are known to have large artifactual biases due to the sampling methods used (Gundel et al. 1995).
Substances in outdoor (or ambient) air are dispersed by atmospheric advection and diffusion. Meteorological conditions, local terrain, and facility designs have an overwhelming influence on the behavior of contaminants in the lower atmosphere. Wind (direction, speed, and turbulence) and atmospheric stability are the most important. The standard models for estimating the local time and spatial distribution of contamination in the atmosphere from point sources are the Gaussian statistical solutions of the atmospheric diffusion equation. These models are obtained from solution of the classical differential equation for time-dependent diffusion in three dimensions. Pasquill (1961) has discussed the physical basis, analytical solutions, and the use of these equations. Turner (1970) and Hanna et al. (1982) have compiled workbooks on applications of these solutions to air pollution problems, including the application of the Gaussian models to area and line sources. There are numerous computer programs available and many papers describing algorithms for assessing the dispersion of point (e.g., stack), line (e.g., roadway) and area (e.g., shopping mall) air pollution sources. The output of a standard Gaussian plume model can be expressed as the ratio of the atmospheric concentration to the source strength release rate. Typical units are µg/m3 per µg/sec, or sec/m3. This ratio is typically estimated using screening-level models such as SCREEN3 (EPA 1995a,b), or more complex, site-specific, models such as the Industrial Source Complex (ISC) models (EPA 1995a). (Such models are easily obtained from the U.S. Environmental Protection Agency at the following website address: http://www.epa.gov/scram001/.) For example, SCREEN3 provides a high-end estimate for the worst-case 1-hour average of this ratio as large as 0.05 sec/m3 for ground-level releases in urban areas, but the ratio typically decreases with the height of release. The annual average concentration-to-source ratio is likely to be about 0.08 (±0.02) times the maximum 1-hour average (EPA 1995b). The ISC models can provide specific estimates for any given location, and can also take account of simple, intermediate, and complex terrain; dry deposition; wet deposition; and plume depletion.
Simpler approaches to estimating the dispersion of substances in the atmosphere may be based on the application of a mass balance to a volume element, parcel, or box of air. This gives rise to the “box” models. In this approach, the region to be studied is divided into cells or boxes. The concentration in each box is assumed to be uniform and is a function of the box volume, the rate at which material is being imported, emission rates within the box, and the rate at which material is exported from the box. Such simplified approaches may be more appropriate than the Gaussian plume models in circumstances where dispersion