Integration of AVHRR and fine spatial resolution imagery for tropical forest monitoring

A multi-sensor approach to tropical forest monitoring is discussed whereby data from the NOAA AVHRR is used to identify changes in forest cover on a regional to continental scale and to direct more detailed assessments using satellite sensor imagery with a finer spatial resolution. Temporal, spatial and spectral attributes of optical sensors and the procedures of stratified sampling applied to different forms of forest disturbance are outlined. Regression techniques, linear mixture models and fuzzy classification algorithms (which can be used to estimate the proportions of land cover types within AVHRR pixels and which require fine resolution imagery for development and validation) are also reviewed in the context of tropical forest monitoring. The use of thematic information derived from fine resolution data for increasing the monitoring capabilities of the AVHRR is considered. -Author