Vulnerability Peatland Fire in Bengkalis using MaxEnt and QGIS Application

Every year fires occur in peatland Bengkalis Regency that impact for humans, biodiversity and the environment. One way to identify peatland fire vulnerability is to identify hotspots. The aim of the research is to identify vulnerability peatland fires in Bengkalis Regency. This research is using hotspot from VIIRS for 2015-2020. Vulnerability peatland fire is uses Maximum Entropy (MaxEnt) with the independent variables being physiographic, physical characteristics of peat, anthropogenic, climate and NDMI. The accuracy of MaxEnt model is using AUC values. The number of hotspots with high confidence level in Bengkalis Regency for the 2015-2020 was 478, with the highest number of hotspots in Bengkalis District at 148. MaxEnt can predict peatland fires well in Bengkalis Regency with AUC values of 0.864 (training data) and 0.813 (test data). Road density, rainfall, peat type, peat depth and river density are important variables that influence peatland fires in Bengkalis Regency. The areas of peatland fire vulnerability in Bengkalis Regency are not vulnerable at 3,517 km
²
(77.04%), low vulnerability at 1,021 km
²
(22.37%), and high vulnerability at 27 km
²
(0.59%). 22.96% of the Bengkalis Regency area is a high and medium vulnerability peatland fire area, most of which is in the Bengkalis District.