Remote Sensing and GIS Application to Establish a Forest Fire Risk Map for Planning of Forest Fire Prevention and Mitigation in Son La Province, Vietnam

Son La is a mountainous province in the Northwest of Vietnam with many ethnic groups, and has an area of ​​14,125 km², accounting for 4.27% of the total area of ​​Vietnam. The ​​forest land accounts for 73% of the total natural area of ​​the province with 357,000 ha of forest. Among this having 4 areas of special use and the natural reserve forest. Every year, hundreds of forest fires cause huge natural, economic and ecological damages to the province. Due to the climate change, forest fires tend to increase in recent years. In order to prevent the fires, warning maps of the forest fire risk are needed. The research has analyzed mechanism and causes of the forest fires, and built a forest fire-related database with multi-layers of natural, social and economic information, in these, some layers were extracted from the Landsat 7 images. The expert method was applied for assessement and Saaty's Hierarchical Analysis (AHP) methods were applied to determine the weight for separated parameters related to forest fires. The research applied the MCA method to build a multi-indicator function with 9 parameters for establishing the forest fire risk map at the scale of 1:100,000 for provincial levels. In verifying the results by regression correlation analysis, the R2 value reached 0.71.These maps have been used for the purpose of forest fire prevention planning for Son La province.