The Clean Air Act identifies 189 hazardous air pollutants (HAPs), or "air toxics," associated with a wide range of adverse human health effects. The U.S. Environmental Protection Agency has conducted a modeling study with the Assessment System for Population Exposure Nationwide (ASPEN) to gain a greater understanding of the spatial distribution of concentrations of these HAPs resulting from contributions of multiple emission sources. The study estimates year 1990 long-term outdoor concentrations of 148 air toxics for each census tract in the continental United States, utilizing a Gaussian air dispersion modeling approach. Ratios of median national modeled concentrations to estimated emissions indicate that emission totals without consideration of emission source type can be a misleading indicator of air quality. The results also indicate priorities for improvements in modeling methodology and emissions identification. Model performance evaluation suggests a tendency for underprediction of observed concentrations, which is likely due, at least in part, to a number of limitations of the Gaussian modeling formulation. Emissions estimates for HAPs have a high degree of uncertainty and contribute to discrepancies between modeled and monitored concentration estimates. The model's ranking of concentrations among monitoring sites is reasonably good for most of the gaseous HAPs evaluated, with ranking accuracy ranging from 66 to 100%.

Occupational and toxicological studies have demonstrated adverse health effects from exposure to toxic air contaminants. Data on outdoor levels of toxic air contaminants have not been available for most communities in the United States, making it difficult to assess the potential for adverse human health effects from general population exposures. Emissions data from stationary and mobile sources are used in an atmospheric dispersion model to estimate outdoor concentrations of 148 toxic air contaminants for each of the 60,803 census tracts in the contiguous United States for 1990. Outdoor concentrations of air toxics were compared to previously defined benchmark concentrations for cancer and noncancer health effects. Benchmark concentrations are based on standard toxicological references and represent air toxic levels above which health risks may occur. The number of benchmark concentrations exceeded by modeled concentrations ranged from 8 to 32 per census tract, with a mean of 14. Estimated concentrations of benzene, formaldehyde, and 1,3-butadiene were greater than cancer benchmark concentrations in over 90% of the census tracts. Approximately 10% of all census tracts had estimated concentrations of one or more carcinogenic HAPs greater than a 1-in-10,000 risk level. Twenty-two pollutants with chronic toxicity benchmark concentrations had modeled concentrations in excess of these benchmarks, and approximately 200 census tracts had a modeled concentration 100 times the benchmark for at least one of these pollutants. This comprehensive assessment of air toxics concentrations across the United States indicates hazardous air pollutants may pose a potential public health problem.ImagesFigure 1Figure 2Figure 3Figure 4

No abstract

Population-based human exposure models predict the distribution of personal exposures to pollutants of outdoor origin using a variety of inputs, including air pollution concentrations; human activity patterns, such as the amount of time spent outdoors versus indoors, commuting, walking, and indoors at home; microenvironmental infiltration rates; and pollutant removal rates in indoor environments. Typically, exposure models rely upon ambient air concentration inputs from a sparse network of monitoring stations. Here we present a unique methodology for combining multiple types of air quality models (the Community Multi-Scale Air Quality [CMAQ] chemical transport model added to the AERMOD dispersion model) and linking the resulting hourly concentrations to population exposure models (the Hazardous Air Pollutant Exposure Model [HAPEM] or the Stochastic Human Exposure and Dose Simulation [SHEDS] model) to enhance estimates of air pollution exposures that vary temporally (annual and seasonal) and spatially (at census-block-group resolution) in an urban area. The results indicate that there is a strong spatial gradient in the predicted mean exposure concentrations near roadways and industrial facilities that can vary by almost a factor of 2 across the urban area studied. At the high end of the exposure distribution (95th percentile), exposures are higher in the central district than in the suburbs. This is mostly due to the importance of personal mobility factors whereby individuals living in the central area often move between microenvironments with high concentrations, as opposed to individuals residing at the outskirts of the city. Also, our results indicate 20-30% differences due to commuting patterns and almost a factor of 2 difference because of near-roadway effects. These differences are smaller for the median exposures, indicating the highly variable nature of the reflected ambient concentrations. In conjunction with local data on emission sources, microenvironmental factors, and behavioral and socioeconomic characteristics, the combined source-to-exposure modeling methodology presented in this paper can improve the assessment of exposures in future community air pollution health studies.

No abstract