Flood Inundation Mapping and Impact Assessment Using Multi-Temporal Optical and SAR Satellite Data: a Case Study of 2017 Flood in Darbhanga District, Bihar, India

Flooding is a recurrent hazard in east Gangetic plains, largely on account of natural factors that pose risks to life and property. Bagmati and Burhi Gandak rivers draining parts of North Bihar causes substantial flooding owing to higher rainfall. This comprehensive study was carried out to map near real-time flood inundation using multi-temporal Sentinel-1A (SAR) and Moderate-resolution Imaging Spectroradiometer Near Real-Time (MODIS NRT) flood data (Optical and 3-day composite) over Darbhanga district of North Bihar during August and September 2017. Floodwater pixels were extracted using the binarization technique, wherein the threshold was applied as −22.5, −23.4, −23.8 and − 22.7 over VH polarization image. The key results revealed that during peak flooding stage (23rd August), 13% of areas were submerged based on SAR data, whereas overestimation by >20% was estimated using MODIS data. As shown in the composite flood inundated map, the inundated patches are quite similar in both the optical and SAR based data. Notably, there were higher flood patches observed in the central, northern, and western parts of the district due to the presence of more water channels in those regions. Our findings suggested that agriculture patches of ~392 sq.km area were inundated due to flood followed by vegetation clutters (16.07 sq.km) and urban (8.46 sq.km). These results indicated the impact of floodwater on agriculture and urban patches. These findings are crucial for policymakers to assess flood impacts. It can be inferred that flood prognosis using SAR data will lead to spatial accuracy and can be improved when coupling with various hydro-meteorological parameters and hydrological models.