Les grands domaines d'application actuels et futurs de SPOT dans le secteur forestier

The use of high resolution satellite data in forestry applications has been continuously developping since the launch of SPOT 1 in 1986. In particular, the first generation of SPOT satellites with a 20-meter resolution in multispectral mode and 10-meter resolution in panchromatic mode provide valuable information for forest resources assessment and monitoring. This paper is reviewing some of the main operational applications of high resolution satellite data in forestry, in tropical, temperate and boreal environments. In areas where the forest situation is already well assessed, e.g. where national or regional forest inventories are implemented, SPOT data are particularly adapted for rapidly detecting major forest cover changes and for updating forest type maps. On the other hand, when little information is available on the current forest situation, e.G. in some tropical countries where no forest assessment and monitoring system has been developed in spite of fast degradation processess, there is a great need in assessing the resources by implementing a comprehensive forest mapping exercise, where the delineation of forest stands can be used as a basis for stratification in a forest inventory design. Selected examples are illustrating this review, and the specificity of multispectral and panchromatic data in forestry applications is presented. Important developments with the new generation of SPOT satellites are expected in the coming years. On one hand, the low resolution VEGETATION instrument on board SPOT 4 will provide effective means for global forest survey and monitoring. The TREES project carried out by the joint Research Centre is already giving very valuable experience in this field. The availability of a middle infrared band on SPOT 4 HRVIR sensor will improve the maping of clear-cuts and thinnings over some types of forests. The planned HRG instruments on board SPOT 5 will open new expectations in forest type mapping and monitoring. In particular, the 5-meter resolution in P mode is likely to provide greater possibilities in the discrimination of forest cover class.