Development of an operational flood early warning system for Black Volta River Basin, West Africa

Floods are the most frequent disaster causing global economic losses in billions and pose a significant threat to modern civilization. The UNDRR strongly advocates for flood early warning system (FEWS) with scientific rationale for all nations by 2027, acknowledging that developing countries face financial and human resource challenges in adopting advanced FEWS infrastructure. This study is focused on the development of a FEWS for the Black Volta Basin (BVB) in West Africa with free data resources and open-source modeling infrastructure. It is based on the approach of integrating the Deltares wflow_sbm hydrologic model and the LISFLOOD-FP hydrodynamic model for forecasting flood and inundation maps. The wflow_sbm is calibrated (1990–1997, NSE value = 0.71) and validated (1998–2007, NSE value = 0.72) by using station-based gridded rainfall from the West African Science Service Centre on Climate Change and Adapted Land Use (WASCAL) and discharge time series from Global Runoff Data Centre (GRDC) portals. Based on the calibrated parameters, wflow_sbm model is utilized to produce hydrograph for the years 2001–2022 with raw and bias-corrected GPM-IMERG rainfall inputs, where the discharge with the latter is found to outperform that from the former. The peak flood event from the produced hydrograph by the wflow model is fed into a 2D hydraulic model, LISFLOOD-FP model, to simulate the flood extent. Evaluation of modeled inundation modeling by comparing with satellite inundation observation during flood 2022 case resulted in an acceptable range (F = 0.527). Hydrograph for the flood 2022 case is overlapped with hydrographs from GEFSv12 weather forecast inputs in 1 day, 2 days, and 3 days. It is found that the absolute error percentage for 1 day throughout most of the season is forecasted under 10% including the peak of the flood. Forecasts lead time of 2 and 3 days are observed to have degraded accuracy as compared to 1-day forecasts due to higher uncertainties. Identification of the onset of hydrograph inclination is also found to underperform by GEFSv12 inputs and possible causes are discussed. The aim of this work is to promote FEWS with limited resources in African river basins, considering the problem of data scarcity.