Click for next page ( 2


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 1
Summary A primary objective of air quality management in the United States has been to reduce human exposure to carbon monoxide (CO) and other pollut- ants produced from incomplete combustion. Elevated ambient CO concen- trations are due mainly to incomplete combustion of gasoline by light-duty vehicles, such as passenger cars and pickup trucks. CO controls are working. Problems with ambient CO were widespread when automobile emissions regulations began in the 1960s. When the health-based National Ambient Air Quality Standards (NAAQS) for CO were promulgated in 1971, more than 90% of ambient monitors indicated violations. Since then, motor-vehicle emissions controls have greatly reduced ambient CO concentrations. Over the last 5 years, the number of monitors showing CO violations has fallen to only a few, and the monitors that continue to show violations do so much less frequently. For example, Denver, Colorado, which had a persistent CO problem and registered as many as 200 days with violations in the ~ 960s, has not had a violation since 995 . Fairbanks, Alaska, reduced the number of days with violations from The standards for ambient air concentrations of CO were set at 9 parts per million (ppm) for an 8-hour average and 35 ppm for a 1-hour average. These stan- dards were set to protect public health with "an adequate margin of safety," as specified in the Clean Air Act. A violation of an NAAQS occurs on the second exceedance and all subsequent exceedances ofthe standard in a calender year. Only the 8-hour standard of 9 ppm has been exceeded recently in a few locations in the country.

OCR for page 1
~ Managing CO in Meteorological and Topographical Problem Areas well over 100 per year in the early 1970s to zero over the last 2 years. Thus, CO regulation has been one of the greatest success stories in air pollution control, reducing the problem, once widespread, to a few difficult areas. As a result, the focus of U.S. air quality management has shifted to characterizing and controlling other pollutants, such as tropospheric ozone, fine particulate matter (PM2 5),2 and air tonics. However, some locations, such as Anchorage, Alaska, and Lynwood and Calexico, California, continue to be susceptible to occasional violations of the NAAQS for CO. These areas are typically subject to problematic meteorological and topographical conditions that produce severe atmo- spheric inversions in winter.3 Although CO emissions from light-duty vehicles are projectedto decrease in the future, atmospheric inversions and Tow windspeeds prevalent in some locations during winter are extremely effective in trapping the products of incomplete combustion, including CO, emitted at ground level. For example, Fairbanks, Alaska, is subject to extreme atmospheric inversions, at times experiencing inversion strengths as much as 30C per 100 meters of altitude. In addition, Fairbanks is situ- ated in a three-sided bowl, surrounded by the Yukon-Tanana uplands, which can inhibit air circulation. Although it is not heavily populated and has no maj or air-pollution producing industries, Fairbanks ' s meteorological and topographical characteristics make the city susceptible to high ambient CO concentrations in winter. The continuing vulnerability of a few locations to high CO concentra- tions prompted Congress, in its fiscal 2001 appropriations report for the U.S. Environmental Protection Agency (EPA), to ask the National Acad- emy of Sciences to study CO episodes in meteorological and topographical problem areas. The study was requested to address potential approaches to predicting, assessing, and managing episodes of high CO concentrations in such areas. In particular, the committee was to address: 2PM2 5 is a subset of particulate matter that includes particles with an aerody- namic equivalent diameter less than or equal to a nominal 2.5 micrometers. Inversions occur when the temperature of the atmosphere increases with alti- tude. Combined with low windspeeds, this prevents air circulation, because colder air is trapped near the ground by the warmer air above. A temperature increase of several degrees celcius per 100 meters is considered a strong inversion. The stan- dard lapse rate for the troposphere is a decrease of about 6.5C per kilometer (or about 3.6F per 1,000 feet).

