RESUMO
In this paper the mesoscale application of the 3D Lagrangian particle dispersion model LAPMOD has been assessed for a field tracer test performed in a short-range complex terrain. The meteorological input was provided through the diagnostic model CALMET, the meteorological pre-processor of the CALPUFF model. The CALMET/LAPMOD coupled system was used to simulate the hourly averaged ground level concentration at 47 discrete receptors. The LAPMOD model has a general tendency to slightly underestimate the hourly averaged ground-level concentrations. A Q-Q plot shows that the predicted concentration distribution has a good comparison with observed one. The Robust Highest Concentration (RHC) indicates that the LAPMOD model slightly underestimates the simulated peak concentration in short-term release conditions. The Fractional Bias (FB), Normalized Mean Squared Error (NMSE), Factor of Two (FAC2), Factor of Four (FAC4) and Factor of Exceedance (FOEX) statistical indices were calculated. The predicted results by LAPMOD are generally in good agreement with observed ones and the model is justified for the use in complex terrain for short-term near-field applications.