Tracking Down Tornado Emergency Products

[Patrick Marsh]

I'm trying to track down all tornado emergencies issued between 1999 and 2005. Ideally I'm trying to get a copy of the actual product (complete with headers). If you think you might be able to help me, please read:

http://www.patricktmarsh.com/2011/11/the-elusive-tornado-emergency-product/

I will be doing some verification on tornado emergencies and need the data prior to 2005 to complete my archive. (I am confident I have all since 2005.)

You can see which tornado emergencies I believe to exist, but don't have copies for, here:

http://www.patricktmarsh.com/research/tornado-emergency/


Thanks!

 

[Patrick Marsh]

Thanks to the power (and help!) of the Internet, I believe I've tracked down all the tornado emergencies. There appears to have been 143 TEs issued since 1999 with 60 of them coming this year (2011).

Top top three offices are:

1. BMX (35; Birmingham, AL)
2. HUN (22; Huntsville, AL)
3. JAN (18; Jackson, MS)

I'll be doing more data mining with the dataset over the course of the next few weeks. You'll be able to keep up with the progress on my website (http://www.patricktmarsh.com) since I don't want to bombard ST with links to my website and the figures will be too large to post here. If you have something you'd like to see from the dataset, don't hesitate to let me know. Eventually I'll find a way to post the dataset online for all to enjoy.

 

[Mike Johnston]

Patrick,

Not trying to bombard you with questions, either. But, since you were looking for headers, I must ask: were there any instances where "tornado emergency" was actually in the header of the product? I was under the impression that "tornado emergency" is not an official product unto itself, but rather enhanced wording which is sometimes included in the text of a tornado warning.

Also, if you don't mind me asking, what is the objective of your research?

 

[Patrick Marsh]

Hi, Mike,

There was never an instance where the Tornado Emergency was declared in the true header (such as the header below from a recent tornado warning).

515
WFUS51 KILN 150207
TORILN
OHC045-089-150230-
/O.NEW.KILN.TO.W.0062.111115T0207Z-111115T0230Z/

BULLETIN - EAS ACTIVATION REQUESTED
TORNADO WARNING
NATIONAL WEATHER SERVICE WILMINGTON OH
907 PM EST MON NOV 14 2011

However, there are instances where the "* TORNADO WARNING" line was replaced with "* TORNADO EMERGENCY". This is actually in violation of NWS directives and resulted in many dissemination partners incorrectly processing these warnings.

THE NATIONAL WEATHER SERVICE IN WILMINGTON HAS ISSUED A

* TORNADO WARNING FOR...
NORTHEASTERN FAIRFIELD COUNTY IN CENTRAL OHIO...
SOUTHEASTERN LICKING COUNTY IN CENTRAL OHIO...

* UNTIL 930 PM EST.

As for why I am doing this research? Simple answer is "curiosity". Long answer is that I'm completely against the notion of tornado emergencies -- a position that I've not shied away from sharing on StormTrack many, many times. Tornado emergencies are often touted as a very dangerous situation but I'm aware of tornado emergencies that have been issued for landspouts, weak tornadoes, and even for tornadoes that never occurred. Ultimately I feel that tornado emergencies weaken the tornado warning, but I can't prove that with this dataset. However, I do hope to show that the verification statistics aren't what they should be for such a high-end product.

 

[Jeff Turner]

"Long answer is that I'm completely against the notion of tornado emergencies". Being a layman I must say some types of NWS products seem to overlap. Warnings versus special weather statements are one example. Such fine lines in criteria for product issuance can confuse the masses who are not interested in weather like the folks here are. I will say that a tornado emergency might be issued when there is 1) a massive and strong twister -- EF4/EF5 and 2) it will hit a town or city. That is my simplistic rule for one but a strong twister out in a sparsely populated area while dangerous and deadly is not an "emergnecy" when compared to other scenarios.

