Characterization of the land use/cover in the Congo rainforest with optical and microwave data

In South Carmeroon the deforestation is mainly caused by slash-and-burn agriculture. This practice is conducted on small plots within the forest, leading to a diffuse pattern of deforested areas. In such a complex farming system, with cacao plantations under tree cover and inter-cropping, the significance of high resolution satellite data can be evaluated in terms of discrimination of the main land covers: savannahs, degraded forest and dense forest. This has been done on SPOT, ERS (received at the German transportable station in Libreville) and JERS SAR images. The classification algorithm used for SAR images is based on the statistical properties of the land covers (PAPRI algorithm). The results are presented, compared to SPOT classified images and a vegetation map, and the synergy between the different sensors is discussed. (Résumé d'auteur)