Cyclone-induced vegetation loss in the Sundarbans and storm surge inundation in the coastal districts of Bangladesh: A multi-cyclone assessment using Sentinel-1 and Sentinel-2 imagery

The coastal districts of Bangladesh and the Sundarbans mangrove forest region are experiencing increasing cyclone impacts, as well as sea level rise and ocean warming. The study uses Sentinel-2 multispectral imagery with a resolution of ~10 m and Sentinel-1 Synthetic Aperture Radar (SAR) to estimate vegetation loss and inundation by storm surge due to Cyclones Fani (2019), Bulbul (2019), and Amphan (2020). Changes in vegetation and storm surge inundation were quantified by NDVI and mapped with Sentinel-1 VV backscatter and scene-specific threshold-based classification, respectively. The cyclone-induced dense canopy loss is the highest in Cyclone Fani (~554 km 2 ), and the storm surge inundation is the highest in Cyclone Amphan (~1416 km 2 ) over seven coastal districts, as revealed in the results. The accuracy of classification was over 92%, with kappa >0.80 for methodological robustness. The multi-cyclone approach is multi-sensor and provides a replicable procedure for deltaic risk assessment. In the interest of enhancing resilience, we recommend focusing on cyclone-resistant mangrove reforestation, institutionalizing the tracking of flood through SAR and making geospatial analytics an integral part of local disaster preparedness. There is also a need for hybrid embankment systems and enhanced drainage infrastructure in the long-term for adaptation support by policymakers.