Forecasting short term convective mode and evolution for severe storms initiated along synoptic boundaries

Forecasting convective mode (discrete cells versus lines) and evolution continue to represent unique forecasting challenges. A comprehensive literature search suggests that there has been limited focused research to address these challenges, some of which were based on numerical modeling studies in idealized conditions. A more skillful forecast of severe weather would likely result from a correct anticipation of convective organization and evolution. Historically, buoyancy and vertical shear have been shown to be discriminators for supercell versus non-supercell modes. Other parameters, however, appear to play a substantial role in determining how thunderstorms will evolve once initiated. The primary objective of this research is to investigate which environmental parameters exhibit skill in discriminating between situations where thunderstorms evolve into lines and those where storms remain discrete during the few hours after initiating along boundaries such as cold fronts, dry lines and pre-frontal convergence zones.

The data included in this study are associated with severe weather events east of the rockies during the fall of 2003 through the summer of 2004. The data base includes cases where: 1) surface-based thunderstorms initiated along a synoptic boundary, 2) the storms produced one or more reports of large hail, damaging wind, and or tornadoes, and 3) the 0-6 km AGL vertical shear was 30 kt or greater. Data used to populate the database was collected from observed or RUC model proximity soundings, radar, satellite, surface observations, and objective analyses based on a combination of RUC and observed data. Preliminary results suggest that several atmospheric variables appear to show some skill in determining how storms will evolve once initiated along a synoptic boundary. The orientation of the mean wind and vertical shear in the 2-6 or 2-8 km AGL layer, with respect to the initiating boundary, appears to demonstrate the best skill. Using this information, a normalized mode parameter has been developed that incorporates the individual parameters that exhibit the most discrimination skill. Based on the preliminary results, values of the mode parameter greater than 1 suggest that storms are more likely to remain discrete within the first few hours after initiation. Values less than 1 suggest that storms are more likely to evolve into lines or mixed modes involving line and discrete cells. The relatively small size of the initial database (less than 40 cases) currently limits the statistical significance of the results, but more robust results are anticipated as additional cases are collected.