Impacts of VORTEX on operational severe storm forecasting

The Verification of the Origins of Rotation in Tornadoes Experiment (VORTEX) continues to provide meteorologists at the Storm Prediction Center (SPC) with numerous insights into the complexity of severe storm forecasting, with major impacts of VORTEX focusing on tornado forecasting. Previous work by Maddox, Hoxit, and Chappell identified an association between tornadoes and thunderstorms interacting with surface thermal boundaries. Their findings are corroborated by recent publications documenting that a majority (~70%) of tornado occurrences were in association with low-level boundaries in the southern Great Plains during the springs of 1994 and 1995. It has been speculated that the distribution of buoyancy and storm-relative helicity (SRH) are critical elements in anticipating if storms are likely to become tornadic when interacting with boundaries. However, VORTEX case studies clearly illustrate that SRH can vary by an order of magnitude on spatial and temporal scales of 50-100 km and less than 3 hours, respectively, reaffirming the previous work of Davies-Jones. Such variability brings into question the representativeness of specific observations, and underscores the difficulty of forecasting tornadic thunderstorms. Finally, several SPC forecasters produced experimental probability forecasts in conjunction with VORTEX field operations. These initial probability forecasts allowed more flexibility in conveying information than categorical forecasts, and have served as motivation for an ongoing SPC probability experiment including national forecasts of tornadoes, large hail, and damaging winds.

Additional, important findings will undoubtedly arise from future analysis of the VORTEX data sets. During the next five to ten years, it is hoped that some of the following questions can be addressed: 1) Are there multiple and/or dominant modes of tornadogenesis in specific atmospheric regimes? If so, conceptual models of critical storm structure and evolution could be adopted to specific forecast techniques. 2) Can rough parameter thresholds be established in an effort to discriminate between supercells that produce multiple or long-lived tornadoes, and supercells that are largely nontornadic? 3) What types of boundaries are most likely to be associated with tornadic supercells, and what data are necessary to identify these boundaries? 4) Can detailed conceptual models of convective initiation be developed for operational use? Current operational forecast models rely on convective parameterization schemes that have proven unreliable, and new interpretations of observed data may be necessary.