Mapping of Risk Fire in Peat Land based on Hotspot Sequential Pattern Mining in District of Pulang Pisau, Kalimantan Tengah

Abstract

Land and forest fire is one of the major causes of Indonesia’s deforestation, who has a significant impact to the environment, loss of conservation, air pollution and economic loss. This research develop a spatial modelling of factors that can drive the forest fire. Spatial model of vulnerability of land and forest fire is developed by the composite mapping analysis method. Hotspot datasets used in this research is the results of data mining processing, with sequential pattern mining technique which to find the relationships between the occurances of sequential event and its pattern that often appear. From the six variables influences land and forest fire there are four variables that have significant impacts in the study area, as follows: forest zone, depth of peatland, distance of irrigation, and distance of road. The fire in the area of study occurs many times in the peatland area with the depth of 400-800 cm. Land and forest fire occurs frequently in 100-900 meters from irrigation and land and forest fire also occurs frequently in 1-4 km form the road. Land and forest fire occurs frequently in protected forest.