca2+, Mg2+, Na+, and K+; dry deposition, including HNO3 and SO2 and particulate SO42–; NO3–, NH4+; measurements designed to provide information at greater temporal resolution Gaseous Pollutant Monitoring Network (GPMN) 1986 NPS ~30 national parks (multiple stations in some) Continuous O3 and meteorological data; some sites measure continuous SO2, CO, NO/NOy, and periodically VOCs aNADP has several hundred cooperators that provide monetary and in-kind support, including several federal, state, local, tribal, and nongovernmental programs. SOURCES: NAMS/SLAMS 2002; PAMS 2002; GPMN 2002; IMPROVE 2002; AIRMoN 2002; CASTNet 2002; NADP/MDN 2002; NADP/NTN 2002.

OCR for page 216
Air Quality Management in the United States FIGURE 6-2 Locations of surface O3 monitoring sites and ozonesonde sites in North America. SOURCE: NARSTO 2000. Reprinted with permission; copyright 2000, EPRI, Palo Alto, CA. Photochemical Assessment Monitoring Stations The CAA Amendments of 1990 required EPA, in partnership with state and local agencies, to carry out more extensive monitoring of O3 and its precursors in areas with persistent exceedances of the O3 NAAQS (those O3 nonattainment areas that are classified as severe or worse). In response, EPA established a network of photochemical assessment monitoring stations (PAMS) in 24 urban areas (see Figure 6-3) to collect detailed data for VOCs, NOx, O3, and local meteorology. The chief objective of PAMS data collection is to provide an air quality database that will assist air pollution control agencies to assess and refine O3 control strategies and specifically to evaluate the trends in and effectiveness of controls implemented on VOC and NOx emissions in an area and to evaluate photochemical models being used by state and local agencies to carry out the attainment demonstrations required for their SIPs (see Chapter 3). In principle, the data from the PAMS network could be extremely useful for the regulatory and scientific communities. However, it appears that the full potential of the data has yet to be realized (NARSTO 2000). Questions concerning the accuracy and specificity of the VOC and NOx concentrations obtained from the PAMS instrumentation have limited the ability of researchers to use the data to empirically assess the relationships between ambient VOC and NOx concentrations and O3 formation (Parrish et al. 1998; Cardelino and Chameides 2000). In spite of those limitations,

OCR for page 216
Air Quality Management in the United States FIGURE 6-3 The PAMS network. Top Panel: Map showing the locations of PAMS sites in 1998. Lower Panel: A schematic showing the general design of the network within an urban area with upwind and downwind sites as well as sites near the largest sources of precursor emissions and sites where O3 concentrations are typically at a maximum. SOURCE: PAMS 2002.

OCR for page 216
Air Quality Management in the United States the PAMS data sets can probably still be used to evaluate trends, but this type of evaluation generally requires a record of measurements of a decade or more, and the PAMS network is just now reaching that level of maturity. Gaseous Pollutant Monitoring Program Complementing the aforementioned urban-focused networks is the Gaseous Pollutant Monitoring Program (GPMP) operated by the U.S. National Park Service. The goal of this network is to provide data on baseline and trend concentrations of O3 and other pollutants in national parks, data that can be used to assist in the development of policies to protect park resources. Interagency Monitoring of Protected Visual Environments The Interagency Monitoring of Protected Visual Environments (IMPROVE) is a monitoring system established to assess visibility levels and trends and to identify sources of visibility impairment primarily in national parks and wilderness areas. Through the IMPROVE program, annual and seasonal spatial patterns in PM and light extinction can be assessed (Box 6-1). BOX 6-1 Monitoring Visibility Visibility impairment from air pollution arises primarily from the scattering and absorption of light by suspended particulate matter (PM) with an aerometric diameter less than 2.5 μm. The contribution of human activities to visibility impairment in wilderness areas and national parks is assessed by monitoring the concentration and composition of PM in these areas and deriving so-called light extinction coefficients and visibility ranges from these measurements. Visibility trends thus derived vary by region and are not fully intercomparable because of the infrequent PM sampling schedules typically used. In the eastern United States, visibility has shown some improvement in the last decade but remains seriously degraded. The mean visual range is about 24 kilometers (km), compared with visibility in a “pristine” atmosphere in the range of 75–150 km. In the western United States, visibility levels do not appear to have changed significantly in the past decade; the mean visual range is 80 km compared with natural visibility of 200–300 km. (Visibility is less in the eastern United States even in the absence of human activities because of the higher humidity, which enhances the ability of particles to scatter light.) There is evidence that over the period 1988–1998, visibility declined in some national parks because of area increases in sulfur emissions (Sisler and Malm 2000). As discussed in the preceding chapters, EPA adopted a new regional haze program in 1999 to help address this problem.

