Short-term (0-6 h) Probabilistic Extreme Rainfall Forecasts Using a Prototype Warn-on-Forecast System

The National Oceanic and Atmospheric Administration's (NOAA) National Severe Storms Laboratory (NSSL) is actively developing prototype Warn-on-Forecast (WoF) systems with the goal to incorporate NWP model based forecast guidance of severe convective hazards in NWS warning decision process. While the main focus of WoF is on tornadoes, the technology and science that is being developed to achieve the WoF goal will likely improve the prediction of other convective weather threats like flash floods, large hail and damaging winds. On average, floods, especially flash floods, cause more fatalities in the USA each year than tornadoes, hurricanes lightning, or high winds. To improve prediction of flash floods and nowcasting of convective precipitation, accurate (i.e. amount, location, and timing) rainfall forecasts from NWP models is crucial. This presentation will focus on the capability of a prototype WoF system in predicting flash flood producing, convectively driven extreme rainfall events. Retrospective short-term (0-6 h) probabilistic ensemble forecasts from several recent record breaking heavy rainfall and flash flood events will be generated. To evaluate how the frequent-update-cycle storm-scale ensemble data assimilation and forecast system performs in predicting these events, we will use both qualitative and quantitative verification metrics.