Preliminary Evaluation of Aermod Using Site Specific Stack and Ambient Sampling Data

A cotton ginning industry-supported project was initiated in 2008 to develop a robust particulate matter (PM) dispersion modeling dataset that could be used for evaluating current and future PM dispersion models. This paper compares total PM data collected by the industry-supported study at one gin to outputs from EPA’s AERMOD ambient air modelling software. Ambient air samplers were placed at 35 locations that were arranged around a gin in three concentric circles. Stack emissions data, on-site weather data, emissions points, and structure data were used as inputs in AERMOD. Modeled outputs were compared to the collected total PM concentrations at the specified locations around the gin. The percent difference of modeled from collected concentrations were mapped in order to analyze its spatial distribution. This analysis found that the model tended to underestimate concentrations close to the emission source. As distance increased from the emissions source, modeled and sampled data began to align then flipped, with the model overestimating nearly all concentrations at distances greater than 100 meters from the source.