Improving water level monitoring using CYGNSS DDMs and machine learning techniques

Water level monitoring plays a vital role in hydrology, as it not onlyfacilitates the optimization of water resource management but alsosupports flood disaster early warning. Traditional methods mostlyrely on in situ water level and empirical curves, which are unsuitablefor complex water bodies. The CYGNSS (Cyclone Global NavigationSatellite System) satellite system offers unique sensitivity to surfacewater, showing significant advantages in water level monitoring.This study leverages the statistical features of CYGNSS Delay-Doppler Maps (DDMs) to develop a water level estimation modelbased on machine learning (ML) techniques. Evaluations againstin situ data show the model accurately estimates water level varia-tions, with the Root Mean Square Error (RMSE) improving from0.735 m to 0.470 m (~36% decrease) after incorporating DDM fea-tures. To further evaluate model performance, additional tests wereconducted on independent spatiotemporal stations, which con-firmed its robustness and generalizability. This study improves real-time water level monitoring accuracy and offers novel approachesfor regional and global hydrological monitoring and disaster early warning.