A GIS-based atmospheric dispersion model

Atmospheric pollution due to the use of agricultural pesticide is a major concern today, regarding both public health, sustainable agriculture and ecosystems quality monitoring. This work aims to model pesticide atmospheric dispersion and to propose an useful air pollution prediction tool, using fluid mechanics equations and open-source GIS programming capabilities. This work focuses on pesticide spray drift modelling for viticulture applications from the plot to the watershed scales. Our goal is to provide a GIS based atmospheric transport simulation platform at very low calculation cost using solution space reduction and reduced order modelling. This is necessary to make affordable simulation and very helpful to quickly perform georeferenced pollution predictions, which can then be used as a basis for risk analysis. The coupling of fluid mechanic equations with GIS is first described, regarding the enhancement of the mathematical model according to GIS parameters like topography and scale variations. Indeed, ground variations are implemented in the model using digital elevation model (DEM). The use of DEM is an asset to replace former Cartesian outputs by georeferenced raster pesticide cloud and also, it allows to use the model on any area. Furthermore, the implementation of topography in the spatialized Gaussian plume model allows scale changes. The plateform has been developed as a plugin for Quantum GIS open source software, which offers many spatial ha l-0 04 68 86 3, v er si on 1 31 M ar 2 01 0 Author manuscript, published in "STIC & Environnement, Calais : France (2009)"