Progress and Challenges with Warn-on-Forecast

Warn-on-Forecast is a ten-year National Oceanic and Atmospheric Administration project to extend severe weather warning lead times by incorporating forecasts from a convection-permitting ensemble modeling system into the warning decision process. Initial results with convection-permitting ensembles that assimilate radar observations have shown that while reasonable thunderstorm analyses are produced, obtaining accurate very short-range thunderstorm forecasts is more challenging. The motivation for Warn-on-Forecast will be outlined. Results from several warn-on-forecast data assimilation and prediction experiments and from real-time testing of an analysis system within the Hazardous Weather Testbed will be shown to highlight where progress has been made and where challenges remain. While other presentations at the conference will provide detailed information on the case studies, this overview will outline the common themes arising from these experiments and provide a vision for the future.