Verification of Severe Weather Proxies from the NSSL-WRF for Hail Forecasting

With convection allowing models (CAMs) becoming more prevalent and finer in resolution, expansive datasets exist for information extraction of simulated storm attributes. Because convective processes are becoming increasingly resolved in CAMs, the question is whether useful forecast information can be extracted to improve upon existing severe weather proxies.   First of all, however, the skill of existing severe weather proxies for hail forecasting must be established. The NSSL-WRF is a 4-km CAM that is used operationally at the Storm Prediction Center (SPC). While some forecast verification  studies have been performed with the NSSL-WRF, none of the efforts have focused specifically on severe hail prediction. The current work will focus on providing a long-term (2012-2015) neighborhood forecast verification of the NSSL-WRF for severe hail using local storm reports and radar-derived maximum estimated size of hail (MESH) as the observational datasets.   The verification results of the various severe weather proxies will be compared along with the impact of the observational dataset used for verification.  In addition, the differences in verification results by neighborhood size and season will explored as well.