Inter comparison of post-fire burn severity indices of Landsat-8 and Sentinel-2 imagery using Google Earth Engine

Forest fires are significant catastrophic events that affect the landscape and vegetation in forested lands. They cause loss of biodiversity, land degradation & ecological imbalance. As the forest fires cause extreme damage to the habitat, it is of utmost necessity to assess the impact of fire on canopy/vegetation. Post-fire assessment is an essential element for finding the effects of fire on vegetation and implementing mitigation strategies. In this article, a Post-fire burn severity assessment was carried out with high-resolution multi-spectral images such as Sentinel-2 and Landsat-8 employing Google Earth Engine (GEE) to locate the burnt areas and fire severity. Three commonly used fire severity indices based on pre-fire Normalized Burn Ratio (NBR) and post-fire NBR, namely differenced Normalized Burn Ratio (dNBR), Relativized Burn Ratio (RBR), and Relativized dNBR (RdNBR) are computed and compared based on their accuracy with the active fire points provided by MODIS & VIIRS. Both Sentinel-2 and Landsat-8 exhibited a similar trend in mapping burn severity. The RdNBR resulted in high accuracy over heterogeneous landscapes with 61.52% for Sentinel-2 and 64.1% for Landsat-8 followed by dNBR (41.67% for Sentinel-2 and 47.44% for Landsat-8) and weak performance by RBR with 32.69% for Sentinel-2 and 26.92% for Landsat-8. Hence RdNBR burn severity maps are considered highly appropriate for mapping burnt areas. Even though severity analysis from both Sentinel-2 and Landsat-8 is at an acceptable level, the Landsat based burn severity maps provided an adequate assessment of the degree of damage.