Predictability of convective rainfall events during different large-scale weather regimes in convection-allowing and convection-parameterizing simulations

A case study analysis of a regional severe weather outbreak that occurred during the early afternoon of 01 June 2007 in eastern and southeast Iowa, as well as northwest Illinois, will be performed. Examination of convection-allowing (CA) and convection-parameterizing (CP) ensemble forecasts indicates that the two ensembles behaved very differently, and that in this particular case, the CA ensemble produced much better forecasts. Preliminary analyses suggest that the superior forecasts in the CA ensemble resulted from the ability of all the members to successfully simulate a large nocturnal MCS that developed in central Kansas during the early morning hours of 01 June 2007. The observed MCS produced a mesoscale convective vortex (MCV) which rotated into eastern Iowa around the periphery of a large cut-off 500-hPa low that was centered over the Dakotas. The synoptic-scale forcing and enhancement of mid-level winds associated with the MCV prouced favorable conditions for severe weather in eastern Iowa and western Illinois during the early afternoon on 01 June 2007 and likely provided the forcing for the development of the severe storms. Unlike the CP ensemble members, it appears that the CA ensemble members were able to successfully simulate the MCV produced by the MCS and the resulting influences on the larger scale conditions that were associated with the regional severe weather outbreak.

It is hypothesized that the large scale conditions associated with the strong flow around the cut-off 500 hPa low were associated with a high degree of predictability for the nocturnal MCS and resulting MCV in the CA simulations, and that, because of the inability of the CP ensemble members to simulate propagating convective systems, there was a large difference between the two sets of forecasts for this event. Thus, in addition to the case study analysis, a large set of experimental CA forecasts and operational CP forecasts covering a multi-year period will be used to compare the ability of these forecasts to predict large scale flow and severe weather parameters during different warm season weather regimes when convection was present in the model domain before the forecast period of interest.