Translating the potential of hydrological forecasts into improved decision making in African regions

Thursday, 17 December 2015: 14:40
3003 (Moscone West)
Justin Sheffield1, Xiaogang He1, Niko Wanders2, Eric F Wood1, Abdou Ali3, Luke Olang4, Lyndon D Estes1 and Kelly K Caylor1, (1)Princeton University, Princeton, NJ, United States, (2)Princeton University, Civil & Environmental Engineering, Princeton, NJ, United States, (3)AGRHYMET Regional Center, Niamey, Niger, (4)ICPAC, Nairobi, Kenya
Hydrological forecasts at local scale and seasonal time scales have the potential to inform decision-making by individuals and institutions to improve management of water resources and enhance food security. Much progress has been made in recent years in understanding climate variability and its predictability over African regions. However, there remain many challenges in translating large-scale evaluations and forecasts into locally relevant information. This is hampered by lack of on the ground data of hydrological and agricultural states, and the generally low skill of climate forecasts at time scales beyond one or two weeks. Additionally, the uptake of forecasts is not prevalent because of lack of capacity, and institutional and cultural barriers to using new and uncertain information. New technologies for monitoring and forecasting relevant hydrological variables, and novel approaches to understanding how this information may be used within decision making processes, have the potential to make substantial progress in addressing these challenges.

We present a quasi-operational drought and flood monitoring and forecasting system and its use in understanding the potential of hydrological forecasts for improved decision-making. The system monitors in near real-time the terrestrial water cycle for the African continent based on remote sensing data and land surface hydrological modeling. The monitoring forms initial conditions for hydrological forecasts at short time scale, aimed at flood forecasting, and seasonal scale aimed at drought and crop yield forecasts. The flood forecasts are driven by precipitation and temperature forecasts from the Global Forecast System (GFS). The drought forecasts are driven by climate forecasts from the North American Multi-Model Ensemble (NMME). The seasonal forecast skill is modest and seasonally/regionally dependent with part of the skill coming from persistence in initial land surface conditions. We discuss the use of the system in the context of operational implementation by regional climate centers and in local evaluations of the predictions. We also discuss how forecasts might be used within the decision making process of smallholder farmers and the practicalities of disseminating information from operational centers to local actors.