Flood Mapping Using Combined Sentinel-1 Amplitude and Interferometric Coherence

Synthetic Aperture Radar has become an operational tool for near-real-time flood mapping, crucial for policymakers and disaster planners around the world. However, it faces challenges in dry, sandy areas where water and land have similar backscatter characteristics, rendering traditional amplitude-based methods ineffective. This study introduces an approach to overcome these limitations by combining the amplitude and coherence information from SAR for flood detection in arid regions. Focusing on a case study in Ormara, Pakistan, we demonstrate that this method substantially enhances flood mapping accuracy in challenging terrain. Our results reveal a $\mathbf{2 5 \%}$ improvement in accuracy compared to amplitude-only information. This advancement improves our ability to map floods in dry arid environments, potentially improving disaster response and mitigation in dry arid regions.