Rapid Forest Fire Detection Using Relative Difference in NDVI from Sentinel-2 Images in Nuristn, Afghanistan

Detection, monitoring, and evaluations of the scope and severity of the wildfire are critical for post-disaster management efforts. In this study, we use the rdNDVI technique to identify a wildfire that occurred on 24 May 2022 in Nuristan, Afghanistan, using pre- and post-event Sentinel-2 images based on Google Earth Engine (GEE) to assess the potential of Sentinel-2 and GEE to capture forest fire-induced land-use and land-cover changes. Wildfire detection, monitoring, and accuracy evaluation are presented. Wildfires ravaged using current methodology was estimated 58 acres. With a single free post-event image, the recommended technique can accurately identify forest fires with a 90% accuracy and 0.79 Kappa coefficient. For fire risk assessment and fire damage estimation this effective framework's outcome is beneficial.