The Severe Hazards in Environments with Reduced Buoyancy parameter, particularly those formulations utilizing the effective shear magnitude (SHERBE) and 0-3 km shear magnitude (SHERBS3), has been shown in previous blogs and presentations to improve upon the forecasting skill of existing composite parameters in high shear, low CAPE (HSLC) environments. Though designed specifically with HSLC environments in mind, recent testing has shown that the SHERBS3 and especially the SHERBE exhibit skill at discriminating between significant severe and non-severe convection across all environments.
Figures 1 and 2 below show the familiar TSS vs. threshold plots for NWS Southern Region and Eastern Region, respectively. In these figures, the TSS is measured as the parameters’ ability to discriminate between all significant severe reports and nulls, regardless of environment. Figures 3 and 4 are similar, except that the TSS is calculated discriminating between all significant tornadoes and nulls, with no environmental constraints.
There are several noteworthy features of these plots. First, regardless of environment, the SHERBE is the most skillful composite parameter at discriminating between significant severe reports and nulls in both Southern and Eastern Region. Not only is it the most skillful parameter, it also shows a consistent optimal threshold relative to other existing composite parameters, including the SHERBS3. The SHERBS3 and Significant Tornado Parameter (STP) have comparable or slightly higher skill when discriminating between significant tornadoes and nulls; however, their thresholds must be adjusted downward for this to be true. At their conventional thresholds, the SHERBE outperforms the SHERBS3 and STP (along with all other tested parameters) at discriminating between significant tornadoes and nulls. Finally, even when using no environmental constraints, the STP’s optimal threshold is below 1, suggesting that the threshold of 1 from Thompson et al. (2004) may not be ideal for our CSTAR domain.
These findings are robust, as they provide evidence that the SHERBS3 and particularly SHERBE can be used with confidence across all environments within our CSTAR domain. In the future, we will cite these findings in an attempt to get the SHERBS3 and SHERBE plotted real-time on platforms such as the SPC Mesoanalysis.