ABSTRACT
Many pollutants show extremely complex dispersion patterns, especially in environments such as towns and cities where there are a large number of emission sources and variations in environmental conditions. This complexity means that it is often very difficult to model or measure pollutant patterns and trends, and thus to predict levels of human exposure. Nevertheless, modelling can compliment monitoring programmes and be an effective A Q M tool. The types of model available to predict the impact of source emissions on ambient air quality or air sensitive receptors (ASR) depends on the user's needs and the morphology of the emitting source. Models can be simple or complex, the latter generally requiring onerous data sets. Models may also be point, area, volume or linear in nature as well as being either statistical, mathematical or physical. Mathematical modelling represents the real world in mathematical terms. Statistical models utilise relationships between two or more variables. Physical models are scaled down representations of the actual modelled feature. Modelling seeks a more precise understanding of the significant behaviour of air pollution measurements. It gives the means to check for consistency in the emissions inventories, observed meteorology and measured concentrations of the air pollutant (Middleton, 1996).
