Using a Frequently Updated, Storm-Scale Ensemble Prediction System for Short-Term Prediction of Heavy Rainfall

The Warn-on-Forecast project (WoF) has made significant progress in storm-scale assimilation, analysis, and prediction since its inception in 2009. This project is motivated by a desire to provide National Weather Service forecasters with numerical guidance that will enable longer lead times, greater accuracy, and probabilistic rather than deterministic forecasts and warnings of severe convective hazards. This will be achieved by assimilating incipient convective storms into ensemble prediction systems (EPSs) and predicting their evolution and associated storm-scale hazards. Much of the early work on this project has focused on tornadoes, which can be the most violent convective hazard and one of the most challenging phenomena to predict, but it is clear that the newly developed EPSs emerging from this project have significant potential to help with prediction of other convective hazards as well.

In this presentation, we examine retrospective 0-6 h ensemble forecasts of the 31 May 2013 tornado and flash flood event over central Oklahoma, with a focus on prediction of heavy rainfall. Precipitation forecasts from the EPS are verified against NCEP’s Stage-IV analyses, using common probabilistic-forecast verification metrics in addition to a novel feature-based storm-tracking algorithm.