 

[William Monfredo]

Just to throw my two-cents worth in here: It's always just been assumed that the enhanced wording of tornado emergency has added to the lives saved. This has been stated at weather conferences. And this may be true, or not. To date, I have seen no studies bolstering the notion that it has been responsible for say, an extra, 30 percent of people taking good shelter and therefore, saving an extra 15 % or so.

 

[rdale]

It's always just been assumed that the enhanced wording of tornado emergency has added to the lives saved.

I think that's a horrible assumption...

 

[Shane Adams]

The Tornado Emergency is simply a meteorological way to express not only (1) dire straits but (2) exceptionally high certainty of the event actually affecting the warned area (no RIT or other false alarm scenario). It's a brilliant tool. But you can only lead a horse to water. What the public does with information produced by the NWS is not in the control of the NWS. This doesn't automatically qualify the TE as a "failure" or overkill or even unnecessary. People who didn't pay attention in 1990 won't pay attention in 2012, and would've/will die the same way if they'd been/are struck. Does this fact wipe out 22 years' of technical advancements in warning systems?

 

[rdale]

The Tornado Emergency is simply a meteorological way to express not only (1) dire straits but (2) exceptionally high certainty of the event actually affecting the warned area (no RIT or other false alarm scenario).

In theory - yes. But as Patrick noted - that's not the case. FAR is still up there for TEs. If they are going to do anything (big if) they need to have FAR of 0.

 

[Patrick Marsh]

Shane, what you speak of is not how the Tornado Emergency has actually evolved. I can give several examples of Tornado Emergencies that did not have tornadoes associated with them. In fact the second ever Tornado Emergency was issued for a landspout.

By definition the Tornado Emergency should have a FAR of 0. Anything greater than that suggests the Tornado Emergency language is overused and leads us back to the FAR issue being discussed in another thread.

 

[Shane Adams]

Shane, what you speak of is not how the Tornado Emergency has actually evolved. I can give several examples of Tornado Emergencies that did not have tornadoes associated with them. In fact the second ever Tornado Emergency was issued for a landspout.

By definition the Tornado Emergency should have a FAR of 0. Anything greater than that suggests the Tornado Emergency language is overused and leads us back to the FAR issue being discussed in another thread.

I agree. I should've included "when used properly" in my comments. The TE as it was created and used (May 3, 1999) is an exemplary warning tool. However, there are more than a few instances (obviously) of improper use of the Tornado Emergency...and this abuse of course causes it to be lumped in with all other imperfect warning processes. A case of perhaps the tool itself being superior to the operator in certain situations?

 

[Mike Smith]

One of the longest and most vigorously debated threads in the history of Stormtrack is here: www.stormtrack.org/forum/archive/index.php/t-12361.html

It is on the topic of tornado emergencies in the immediate wake of the Greensburg tornado. I was one of a small minority arguing against the use of the Tornado Emergency. This is what I wrote four days after Greensburg:

[This] is the crux of the matter as far as I am concerned: We all agree that Friday's TE was fine. It was issued on a classic hook with gate-to-gate shear off the chart. DDC got praise for issuing it.

The next evening a far weaker signature approached Great Bend. ICT NWS (for which I have great respect) appeared to feel compelled to issue a "tornado emergency." It "busted."

The first ever PDS tornado watch of which I am aware was April 26, 1991, which produced Andover, Red Rock and Cowley Co., all of which were F4 or F5. At first, PDS's were rare.

Now, PDS tornado watches are issued much more frequently than they were at first. On Saturday, SPC issued five (more than used to be issued in an entire year), none of which verified from the point of view of long-track F4, F5's (which was the original intent of the PDS). The tornado watch for Greensburg Friday was an "ordinary" tornado watch -- but an extraordinary tornado occurred. Because it was an "ordinary" tornado watch did we want the public to be less aware? Do we really have that much meteorological reliability (which I define as consistent skill)?