OCR for page 1
Summary ~ . Types of emission sources and operating conditions that contribute most to episodes of high ambient CO. Scientific bases of current and potential additional approaches for developing and implementing plans to manage CO air quality, including the possibility of new catalyst technology, alternative fuels, and cold-start technology, as well as traffic and other management programs for motor- vehicle sources. Control of stationary-source contributions to CO air qual- ity also was to be considered. The effectiveness of CO emissions control programs, including comparisons among areas with and without unusual topographical or mete- orological conditions. Relationships between monitored episodes of high ambient CO concentrations and personal human exposure. . The public-health impact of such episodes. Statistically robust alternative methods to assist in tracking prog- ress in reducing CO that bear a relation to the CO concentrations consid- ered harmful to human health. In response, the National Research Council convened the Committee on Carbon Monoxide Episodes in Meteorological and Topographical Prob- lem Areas, which prepared this report. Fairbanks, Alaska, was identified as the subject for a case study in an interim report, which was completed in 2002. The following report is the final report requested by Congress. FINDINGS AND RECOMMENDATIONS Vulnerability to Future Violations Findings Because of a number of factors including differences in topography and temporal variability of local meteorology and emissions rates some areas are especially vulnerable to violations of the 8-hour NAAQS for CO. In geographical areas that have achieved attainment of the NAAQS, it might still be possible for ambient concentrations of CO to sporadically exceed the standard under unfavorable conditions, such as strong winter inversions. This vulnerability, defined as the probability for violation in a future year, depends on both the current CO levels and the variability of air

OCR for page 1
4 Managing CO in Meteorological and Topographical Problem Areas quality indicators. An area in attainment might still be substantially vuIner- able if the variability of its air quality is high. There is evidence that local meteorological conditions conducive to high CO concentrations are sometimes associated with large-scare meteoro- logical and cTimatological phenomena. For example, all recent exceed- ances of the NAAQS for CO in Fairbanks have occurred with a low-pres- sure system in the Gulf of Alaska with cyclonic flow extending over Fair- banks. Although the role that this low-pressure system plays is unclear, it might produce warm winds aloft that reinforce inversions near the ground. In Denver, Colorado, the presence of Tong-term snow cover and light winds can produce conditions conducive to CO buildup in the ambient air. Snow cover diminishes ground-level solar heating, intensifying inversions, and light winds reduce horizontal dispersion. However, over the past decade, Denver has not experienced the combination of these meteorological fac- tors, reducing the city's susceptibility to high CO concentrations. Changes in the frequency of some large-scale meteorological and cTimatological events, such as the frequency of low-pressure systems in the Gulf of Alaska, will influence vulnerability to CO violations. Recommendations Air quality managers typically recognize whether their region is espe- cially vulnerable to future CO violations as a result of increases in vehicle activity, the spatial and temporal variability of meteorology, and problem- atic topography. However, in some cases, air quality planning does not encompass the worst-case combinations of emissions and meteorology. Achieving sufficient emissions reductions to account for these conditions is prudent, particularly in areas with high population growth and/or high meteorological variability, to further reduce the risk of violations. In addi- tion, given that the form of the CO standard defines a violation as the sec- ond and all subsequent exceedances in a calendar year, regions are suscep- tible to violating the standard due to extreme meteorological conditions, which contributes to the difficulties that meteorological and topographical problem areas have in reaching and maintaining attainment. It is also im- portant to investigate how large-scare and local meteorological and climato- logical phenomena can affect the susceptibility of a location to CO buildup in ambient air. Air quality managers should recognize that their regions might be espe- cially vulnerable to future CO violations because of increases in vehicle

OCR for page 1
Summary activity, spatial and temporal variability of meteorology, and problematic topography. The meteorological conditions assumed in current regulatory air quality planning might not encompass the worst-case conditions. Achieving additional emissions reductions is prudent to further reduce the risk of violations, particularly in areas with high population growth and/or high meteorological variability. It also might be important to investigate how large-scale meteorological and climatological phenomena can affect the susceptibility of a location to CO buildup in ambient air. Health Effects Findings In patients diagnosed with coronary artery disease, CO alone has been shown to exacerbate exercise-induced chest pain (angina) in controlled laboratory experiments. Those studies serve as an important part of the basis for the NAAQS. In addition, epidemiological studies have correlated high CO concentrations with other adverse human health effects, such as heart disease, childhood developmental abnormalities, and fetal loss. Some of these effects have been correlated with ambient CO levels below the NAAQS. However, CO is not produced alone, and epidemiological studies have difficulty separating the effects of CO from those of other pollutants that are often associated with CO (such as benzene, 1,3-butadiene, alde- hydes, and various components of PM2.s). Though changes in ambient levels of CO may sometimes correlate with health effects other than exercise-induced angina, there is insufficient evidence that CO is the single direct causative agent in the other effects. Thus, reducing CO alone may or may not reduce the incidence of heart disease, childhood developmental abnormalities, and fetal loss. A significant collateral benefit from reducing CO vehicle emissions standards has been the substantial reduction in accidental deaths due to acute CO poisoning. CO is unique among criteria pollutants4 because it is relevant to both ambient air quality management and public safety. Expo- Criteria pollutants are air pollutants emitted from numerous or diverse station- ary or mobile sources for which NAAQS have been set to protect human health and public welfare. The other criteria pollutants are ozone, particulate matter, sulfur dioxide, nitrogen dioxide, and lead.

