Within the U.S. Environmental Protection Agency (EPA), the Office of Air Quality Planning and Standards (OAQPS) has the responsibility for establishing and revising National Ambient Air Quality Standards (NAAQS). In assessing alternative NAAQS proposed for a particular pollutant, OAQPS evaluates the risks to human health of air quality meeting each of the standards under consideration. This determination of risk requires accurate estimates of the number of persons exposed at various pollutant concentrations for specified periods of time. The estimates may be specific to an urbanized area such as Los Angeles or apply to the entire nation.
Several researchers have recommended that such estimates be obtained by simulating the movements of people through zones of varying air quality so as to approximate the actual exposure patterns of people living within a defined study area. 1,2,3 OAQPS has implemented this approach through an evolving methodology referred to as the NAAQS Exposure Model or NEM. From 1979 to 1988 OAQPS developed and applied pollutant-specific versions of NEM to ozone,4 total suspended particulate matter,5 and carbon monoxide.6 These versions of NEM are referred to as "deterministic" versions in that no attempt was made to model random processes within the exposure simulation. An overview of the evolution of the deterministic version of NEM is provided by Paul et al.7