Flood Extent Mapping of the 2022 Sindh Flood: A Remote Sensing Analysis Using SAR, Optical Data, and Google Engine Platform

Floods are the most frequently occurring natural disasters worldwide, caused by extreme rainfall, river overflow, high runoff, glacier melting, and flood infrastructure failure. Climate change has further increased the frequency and severity of floods. While traditional methods are limited over large areas, remote sensing provides a viable solution to analyze floods. However, challenges such as cloudy weather conditions can hinder accurate flood detection. To address these problems, this study used optical and SAR remote sensing indices in the Google Earth Engine (GEE) cloud computing platform. This paper analyzes the impacts of the 2022 flood in Qambar Shahdadkot and Dadu districts, Sindh Province, Pakistan, on population and cropland area. SAR and optical remote sensing have been used to delineate flood extent. Population data has been collected from the JRC Global Human Settlement Population Layer, which has a 250 m resolution. To estimate the cropland affected, the MODIS Land Cover Type product has been used with a spatial resolution of 500 m. The resultant analysis indicated that approximately 50% of Qambar Shahdadkot was inundated, covering an area of 2,620 km2, while Dadu was around 25%, with damage to cropland standing at 73% and 40% of agricultural areas in the respective districts. Population exposure analysis identifies the most densely populated areas in the districts that remained submerged for more than 3 months; the result has been the massive displacement and loss of livelihood for thousands of people. Though this study used the 2022 flood as a case study, the findings and the methodology can be adopted and applied by other researchers to analyze any flood in different regions to get near real-time information for emergencies during the flood.