On the Predictability of Mesoscale Convective Systems: Experiences of the Plains Elevated Convection at Night (PECAN) Forecasting Team

Now that convection-allowing grid spacing is used routinely in models to assist forecasters, it has become important to evaluate how predictable some of the finer-scale details of mesoscale convective systems (MCSs) are. Prior work has shown that model skill varies depending on the morphology of the MCSs. For instance, cellular morphologies are usually predicted more accurately than linear systems, with trailing-stratiform squall lines and bow echoes being especially poorly predicted by the Weather Research and Forecasting model. This presentation will focus on two aspects of MCS predictability. First, results from Weather Research and Forecasting model ensemble simulations of two bow echo events will be discussed to investigate the inherent predictability of this most poorly predicted morphology. Emphasis will be on the impact of refinement of horizontal grid spacing from 3 to 1 km. Our results suggest that a decrease in grid spacing within ensembles from 3 to 1 km does not affect spread in various fields examined, while results regarding skill of the forecast are mixed, suggesting a limit to predictability has already been reached for these systems. These results will be supplemented with some similar comparisons made using the NMMB model used operationally at NCEP.

Secondly, we will offer a practical look at predictability of MCSs and some finer scale details of these events by considering the experiences of the forecast team that was tasked with providing guidance to the science groups participating in the summer 2015 Plains Elevated Convection at Night (PECAN) experiment. During that project, the forecasters supplied probability forecasts for MCS occurrence in 3-hourly windows, and bores and pristine elevated nocturnal convective initiation in 6-hourly windows. Several bow echo events that were ultimately sampled by the mobile instrumentation networks during the projects will be emphasized.