OCR for page 216
Air Quality Management in the United States stressors, such as wet and dry deposition of sulfates, nitrates, other sources of acidification, and ammonium, as well as to track changing climatic and forest stand conditions. Sensitive Surface Waters and Estuarine Systems Deposition of air pollutants can also perturb the chemistry and ecology of sensitive surface waters. Acid deposition is of concern for acid-sensitive lakes and streams, primarily in the Northeast and the mid-Atlantic high-lands (see discussion in Box 6-6). Further, deposition of nitrates and ammonium has been shown to be a major source of nutrients and associated eutrophication of estuarine systems along the eastern seaboard of the United States (Castro and Driscoll 2002). There is also considerable interest in the effects of atmospheric mercury deposition on coastal and other ecosystems. A number of programs are in place to monitor aspects of this problem. The most comprehensive atmospheric deposition monitoring networks are NADP/NTN, NADP/MDN, CASTNet, and AIRMoN (for example, see Figure 6-6). However, these networks do not track dry gaseous ammonia (NH3) deposition, which could be important in estuaries that drain agricultural watersheds (for example, Delaware Inland Bay, Chesapeake Bay, and Neuse River Estuary). Another shortcoming of these programs is that they do not measure organic nitrogen deposition, which may account for 10–30% of the total nitrogen budget to East Coast estuaries (Peierls and Paerl 1997; Whitall and Paerl 2001). In addition to the networks noted above, EPA administers the temporally integrated monitoring of ecosystems (TIME) and long-term monitoring (LTM) programs, which monitor lake and stream chemistry and document changes in response to changing emissions and acidic deposition. There also are several national coastal monitoring networks (for example, the EPA national estuarine program and NOAA’s National Estuarine Research Reserve System [NERRS]), but none directly assesses the effects of atmospheric deposition of pollutants. The NERRS has 20 coastal sites (and one Great Lakes site) and includes a systemwide monitoring program through which data for pH, salinity, dissolved oxygen, temperature, and turbidity are collected (NERRS 2001). No attempt is made to relate the oxygen status of estuaries directly to nutrient loading from either atmospheric or watershed sources except at very basic levels. Similarly, a study that documents estuarine water quality problems related to runoff was published by NOAA’s NERRS (Bricker et al. 1999), but no direct link to atmospheric deposition was made. The National Estuaries Program (NEP) was established to allow local groups to take responsibility for tracking estuarine function, and perhaps as a result, there is no systemwide effort to document the effects of nitrogen or

