The short-range hydrologic forecasts from the NWM are driven by inputs from the high-resolution, convection-allowing High Resolution Rapid Refresh (HRRR) model. This NOAA Testbed project began by using a limited set of high-impact flood case studies to design a time-lagged NWM ensemble in which the hourly short-range NWM forecasts are consecutively combined for overlapping periods (i.e. valid times) to create time-lagged ensembles. After testing on case studies, real-time experimental products were created in a demonstration effort at the National Water Center (NWC).
Through the NOAA Hydrometeorology Testbed, experimental probabilistic case study prototypes were first demonstrated to forecasters at the Weather Prediction Center’s 2016 Flash Flood and Intense Rainfall (FFaIR) experiment. After incorporating forecaster feedback collected during the 2016 experiment, real-time probabilistic NWM forecasts and displays were demonstrated and evaluated more fully at the 2017 FFaIR experiment. This talk will highlight the process of developing, testing, and evaluating the probabilistic products, in addition to preliminary ensemble verification results. We will discuss FFaIR case studies of interest, and will summarize forecaster feedback as input to ongoing NWM flash flood forecast guidance improvement efforts.