Questions regarding the effects of future anthropogenic climate change on severe convective storms remain unanswered owing to: (i) the inability of typical global climate models (GCM) and even regional climate models (RCM) to explicitly resolve such events, and (ii) limitations in simple physical arguments. Ultimately, some means of downscaling is required. Toward this end, we have developed a dynamical downscaling approach that culminates in the use of a convection permitting dynamical model.

Our scale-spanning, multiple-model system consists of a GCM (or global dataset), which drives long-term RCM integrations, which in turn drive short-term mesoscale and nested convection-permitting model integrations via the Weather Research and Forecasting (WRF) model; an alternative modeling system consists only of the global climate and mesoscale/cloud-resolving models. Retrospective simulations of two events have been used to test both modeling approaches as well as our use of model data to generate proxies of severe-storm reports. Our testing now focuses on estimating climate statistics of severe storms over limited areas and extended time periods. Results and unresolved model-related issues will be discussed.