Hydrological Dynamics of Large Tropical Savanna Wetland Through Sentinel-1 SAR Imagery: Pantanal Ramsar Site Case Study

Seasonal tropical wetlands such as the Brazilian Pantanal are increasingly threatened by climate change and extreme events, creating a need for robust monitoring tools that capture hydrological dynamics at high spatial and temporal resolution. This study evaluates Sentinel-1 Synthetic Aperture Radar (SAR) imagery to map and monitor flooding within the Ramsar-designated SESC Pantanal Reserve from 2017 to 2020. Ground Range Detected (GRD) VV-polarized scenes were pre-processed with radio-metric terrain normalization and speckle filtering (Lee filter, 5×5 window) to improve separability of water and non-water surfaces. Flooded areas were first extracted using Otsu’s histogram thresholding and validated with high-resolution optical imagery (PlanetScope and Landsat-8). A supervised Random Forest classifier then refined land-cover discrimination into three classes (open water/flood, open land/vegetation, and others), with temporal consistency supported by Cuiabá River hydrological data. Results indicate strong interannual variability in flood extent, with March 2017 inun-dating 34.7% of the reserve compared with 0.75% in March 2020, and peak inundation in April 2017 (79.9%). Overall, Sentinel-1 SAR effectively delineated open water and flooded vegetation under persistent cloud cover, highlighting its value for comple-menting existing products (e.g., MapBiomas), strengthening wetland management, and supporting scalable flood monitoring in other tropical, flood-prone Ramsar sites.