National Weather Service Announces Unfortunate Tornado Warning Experiment

[Mike Smith]

As I understand the PDD, the idea is to change the verbage so that the affected areas / people take the correct action based on the severity of the impending storm. My question for Mike is; what solutions, thoughts, theory do you have that is different than how the PDD addresses this problem? Am I being clear?

I want people to take shelter (basement, no basement = get in bathtub in middle of house, etc.) whenever there is a tornado warning. We do not need the higher categories.

I fear, once word of this gets out, that people will stop taking shelter for "ordinary" tornado warnings.

With regard to the enhanced SVR, what do you want a school principal to do when she hears, "severe thunderstorm warning, a tornado is possible"?

I fear mass confusion and mass over warning = less effective warning system and more people killed.

 

[Jake Orosi]

I want people to take shelter (basement, no basement = get in bathtub in middle of house, etc.) whenever there is a tornado warning. We do not need the higher categories.

I fear, once word of this gets out, that people will stop taking shelter for "ordinary" tornado warnings.

In a perfect world I'd see where you were coming from; but the reality is that people already ignore tornado warnings quite often. Everybody should take them seriously; but as it stands now they don't.

I foresee this new system as a method of getting those people's attention who would ordinarily not care: "This is a really, really bad tornado and everyone needs to drop everything and hide NOW. This includes YOU, doofus."

 

[rdale]

With regard to the enhanced SVR, what do you want a school principal to do when she hears, "severe thunderstorm warning, a tornado is possible"?

Pay close attention to any followup tornado warning. I don't want every school shut down in a 12 county path just because winds might gust to 65mph in an eddy.

 

[Elaine Spencer]

Although I am not a degreed meteorologist or storm chaser, just an ordinary person with an interest in weather, I would like to offer a few thoughts on the experiment in question.

My suggestion would be to have just two official tornado warning tiers – regular and PDS, which dovetails nicely with the SPC's regular and PDS tornado watches. A PDS warning would be issued whenever a damaging tornado is confirmed by multiple spotter reports and appears likely to hit a community, a large facility such as an airport or shopping mall, or a large gathering (state/county fair, outdoor concert, festival, sports event) within 20-30 minutes. That kind of situation, I believe, can accurately be described as "particularly dangerous". (There could also be PDS severe thunderstorm warnings for high-end derechoes with winds over 100 mph or extreme damaging hail. SPC does, on rare occasions, issue PDS SVR watches when conditions are right for extreme winds and hail but not tornadoes.

The Tornado Emergency message, however, should remain an informal and optional product that local NWS offices use at their discretion. The reason I prefer it not be an official product is, where do you draw the line between a PDS warning and a Tornado Emergency? Does it depend on how big the tornado is (which doesn't necessarily indicate how intense it is)? Or on how big a community is impacted, which raises the question: is a tornado headed for a community of 500 or 1,000 people any less of an emergency to the people in its path than one headed for a city of 20,000 or 200,000? Or on the presence of a debris ball – which not all damaging or killer tornadoes have

I realize these products are intended more for media people and emergency managers than for the general public, but the general public will still see or hear them via internet and NOAA Weather Radio. And even if it is a product intended for professionals only, why not keep it simple? The PDS wording would simply be an extra flag telling them to pay closer attention to the situation and to convey the message to the public in whatever way they deem appropriate.

 

[Keith Wilson]

There is a thread in here discussing the need for better advanced warning time, but that seemed to boil down to being impracticle right now due to the negative effects that earlier warning would have on the accuracy % of the larger warning area actually experiencing severe conditions.

That being said, if longer warning times aren't practical, then maybe more detailed information pertaining to the specific threat levels within a warning is the direction that needs to be explored. Given the events of last year, I would rather see them try this sooner than later and I hope that it experiences at least some measure of success to build on.
Granted, a better job could have been done getting the word out about the changes.

The wild card in the experiment is the accuracy of the information that they base the threat levels on.

 

[Keith Wilson]

Edit: Somehow managed a double post. My apologies.

 

[Mike Smith]

I have revised a blog post from two weeks ago pertaining to overwarning and it is here: http://www.mikesmithenterprises.com/2012/02/error-on-the-side-of-safety/

When I wrote Warnings I had no idea how serious the overuse of sirens was. It has gotten worse since JLN with four jurisdictions announcing the sounding of sirens for SVR (in each case they previously sounded them for TOR only). Now, with the sentence, "a tornado is possible" added to some SVR's the false-alarms will get even worse.

From a warning effectiveness standpoint we are in danger of "repealing" the progress we have made the last twenty years due to the combination of overwarning and confusion.

