Real-Time Evaluation of Extended Guidance Produced by a Global Convection-Permitting Model

The National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS) has recently begun an effort to evaluate and select the Next Generation Global Prediction System (NGGPS) for the U.S. The global model within this system is required to have the ability to run at non-hydrostatic scales, e.g., where Δx < 4 km, and NCAR's Model for Prediction Across Scales (MPAS) is one of the models being considered. The MPAS model is unique in that it can use variable-resolution meshes to economically span hydrostatic to nonhydrostatic scales, and its unstructured Voronoi horizontal mesh (nominally hexagons) is smoothly varying, thus eliminating many downscaling problems encountered using one- or two-way traditional grid nesting. We will be using the MPAS model, employing a variable mesh with 3 km CONUS resolution, to produce five-day forecasts for the 2015 Spring Forecast Experiment conducted at the NOAA Storm Prediction Center. We are evaluating these forecasts as part of SPC's ongoing efforts to provide extended range guidance for severe weather out to day five. MPAS will be initialized at 00 UTC each evening and 120 hour forecasts will be produced. Standard convective-scale fields and diagnostics will be made available for evaluation by HWT staff and SPC forecasters during the experiment in addition to the standard products from the global forecasts.

We will present preliminary results from the MPAS forecasts and focus on an evaluation of the forecast skill in capturing convection in the CONUS region in both the near (days 0-2) and extended (days 3-5) range. We also will show results illustrating the smoothly varying solutions produced in the MPAS forecasts in the mesh transition regions.

An important aspect of this modeling system is that the model physics be scale-aware, particularly the deep convection parameterization. These MPAS simulations employ the Grell-Freitas scale-aware convection scheme. Preliminary tests show that the scheme produces a gradual transition in the deep convection, from the deep unstable convection being handled entirely by the convection scheme on the coarse mesh region (Δx > 15 km), to the deep convection being almost entirely explicit on the 3 km CONUS region of the mesh. We will present results illustrating this behavior in our presentation.