OCR for page 1
6 Managing CO in Meteorological and Topographical Problem Areas sures to mobile-source emissions cross these two areas of public-health management. Using computerized death-certificate data, a recent study by researchers at the Centers for Disease Control and Prevention estimated that over 11,000 deaths from accidental CO poisoning have been avoided over the 1 968- ~ 99 ~ time period because of the more stringent vehicle-emis- sions standards. This collateral benefit is not accounted for in EPA's recent report to Congress on the benefits and costs of the Clean Air Act. Recommendations To reduce the potential adverse health effects of CO, the few remaining areas not in attainment need to continue making progress towards meeting and maintaining the CO standard. Public-health issues associated with ambient CO should be emphasized through enhanced public-awareness campaigns. Further study to reveal the effects of CO on the fetus and to separate the effects of CO from its copollutants is encouraged. Also, there should be more toxicology studies of the automobile exhaust mixture. Management and Control of CO Management of CO Fin dings To reach attainment, communities vulnerable to exceeding the health- based NAAQS for CO can implement various local measures to comple- ment federal vehicle emissions standards. These include but are not limited to vehicle emissions inspection and maintenance (~/M) programs, the use of cold weather engine-block heaters in vehicles, and the use of low-sulfur gasoline and oxygenated fuels. These measures reduce CO emissions by reducing the number of malfunctioning vehicles (~/M programs), reducing the length of time before a vehicle's emissions control catalyst is fully functional (cold weather engine-block heaters), and improving the eff~- ciency of the vehicle's emissions control catalyst (Iow-sulfur gasoline). Other measures include mass transit initiatives, traffic management, and bans on wood burning. Air quality issues have tended to be assessed and regulated independ- ently. As such, CO management frequently occurs in isolation, even

OCR for page 1
Summary though other pollutants have similar emissions sources. The committee recognizes that few areas in the country continue to violate CO standards and that the focus of air quality management in the near future will not be on CO but on attaining new PM2 s and tropospheric ozone air quality stan- dards and reducing air tonics. Recommendations . Communities with special CO problems should be encouraged to de- sign locally effective programs. Federal and state assistance should be provided to these communities for characterization and implementation of management options. This should include assistance to improve non-road and stationary-source emissions characterizations. Because the CO stan- dards are health-based, all communities need to be diligent in working toward attaining and maintaining the CO standards. In addition, the pro- grams implemented to reduce CO emissions should be reassessed periodi- cally. This reassessment should include evaluation of their impact on CO, as well as other pollutants, and their impact at low temperatures. CO management should be better integrated into air quality manage- ment. Although the focus of air quality management in the near future will be on other air pollution issues, winter inversion conditions not only affect CO buildup but can also be related to higher concentrations of PM2 5 and some air tonics. In addition, the primary source of CO, fuel-rich operations of light-duty vehicles, is a major source of other pollutants of concern. The committee therefore recommends that EPA assess the relationship of CO to these other pollutants when the CO criteria are updated. Federal Tier 2 and Cold-Start Emissions Standards Findings Federal new-vehicle emissions standards have been effective in reduc- ing CO emissions, including emissions from vehicles operated in cold climates. Emissions from passenger vehicles have been reduced from their pre-contro] levels of over 80 grams per mile (g/mi) in the late 1 960s to the 3.4 g/mi standard implemented in 1981. Also, progressively Tower stan- dards for hydrocarbon emissions, as well as other requirements, have tended to decrease CO emissions levels. Since 1994, new cars have been

