Indonesia has quite a large forest, and some forests often experience fires. These fires typically occur due to several factors, including high solar heat, drought in peat forests, and the practice of clearing land by burning. This research focuses on West Kalimantan, one of the areas that experiences the most frequent forest fires. To achieve this, the study employs K-Means Clustering and Self-Organizing Map (SOM) algorithms, integrated with Geographic Information System (GIS) tools, to process satellite imagery from NASAs Terra and Aqua satellites. Key parameters include geographic coordinates (latitude and longitude), brightness temperature, and hotspot confidence levels. The clustering results identified two primary groups, with Cluster 2 representing the group with the highest thermal activity and fire risk. This cluster recorded a peak brightness temperature of 432.42 K and achieved a silhouette score of 0.71, indicating high clustering validity. GIS-based mapping revealed that the Sambas region had the highest concentration of hotspots, accounting for 36.01% of all detected points. These findings underscore the importance of targeted fire prevention efforts, particularly in high-risk zones with dense vegetation and frequent fire incidents.
Copyrights © 2025