STORM TRACK: January 31, 1984 (Volume 7 Issue 2)

Feature #2 in the last ST struck a responsive chord among several readers. You may recall that it discussed limitations and oocasional inaccuracies in Storm Data, concerning tornado reports. The following comments were received, including one that preceded the article but expresses a similar concern.

Dr. Kenneth C. Crawford, MIC - Area Manager, Oklahoma City NWS Forecast, Office: "I read with interest the comments made about Oklahoma Storm Data in your November 30 issue of Storm Track. May I say that I fully appreciate your comments regarding the discrepancies found with our entries into Storm Data. But, let me share with you what we use to make decisions about entries into this widely-used publication.

Basically, our information comes from three sources: (1) reports we seek out in real time through phone calls and amateur radio operators; (2) reports that are passed on to us in near real time from Highway Patrol units, local police, Civil Defense, the broadcast media, etc.; and (3) reports we cull from 'weather clippings' sent to us several days later by a local newspaper clipping service. The myriad of storm chasers across Oklahoma occasionally volunteer their reports to us, but that information source is somewhat 'hit or miss'.

Then, some 3-5 weeks after the event, our Preparedness Meteorologist will sit down, organize these reports, and try to put together a summary that hopefully bears some resemblance to the truth. The flaws in this system are obvious: (1) writing a summary weeks later is not good, (2) not fully utilizing chase reports needs to be overcome, and (3) relying heavily on newspaper clippings is just short of ludicrous. However, this is the only system we have. With more resources, I am confident we could do better. The chances are not good that either of us will see any early improvement in the accuracy of Storm Data.

So, where do we go from here? Obviously, we need to use the chase reports more. Perhaps you can help by asking all avid chasers to share their observations with a neighboring WSFO (Weather Service Forecast Office), rather than wait and see if we 'discover' each event. We also need to work more closely with the Civil Defense (a strong point in Oklahoma) to stress the importance of such reports. The same goes with law enforcement offices. I am not sure that we can make much early progress with the press, but we will try."

Ken's comments go right to the core of the problem and one of the best sources of information: chasers. In fact, the following letter from Alan Moller anticipates a standardized reporting form, which the Editor was going to include in this issue.

Alan Moller, NWS Meteorologist, Fort Worth: "Dave, your Feature #2 "Storm Data, Some Limitations" aroused me! You were right on with your comments about its limitations! I'm a strong believer in Storm Data, but not all NWS offices, which are responsible for providing data to it, can do an exhaustive job. In many offices, this is due to lack of manpower and time. It is possible only to survey the very major (or very local) events; otherwise, we must depend on newspaper clipping services (usually mediocre service, at best) and storm inquiries that are mailed to local officials by NWS personnel.

More emphasis (we don't know yet how much more) is being placed on Storm Data, because we are verifying our warnings with it. Thus, offices doing a poor job stand to lose on their verification scores.

And, storm chasers can help. By logging their reports, times and locations, estimates of the tornado FPP scale will also help. Several years ago, many ... in the NWS would not have accepted this info, but things are changing. [The observational skills and experience of] storm chasers generally are being more readily accepted today (as will be their storm reports).

I have enclosed a Storm Inquiry sheet for your use. My question to you and other chasers is: Would it be feasible to fill out these sheets and send them to me in Fort Worth or to appropriate WSFO's (If sent to me, I will take care of WSFO distribution). I am willing to do this to ensure that all known reports are considered for inclusion in Storm Data."

A copy of one of the "Storm Inquiry" sheets from Al Moller is attached. Also attached is a revised copy from the Editor with some minor additions and an address section at the top, showing the respective WSFO's in Oklahoma and Texas (include or substitute those from other states, if you wish). You can reproduce this form with either address blocked out, or leave both on. When each report is completed, just fold it twice with the address side out (cross out the inapplicable address, if both appear), tape shut top & bottom, stamp and mail. Simple, quick -- efficient! Make copies for a chase friend. Let's make this a habit. We have a responsibility to improve the quality of data used by research scientists. Let's start paying back some of the services we receive and from which we benefit.

It might also be desirable, if not already done, for NWS to develop a standard reporting form and make it available in all NWS and FAA field offices. If so, it should also appear in Storm Data, to help standardize reports and make readers more aware of their responsibilities.

Al continued with an additional comment "that we should attempt a subjective classification, other than an arbitrary 'F0' when photographs or tornadic observations are available from veteran storm chasers." He goes on to offer a subjective reference scale to do this, which I wish to add to in a future ST. Do you have any suggestions on how to estimate the F-scale capability of a tornado that leaves no damage record, based on visual references such as size, shape, etc? In a separate letter below, Tim Marshall takes exception on this point, albeit unaware of other correspondence shown here. Write to ST and share your experiences or opinions.

Tim Marshall: "The number of unreported tornadoes is a function of many variables. Popu- lation density, visibility, land use type, Ham radio density, Storm Chaser density, etc. -- are all factors in this equation. Initially, even Pampa went into the records as three F0 tornadoes, before I informed the NWS of eight; two of which were F3's, and one which was a mile wide!

