Dynamic mapping of snow-free vegetation and bare soil albedos at global 1 km scale from 10-year analysis of MODIS satellite products

Land surface albedo is considered to be a fundamental quantity for accurate assessment of the surface energy budget. Whereas satellite products usually provide total surface albedo, land surface models often require a separation of snow-free surface albedos of the vegetation canopy and the underlying bare soil. In this paper, visible (0.3-0.7 mu m) and near infrared (0.7-5.0 mu m) vegetation and bare soil albedos are retrieved from a decade of MODIS data at global scale by using a Kalman Filter-based method. The retrieved quantities show good agreement with those derived in previous studies. The main advantage of the proposed method over other approaches is that it provides albedo components efficiently throughout the year, thus allowing the seasonal cycles to be captured. This study also exploits the relationship linking the retrieved bare soil albedo to soil moisture in order to reinforce indirectly the quality of the albedo splitting. The proposed method makes it possible to update the bare soil and vegetation albedo fields in the ECOCLIMAP land cover map, the database of physiographic parameters used by Meteo-France for climate and NWP modelling. The mean relative difference between the new and the old ECOCLIMAP shortwave albedos (0.3-5.0 mu m) is 8.5% at the global scale if only albedos lower than 0.5 are considered, and 41.2% otherwise. This update of the ECOCLIMAP database will have numerous NWP and climate applications for carbon, energy, and water cycle modelling. (C) 2013 Elsevier Inc. All rights reserved.