Evaluation of the performance of remotely sensed rainfall datasets for flood simulation in the transboundary Mono River catchment, Togo and Benin

This study focused on the Mono River Basin in west Africa. The lack of extensive and functional measurement networks for flood monitoring, introduces satellite-based rainfall datasets as an alternative which needs however to be evaluated beforehand. investigated performance four satellite gauge-based products – Climate Hazards Group Infrared Precipitation with Station data (CHIRPS), Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record (PERSIANN), Tropical Applications Meteorology Satellite ground-based observations (TAMSAT), Global Climatology Centre full daily (GPCC) grid-to-point hydrologic modelling approaches at different time scales over basin. With assessment, results show poor performances annual while seasonal cycles were well reproduced Nash-Sutcliffe efficiency (NSE) equal or higher than 0.94, correlation coefficient above 0.9. All assessed exhibited high probability detection (POD) low false alarm ratio (FAR) dekadal scale. Based NSE values modelling, best achieved by PERSIANN, followed GPCC TAMSAT, but CHIRPS performed worst negative values. By filling gaps gauge products, we noticed that missing does not necessarily improve quality may needed case basin if interpolation methods like kriging are applied.