Application of Remote Sensing and Geographycal Information System on Prediction of Peat Fire in Bengkalis District Riau Province
Achmad Siddik Thoha1), Lailan Syaufina2), Tania June3)
District of Bengkalis Riau Province has been known as one of the most frequently fire prone areas occurrence in Indonesia. According to Ministry of Environment Indonesia, Forest conversion into plantation caused risk to pea and land fire.. Fire occurrences mostly found on peat land as result of land clearing activities. To decrease damage and environment impact from peat fire, it is important to identify and predict peat fire occurrence. The objectives of research were to compare accuracy of hotspot from data supply sources, to explain effect of biophysical and human activities on the fire indicated location and to establish peat fire prediction using variables of biophysical and human activiies factor. Methods used in the study were descriptive statistical and spatial analysis of hotspot data. Fire prediction model were established from binary logistic regression then implemented to spatial model of peat fire prediction. The accepted variables of fire prediction model are peat depth and NDVI with level of significance was 0.05 and distance from forest concession area (HPH/HTI) with level of significance was 0.10 Correlation between dependent variables (fire occurrence) and these variables was negative which shows the decreasing peat depth, NDVI and distance from HPH/HTI may result in the increasing of fire occurrence.
Keywords : prediction, peatland fire, hotspot
Please contact me for full text and graphic