OCR for page 216
Air Quality Management in the United States mercury deposition. Nitrogen loading as a water quality problem has been addressed for several estuaries (for example, Long Island Sound and Tampa Bay) in the program but only to assess inputs from specific major stationary sources of concern. The best examples of the integration of atmospheric deposition measurements with coastal effects can be found at specific sites where the research community and government agencies are actively engaged in understanding these problems. These efforts are usually multidisciplinary and involve cooperative efforts between university researchers and state or federal scientists to look holistically at a system plagued by problems, such as coastal eutrophication, associated with atmospheric deposition. For example, Neuse River estuary modeling and monitoring program involves researchers from several North Carolina universities as well as cooperation from the North Carolina Department of Environment and Natural Resources and Weyerhauser Corporation. This program has quantified not only the fluxes of atmospherically deposited nitrogen to the system (Whitall and Paerl 2001) but also the ecological effects of nitrogen loading from the atmosphere and from other sources (Peierls and Paerl 1997; Paerl et al. 1998). Similar efforts have been made for the Chesapeake Bay (Russell et al. 1998; CBADS 2001), including the development and application of the Chesapeake Bay eutrophication model (Cerco 2000). Agriculture Because of their direct economic import, the effects of air pollutants on agricultural crops are of particular concern. Probably the most comprehensive assessment of the agricultural losses incurred from exposure to air pollutants was conducted in the 1980s as part of the U.S. National Crop Loss Assessment Network (NCLAN) (Heck et al. 1988; Preston and Tingey 1988). NCLAN goals were (1) to conduct experiments using chambers with tops open to the atmosphere (see Figure 2-12 in Chapter 2) to relate doses of O3 to yields of economically important crops in several major areas in the United States; (2) to estimate actual crop losses over the United States by combining the O3-dose-to-crop-yield information with the data on crop acreage and pollutant levels in each county; (3) to assess dollar losses each year from these pollutant effects; and (4) to create models that relate yields to level of pollutant, water stress, stage of crop development, and temperature, using the results to determine the NAAQS based on injury thresholds. There are three noteworthy aspects of the NCLAN study: (1) the data from this study are over two decades old and still represent the most comprehensive information on O3 effects on crops and are widely used to assess crop losses in the United States from air pollution; (2) the study found that crop yields are depressed substantially when O3 concentrations reach about

OCR for page 216
Air Quality Management in the United States 50–70 parts per billion by volume (ppbv), a level not uncommon in rural areas of the United States during the summer; and (3) the study indicated that crops are best protected from O3 damage by an air quality standard that limits the integrated exposure over the 3–4 month period that the crop is growing rather than the relatively short-term 1-hr or 8-hr primary NAAQS used to protect human health (EPA 1996b). In part, because of the NCLAN study, EPA recommended an alternative secondary standard in its 1996 staff paper to review the O3 NAAQS. This standard would regulate O3 in a seasonal, cumulative manner and was designed to be more protective of vegetation. However, it was never implemented (Heck et al. 1998). The Agricultural Research Service (ARS) in the U.S. Department of Agriculture (USDA) has the Air Quality-Plant Growth and Development Research Unit to carry out the following objectives: determine the separate and combined effects of O3 and elevated CO2 on growth and yield of selected agronomic species; determine whether plant, pest, and pathogen interactions are altered by exposure of plants to these pollutants; and develop techniques for mitigating the problems (USDA-ARS 2002). The effects of O3 and CO2 are studied individually, in combination, and in interaction with other factors associated with changes in global climate, and research is conducted under field, greenhouse, and laboratory conditions. The research unit is also working with crop-growth models for evaluation of air quality impacts on production. The future plans of the unit are to investigate the extent to which rooting media affect soybean response to mixtures of O3 and revised CO2 and to investigate the effects of increased concentrations of CO2 and O3 on a few insect pest populations. Integrated Ecosystem Studies A major problem of impacts research is the difficulty of predicting ecosystem-level responses from short-term studies of young trees grown under controlled conditions. More realistic studies, such as FACE experiments, are needed (Karnosky et al. 2001). To assess the effect of AQM on ecosystems, programs are needed that integrate measurements and analysis across terrestrial and aquatic systems, integrate disciplines, and integrate information across vegetation types and climatic and physiographic regions. To date, a number of research programs have been initiated that attempt to achieve this level of integration. However, for the most part, they are focused on unraveling the links between climate change and ecosystem function and are not considering air pollution per se as a perturbing factor. Some studies of note are described below: AmeriFlux is a network of more than 100 sites dedicated to studying the influence of climate variation, vegetation developmental stage, vegeta-

