Forecasts of Convective Mode using the NCAR WRF-ARW Convective Scale Ensemble

The NCAR/MMM real-time convective ensemble offers a glimpse into the future for the prediction of hazardous convective weather, providing forecasters valuable guidance as to anticipated convective mode and a new, more probabilistic perspective for anticipating potentially hazardous weather events such as tornadoes, hail, high winds, and flash floods. In this talk, I will present recent examples of both these unique capabilities as well as examples of forecast failures.  Some of the more notable events to be discussed include the 26 April 2016 severe weather outbreak (for which a somewhat controversial “particularly dangerous situation’’ (PDS) tornado watch was issued), a fairly rare derecho on 20 May 2016 along the Gulf coast, and the 22-23 June 2016 severe weather and flash flooding event in the Ohio valley and West Virginia. Such cases help to clarify both the inherent capabilities as well as limitations of such high-resolution ensemble systems, ultimately helping to clarify future research needs. Of particular note for the present system is a potential systematic northward bias for convective precipitation that is currently being investigated. Another ongoing issue is the apparent lack of sufficient ensemble spread, perhaps most notable as a tendency for significant forecast failure events to be completely missed by the envelope of ensemble solutions.