Detection of forest changes over French Guiana using ERS-1 and ASAR iamgery

ERS-1 and ASAR imagery were used for mapping and monitoring the forest area of French Guiana for 1992/1993, 2003/2004, and 2006/2007. Imagery was ortho-rectified using the SRTM DEM. A least square algorithm was used to improve the residual geometric inaccuracies. For the ERS data, significantly better geometric performance could be achieved using the PGS processor (the so called Envisat processor) data than with the standard VMP processor. The image analysis used three SAR based features, which were computed of two to five acquisitions from the same location. The most effective feature in change detection is annual average amplitude. The planned future operative system uses annual SAR data acquisitions and sample point inventory using high-resolution optical data twice in decade. The study was part of the GMES (Global Monitoring for the Environment and Security) Service Element Forest Monitoring.