Melbourne NWS in August, 2005, received praise for issuing a tornado warning for the 100 mph winds associated with the decaying eye of Hurricane Charley. It spread across the NWS and morphed into something unfortunate: Telling people in the path of Katrina to go to the lowest floor as a 30 ft. storm surge came in.

These things seem to have a "creep" to them. The first few are great. Then, they start being used more and more often until they become less meaningful. Then, they can continue to morph into something undesirable if a great deal of thought is not given to whether it is a good idea in the first place and, if so, what are the circumstances under which it is appropriate use the new special product. Otherwise, in a few years, TE's might become routine until some NWS office issues a Super Duper Tornado Emergency message.

When you combine the TE concerns above with the additional complexity (are people going to hear about these new products and reprogram their WR-SAME, NWWR heading decoders, etc., in time for a future rare event?) especially in areas where tornadoes are infrequent, to catch the "tornado emergency message"? If they do, will they get disenchanted when their NWR's are waking them up for Statements?

If you restrict TE's to dense population areas, are we saying that a life in a big city is worth more than in a small town?

That is why I believe the polygon tornado warnings, which become official October 1, should be given a chance to work before we make another major change to the tornado warning system.

I do believe many influential and smart people read this board which is why I have posted my comments and spent so much time on this.

Thanks for reading, everyone,

Mike

I would change little today. As I feared, based on Patrick's research, far too many TE's have been busts. We do not have the skill to predict short-term tornado behavior (will it lift?) or direction (the Joplin tornado changed directions from ENE to ESE on the east side of JLN). Thus, tornado emergencies are a bad idea until we have the consistent skill to make them useful.

Mike

 

[Greg Stumpf]

By definition the Tornado Emergency should have a FAR of 0. Anything greater than that suggests the Tornado Emergency language is overused and leads us back to the FAR issue being discussed in another thread.

And the FAR for a TE should be scored differently. In order to verify a TE, it shouldn't require any old tornado any where, but a large and particularly damaging tornado needs to hit a dense population area (city or town). And the warning polygon better not be drawn huge either!

 

[Shane Adams]

The very nature of the successful Tornado Emergency almost dictates that some "collateral damage" is a distinct possibility when properly used. A sustained, violent tornado in a heavily-populated area is the only criteria, therefore, there is the possibility that in any TE situation in which it's used properly, serious injuries or deaths might have already occurred. Joplin this year is a great example of when the TE would be needed. IMO, looking at radar isn't even a pre-requisite; successful TEs are only issued based on real world reports of an ongoing, violent tornado in heavily-populated areas. "Off the charts" velocity couplets over a city without real world reports of an actual tornado is not cause for a TE.

Again. One need only look at the people who (on the spot) created and first-used the TE warning message on May 3, 1999. If it's not that type of situation, don't use it. Even if the tornado dissipates (they don't 'lift') shortly after issuance, at least the fact a tornado WAS on the ground in a densely-populated area is in the text (folks just east of Del City, OK on 5-3-99 or on the east side of Joplin, MO on 5-22-11 would've appreciated the effort). Not pulling the TE trigger for fear the tornado may suddenly go away is a piss-poor excuse IMO for not throwing out the intensely-worded, enhanced Tornado Emergency type warning.

 

[rdale]

A sustained, violent tornado in a heavily-populated area is the only criteria

Just to be clear - that's your suggest criteria. Here is the criteria:

"In exceedingly rare situations, when a severe threat to human life and catastrophic damage from a tornado is imminent or ongoing, the forecaster may use the terminology "TORNADO EMERGENCY FOR [GEOGRAPHIC AREA]" in the third bullet of the warning. Additionally, in such a situation, this terminology should only be used when reliable sources confirm a tornado, or there is clear radar evidence of the existence of a damaging tornado such as the observation of debris."

Nothing quantitative - nothing about population. Just "severe threat" (how many times are tornadoes a non-severe threat?)