What Is “Tornadogenesis”?: A Comparison of Radar Observations, Visual Cues, and Storm Reports

The latitude/longitude point of “genesis” and start time are two vital pieces of information in the tornado report database. There are a plethora of studies that use tornadogenesis points to investigate a wide range of research topics, ranging from investigations of near-tornado surface characteristics, to case studies, to climatologies. The degree of accuracy in identifying the exact point of genesis has direct implications on tornado climatology records – especially the length and duration of the event. Warning statistics are verified using these points, with implications on the success, or failure of the National Weather Service offices issuing the warnings. But how, exactly, does one define when a tornado “becomes a tornado”? The standard practice is to acquire this information from the NCEI’s Storm Events Database, which bases tornadogenesis time upon when EF0 intensity damage can be found. However, not all tornadoes receive formal damage surveys, and even those that do sometimes end up with erroneous information for one reason or another. Periodically, WSR-88D radar observations are used to confirm the presence of debris through the observation of a tornadic debris signature, which indicates ground debris is being lofted by surface-based rotation. Mobile radar Doppler velocity observations provide high-resolution, in situ information about the strength of tornadoes as they form. But, to date, these radar observations have not been utilized in formal storm data reports owing to a wide range of limitations, and because doing so would potentially bias tornado climatology. This is the legacy upon which tornado reports have been built for the past 50 years.

One of the biggest challenges in using mobile radars to confirm tornadogenesis is that a quantitative definition of a tornado does not exist. This glaring absence is a result of the inability of most radars to fully resolve the tornado wind field because the radar beam is wider than the tornado, which causes a degradation in the radial velocity values that depends upon range to the tornado, the radar’s beamwidth, and the size of the tornado. Thus, one cannot ubiquitously apply a quantitative threshold that is accurate in all scenarios.

In this study, cases of tornadogenesis that were acquired from the Rapid scan X-band Polarimetric radar (RaXPol) are compared with visual observations (when available), and storm events tornadogenesis times to determine consistencies and inconsistencies between the various different data sources. Because a quantitative threshold is unknown, we will look at how visual observations (such as a condensation funnel reaching the ground, or swirling dust available from photographs and video of the events) match temporally with various radar-based metrics of tornadogenesis. We will include delta-V values of 30, 35, and 40 m s^-1 over 2 km, 1.5 km, and 1 km, pseudo vorticity values of 0.05, 0.1, and 0.15 s^-1, a qualitative human interpretation of the TVS, and the onset of a TDS in the lowest elevation angle data. Then, we will note the Storm Event Database time designation. Advantages and limitations of each identification technique (radar, visual, and damage) will then be discussed.