Rapid Flood Inundation Mapping During the 2019 Mississippi River Flood Using Sentinel‑1 SAR and Google Earth Engine

Flooding represents one of the most damaging natural hazards worldwide, causing substantial economic losses, infrastructure disruption, and environmental impacts. Timely delineation of inundated areas is critical for disaster response and flood-risk management. Optical remote sensing approaches are frequently hindered by cloud cover during extreme weather events, limiting their effectiveness for near-real-time monitoring. Synthetic Aperture Radar (SAR) provides an effective alternative because of its ability to acquire observations independent of weather and illumination conditions. This study develops a cloud-based flood mapping framework using Sentinel‑1 Synthetic Aperture Radar imagery and Google Earth Engine (GEE) to characterize inundation associated with the historic 2019 Mississippi River flood. The 2019 event affected large portions of the Mississippi River Basin and persisted for several months, resulting in extensive floodplain inundation across multiple states. Sentinel‑1 Ground Range Detected (GRD) imagery acquired before and during peak flooding conditions was processed using a backscatter change-detection approach to identify newly inundated areas. Image ratio analysis, terrain masking, and permanent-water removal techniques were applied to improve flood delineation accuracy and reduce false detections. Flood extent was quantified and integrated with Dynamic World land-cover data to assess impacts on agricultural lands, forests, wetlands, and developed areas. Spatial analysis was conducted to evaluate patterns of flood distribution throughout the study region and to estimate the inundated area by land-cover category. The cloud-computing capabilities of Google Earth Engine enabled rapid processing of large SAR datasets, demonstrating a scalable approach suitable for operational flood monitoring. Results indicate that Sentinel‑1 SAR effectively captures flood dynamics over extensive floodplain environments and provides reliable inundation mapping under cloud-covered conditions. The proposed framework offers a computationally efficient and transferable methodology for large-scale flood assessment and emergency management applications.