HIGH RESOLUTION PUFF POLLUTANT DISPERSION ALGORITHM FOR UNSTEADY AND NON HOMOGENEOUS ATMOSPHERE

Abstract

In a rapidly developing induStrial world, environmental damage, human,suffering and material deterioration have posed serious problems, warranting solution. Land, water and air are three basic pathways for transport of pollutants. The overall ecological deterioration is largely affected by the pollutants carried by air. Modelling of dispersion and diffusion of wind- -borne pollutants depends on several micro-meteorological as well as source parameters. These models have been categorised into (1) the simple empirical models which are subjective due to excessive simplifications and (2) very complex ones which are too difficult for practical use. The 'puff' models have the beneficial characteristics of both of these. This approach discretises the emissions and by it, the complexities of atmospheric circulation patterns can be, to an extent, simulated. The algorithm consists of five computational substeps- 1. Puff generation with puff rise. 2. Advection - according to the prevailing wind field at its location (spatial as well as temporal) 3. Diffusion - according to the turbulence scale of atmosphere 4. Deposition and chemical transformation. 5. Contribution of individual puffs to receptors iv Mathematically, motion of each puff is tracked in a Lagrangian frame attached to its centre and the calculations are transferred to an Eulerian frame attached with the source and having x-axis along the prevailing wind direction, z axis vertically upward and y-axis in the cross wind direction. The puff modelling finds special applications in calm wind conditions where no other model is well suited. The present model attempts to simulate dispersion in unsteady and non homogeneous atmosphere where the emissions may also be unsteady and non-homogeneous. Though it seems to be extreme combination but practically it is of frequent occurance. Furthermore, this model may be useful for automatic monitoring of air quality where the input parameters may be fed from G.I.S. in digitised form.