Surface Parameters Evaluated from Satellite Remote Sensing Images for Pollutant Atmospheric Dispersion Modelling

This contribute deals with the use of surface parameters extracted from satellite remote sensing images for the setup of the input dataset required by pollutants atmospheric dispersion models (PATM). These models need 2D distributions (grids) of many surface parameters to model turbulence parameters, as roughness length, albedo, leaf area index and Bowen ratio. Very often these parameters are set using predefined tables defined as a function of land cover (LC). Usually, this last information is extracted from public datasets, such as, for European countries, the CORINE Land Cover (CLC). Some of these parameters can be computed directly from remote sensing. Moreover, land cover classification evaluated from remote sensing can be used to update existing LC datasets. In this work ASTER images have been used to evaluate, using a supervised classification method, the LC map of the area. This LC is used to update the CLC. Moreover, albedo was directly calculated from the image. The importance of information extracted from remote sensing is evaluated using the SPRAY lagrangian PATM. SPRAY has been used to simulate the dispersion of an inert generic pollutant emitted from two virtual sources on a 30 km x 40 km domain in a study area located at Venice (Northern Italy), where a big industrial site is found (Porto Marghera). Real (measured) meteorological data have been used.