<![CDATA[Forest fires are a recurring issue every year in various countries, especially in those with extensive forests like Indonesia. An initial step in fire prevention is detecting the potential occurrence of fires, which can be achieved by utilizing satellite data, such as MODIS data. In this study, clustering or grouping of MODIS data in Indonesia for the years 2021 and 2022 was conducted using the CLARA algorithm due to its robustness against outliers and efficiency in handling large datasets. The application of clustering with the CLARA algorithm on both datasets resulted in two clusters, and the evaluation using the Silhouette Coefficient yielded values of 0.89 and 0.88 for both years. The analysis revealed that both clusters in both datasets exhibited similar characteristics. In the data for the years 2021 and 2022, the first cluster displayed a moderate to high potential for fire, while the second cluster indicated a low potential for fire. The results of this study can be used as a reference for authorities to identify the level of forest/land fire potential from observed hotspots in Indonesia, thus enabling early prevention measures such as early extinguishment to prevent further spread of the fire.]]>
Copyrights © 2024
