Fifty Years of Near-Field Air Dispersion Modeling Advances and Challenges

Ever since the United States Environmental Protection Agency was established in 1970, it has relied on short-range Gaussian dispersion models for many air pollution exposure analyses. In the 1970s, initial modeling efforts used workbook approaches that focused upon case study and worst-case conditions. Since about 1980, when EPA initiated a program of national consistency in dispersion modeling, it has relied on just two short-range dispersion models: the Industrial Source Complex (ISC) model until 2005, and AERMOD since then. ISC represented scientific approaches from the 1960s and 1970s with Pasquill-Gifford stability classes. The adoption of AERMOD involved the use of Monin-Obukhov similarity theory and other major advances in short-range dispersion modeling.

Advances in several areas have indicated the need for another generational advance in near-field modeling approaches and consistency with multiple-hour transport models. The key limitations in the current short-range modeling approaches are:

• The steady-state assumption limits the model applicability to transport times of just 1 hour. With light wind speeds, this can result in distance limits of just a few km.
• The model’s ability to utilize 3-dimensional wind fields is restricted, further limiting the utility of the modeling.
• Critical dispersion conditions occur in near-calm conditions, for which steady-state models are not designed.
• The ability to account for the chemistry involved in secondary pollutant formation (e.g., ozone and fine particulate) is limited to advanced Lagrangian or photochemical grid models.
• Source characterization issues that are critical for accurate modeling continue to be a challenge. This includes effects of buildings, source heating or cooling, urban/rural effects, etc.
• The role of Computational Fluid Dynamics Modeling to better characterize the near-source plume behavior is worth considering.

Recommendations are provided for resolving these limitations in the next-generation modeling system. In addition, there is a need to take full advantage of and to organize existing field databases suitable for model evaluation. There is also a need for the careful design and development of new evaluation databases for the next-generation dispersion model.