Assessment of AERMOD modeling system for application in complex terrain in Pakistan

Abstract

Atmospheric dispersion models based on Gaussian plume formulation are widely used for regulatory purposes in flat terrain. In this paper, the application of an advanced Gaussian plume modeling system, AERMOD, has been investigated over complex terrain setting through a field tracer test conducted in April 2018 near Islamabad, Pakistan. The model was validated through short term (2 h) release from a point elevated non buoyant source of SF6. The hourly averaged concentrations were measured at 47 discrete sampling points. The mesoscale meteorological model, WRF was coupled with AERMOD to provide hourly surface and upper boundary layer meteorological parameters. The statistical evaluation of model was carried out using Fractional Bias (FB) and Normalized Mean Square Error (NMSE). The quantile-quantile (Q-Q) plot was used to assess model's ability to follow the observed concentration distribution. For this field tracer test the model over-predicted ground level concentrations at the receptors close to the source and underestimated at the receptors located in a valley and far away from the source. Moreover, AERMOD overestimated the concentration at receptors which were at the point of direct impaction of plume and ridge. Overall, the model under-predicted the ground level concentration within a factor of two (FB = 0.63) for this field tracer test. It was found that the model better simulated the higher concentrations which are of primary interest for regulatory purposes.