Poisson Clustering Process on Hotspot in Peatland Area using Kulldorff’s Scan Statistics Method

Abstract

The increase in peatland fire’s intensity has encouraged people to develop methods of preventing wildfire. One of the prevention methods is recognizing the distribution pattern of hotspot as one of forest and land fire indicators. We could determine the area that has high fires density based on distribution patterns so any early prevention steps could be performed in that area. This research proposed to recognize the distribution pattern of hotspot clusters in the peatland areas in Sumatera in the year 2014 using Kulldorff’s Scan Statistics (KSS) method with Poisson model. This approach was specifically designed to detect clusters and assess their significance via Monte Carlo replication. Results showed that the method is reliable to detect the clusters of hotspots which have the accuracy of 95%. Riau and South Sumatera province have the highest density of cluster distributions of the hotspot. Based on the maturity level of peat, cluster distributions of hotspot were mostly found in ‘hemic’ maturity level. Based on peatland thickness, cluster distribution of hotspot was mostly found in ‘very deep’ thickness.