Advancing NASA's seasonal hydrologic forecasting system for improved food insecurity warning in Africa

Hydrologic extremes such as droughts and floods contribute or lead to food insecurity, especially in vulnerable regions of Africa. Monitoring and forecasting such hydrological extremes thus provides an opportunity for early warning of food insecurity. With this in mind, an effort led at NASA's Goddard Space Flight Center has developed a multi-model, remote sensing-based hydrological forecasting and analysis system, referred to as NHyFAS (NASA's Hydrological Forecasting and Analysis System), to support food insecurity early warning efforts of U.S. Agency for International Development's (USAID) Famine Early Warning System Network (FEWS NET). For the past year, NHyFAS has been generating near real-time operational hydrological forecasts and analysis over continental Africa and the Middle East, using NASA's Goddard Earth Observing System Model (GEOS) Seasonal to Sub-seasonal (S2S) forecasts and NASA's Land Information System (LIS).

The work presented here describes an ongoing effort to expand and advance the hydrological forecasting system by including multiple climate model-based forecasts as opposed to using GEOS alone. The multi-model climate forecasts are provided by the North American Multi-Model Ensemble (NMME). The NMME suite currently provides near real-time monthly forecasts from global climate models such as the Climate Forecast System, version 2 (CFSv2); Geophysical Fluid Dynamics Laboratory's (GFDL) forecast-oriented climate model version 2.5, Canadian Coupled Model versions 3 and 4; the NCAR Climate System Model, version 4 (CCSM4); and GEOS S2S.

Previous independent studies have shown that the forecast skill of the NMME ensemble mean, for both precipitation (P) and temperature (T), is equal to or higher than the forecast skill of any single model. It is thus hypothesized that multi-model climate forecasts will lead to improved skill in the hydrologic forecasts generated by NHyFAS. Here we provide a comprehensive skill evaluation of this system over the single model-based forecasts using independent remotely sensed and in situ data as reference. The evaluation focuses in particular on the performance of this system in early warning of drought events that, in the past, have led to major food insecurity in the region.