Coastal flood detection in Šilutė district using Sentinel-1 SAR and comparison with modeled 2022 flood-risk zones

Abstract

Flood monitoring and water level change analysis are essential for assessing climate change impacts and managing flood risks. This study aimed to identify flood‑affected areas in the Šilutė District using Sentinel‑1 SAR change detection and to compare them with modeled flood‑risk zones. Sentinel‑1 GRD VV‑polarized data from February and March 2025 were processed in ESA SNAP and analyzed in QGIS, applying orbit correction, radiometric calibration, speckle filtering, and terrain correction using SRTM DEM. Change detection results were classified into three groups: flooded areas (ΔVV ≤ –3 dB), double‑bounce/inundated vegetation (ΔVV ≥ 3 dB), and non‑flooded zones (–3 dB < ΔVV ≤ 3 dB). Statistical comparison with official flood probability maps (0.1 %, 1 %, and 10 %) revealed a substantial spatial overlap and general consistency between the detected flood‑affected areas and modeled flood hazard zones: the largest flooded areas coincide with the 0.1 % probability zone (31.3 million m²), decreasing to 26.2 million m² in the 10 % zone. Group 2 dominated in high‑risk areas, indicating extensive water interaction with vegetation and infrastructure. These findings confirm that Sentinel‑1 SAR data provide a reliable and spatially consistent tool for flood analysis and can effectively complement traditional hydrometric networks. Future work will integrate Sentinel‑2 multispectral data to improve classification accuracy and enable vegetation impact assessment during inundation events.