Ongoing Refinements to the High-Resolution Ensemble Forecast System (341)

Matthew Pyle, NCEP/EMC, College Park, MD

Benjamin Blake, I.M. Systems Group, Inc. @ NCEP/EMC

Jacob Carley, NCEP/EMC



Ongoing Refinements to the High-Resolution Ensemble Forecast SystemMatthew Pyle, Benjamin Blake, and Jacob CarleyConvection-allowing ensembles are an increasingly important tool in short-term forecasts of high impact weather events, and improving them continues to be an active area of development. This talk describes ongoing developments with the High-Resolution Ensemble Forecast (HREF) system, which is a multi-model convection-allowing ensemble run in NWS operations. The HREF currently combines forecasts from the two most recent cycles of the High-Resolution Window (which includes both the NMMB model and two different WRF-ARW configurations) and the 3 km NAM nest into an eight member system for product generation. The HREF builds upon the success of SPC's well-known Storm Scale Ensemble of Opportunity (SSEO), utilizing a similar ensemble membership and applying comparable product generation techniques. Spread within the HREF, as with the SSEO before it, comes from the sampling of different dynamical models, physics schemes, analyses, and initialization times. This approach, while somewhat ad hoc, provides skillful mean fields and probabilistic guidance that highlights regions of interest for forecasters. Different forms of probabilistic product generation (point, neighborhood maximum) will be described, as both currently are used and each has utility for specific applications. A newer method for defining probabilities, based on the Ensemble Agreement Scale (EAS) approach, also will be examined. A switch to EAS probabilities for certain fields is being considered for a future upgrade, as is adding the extended 36-h HRRR forecasts as a new HREF model component the impact of these two changes on HREF forecast skill will be presented.