STORM TRACK: November 30, 1983 (Volume 7 Issue 1)

The May issue of Storm Data* reported on several different storms or tornadoes that I saw on the 12th, l3th and 17th in Texas, Oklahoma and Kansas. My interest in these accounts was sparked by discrepancies between what was reported and what I saw. This would have passed without further regard on my part, except for a chance dinner/conversation recently with Joe Golden about the paucity of storm reports with which watches are verified. Our discussion centered on those isolated events, away from metropolitan areas or main highways, which only diligent chasers are apt to see. This set me to thinking again about the completeness of Storm Data accounts, which is the only official, national source for this type of information. It is heavily relied on by the research community for various statistical analyses and, considering its limited resources, does an excellent job. However, because it is the only comprehensive reference of this kind, it is appropriate to indicate some of its limitations. Other than minor differences in time or location reports, the following compares significant discrepancies between my observations this past spring and Storm Data's.

Hoadley Tornado Sightings

What does this account reveal? Obviously, the ST Editor needs to do a lot better at submitting reports, especially on the smaller, localized events. That would take care of the missing reports on the 12th and 17th. However, other discrepancies remain, despite reports being made. Storm Data records show that the second tornado of the three seen by the Editor on May 13 had a path length of 17 miles (longest of any of the 30 or so this day). However, I observed and followed this cell almost continuously for the next 35 minutes and saw no further ground contact after the first 1/2 mile. Barring an occasional suction vortex beneath this cell, which may have occurred behind a tree line and out of sight, I question the 17-mile path. This may have been a radar track of a hook echo over this distance but the Storm Data account doesn't clarify how it was determined. Of even more interest to me, however, was the F-scale estimate by Storm Data for the tornado west of Estelline (and -to a lesser extent- for the one near Shrewder). At the risk of being tedious, I have again reproduced from memory (no pictures taken due to camera failure) the appearance of the Estelline vortex, which was shown in the July 31, 1983 ST -- from my location 7.2 miles west of the town.

Note the widely flaring debris/ dust cone on the ground, almost as vide as the tornado length to cloud base. It did occur over an open field, with almost no trees and fewer structures. Without verifiable damage to report, Storm Data recorded it as an F0 storm, and it will forever be classified as the weakest of all tornadoes. Is this right? If it had dropped on top of Estelline instead of Texas prairie, would it still be an F0 storm? Of the three seen this day, I considered it by far the strongest. The point here is that Storm Data, while doing a generally excellent job (certainly the only good source that we have), does sometimes miss an important event or gives misleading information, even when observed and reported.

** Damage rating of tornadoes by T. Fujita: F0 Some damage to chimneys, breaks tree limbs; F1 Peels surface off roofs, mobile homes pushed off foundation or overturned; F2 Roofs torn off frame houses, large trees snapped or uprooted; F3 Roofs and some walls torn off well constructed homes, cars lifted and thrown, most trees uprooted; F4 etc

Based on the author's experience and generalizing from just five tornado experiences over a two week period, there is probably a significantly greater number of large/powerful tornadoes than is recorded. Therefore, scientists should be cautious when drawing conclusions from national statistics on the number of F0, Fl, F2, etc. events. Since F1 or more destructive values at least reflect the presence of some damage indicators, those ranked as F0 may be much less reliable regarding actual storm severity, depending on location away from built-up or wooded areas.

Obviously, in this case, I'm weighing a judgement call (my own) against a scientifically verifiable damage report (when available). I don't expect meteorologists to start incorporating into scientific studies such guess-calls by amateurs or even by other meteorologist chasers, regardless of how many tornadoes either have seen. However, it is a problem for an accurate assessment of U.S. storms. How many real F3s (or worse) are out there, hidden among all those F0 reports, because no damage occurred? Some of John Weaver's "observational" meteorologists (mentioned earlier in this issue) know better! . . . Food for thought.