OCR for page 216
Air Quality Management in the United States tion type, disturbance, and other factors on ecosystem processes controlling the exchange of CO2, water vapor, and energy. Measurements are made of meteorology, CO2, water vapor, and energy exchange, as well as soil and plant processes (respiration, photosynthesis, transpiration, and production) in intact ecosystems. The Free Air CO2 Enrichment (FACE) sites (see discussion in Chapter 2 and Figure 2-13) involve long-term experimental studies on the effects of increased atmospheric CO2 on terrestrial ecosystem processes (for example, photosynthesis, respiration, transpiration, and carbon allocation within plants). Atmospheric CO2 is enriched in natural ecosystems through a pipe distribution system, and ecosystem processes are compared with reference plots. Some of the FACE sites also measure concentrations of O3 or methane or nitrogen deposition. National Science Foundation long-term ecological research (LTER) sites (24 sites) have core areas of research on plant production, population distributions representing trophic structure, soil processes, and disturbance patterns and frequency. A key feature of the first two aforementioned studies is the focus on the effects of changing CO2 concentrations. Comprehensive studies of the effects of air pollutants on terrestrial ecosystems should probably also consider the effects of increased atmospheric CO2. The atmospheric concentrations of CO2 are clearly increasing at a faster rate than has occurred during the evolution of current vegetation (Indermuhle et al. 1999), and it is conceivable that ecosystem response to air pollutant exposures will change as a function of CO2 concentrations. For example, the interaction between increased nitrogen availability from nitrogen deposition, increased atmospheric CO2, and water availability can result in greater effects on carbon uptake and allocation by some species than others (Hungate et al. 1997) and may have implications for changes in plant community composition and biogeochemistry. Action Needed for Enhanced Ecosystem Monitoring, Research, and Risk Assessment Improved and sustained long-term monitoring of ecosystem condition and its relationship to air pollution exposure is essential if the nation’s AQM system is to have a credible capability to protect ecosystems and monitor progress. A unified multiagency network to monitor the ecosystem, air quality, and meteorology will likely be required to accomplish that capability (Farrell and Keating 1998). Intensive ecosystem studies to understand the influence of air pollutants on ecosystem processes and community dynamics have been conducted by independent research programs at academic and research institutions but not as part of a larger

OCR for page 216
Air Quality Management in the United States integrated approach to measuring progress in air quality management. Development of mechanistic ecosystem models that quantify and propagate uncertainty is essential as they represent an integration of direct and indirect effects on ecosystems. Model development should be an integral part of ecosystem studies and modeling. The models should be appropriate for application across regions to estimate response at the scale of air pollutant impacts. Such application requires linking atmospheric models (for example, exposure and meso-scale climate models) with ecosystem process models (for example, carbon, water, nitrogen, and element cycling) and community dynamics models. Finally, selection of measurement end points should be done with the regulatory community to ensure that the measurements will provide a basis for accurate and defensible risk assessment (Laurence and Andersen 2003). For example, end points should go beyond productivity and species composition to include integrity of soil food webs; quantity and quality of water supplied from terrestrial ecosystems; wildlife and recreational values; and transfer and fate of carbon, nutrients, and water within the systems (Laurence and Andersen 2003). In Appendix D, a more detailed discussion of the new and expanded program elements that are needed in ecosystem research and monitoring is presented. The committee notes that similar recommendations can be found in previous reports, such as the EPA “Ecological Research Strategy” (1998e), the White House Office of Science and Technology Policy’s Committee on Environment and Natural Resources proposed framework for integrating the nation’s environmental monitoring and research networks and programs (NSTC 1997a), the NRC’s (2000d) report Ecological Indicators for the Nation, and the Heinz Center’s (2002) report The State of the Nation’s Ecosystems. ASSESSING THE ECONOMIC BENEFITS OF AIR QUALITY IMPROVEMENTS Overview A final step that can be taken to measure the effectiveness of AQM involves an economic assessment of the costs and benefits of the policies and regulations implemented to manage air quality. Because such assessments require that health and welfare benefits, such as fewer cases of asthma in children and increased visibility in our nation’s parks, be assigned a specific monetary value, they are controversial. Nevertheless, economic assessments are intrinsic to the system, because they provide policy-makers with a tangible, quantitative measure of the net gains obtained from AQM (NRC 2002a). Indeed, the 1990 CAA Amendments explicitly direct the EPA administrator to carry out a retrospective assessment followed by

