National Weather Service Announces Unfortunate Tornado Warning Experiment

[Greg Higgins]

"beaudodson - Where do you see (in the PDF posted) that the NWS won't be issuing tornado warnings for squall line events (QLCS events)? I am not seeing that - at least not clearly."

I think part of the problem is from the comment : "rdale - a QLCS event with little eddies on the leading edge are now SVR's." There is nothing in the PDD that address this specific issue other than what someone might want to infer. rdale also introduced the "Best Practices" document regarding siren activations in this thread that has some verbiage that may cause confusion about the PDD.

 

[rdale]

Where do you see (in the PDF posted) that the NWS won't be issuing tornado warnings for squall line events (QLCS events)? I am not seeing that - at least not clearly.

My answer hasn't changed since the first time To issue a tornado warning, "credible evidence" of a tornado must be present. A one-scan shear marker is not credible evidence. That is why they now can issue a SVR with "tornadoes possible" for those circumstances.

Obviously if offices in the experiment continue to issue 9-county 80 minute tornado warnings and verifying them with a shed that lost some shingles, all of the work going into the tiers will be wasted and we're back to square one.

 

[beaudodson]

My answer hasn't changed since the first time To issue a tornado warning, "credible evidence" of a tornado must be present. A one-scan shear marker is not credible evidence. That is why they now can issue a SVR with "tornadoes possible" for those circumstances.

Obviously if offices in the experiment continue to issue 9-county 80 minute tornado warnings and verifying them with a shed that lost some shingles, all of the work going into the tiers will be wasted and we're back to square one.

I think your interpretation is liberal. Not saying you are wrong - but I am saying it is opinion.

I don't believe for one second that these offices are going to stop issuing tornado warnings for QLCS event - squall line event.

"Credible evidence" is in the eye of the beholder. What is the definition of "is" comes to mind. What is the definition of "credible evidence" going to be? Does anyone know that? Maybe not one scan but two scans?

The majority of all tornadoes (or close to it) in this region are QLCS tornadoes. Certainly a sufficient number of them that I don't see them stopping the practice.

I have posed this question to a couple of NWS meteorologists - I will see what they say. I read exactly what you read - that is why I asked the question again. I don't see in this PDF document where it says what your are interpreting it to say.

That being said - we will see what the offices do. I don't think they know, as of yet, how they are going to handle this either.

Also that NWS has always had the option to issue a severe thunderstorm warning and use the sentence "severe thunderstorms can and occasionally do produce tornadoes" - or - "this storm has shown some signs of rotation and could produce a tornado with little or no warning" - at least in our region they can say that and have.

 

[rdale]

The majority of all tornadoes (or close to it) in this region are QLCS tornadoes.

Again - the wavelengths we are talking on have been twisted What I am talking about is the 9 county blanket TOR because of one or two brief GrG shear markers on the leading edge of a squall line. Not something related to the large-scale rotation, not embedded storms/supercells in a line.

http://ams.confex.com/ams/25SLS/techprogram/paper_176213.htm

 

[beaudodson]

Again - the wavelengths we are talking on have been twisted What I am talking about is the 9 county blanket TOR because of one or two brief GrG shear markers on the leading edge of a squall line. Not something related to the large-scale rotation, not embedded storms/supercells in a line.

http://ams.confex.com/ams/25SLS/techprogram/paper_176213.htm

Yes, I have read all of that information.

OK - on the blanket tornado warnings. Perhaps they will issue 3 tornado warnings instead of one large big warning. Perhaps making the polygons even smaller. I would rather see severe thunderstorm warnings used for some of these squall line events - include the tag about severe thunderstorms can and occasionally do produce tornadoes. Hopefully they do that. Sure would be nice to cut down a lot of the warnings.

It would also help if they raised the wind speed criteria for severe thunderstorms. Perhaps to 70 mph - don't know. We have 55-60 mph winds here on a regular basis with little or no real damage. Sure you may occasionally get a tree branch down on the ground - nothing significant.

I don't know - I just don't believe they are going to stop the practice of issuing tornado warnings for the QLCS events - squall line spin ups. That is all I was saying.

I hope you are right and they stop with the six county blanket warnings - I have seen them for more than that in our region. Very large blanket tornado warnings - same with severe thunderstorm warnings (for that matter).

