Employing Remote Sensing Data to Determine the Hydraulic State of Rivers in Data-Scarce Regions: Application to the Sourou River, West Africa

A considerable share of the world’s population lives in riverine and deltaic regions, where access to water, transport, and fertile soils is relatively easy. While seasonal floods can benefit agriculture, extreme flood events threaten lives and infrastructure. These risks are expected to intensify, especially in developing regions of Africa and Asia where vulnerability is high and hydrological data are limited. Conventional flood risk assessments frequently rely on discharge measurements, high-resolution Digital Elevation Models (DEMs), and hydrodynamic or hydrological models. However, these resources are not always available in data-scarce areas due to institutional, technical, or financial constraints. Discharge data may also become unreliable during overbank flows, and obtaining elevation or hydrological data can be time-consuming bureaucratic processes. This emphasizes the need for alternative, open-source methodologies that leverage publicly available Earth observation data with high temporal resolution. This study explores the potential of L-band passive microwave remote sensing (PMRS) from the Soil Moisture Active Passive (SMAP) mission to detect seasonal inundation in the Sourou River floodplain, West Africa. Multiple datasets—including precipitation from GPM, water levels from DAHITI and SWOT, historical discharge records, and Sentinel-2 optical imagery—were used for validation and context. Results show that the CM-ratio method outperforms polarization-based indices in capturing floodplain dynamics, particularly during transitional periods (January and October–November) when flooding is shallow but widespread. Validation with SWOT confirmed strong seasonal agreement, while Sentinel-2 was less reliable in late-season shallow flooding due to vegetation cover. These findings indicate that, with careful calibration, PMRS can provide near-daily floodplain monitoring in data-scarce regions, offering a practical complement to sparse in-situ networks.