Modeling is an important tool for determining the contributioin of specific sources to the overall airshed and is therefore very important in control strategy development. It also provide a means to determine the potential impact of a new, proposed source that has not yet been built. They can also be used to mathematically simulate pollutant transport over urban, regional or global scales. Modeling can be as simple as a spreadsheet calculations (linear proportional rollback modeling) or as complex as photochemical long range transport models which require sophisticated computer systems.

Receptor-oriented models that apportion source contributions on the basis of observations made at the receptor (rather than the source) can also provide valuable information. These models typically use only chemically-speciated PM10 or PM2.5 data (trace element compositon data) to infer source contributions independent of emission inventories and meteorological data.

The joint application of both dispersion and receptor modeling methods is a powerful combination in understanding the dynamics of air pollution.