Contribution of Meteosat Second Generation (MSG) to drought early warning

Résumé En

Several low resolution satellite sensors are now available and can provide near real-time monitoring of vegetation cover seasonal changes on a large scale basis. This paper presents a comparison of NDVI derived from SPOT-VEGETATION and MSG-SEVIRI for the whole continent of Africa. Dekadal NDVI from VEGETATION data are available, obtained from maximum NDVI compositing technique. VEGETATION data from July 1-10,2004, have been re-sampled at 0.025° spatial resolution (= resolution of MSG-SEVIRI data) for whole Africa. In the case of SEVIRI data, a 2-steps procedure is used: first daily compositing is implemented through analysis of thermal data (maximum surface temperature compositing), followed by dekadal NDVI synthesis. Results include the observed correlations and discrepancies of NDVI products of the 2 sensors. Results suggest that, due to higher temporal frequency (1 image each 15 mn), MSG-SEVIRI data can be used to obtain improved NDVI products through better removal of cloud-contaminated pixels