Flood Inundation Mapping and Analysis Using Sentinel- 1 GRD and Sentinel- 2 MSI Datasets

Floods are among the recurrent catastrophic disasters affecting India, particularly the coastal districts of Tamil Nadu during the northeast monsoon season. In December 2024, continuous heavy rainfall caused the Thamirabarani River to overflow, resulting in severe inundation across several regions near the riverbanks. This study presents an integrated multi-sensor approach for delineating and validating the spatial extent of flooding in and around Eral, Thoothukudi District, Tamil Nadu. The proposed approach combines Sentinel-1A dual-polarization SAR (Synthetic Aperture Radar) data with Sentinel-2 MSI optical observations, integrating the advantages of all-weather radar imaging with optical spectral information. Sentinel-1A Level-1 Ground Range Detected (GRD) data acquired in Interferometric Wide Swath (IW) mode with VV and VH polarizations were used to characterize changes in surface backscatter associated with inundation. An Otsu threshold-based backscatter change detection approach was employed to objectively delineate flooded areas, reducing dependence on manually selected thresholds. The resulting SAR-derived flood extent indicated substantially reduced backscatter over inundated agricultural lands and settlements adjoining the floodplain. To enhance the reliability of the delineation, the extracted flood extent was independently validated using Sentinel-2 MSI-derived Normalized Difference Water Index (NDWI) and Modified Normalized Difference Water Index (MNDWI). The validation showed good spatial agreement between SAR-derived and optically identified inundated areas, with approximately 84.3% correspondence in the mapped flooded regions. The key innovation of the study lies in integrating dual-polarization SAR backscatter change detection, objective Otsu thresholding, and independent optical-index validation within a unified framework for rapid and reliable flood delineation under monsoon conditions. The proposed approach demonstrates the potential of multi-sensor satellite observations for effective flood assessment in vulnerable riverine and agricultural regions of Tamil Nadu.