The Value of Utilizing 1-km Horizontal Grid Spacing for the United States Air Force: Assessment of Storm Attributes within the Mesoscale Ensemble Prediction Suite

The 16th Weather Squadron (WS) provides tailored numerical weather model guidance for the United States Air Force (USAF) and its partners. Knowledge on high impact weather hazards (e.g., lightning, tornadoes, hail size, dust, turbulence, icing) is crucial to protect USAF assets and ensure successful missions globally. Given their small-scale resolution, convection-allowing (or permitting) models (CAMs) are utilized by the 16th WS to develop products forecasting these high-impact phenomena. Ensembles of CAMs better capture the inherent uncertainty in predictability (especially at long lead times), particularly for these high-impact, difficult to forecast phenomena. The 16th WS was one of the first to operationalize a CAM ensemble with various design configurations in 2012, referred to as the Mesoscale Ensemble Prediction Suite (MEPS). The MEPS utilizes the Weather Research and Forecasting (WRF) dynamical model core to generate high-resolution forecasts (with 1-km and 4-km grid point spacing) for numerous regional windows.

CAMs and CAM ensembles for many years now have employed horizontal grid spacings on the order of 3-4 km, which is sufficient to resolve many processes key to severe convective storms. These ensembles have enhanced forecasting capability for convective storm mode, initiation, and intensity. Studies examining CAMs with 1-km horizontal grid spacing have frequently found that the forecast improvement provided by moving from 4-km to 1-km horizontal grid spacing has been minimal, particularly when considering the computational cost of increasing the resolution. However, many of these studies have focused on the warm season, when deep, moist convection is responsible for most of the severe convective storm reports. This work examines storm attributes (e.g., simulated composite radar reflectivity, 2–5 km Updraft Helicity, 10 m wind speed) from two sets of CAM ensemble forecasts, one with 4-km horizontal grid spacing and one with 1-km horizontal grid spacing, to explore the question of when 1-km ensembles may create better forecasts for severe storms. Traditional warm-season convection cases as well as cool-season convection cases will be examined, as will regional differences, to determine whether 1-km horizontal grid spacing can add value for forecasting severe convective hazards when storms may be smaller or shallower. Expanding our understanding of when 1-km horizontal grid spacing can benefit forecasts relative to 4-km horizontal grid spacing will allow for better prioritization of computational resources while providing the best possible products to end-users and stakeholders.