Keywords: atmospheric dispersion models, model comparison, model sensitivity, regulatory models, air pollution, environmental pollution, modelling, source emissions, grid resolution, stack height, flue gas temperature, plant capacity, mass flow rate, roughness length, meteorological time series
Comparative sensitivity study for operational short range atmospheric dispersion models
Eight atmospheric dispersion models are compared in a sensitivity study in which the responses towards variations in model input parameters are evaluated. For a single source the values of eight practical model input parameters are varied. The model response is evaluated by examining the maximum value of the 98th percentile of the hourly averaged concentration values. The results are represented by means of a sensitivity indicator relating the standard deviation of the parameter variations to the standard deviation of the resulting maximum 98th percentile variations. The parameters studied in this exercise are grid resolution, stack height, flue gas temperature, source location, plant capacity, mass flow-rate, roughness length and meteorological time series. The participating models are AUSTAL86 (TA-Luft), IFDM (VITO), ISCST2 (EPA), OML (NERI), OPS (RIVM), PLUIMPLUS (TNO), STACKS (KEMA) and UK-ADMS (CERC). The meteorological databases were chosen by the participants, being representative for their typical model applications. Source emission data, however, were prescribed in order to provide the same initial conditions.