Application of Forecast Verification Science to Operational River Forecasting in the U.S. National Weather Service

Abstract

Forecast verification in operational hydrology has been very limited to date, mainly due to the complexity of verifying both forcing input forecasts and hydrologic forecasts on multiple space?time scales. However, forecast verification needs to be the driver in both hydrologic research and operations to help advance the understanding of predictability and help the diverse users better utilize the river forecasts. Therefore, in NOAA's National Weather Service, the Hydrologic Services Program is developing a comprehensive river forecast verification service to routinely and systematically verify all hydrometeorological and hydrologic forecasts. This verification service will include capabilities for archiving forecast and observed data, evaluating logistical properties of the forecast services, computing a variety of verification metrics to evaluate the different aspects of forecast quality, displaying and disseminating verification data and metrics, and analyzing the sources of forecast skill and uncertainty through the use of multiple forecast and hindcast scenarios. This paper describes ongoing and planned verification activities for enhancing the collaboration between the meteorological and hydrologic research and operational communities to quantify forecast improvements based on rigorous forecast verification.