Flood mapping and impact analysis by fusion of remote sensing and open geospatial data: Sindh case study

Abstract

Flooding remains one of the most severe natural hazards in Pakistan, consistently leading to substantial losses in lives, livelihoods, and infrastructure. The country has experienced recurring flood events, with their frequency and intensity increasingly influenced by shifting climate patterns and irregular rainfall. The phenomena got worse over time and in 2022 all provinces of the country were severely impacted. The damage and impact of a flood may be detected, determined, and estimated with the use of remote sensing and available open geographic information system data. This study presents a scalable, efficient flood mapping framework that leverages freely available multi-source satellite data and open geospatial datasets to assess flood impact with high spatial detail. Multisource satellite imagery was utilized to detect inundation extents. Pre-processing of the remote sensing data was conducted using Google Earth Engine, and spatial integration of data layers for flood mapping was performed in ArcGIS. The results demonstrate that the 2022 Pakistan flood was the worst environmental disaster in history. The flood submerged a total area of nearly 25,000 km${}^2$ in the Sindh province, destroying 14,558 villages and leaving behind a trail of devastation. The methodology enables rapid, repeatable, and cost-effective flood damage assessment and is transferable to other regions. By combining cloud-based processing with open data, this framework supports timely decision-making for disaster response, prevention, and policy planning.