Analysing Temporal Hotspot Occurance over Sumatera and Kalimantan

Abstract

A hotspot is an indicator of forest fires which is the temperature is relatively higher than the temperature around. Indonesian forest has a problem that continues to recur almost every year, specifically forest fires. One area in Indonesia that is currently experiencing frequent forest fires is Kalimantan Island. Therefore, it is necessary to predict controlling forest fires by monitoring areas at risk of forest fires. a hotspot is an area that has a higher temperature compared to its surroundings, Hotspots data have attributed to detect the presence of forest fires, namely Latitude, Longitude, Confidence, Brightness Temperature, and Fire Radiative Power. In the prediction, the process can be done by implementing the Fuzzy Subtractive C-Means method using the PHP programming language. This system succeeded in displaying hotspot data into 3 groups through input processing in the form of fire radiative power, confidence, and brightness temperature. The groups were identified into 3 levels of risks, which are high risk, medium risk, and low risk. The data used in this research was the hotspots occurrence over Kalimantan and Sumatra in 2016–2018. This research found that Kalimantan islands produced more number that Sumatera Island over the year. Kalimantan island (38.25%) has also significant number of high-risk clusters than Sumatera Islands (25.08%). Furthermore, this research also found that the most significant number of hotspot occurrence over Sumatera and Kalimantan was August (32.3%), September (21.37%), and October (8.7%). The area with the most occurrence hotspot was Riau (35.09%) and West Kalimantan (45%) for Sumatera and Kalimantan respectively. The algorithm was tasted with the silhouette algorithm that showed a strong (0.74) connectivity of structure coefficient value.