Advances in short range dispersion modeling (Invited Presentation)

This paper reviews recent progress in modeling dispersion from surface and elevated releases at scales of the order of 10 km. The paper will examine the role of a new theoretical tool, Lagrangian Stochastic Simulation (LSS), in providing insights into vertical dispersion in the surface boundary layer. This understanding has been incorporated into recent dispersion models. We will also discuss methods to develop models by interpolating descriptions of the relevant system at certain asymptotic limits; these limits are generally much easier to model than the complete response of the system. We will illustrate the utility of this method in developing a new model, AERMOD, designed to replace the Industrial Source Complex (ISC) model. We show that the technique yields a simple model for dispersion in complex terrain that performs as well as a model that attempts to capture the complexity of the system.

In the second part, we discuss the major gaps in current understanding of dispersion. For example, we know little about modeling dispersion in the urban boundary layer. The final part of the paper discusses the problem of evaluating the performance of models in explaining observations. Because deviations between model estimates and observations are inherently large, statistics of these deviations are not useful in demonstrating model improvement. We discuss a new technique that attempts to solve the problem.