We tested the use of CAMs to forecast severe weather threats associated with nocturnal supercells. Since nocturnal supercells can be associated with just large hail (hail-only) or both large hail and damaging winds (hail-and-wind), we ran simulations with 3-km grid spacing of nocturnal supercells over the central and northern plains of the United States 11 hail-only cases and 11 hail-and-wind cases. Our results show that the WRF-ARW successfully differentiated between storms that have only a hail threat from those that have both a hail and wind threat. Updraft helicity, updraft and downdraft strength, maximum reflectivity and 10-m wind speeds associated with simulated convection were significantly stronger in hail-and-wind cases than hail-only cases (p-value < 0.01).
We also investigate the impact of different parameterizations and ran each case with four additional combinations of microphysics (MP) and planetary boundary layer (PBL) schemes. We find that differences between hail-only and hail-and-wind cases remain regardless of which combination of parameterizations was used. But comparing storm-scale diagnostics on the same set of cases (i.e., hail-only cases) using different parameterizations also resulted in differences. In some cases, the differences between the same set of cases using different parameterizations were nearly as large as those between the different groups of cases (hail-only vs. hail-and-wind) using the same parameterization.
These results show that development of critical values of updraft helicity or other storm-scale diagnostics indicative of severe weather within CAMs will have to be derived independently for different combinations of PBL and MP schemes. Forecasters will also need to be aware of the parameterizations being used in CAMs to better interpret the likelihood of the type of severe weather associated with nocturnal supercells.