OCR for page 1
8 Managing CO in Meteorological arid Topographical Problem Areas required to meet a winter cold-start CO limit, which reduces emissions from vehicles started at cold temperatures (20F). New Tier 2 and California certification standards5 are expected to further reduce hydrocarbon and nitrogen oxides (NOX) emissions. Average CO emissions from vehicles certified to Tier 2 standards are expected to decrease due to the technological improvements in emissions control sys- tems needed to meet the standards, especially the Tower hydrocarbon stan- dards and lower CO standards on some light-duty trucks. In CO problem areas, the decrease in emissions resulting from Federal Tier 2 standards will depend on the mix of vehicles sold and used in these areas. As such, the following uncertainties arise in predicting the continued motor-vehicle emissions reductions in CO problem areas: The sales strategy used by manufacturers to comply with average NOX limits. [f manufacturers tend to sell higher-emitting vehicles in CO problem areas, the improvements in CO emissions will not be as large as those predicted based on national averages.6 The impact of the emissions averaging and trading provisions (which allow vehicle manufacturers to generate, trade, and bank emissions credits) on the fleet of vehicles operating in CO problem areas.7 The effects of Tier 2 requirements on CO emissions at temperatures below 20F. 5Federal Tier 2 emissions standards will be introduced for passenger cars in model-year 2004 and fully implemented by model-year 2007. The standards will require that vehicle manufacturers meet a fleet average NOX limit of 0.07 grams per mile (g/mi) along with lower standards for hydrocarbons. California, which is allowed under the Clean Air Act to adopt its own vehicle emissions standards, has already implemented a similar set of emissions limits. Both the Tier 2 and Califor- nia standards are for vehicles certified at 68-86F. Since 1994, new cars also have been subject to a cold-start CO standard, which requires cars and most light-duty trucks to meet a CO limit of 10 g/mi on a certification test run at 20F. The Clean Air Act Amendments of 1990 include a provision for more stringent cold-start standards to be set if needed. 6The Tier 2 standard is a sales-weighted fleet-averaged standard. Thus, some vehicles that are sold will have emissions greater than the standard, some less than the standard. 7The Tier 2 standard has provisions that allow manufacturers bettering the fleet- averaged standard to generate tradable emissions credits that can be sold to manu- facturers who have not met the fleet-averaged standard. Manufacturers can also bank credits for use in later years.

OCR for page 1
Summa7y 9 The ability of EPA's MOBILE emissions rate model to adequately account for the effects of Tier 2 standards on CO emissions rates.8 Recommendations In the absence of compelling evidence, the committee does not recom- mend tightening the national cold-start standard below 10.0 g/mi or requir- ing that the 10.0 g/mi standard be met at a lower temperature. However, supplemental emissions testing should be undertaken at temperatures below 20F to determine to what extent CO emissions systematically increase as ambient temperature decreases. Testing data should be obtained and ana- Tyzed at 0F and ~ 0F, and should include CO as well as other pollutants (air toxics and PM). The extent of the anticipated reduction in CO emissions from Tier 2 vehicles needs to be confirmed through analysis of data, including those from cold starts at 0F and 10F. Again, testing should include CO as well as other pollutants. Tfthe analysis of Tier 2 end prior controls indicates that all locations will attain the 8-hour CO standard, more stringent federal CO vehicle emissions standards will be unnecessary. The results of all emis- sions testing must be incorporated into EPA's MOBILE model to accu- rately estimate future CO emissions. The effects on CO problem areas of the sales strategy used by manufacturers to meet the NOX limits and of the trading and banking provisions also need to be assessed and incorporated into emissions modeling. High-Emitting Vehicles Findings A relatively small number of high-emitting vehicles contribute dispro- portionately to CO and other motor-vehicle emissions. The vehicle fleets operating in and around places with high local concentrations of CO (hot spots) often include a higher proportion of high-emitting vehicles compared with the surrounding region. Elimination or repair of high emitters would The MOBILE model is used to estimate current and future on-road motor- vehicle emissions. MOBILE6 is the current version of that model.

