Delineating Flood Zones upon Employing Synthetic Aperture Data for the 2020 Flood in Bangladesh

Abstract

Delineating a flood map is critical to perceive the potential risks of the event at diverse communities living both in urban and rural settings in Bangladesh. A timely generated flood map can help determine the losses of properties, calculate payment options from insurances, and set up mitigation measures when required. Application of satellite remote sensing (RS) and geographic information systems (GIS) are common these days to determine inundated areas, and to calculate possible losses of economies at scale. However, challenges remain while considering the available options for collecting satellite imageries obtained during the monsoon season with more than 70% cloud coverages found in the data. As a result, active synthetic aperture radar (SAR) sensors are a better choice to utilize the data in delineating the inundated areas. In doing so, this scientific paper sets up a few objectives to (1) prepare a flood map of Bangladesh using SAR remote-sensing data available from Sentinel-1 satellite; and (2) generate the inundated maps using cloud-based product, i.e., Google Earth Engine (GEE) in categorizing flood-affected districts of Bangladesh in 2020. Results have demonstrated that approximately 11% area of Bangladesh has been affected by the 2020 flood mainly located in the north-central and north-eastern part of the country. Moreover, the old Brahmaputra floodplain, Tista floodplain, lower Ganges-River floodplain, and Karataya-Bangali floodplain have been severely affected by the flood. Note that, the GEE-based automated processing systems adopted in this study have enhanced the computational time while obtaining freely available satellite data to generate mitigation strategies for the betterment of the communities suffered by the flood event.