Evaluating NARCCAP Models for Severe Storm Environment Variables

Information about the future climate pertaining to the likelihood of extreme weather is important, but difficult to ascertain.  One method to infer about future extreme weather is to investigate large-scale indicators for severe storms.  In this study, concurrently high values are convective available potential energy (CAPE) and 0 - 6 km vertical wind shear (S) are calculated from the North American Regional Reanalysis (NARR) and output from several climate models from the North American Climate Change Assessment Project (NARCCAP) via the product between maximum updraft velocity (Wmax = the square root of twice CAPE) and S (provided CAPE > 100 J kg^-2 and 5 ms^-1 < S < 50 ms^-1; the product is set to zero otherwise), dubbed here as WmSh.  Because climate models do not attempt to match the daily values of the weather, distributional properties concerning high WmSh are considered.  In particular, the spatial patterns of ω, the frequency of WmSh values larger than 225 m²s^-2 conditioned on high field energy, is evaluated for each climate model output using cutting edge spatial weather forecast verification methods.  It is found that the models tend to under-predict the spatial extent of high frequency areas of ω, but generally characterize the spatial location and shape of these high-frequency areas reasonably well.