OCR for page 1
10 Managing CO in Meteorological and Topographical Problem Areas likely reduce the severity of CO hot spots and reduce motor-vehicle emis- sions overall. Recommendations Air quality management agencies should identify high-emitting vehi- cles and target them for repair or removal from the fleet. Enhanced onboard diagnostic testing programs, tailpipe testing, motor-vehicle emis- sions profiling, and/or remote sensing can identify these vehicles. How- ever, programs designed to mandate repair or removal of high-emitting vehicles might raise issues of fairness, because high emitters are often ownedby people with limited economic means. The cost-effectiveness and equity impacts of policies that provide incentives for owners of high-emit- ting vehicles to seek repairs or vehicle replacement, such as repair assis- tance programs, should be explored. There should also be additional Tow- temperature testing to evaluate the effectiveness of programs aimed at controlling high-emitting vehicles. This evaluation should include the impacts on CO as well as other emissions. Oxygenated Fuels Findings An oxygenated fuel is a gasoline containing an oxygenate (typically methyl tertia~y-buty] ether ~MTBE] or ethanol) intended to reduce produc- tion of CO. Oxygenated fuels program benefits are declining in effective- ness as more modern vehicles enter the fleet. EPA's MOBILE model pre- dicts CO emissions reductions from oxygenated fuels of 3-7% for the 2010- 2015 fleet because of reduced emissions from pre-1994 vehicles, cold starts, and malfunctioning vehicles. There is still uncertainty about the overall effectiveness of oxygenated fuels, especially at temperatures below 20F. Recommendations EPA should undertake a science and policy review of the current oxy- genated fuels programs to determine the conditions under which these

OCR for page 1
Summa7y 1 1 programs are cost-ettective. The review should also determine when changes in fleet technologies will render these programs ineffective. Low- temperature testing, especially below 20F, is recommended. Oxygenated fuels programs should be implemented only when they provide cost-effec- tive reductions in CO that help areas come into compliance or prevent areas that have attained the NAAQS from falling back into nonattainment. Public Education Findings On the basis of its review of programs in Fairbanks, Alaska, the com- mittee is concerned that public education campaigns have not sufficiently emphasized the adverse health effects associated with exposure to high ambient concentrations of CO. Also, the public is not fully aware of the link between transportation choices and overall air quality. As a result, public acceptance of and participation in locally proposed programs to achieve and maintain attainment of the NAAQS for CO is often poor. Recommendations Public-health education to improve public acceptance and compliance should tee a component ofall local emissions-reduction programs. Commu- nities should use surveys and focus groups to regularly evaluate the effec- tiveness of public education programs and the impact they have on the success of CO emissions control. CO Assessment CO As an Indicator of Motor-Vehicle Pollutants Findings In urban environments, ambient CO concentrations are a strong indica- tor of motor-vehicle emissions. EPA estimates that as much as 95/O of all CO emissions in some cities can be from automobile exhaust. Spatial and

OCR for page 1
12 Managing CO in Meteorological and Topographical Problem Areas temporal variability in motor-vehicle activity and atmospheric dispersion characteristics can lead to CO hot spots. CO is useful as a gauge of human exposure to other directly emitted mobile-source pollutants, such as air tonics and PM2 s. However, CO is not a perfect indicator of all mobile-source pollutant emissions, because CO reacts more slowly than many other pollutants, and the ratio of CO to copollutants varies by emissions source. The atmospheric conditions that produce high CO concentrations are different from those that produce high concentrations of photochemical pollutants. Despite these caveats, mea- surements in the Los Angeles area and elsewhere have shown strong corre- lations between ambient CO and benzene concentrations. A strong correla- tion between CO and concentrations of the relatively short-lived 1 ,3-buta- diene also was observed. Recommendations The committee has several recommendations in regard to the use of CO to represent the distribution of other pollutants. CO can be used to demon- strate the spatial distribution of some mobile-source pollutants, to identify hot spots, and to improve model representation of relationships between transportation activity and emissions. CO can also be used to approximate the concentrations of some air tonics arising from motor-vehicle exhaust emissions, such as benzene, I,3-butadiene, and perhaps directly emitted PM2 s CO is most useful as an indicator in the microscale setting where concentrations of pollutants vary dramatically over short distances (e.g., with distance from a roadway). it is less reliable in representing regional distributions of these pollutants and is probably a poor indicator of motor- vehicle air tonics, such as formaldehyde and acetaldehyde, that react rap- idly and have substantial sources in the atmosphere. Spatial Distribution of CO Findings Although ambient CO concentrations have dropped considerably throughout the country, the number of monitors is inadequate to character- ize CO distribution and identify all locations of high CO concentrations. There may be hot spots within cities that have already attained the NAAQS

