Traditionally, little attention has been focused on the systematic verification of operational hydrologic forecasts. This paper summarizes the results of forecast verification from 15 river basins in the United States. The verification scores for these forecast locations do not show improvement over the periods of record despite a number of forecast process improvements. In considering a root cause for these results, the authors note that the current paradigm for designing hydrologic forecast process improvements is driven by expert opinion and not by objective verification measures. The authors suggest that this paradigm should be modified and objective verification metrics should become the primary driver for hydrologic forecast process improvements.
Systems Engineering Center, National Weather Service, NOAA, Silver Spring, Maryland
Department of Civil and Environmental Engineering, University of California, Irvine, California
Office of Hydrologic Development, National Weather Service, NOAA, Silver Spring, Maryland
Arkansas-Red Basin River Forecast Center, National Weather Service, NOAA, Tulsa, Oklahoma