OCR for page 216
Air Quality Management in the United States biannual prospective assessments of the costs and benefits of implementation of the CAA. In an economic assessment, costs and benefits are estimated relative to a baseline. For example in EPA’s most recent retrospective assessment, the baseline was modeled to replicate conditions in the absence of implementation of the CAA (EPA 1997). Progress in air quality and the resulting human health and welfare benefits were then estimated using the air quality data and methods described in the preceding sections of this chapter. In EPA’s first prospective assessment (EPA 1997), conditions in 1990 were chosen as the baseline, and models, instead of air quality data, were used to estimate the air quality benefits from continued implementation of the CAA. A similar approach is taken in the Office of Management and Budget’s (OMB’s) annual cost-benefit analyses of AQM and other federal regulatory programs (for example, OMB 2003a), in assessments of the Acid Rain Program (Burtraw et al. 1998), and in many regionally specific assessments (for example, Winer et al. 1989; Hall et al. 1989, 1992; SCAQMD, 1996; Krupnick et al. 1996; Lurmann et al. 1999; Burtraw et al. 2001a). An alternative approach uses a shorter-term episode or event to establish a baseline. For example, Ransom and Pope (1995) used air quality and epidemiological data from the Utah Valley during two intervals in the late 1980s when a steel mill was shut down (because of a labor dispute) to estimate the health costs incurred by the 200,000 residents of the Utah Valley because of their exposure to normally increased concentrations of PM10. Most economic assessments of AQM in the United States conclude that the economic benefits of implementation of the CAA exceed the economic costs. For example, EPA estimated that the economic benefits of implementation of the CAA between 1970 and 1990 exceeded the costs by a range of about $5–50 trillion (EPA 1997). The estimates from EPA’s prospective analysis for 1990–2010 had smaller but still positive net monetary benefits (EPA 1999a).6 OMB (2003a) estimated a range in the monetary benefits of regulation for 1992 to 2002 to be approximately $121 to $193 billion and a range in costs to be $23 to $27 billion. In virtually all of these assessments, the major monetary benefit arises from deferred mortality associated with reductions in atmospheric PM10 and PM2.5. As discussed below, these estimates raise some challenges to the robustness of the estimated net monetary benefits. 6   Some critics have questioned the existing framework of estimating the benefits and costs of environmental improvements. Parry (1995) and Goulder et al. (1999) point to hidden costs of additional regulations in society (see also the special issue of Journal of Risk and Uncertainty, Vol 8, No. 1, 1994). Williams (2002) points to the hidden benefits of additional regulations. An important issue for EPA is whether it continues to remain open to these criticisms and new ideas.

OCR for page 216
Air Quality Management in the United States Economic Assessments Economic assessments (as they are now generally done) attempt to provide a transparent process for relating economic benefits to a series of specific measures to reduce pollutant emissions, population and ecosystem exposure, and adverse health and welfare effects. Consequently, a wealth of information is provided to decision-makers about the probable future gains from better air quality. Aggregate analyses, such as those mandated in the 1990 CAA Amendments and carried out by EPA, also provide a sense of scale and can support assessments of the relative benefits of alternative regulatory policies (for example, focusing resources on reducing ambient concentrations of criteria pollutants versus hot-spot concentrations of HAPs). The cost-benefit analyses carried out thus far by EPA in response to the directive of the 1990 CAA Amendments have undergone extensive peer review, and EPA’s Science Advisory Board (EPA/SAB 1997) concluded that the overall results were sound and useful for broad policy purposes. The qualitative consistency between the retrospective analyses of EPA and OMB also lends credibility to the analysis. However, concerns remain about aspects of the methods used; some are discussed below. The estimates of the health benefits from improved air quality were based on the same health-effects literature that was used to develop the NAAQS; thus, the assessment does not represent an independent confirmation that such benefits were attained. The dominant contributor to the benefits estimates was the estimate of deferred mortality from reduced population exposure to PM. Thus, the aggregate benefits of the CAA are sensitive to the dose-response relationship adopted for PM and the monetary value assigned to deferred mortality. Different choices for either of those parameters drawn from the mainstream literature can change the net benefits estimated by a factor of nine (Burtraw et al. 2003). The principal study used for benefits estimation to date (Pope et al. 1995) was subjected to intensive reanalysis by an independent team of analysts who in large measure confirmed the results (HEI 2000). However, new analyses of the dose-response relationships conducted in different populations will provide enhanced understanding of these relationships. Several such studies are now under way. The measure of effects used in these analyses—the numbers of lives lost to exposure—is controversial. One important source of controversy is the epidemiological evidence that the very young, the old, and the infirm are most susceptible to environmental exposures, and economic valuation of changes in health status is often based on prime-age adults. Economic estimates based on willingness to pay to avoid small risk to health status

