The NCAR Ensemble Project: Lessons Learned and Future Plans

Recently, NCAR has been actively developing experimental, convective-scale, real-time, ensemble prediction systems, to learn about best practices in storm-scale ensemble design. A current demonstration system, known as the NCAR ensemble, is focused on next-day high-impact weather prediction that relies on a mesoscale (15 km horizontal grid spacing) ensemble analysis produced by an ensemble Kalman filter. The NCAR ensemble project, begun in April 2015, has demonstrated the feasibility of a real-time, conterminous U.S. coverage, ensemble analysis and forecast system for high-impact, next-day, short-range weather predictions. These predictions have been included each Spring in the NOAA Hazardous Weather Testbed activities, while the full multi-year dataset has garnered considerable interest from within the research community. Moving forward, research activities will be focused on the development of a high-resolution ensemble analysis system (3-km horizontal grid spacing) that includes assimilation of cloud and precipitation information to markedly improve short-term (1-12 h) high-impact weather prediction. The talk will provide a brief overview of NCAR’s ensemble analysis and forecast system design strengths and deficiencies for high-impact weather applications along with our progress in developing a high-resolution ensemble analysis over the conterminous United States.