As far as F-scale is concerned, you're talking about another problem. Remember, the F-scale classification for tornadoes was designed to determine intensity of damage to structures, especially for residences. There is no F-scale rating system for a mesquite tree! Also, there is no way to rate a tornado strictly by its appearance. So, we have a dilemma. We can't accurately place an F-scale on a tornado, unless it hits a structure. ... Basically, the F-scale problem is that wind speeds have NEVER been verified. My thesis approaches the problem through engineering calculations of wind speeds on damaged structures and then comparing the results to the F-scale wind speed, The result: Confidence in the F-scale rating is within 1 F-scale for most residential structures, because of their high variability in material strength, construction practices, variation of wind pressures due to structural orientation, etc. The perspective of the tornado from an engineering point of view is available in NSSL-82 ... I recommend that every chaser read it. It's not mathematical and is easy to read. Its available from NTIS."

Mike Watts; "I noticed that the entry for the Medicine Lodge storm of May 17 in Storm Data was off, also. The first tornado was not included, and the time given for the tornado 10 miles southwest of Medicine Lodge was wrong. Jeff Lazalier and I phoned our observations to one of the television stations in Wichita, which read them on the air. I though that this would be sufficient notice to get the info to NWS, but it was not. The weather service may not like that channel.

The same problem occurred earlier that month. I photographed a tornado north of Odessa, Missouri (about 30 miles west of Kansas City) on May 1. It was in full view of a Highway Patrol unit and Interstate 70 (which happened to be about the only road above the water table that day). This storm is not listed, either.

I resolve to be more conscientious in notifying the weather service of these sightings, although this may not help. Not all of these offices know what a reliable observer I am." (Good point -and an endemic problem among other chasers. -Editor) "However, this is understandable, since I don't believe a lot of what I hear over the phone, either. Whenever I have used Storm Data, I've always hoped that, in terms of numbers, the missing reports were balanced by the bogus obs."

Steve Levine (November 5 letter): "I have just spent this entire week ... compiling l983 storm data for ... publication by the Blue Hills Weather Observatory and Club. The frustrating thing is that the local NWS office has [apparently overlooked several] storms with suspiciously tornadic damage (e.g. twisted off tree tops, trees twisted out of the ground, and eyewitness accounts of rotating phenomena). It is possible, and even likely, that up to 6 locations in Massachusetts received F0 to F1 tornadoes in July alone! However, some of the strongest candidates (based on rotating damage reports, eyewitness accounts and even a hook echo on radar) were left uninvestigated and thus not officially confirmed. In another area, where relatively minor damage was done ... investigation was made and a tornado confirmed! In general, tornadoes here in New England are very brief and often embedded, hidden in the general thunderstorm rainshield. This is what makes eyewitness accounts and reports of any twisting or rotating damage so important to officially investigate and confirm."

Fred Varian: The 2nd Annual STORM-COM begins February 18 at Mountain View College, Dallas to assist and instruct local communities and military installations in establishing, or improving, Skywarn Spotter programs TO SAVE LIVES! This unique conference will convene on Saturday, from 8-5, in a general seminar type meeting. It will be presented by experienced Skywarn veterans, many of whom pioneered the earliest such programs in the nation. Discussion topics will include spotting techniques and the uses of Doppler radar. Among those invited are local emergency operations officials, community spotters, storm chasers, Hams and any interested parties. Last year's conference drew 200 from three states. This is the only conference of its kind known to STORM-COM Organizers. The conference is free and is sponsored by the Fort Worth National Weather Servics Office, the Dallas Office of Emergency Preparadness, and the "ARES" and "RACES" amateur radio associations. Those interested in attending should contact Dave Martin or Alan Moller at the Fort Worth NWSFO (Phone: 817-334-3401) or Fred Varian (Phone: 214-644-5336).

Roger Jensen; one of the most experienced chasers (who was on the road before many of you were born!) would like to acquire some good prints of: (1) multiple exposure and sheet lightning (filling most of picture, without power-lines, poles, etc. obstructing view); (2) large hail stones, especially with scale references in picture, such as a ruler or baseball; (3) large wall clouds with maximum vertical development; (4) large, cone shaped tornadoes; and (5) pictures of the June 3, 1980 Grand Island tornadoes. Roger is seeking 7" x 10" or 8' x 10" color prints and is willing to pay to have them done or will do them himself, if you'll send him the negatives/slides. If you'd rather swap pictures than take cash, he has some excellent prints from his many years of chasing. In fact, he is probably the leading storm photographer of northern plains storms (look for the characteristio climatic differences in his pics, e.g. high bases, crisp contrasts and well defined storm structures). However, before investing in any large prints, he would like to preview smaller, sample prints or slides if you'll send them to him. Oh yes, he will "absolutely guarantee ... that all such slides/negatives will be returned ...in the... same condition" as sent to him.