OCR for page 216
Air Quality Management in the United States indicate that, at least for valuing the benefits of environmental improvements for older populations, the usual economic measures are fairly accurate (Krupnick et al. 2002). Other metrics to monetary valuation hinge on cost-effectiveness analysis, which is designed to compare a set of regulatory actions with the same primary outcome, such as number of years of life lost (for example, see NRC 2002a) or the disability-adjusted life-years (DALYs) lost (Murray and Lopez 1996),7 but a universally accepted method has yet to emerge. Gaps in data and knowledge limited the scope of the analysis. HAPs were excluded from the analysis because of a scarcity of information on exposure and health effects. Benefits related to improvements in ecosystem function were excluded because of uncertainties in the monetary value of such improvements. Similarly, the benefits gained from improvements in lung function in children were omitted because a dollar value could not be reliably attached to this outcome. Finally, because monitoring is concentrated in urban areas, the exposure characterization of rural populations is not robust, and consequently, the benefits to rural populations could be either overvalued or undervalued. Any attempt to compare monetized costs with monetized benefits, as in the EPA study, is subject to uncertainty and missing information. Not all costs and benefits can be equally well quantified and monetized. Also, elicitation of estimates can be expensive and analysis often must rely on the transfer of estimates from other settings that are imperfect substitutes. Further, aggregation of measured individual preferences to achieve an estimate of social costs and benefits can involve political decisions that are implicitly controversial. The usefulness of economic assessments such as those carried out by EPA could be enhanced if the connection between specific air quality policies and the dollar values of discrete health and welfare benefits were more transparent (Hall 1996; Krupnick and Morgenstern 2002). For example, a source-specific multipollutant approach could be adopted that would assess the costs and monetized benefits associated with implementing emission controls on specific types of sources (for example, utilities and mobile sources). Alternatively, specific individual measures could be evaluated on 7   These kinds of measures have been important in decisions about allocating public health resources but are controversial with regard to measuring the benefits of social programs. Approaches using estimates of life-years lost and disability-adjusted life years (DALY) are controversial, because they associate a nonmonetary value of change in health status that differs for different members of the population. For example, a senior citizen may have fewer life-years lost than a prime-age adult because of an environmental exposure. Similarly, the DALY approach suggests that older citizens or citizens with a preexisting disease, or disabilities, suffer a smaller loss from premature mortality than younger or healthier citizens (Heinzerling 2000).

OCR for page 216
Air Quality Management in the United States BOX 6-7 Pollution Abatement Cost and Expenditures (PACE) Survey The PACE survey was conducted annually by the U.S. Bureau of the Census from 1973 through 1994, when it was suspended by the bureau for budgetary reasons. In 2000, another survey was carried out. However, time and resource constraints have delayed implementation of subsequent PACE survey cycles. The PACE survey data provide a distinctive tool for evaluating the costs of compliance with environmental regulations. EPA has used the PACE data in its reports on the cost of clean air, Section 812 clean air retrospective cost analysis, numerous sector-specific studies, regulatory impact analyses, analyses of recycling activities, and national studies of environmental protection activities. The PACE survey provides three levels of data. First, the published PACE survey provides aggregate data on pollution abatement spending, both for new capital expenditures and for operating costs. Second, the PACE survey provides abatement spending data at the industry and state level. EPA has used these data for specific sector studies and regulatory impact analyses. Third, the facility-level data collected for the PACE survey can be linked with other census-collected data for those plants and accessed by researchers working at census research data centers around the nation under strict controls to maintain confidentiality. a disagregated basis, because large net benefits might be masking ineffective or inefficient programs or regulations. A regional approach could identify control options that produce the largest benefits in specific parts of the country. Economic assessment, such as that carried out by EPA, requires a reliable assessment of the costs of implementation. The United States probably leads the world in the availability of information about the cost of environmental compliance through data collected in the Pollution Abatement Cost and Expenditures (PACE) Survey. The survey has provided the basis for a large number of scholarly studies that have contributed to improving the effectiveness of environmental regulation. In recent years, the survey has been intermittent due to interruptions in funding. The PACE survey is discussed further in Box 6-7. SUMMARY Strengths of Techniques for Tracking Progress in AQM Continuous emissions monitoring (CEM) systems on electric utilities regulated under the acid deposition program of the CAA have documented substantial reductions in SO2 emissions.

