In this work the importance of the land cover-usage data retrieved by remote sensing images in support of air dispersion models is shown. Models for the simulation of the dispersion of pollutants emitted in atmosphere are considered very useful tools for studying, monitoring and managing the environment. One of the most important input for these models is the land cover-usage, that must be furnished for each cell of the computational grid. In general, this information is retrieved from shared data set that are not sufficiently upgraded and not enough spatially detailed in order to obtain an accurate simulation of atmospheric dispersion. In this frame, land usage-cover retrieved from remote sensing image could represent a very useful support. The AIR INDUSTRY dispersion package is used to model the atmospheric dispersion of the suspended particulate matter emitted from a disposal area located in the main landfill site of the city of Modena (Italy), using two different land cover-usage data set: that extracted from the European Corine Land Cover data base, and that obtained by the integration of the previous with information retrieved from a satellite image acquired by the ETM+ sensor. The comparison of the pollutant concentration fields obtained for the two cases shows important differences, both in the area interested by pollution and in concentration levels. Nevertheless, this work highlight also the necessity of making more tests, for different case studies, before drawing final conclusion.