Detecting Post-fire Burn Severity Level using Sentinel-2 and MODIS Satellite data

Abstract

Background Forest fires adversely affect forest ecosystem by altering its composition, structure, regeneration, and recovery potential of its landscape. The present study investigates forest fire hotspots and examines the relationship between these fire events and deforestation in Tehsil Dhansar, District Sherani, Balochistan. This study proposed a three-step research methodology to achieve its objectives. Firstly, it aims to assess the severity level of the forest burn resulting from the fire event. Secondly, it analyzes the extent of vegetation loss caused by the fire. Thirdly, the study identifies forest fire hotspots using Sentinel-2A images and MODIS Fire Radiative Power (FRP) data. The analysis involves utilizing Normalized Burn Ratio (NBR), Normalized Difference Vegetation Index (NDVI), and Hot Spot Analysis (Getis-Ord Gi*) to gain comprehensive insights into the pre- and post-fire situation accurately. By defining classes, the study achieves a better understanding of the extent of burnt areas and vegetation loss. Results The findings show that 0.03% of Tehsil Dhansar is found to have low to medium burn severity levels during any forest fire event. It is also revealed that the forest remained dominant in the same region and frequency of occurrence of forest fire events is increasing by 1.6% with each passing year. Conclusion The current study's findings in the region famous for the world's oldest forest have significant potential for similar landscapes worldwide, primarily characterized by dry deciduous forests and juniper forests well adapted to arid and semi-arid environments. Given these findings, further studies in the same location should prioritize obtaining precise in-situ measurements to deepen our understanding of the situation.