Synoptic‐scale and mesoscale controls for tornadogenesis on cold fronts: A generalised measure of tornado risk and identification of synoptic types

Abstract

AbstractEnvironments of tornadic and non‐tornadic narrow cold‐frontal rain bands (NCFRs) are investigated using ERA‐Interim reanalyses for a sample of 114 events over the United Kingdom and Ireland (44 tornadic). The results offer a practical tool for prediction of the likelihood of tornadoes in these potentially high‐impact events. Of 22 analysed parameters, a bulk measure of shear vorticity, and the front‐normal wind component on the cold side of the front, yield the best discrimination between event classes, showing significantly larger values in tornadic events. A generalised measure of tornado probability, p[TN], is obtained using the distribution of points within the two‐dimensional parameter space defined by these parameters. Synoptic situations commonly associated with tornadic NCFRs are identified and conceptual models describing the large‐scale evolution are developed. Most events are associated with developing secondary cyclones (i.e., frontal waves) along trailing cold fronts (≥54.5%), generally within west to southwesterly large‐scale flow. Another significant class of event corresponds to situations where a strong mid‐ to upper‐level jet streak cuts across the front within an amplifying large‐scale flow pattern (upstream ridge building and downstream trough extension), generally within northwesterly flow (27.3%). In frontal waves, tornadoes occurred relatively early in the wave's development and just down‐front of the wave centre, where rapid increases in p[TN] occurred as the wave amplified. In northwesterly flow cases, tornadoes occurred along a well‐defined NCFR bulge close to where the mid‐ to upper‐level jet streak and an associated positive potential‐vorticity anomaly intersected the front. Analysis of a high‐tornadic subset of tornadic events (NCFRs producing ≥7 tornadoes) revealed an even stronger association with frontal waves (72.2% of cases), suggesting that the highest‐impact events are usually associated with secondary cyclogenesis. The possible relevance of identified environmental parameters to candidate vortex‐genesis and tornadogenesis mechanisms within NCFRs and quasi‐linear convective systems is discussed.