Forest Fire Hazard Model Using Remote Sensing and Geographic Information Systems: Toward understanding of Land and Forest Degradation in Lowland areas of East Kalimantan, Indonesia

A heavy scale of forest fire occurred during the year of 1997/1998 in Indonesia, which burnt millions hectares of the tropical rain forest. The Indonesian East Kalimantan province was the most severely affected by this forest fire, which have destructed a natural forest and induced accelerated soil erosion. A repetitive forest fires in this province lead to bring a land and forest degradation process. This study develops a forest fire hazard model using remote sensing technique integrated with Geographic Information Systems (GIS). It also attempts to establish a method for land and forest degradation hazard mapping in lowland area of East Kalimantan based on analyzing of forest fire hazard with respect to soil erosion hazard. The following variables were derived for a study area: vegetation fuel type, terrain, road, and bare soil. Those variables were weighted according to their impact on the forest fire hazard or accelerating soil erosion. Supervised classification and NDVI analysis are applied to Landsat TM to derive a fuel type. Tasseled cap transformation, the simplified end member technique and the GIS neighborhood analysis was carried out to identify bare soil areas, which consist of maximum approximating of a pure bare soil on 3 by 3 pixel size. The final result offers a model of forest fire and soil erosion hazards. The models of forest fire hazard show reliable from perspective of the proportion of areas affected by high burnt areas of forest fire in 1997/1998. A derived land and forest degradation map has been offered. Based on this map it is possible to identify and monitor areas affected by forest fire or to evaluate the pattern of land and forest degradation.