Abstract:
Indonesia has undergone significant land cover changes in recent decades, particularly in its vegetated regions. As a country abundant in tropical rainforests, it grapples with local issues such as forest and land fires, particularly prevalent in Sumatra and Kalimantan. Among the affected areas are the forests of Central Kalimantan, integral to the global carbon ecosystem, a subject of ongoing debate. To address these challenges, this research aims to analyze burned areas utilizing the Normalized Burn Ratio (NBR) and the Random Forest (RF) method. By employing NBR and RF, the study seeks to identify and assess the extent of burned areas in these regions. The findings indicate that the accuracy test conducted on the burned area using both NBR and Random Forest classification yielded a notable accuracy rate of 86.67%. This accuracy reflects the degree of alignment between burned areas identified through Random Forest classification and those identified via NBR analysis. This research underscores the importance of employing advanced remote sensing techniques like NBR and RF to effectively monitor and analyze forest and land fires in Indonesia, particularly in regions like Central Kalimantan. The high accuracy achieved demonstrates the potential of these methods in accurately mapping burned areas, thereby aiding in managing and mitigating fire-related challenges in these ecologically significant areas.