OCR for page 1
Summary 13 for CO or have not previously registered high ambient CO concentrations. The locations of these hot spots may raise social equity issues regarding exposure to mobile-source-related pollutants. The monitors that have registered CO concentrations in excess of the NAAQS since 1995 are predominantly located in Tower-income areas with greater minority populations. However, unidentified hot spots might exist in any location. Current data are insufficient to adequately characterize the relationships between hot spot locations, population characteristics, and health impacts. Recommendations EPA should employ air quality modeling and saturation studies9 in CO problem areas to better characterize the spatial distribution of CO and the populations affected. The information garnered can be used to improve site selection for permanent monitoring, to improve model performance, and to address possible environmental equity issues. Programs targeted to local conditions can be developed using this information. These results should also be linked to health impact studies in these locations. In particular, EPA should try to better understand the upper end (higher CO levels) ofthe distribution of ambient exposures to motor-vehicle emissions that occur in most CO hot spots. Permanent Monitoring Findings Although fewer and fewer locations are experiencing CO concentra- tions above or near the NAAQS, the continued operation of most current Saturation studies typically rely on portable monitors that "saturate" a geo- graphical area with samplers to assess the air quality in places where high concen- trations of pollutants are possible. Monitors can be deployed at temporary fixed- site locations or in mobile sampling vehicles. These studies are helpful to air pollu- tion control agencies for evaluating their ambient air monitoring networks, charac- terizing pollutant concentrations over the entire saturation study area, and locating hot spots or high pollutant impact points. Personal monitoring could be incorpo- rated into such studies to relate ambient concentrations to personal exposure.

OCR for page 1
14 Managing CO in Meteorological and Topographical Problem Areas ambient monitors remains essential for Tong-term assessments of air quality and health impacts. Recommendations Because of the value of CO monitoring information for air quality management in general, agencies should resist removing CO monitors in locations not expected to show violations. Instead, they should consider continuing operations at existing CO monitoring sites, noting that when monitors are co-Iocated the incremental costs of continued operation may be relatively small compared with the data's usefulness for purposes be- yond demonstrating attainment. The number and placement of permanent monitors also need to reflect changes in growth and development patterns to accurately assess the local air pollution situation. However, communi- ties that have attained the CO standard with an adequate level of protection of safety might not be willing to pay to obtain data from these monitors. Support from federal and other sources might be needed to continue moni- toring operations. Ambient CO Modeling Findlings Emissions and air quality models are important tools for air quality planning. Models help forecast changes in the mass of pollutants emitted resulting from controls and severe air pollution events. Models are also used to demonstrate attainment of the CO NAAQS, evaluate the effects of new construction projects that greatly increase emissions, and research the causes of pollution episodes and how to predict them. However, the spatial and temporal resolution of models typically used in CO management at this time is too coarse to capture the variability in pollutant concentrations, which is necessary to identify local hot spots and accurately represent unusual meteorological conditions. Statistical forecasting models have been used to assess the probability of future high-CO episodes. The approach was used in Denver, Colorado, to assess the probability of having CO concentrations in excess of the NAAQS after the alteration of the oxygenated fuels program. This model

OCR for page 1
Summary 15 takes into account the historical variability in CO concentrations resulting from meteorology and unusual traffic events. Recommendations More sophisticated, physically comprehensive models that can simulate how CO concentrations vary in time and space should be developed, ap- plied, and evaluated. Ongoing research should be continued. Such models would be used for air quality planning and forecasting and for assessing human exposure to high concentrations of CO and related pollutants. Be- cause CO is a relatively unreactive pollutant, the ability to better represent CO's temporal and spatial distribution provides an effective diagnosis of atmospheric dispersionpatterns. Mode! improvements would have applica- tions for other air quality management issues and would offer the potential to better understand the dispersion of chemical, biological, and radiological materials. Most important, improved models will permit more effective and realistic planning, leading to better-informed decisions by administra- tors. Mode] development should occur in concert with improved monitor- ing to enable mode] evaluation. In addition, the statistical forecasting mod- els should be improved.