Developing an operational impact-based flood forecasting system for the Greater Horn of Africa region

Every year Africa is affected by extreme weather related hazards which, combined with high levels of vulnerability and increasing population exposure, result in considerable impacts to people and assets. Here we present recent activities in the development of an African Multi Hazard Early Warning System for disaster risk reduction, a multi-year project funded by the Italian Government through the United Nations Office for Disaster Risk Reduction. After a brief introduction on the main project goals, the presentation will give insights on one of its key activities, focused on strengthening impact-based flood monitoring and forecasting capabilities for the Greater Horn of Africa region. This ongoing activity foresees the implementation of a probabilistic impact-based flood forecasting system based on the Flood PRObabilistic Operative Forecasting System (Flood-PROOFS) developed by CIMA Foundation and run routinely in several world regions. Flood-PROOFS has as its core the Continuum distributed hydrological model, which takes as input meteorological variables and several other static and dynamic data to simulate the hydrological processes in the focus region. For this application, Continuum was set up at hourly time step on 17 hydrologically consistent domains, with grid resolutions ranging between 250m and 3.3km, ultimately covering a surface of 6.82 million km². Given the relative scarcity of in situ data, model calibration is based on observed river discharges at 130+ river gauging stations, as well as on GLEAM satellite evaporation and soil moisture products. A 40-year continuous hydrological reanalysis is produced by forcing the model with ERA5 atmospheric reanalysis bias corrected for precipitation and temperature. Daily runs include model updates with satellite precipitation estimates and 5-day forecasts forced by the Global Forecast System. Ongoing activities are expanding the current deterministic setup to ensemble forecasts, as well as coupling hydrological forecasts in real time with state of the art global inundation maps, to estimate impact-based flood warnings by integrating information on exposure, vulnerability and coping capacity. Such information is used in the operational monitoring and early preparedness versus impending disasters, as well as to design prevention and mitigation measures, which is fundamental to prioritize and optimize the use of available resources for disaster response.