OCR for page 216
Air Quality Management in the United States National atmospheric deposition monitoring stations have documented a reduction in sulfate deposition in the eastern United States. Air quality monitoring networks are a significant national resource and have provided qualitative confirmation of emission-inventory estimates that indicate that pollutant emissions in the United States (and especially in urban areas) have decreased over the past three decades. Analyses of short-term episodes with significant changes in pollutant concentrations confirm that health benefits accrue when air quality is improved. Biomarkers may provide a useful surrogate for documenting trends in population exposure to pollutants over the longer term. EPA’s congressionally mandated economic assessments of the costs and benefits of AQM in the United States are peer reviewed and appear to represent the state of the science. Limitations of Techniques for Tracking Progress in AQM8 The nation’s AQM system has not developed a comprehensive and quantitative program to track emissions and emission trends. Although improvements have been made, accessibility to actual data acquired from the monitoring networks is limited. The AQM system has not completed a comprehensive program to monitor HAPs so that population exposure and concentration trends can be tracked. With the exception of CEM, there is limited ability to quantify stationary sources emissions. The air quality monitoring network for criteria pollutants is dominated by urban sites, limiting its ability to address a number of important issues. Some of the instruments and methods used in the nation’s air quality monitoring networks are inadequate to meet the objectives of the monitoring. Methods used by EPA to calculate pollutant trends from data collected from air quality monitoring networks could be improved. The AQM system has not developed a method and related program to document independently improvements in health and welfare outcomes achieved from improvements in air quality (see Box 6-8). The AQM system has not developed a cohesive program capable of reliably reporting the status of ecosystem effects of air pollution and the response to changing air pollution conditions across regions and the nation. 8   Recommendations are provided in Chapter 7.

OCR for page 216
Air Quality Management in the United States BOX 6-8 The State of the Environment Report—A Sign of a New Paradigm Emerging at EPA? The list of limitations in tracking progress at the end of this chapter presents a sobering picture. While significant resources have been expended in the United States to identify air quality problems and to reduce the pollutant emissions believed to be fostering these problems, there appears to have been a far less concerted effort to track and document objectively and comprehensively the real-world benefits of AQM. Fortunately, a new paradigm appears to be emerging at EPA that recognizes the importance of such an effort. As this committee was completing its work, EPA released its Draft Report on the Environment (2003n). The purpose of the report was to identify and quantify environmental indicators “to better measure and report on progress toward environmental and human health goals and to ensure the Agency’s accountability to the public.” The report represents an important addition to an understanding of the state of the environment and of the limitations of data available for assessing progress. Few data are available that can be used to assess the impacts of AQM measures in the United States on specific human and ecological health goals. Even with large data collection efforts, the establishment of trends attributable to AQM will continue to be a formidable task for diseases such as cancer and asthma because of other strong risk factors. For those environmental indicators that have data, a continuous tracking of the indicators over a span of a decade or more will be needed to establish a trend and its relationship to AQM activities. Thus, long-term support for the activity initiated with the production of EPA’s Draft Report on the Environment is recommended (1) to ensure that EPA is able to produce a series of reports on a biannual or triannual basis, and (2) to provide the scientific and technical basis for environmental indicators that link human and ecological health outcomes with air quality. Cost-benefit analyses of AQM carried out by EPA and others are limited by a lack of relevant data (for example, on the health effects of HAPs) and a reliance on controversial value judgments. Data on costs, such as the PACE survey, are necessary to monitor the costs of CAA compliance and to identify cost-effective policies. However, inconsistent levels of funding in the past have undermined the ability of the PACE survey to provide a long-term data set and to serve as a tool to estimate costs.