 

[Kelly Sugden]

I have revised a blog post from two weeks ago pertaining to overwarning and it is here: http://www.mikesmithenterprises.com/2012/02/error-on-the-side-of-safety/

When I wrote Warnings I had no idea how serious the overuse of sirens was. It has gotten worse since JLN with four jurisdictions announcing the sounding of sirens for SVR (in each case they previously sounded them for TOR only). Now, with the sentence, "a tornado is possible" added to some SVR's the false-alarms will get even worse.

From a warning effectiveness standpoint we are in danger of "repealing" the progress we have made the last twenty years due to the combination of overwarning and confusion.

Mike, did you get Jeff Hutton's message yesterday?

 

[Mike Smith]

Mike, did you get Jeff Hutton's message yesterday?

No, I did not. Email? PM on Stormtrack?

 

[Kelly Sugden]

No, I did not. Email? PM on Stormtrack?

Well to be honest, I'm not sure which form of communications, but he told me he would talk to you about this issue yesterday at the wfo. Btw, did you see my message from which central region offices are participating. DDC and GLD are not. DDC would only participate indirectly if we had to backup ICT, which is unlikely.

 

[Mike Smith]

Btw, did you see my message from which central region offices are participating. DDC and GLD are not. DDC would only participate indirectly if we had to backup ICT, which is unlikely.

Yes, and I also received messages from Chad and Chance. So, I posted a correction on my Facebook page and here: http://www.mikesmithenterprises.com...cations-included-in-the-tornado-warning-test/

Thanks for helping set the record straight!

I'm open to hearing from Jeff or anyone else.

 

[rdale]

Now, with the sentence, "a tornado is possible" added to some SVR's the false-alarms will get even worse.

Actually no, it should get better. The Best Practices document says not to sound sirens on SVR's uniless 75mph winds are observed. So the QLCS-type systems that used to get TORs and sirens, will now get SVRs and no sirens. I understand it will be discussed more at NSWW, but unfortunately I am not able to attend the forum.

http://skywatch.org/ows.pdf

 

[Mike Smith]

Some Thoughts from 5 Years Ago on This Topic

Rob, thanks for bringing the "best practices" to my attention. Well done!

While I might have set the winds up a little higher (80-85mph), a 75 mph "floor" is a step in the right direction. As you may know, I would rather see the sirens used only for tornadoes, tsunamis (as they do in Hawaii) and genuine life-threatening situations. Since I started focusing on this issue, I've learned there are jurisdictions that sound them -- in the middle of the night -- for heavy snow warnings!

I thought it might be useful to reproduce something I posted on Stormtrack four days after Greensburg (Friday, May 4, 2007) where I took the (extremely unpopular) position that tornado emergencies (TE) were not a good idea. The original post is in italics and then I'll add a couple more comments at the end:

Here 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 watch). 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

We now have the "super duper tornado warning" as of April 1, more or less as I predicted, which can be issued by the STL, KC, SGF, TOP, and ICT NWS offices. But, I ask the same question I did nearly 5 years ago: What do we want people to do differently? If the answer is "nothing" (i.e., we want them to go to the basement any tornado warning), then this is complexity without a benefit.

The "creep" I described has certainly occurred: Very few TE's since Greensburg have verified. Patrick Marsh is preparing a formal verification that I know he plans to post when it is complete.

The polygons have not worked as I hoped because EM's are largely ignoring them, www.mikesmithenterprises.com/2012/02/error-on-the-side-of-safety/ , NWR doesn't handle them, and the media is not using them as effectively as I had hoped. I still believe polygons are a great idea when they are effectively used.

I see no reason why the "creep" tendency will not infect the SVR with "a tornado is possible" and more and more "tornado warnings on steroids." By doing so, we risk confusion and complacency (people waiting to take shelter until a "super duper" tornado warning is issued).

 

[rdale]

Mike - many policies that we saw already had 70, and while 80 sounds great, if you change things too much then EMs would just throw it all away.

I certainly see your "creep" worry. If QLCS tornadoes start becoming Tier 1 TORs "just to be safe" then you'll hear an uproar. If every spotter reported tornado automatically triggers Tier 3 because it will pass near a town of 150, I'll be the first to make noise. The fact that they did this without mentioning it at any Weather Ready Nation event, with any IWT (Missouri has the first one and theirs is great) and without input from social scientists is disturbing.

But at least it's a step...

 

[Elaine Spencer]

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.

The SPC's online history indicates that the first PDS Tornado Watch was issued on April 2, 1982, for an outbreak in the Plains:

http://www.spc.noaa.gov/history/timeline.html

Wikipedia's article on PDS Tornado Watches (http://en.wikipedia.org/wiki/Particularly_dangerous_situation) notes that PDS wording can be applied to SVR watches (which is still very rare) and even to flash flood watches (at the discretion of the local NWS, not SPC).