Verification of Pre-convective Environments in WRF 3-km forecasts using Sounding Data from the Verification of the Origins of Rotations in Tornadoes Experiment (VORTEX 2)

The Weather Researching and Forecasting (WRF) model has been used operationally to predict weather. Previous studies have mentioned the difficulty of predicting convective weather accurately in Numerical Weather Prediction (NWP) models. The focus of this study is to examine the pre-convective environment in high-resolution WRF forecasts using sounding data. Pre-storm environments in the WRF 3-km forecasts from the NOAA Hazardous Weather Testbed (HWT) 2010 Spring Experiment will be compared to sounding and observed radar data from the Verification of the Origins of Rotation in Tornadoes Experiment (VORTEX2). Data from June 7, 9-11, and 13-15, of 2010 of the VORTEX2 field campaign were selected for comparison with model forecasts. The examined WRF 3-km forecasts were initialized at 0000 UTC on these days. Sounding data taken before storms initiated in the model and observations were compared to observe differences in the pre-convective environment. Significant differences included consistently warm profiles in the boundary layer and a weak or missing capping inversion; however, above the boundary layer, the model temperature and wind profiles were fairly accurate. The differences seen in the profiles and radar composites suggest that a good long-term forecast (18-24 hr) is dependent upon a good short-term forecast (0-12 hr). Model vertical resolution may have been an issue in resolving capping inversions. The VORTEX2 data set proved valuable in evaluating model representation of the pre-convective environment. These findings reveal the need for further investigation into the high-resolution WRF forecast capabilities to represent the environment, pre-convective and/or convective.