HR: 1330h
AN: H12B-0970    [PDF]
TI: Ensemble Streamflow Forecast Verification: Putting Research Into Practice
AU: * Franz, K J
EM: franzk@uci.edu
AF: University of California, Irvine, Civil and Environmental Engineering, E/4130 Engineering Gateway, Irvine, CA 92697 United States
AU: Sorooshian, S
AF: University of California, Irvine, Civil and Environmental Engineering, E/4130 Engineering Gateway, Irvine, CA 92697 United States
AU: Welles, E
AF: National Weather Service, Hydrology Lab, 1325 East-West Hwy, Silver Spring, MD 20910 United States
AU: Brandon, D
AF: National Weather Service, Colorado Basin River Forecast Center, 2242 W. North Temple, Salt Lake City, UT 84116 United States
AU: Townsend, S
AF: Riverside Technology, Inc., 2290 E. Prospect Rd Ste. 1, Fort Collins, CO 80525 United States
AU: Teixeira, L
AF: Riverside Technology, Inc., 2290 E. Prospect Rd Ste. 1, Fort Collins, CO 80525 United States
AB: The application of probabilistic forecast evaluation methods to synthetic hindcasts was demonstrated in a previous study. The goal of this initial research was to assess the potential for using specific verification methods for National Weather Service Ensemble Streamflow Predictions (NWS ESP). Results from this work indicated that the measures studied (ranked probability skill score, discrimination, and reliability) provided a comprehensive evaluation of forecast characteristics. However, it remained unclear whether the statistics could be put into practical application. The second phase and current phase of this study highlights the problems, issues, and obstacles that come to light when trying to apply the developed statistical measures to operational forecasts. With respect to the NWS ESP forecasts, data archives presented a major obstacle to implementing the procedures on one type of forecast data set. Variations in forecast formats across the country present another challenge to be faced as formalized software development begins. Initial progress indicates that once implemented, this initial set of statistical measures will provide the NWS forecasters and users with consistent comprehensive forecast performance data.
DE: 1800 HYDROLOGY
DE: 1899 General or miscellaneous
SC: Hydrology [H]
MN: 2003 Fall Meeting