Ensemble-based long-range water resources predictions have been generated and relied upon in the Western U.S. for many years. In recent years, this methodology has become foundational for National Weather Service (NWS) River Forecast Centers (RFCs), replacing the traditional regression-based procedures. Over the past several years, the NWS has made a concerted effort to develop and implement the Hydrologic Ensemble Forecasting Service (HEFS) at its RFCs which applies this technology in all time domains from hours to seasons.
Movement of this technology into the shorter time domains has been challenging. Details that are not important at long range become extremely important in the short range. Fully leveraging the skill in weather and climate forecasts, representing extreme events, and retaining coherency between probabilistic and single-value predictions have proven particularly difficult. Given that reliable predictions can be operationally generated, communicating the information to potential users is challenging as well. Nonetheless, the value of providing meaningful and actionable risk-based information to the emergency services, flood management, and hydropower is extremely high and directly contributes to the NWS vision of a Weather Ready Nation.
This paper outlines recent efforts to meet the challenges of generating reliable probabilistic hydrologic predictions in real-time operations.
1NOAA/National Weather Service, California-Nevada River Forecast Center
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner