Applicability of a high-resolution meso-scale meteorological model to a near-field-scale atmospheric dispersion problem

This study examines the feasibility of numerical simulations using a meso-scale meteorological model (NuWFAS: Numerical Weather Forecasting and Analysis System) for a near-field-scale atmospheric dispersion problem. A series of observation data from a field tracer experiment was used for the validation of the model. In the experiments, the tracer was released from a tower at a height of 95 m. The receptors were located on the arc lines with distances from the source of 400, 750, 1500, and 3000 m. The numerical simulations were implemented with two different minimum spatial resolutions of 100 m and 300 m. The meteorological fields were reproduced with a reasonable accuracy, showing the less dependency on the mesh sizes of the simulation. The dispersion fields were also less dependent on the spatial resolutions except for the stable atmospheric conditions. In stable conditions, the smaller spatial resolution leads to the higher surface concentrations due to the larger turbulent diffusions. In most cases, the predicted surface concentrations agreed with the observation within the factor of ten. However, the simulation tends to underestimate the surface concentrations in stable conditions, whereas it overestimates in unstable conditions. Our study revealed that the limitation of the model in estimating the turbulent diffusion coefficients for thermally stratified conditions is the one cause of these trends. The current model underestimates the influences of atmospheric stability, which is one of the most important factors for the near-field-scale atmospheric dispersion. (author)