Exploitation of surface albedo derived from the Meteosat data to characterize land surface changes

An advanced algorithm to characterize land surface properties on the basis of Meteosat-5 data has been developed and applied to year 1996 observations. Surface albedo values decontaminated from atmospheric effects and computed for a fixed Sun location have been derived for every day of 1996. A simple composite procedure has been applied over consecutive 10-day and 30-day periods to produce geographically complete maps of surface albedo. The sensitivity of Meteosat-5 observations to various land cover change scenarios has been assessed using models to simulate radiation transfer of solar radiation in soil-vegetation systems. The results suggest that the observed seasonal albedo changes cannot be explained by seasonally-driven phenological cycles of tropical vegetation so that a more complex set of mechanisms incorporating the effects of anthropogenic activities must be devised. Additional information derived from ATSR-2 measurements support the finding that the observed large and unexpected variations in surface albedo at a continental scale over Africa could be explained by intense biomass burning activities during the dry seasons.