Main Article Content
Early warning system is an important component in land/forest fire management. Since Kalimantan is one of prone areas to fires in Indonesia, land/forest fire hazard mapping for the area is essential to provide early warning information. Methods on static fire hazard mapping have been established using geographic information system. Land/forest fire hazard mapping could be established based on spatial biophysical parameters such as rainfall, vegetation condition, land cover, and land type. Since most parameters can be derived from satellite data and some of them are predictable, a dynamical land/forest fire hazard maps can be generated. The objective of this research was to construct a model of forest fire hazard mapping for Kalimantan. Spatial data used consisted of spatial rainfall maps, Normalized Difference Vegetation Index (NDVI) maps derived from NOAA-AVHRR data, land cover maps from Landsat TM data, and land type map. The results show that contributions of rainfall and NDVI to fire hazards should be higher than land cover and land type. The weights of NDVI, rainfall, land cover, and land type are 0.35, 0.30, 0.20, and 0.15 respectively. For the case study of 1997 – 2002 periods, it has been shown that most hotspots are located in areas with forest fire hazard of high level.