I hope someone keeps track of each tornado warning that these offices issue this spring - let's us know how each one turned out. How many times did the tier level in the warning change - how did the NWS handle the ramp up and ramp down of the tiers. Also was there damage - how often did the tornado emergency or catastrophic damage verify.

Scoring these tornado warnings may be a bit easier than in the past. However, that will not tell us how people responded to them. Not sure how we find that out.

I would love to know how many television stations have been contacted about the subject of tiered warnings. Apparently not all of them.

Again, to be clear, I am 100% in support of some changes being made. I have always thought the NWS should put as much information into the tornado warning as possible. I like the TAG idea - tornado confirmed or not confirmed - so on. There are some nice ideas being floated.

How on-air mets handle this - I have no idea. How the public will respond to this - I have no idea.

 

[Mike Johnston]

Well, if they're going to experiment, I do hope there is a comprehensive follow-up, with stringent and detailed verification statistics for each warning level. Actual warning forecaster performance needs to be carefully scrutinized in each and every case. One thing that's a little concerning is the small number of offices involved. Sure, they're all located in relatively tornado-prone areas, but each season is different and you can't count on a decent sample size from 5 or 6 offices. (Remember the first year of Vortex 2?)

I remember, I think it was in 2005, going to a spotter training class in Kansas City and the presenting NWS met from Pleasant Hill announced two important changes - one was the polygon warning areas and the other was raising the SVR hail criteria to 1 inch. As I recall, one of these changes was experimental and limited to a small number of offices very similar (if not the exact same) as the office locations in the instant experiment discussed here. The other change may have been universally adopted that year. Does anyone recall when and where either of these changes were actually made?

 

[GPhillips]

Does anyone recall when and where either of these changes were actually made?

The 1" hail experiment started in KS in 2005 with offices who had KS counties in their forecast area. The other thing that started in 2005 was the VTEC coding on watches/warnings/advisories and followup statements. That was nationwide. Storm-based warnings started October 1st 2007.

 

[J Tyler]

Excellent, respectful discussion. Great points on both sides. I am one who wants a different warning product for Greensburg than a standard tornado warning. When you live in these areas, you hear a LOT of tornado warnings in your life, and your house is still there. People in Greensburg will tell you, they KNEW this was different, they KNEW this was life threatening. I spent a day hanging around The Greenbean (I think thats the name of the little restaraunt there downtown), just chatting waiting for a late storm system. The regular topic was the very different wording out of the Dodge City office and the Wichita tv met who was "really different" that night. People who rarely went to shelter did so that night, because of what they were told.

I *DO* like the concerns that are being brought up in this discussion. There are some very good points that need to be considered as this process moves forward.

Finally, I really like this at the end of the product:
TORNADO...OBSERVED
TORNADO DAMAGE THREAT...CATASTROPHIC
HAIL...2.75IN

 

[beaudodson]

I did get a reply back from someone at the NWS - they said there will be no restrictions placed on issuing tornado warnings. Tornado warnings will still be issued for QLCS type events - as they see fit.

If this isn't the case and another NWS meteorologist comes in here and says something different - then there is clearly a miscommunication among NWS meteorolgists as to how this experiment will be conducted.

 

[rdale]

It's not a restriction, I think you're just confused on what they're doing and the terminology. We'll just wait and see

 

[beaudodson]

It's not a restriction, I think you're just confused on what they're doing and the terminology. We'll just wait and see

Hmm - I am just going off your comment on the previous page. According to some in the NWS that is not the case.

On the previous page I said this
"I am not sure (Rdale) that this is going to stop tornado warnings for squall lines - QLCS events. Why do you think that?"

Then you said this
"Because that is what the directive says

For squall lines with embedded brief circulations, "Finally, the option to include a tag line which explicitly states "TORNADO POSSIBLE" will be available."

You can't do that for a tornado warning, since the low end states "This warning type will be selected for cases in which there is credible evidence of a tornado."

Nobody would claim that a one-scan 60kt GtG shear which disappears in 4 minutes is "credible evidence" - which is why QLCS eddies go into the SVR category now (like they were for 50 years prior to hi-res 88D.)"

End comments/quotes

Not sure where I would be confused on that - and if I am confused on the subject then apparently a lot of other people are, as well.

Hopefully you are right about the blanket warnings - hopefully those are reduced. As you said - we will see how it goes.

Not saying those offices experimenting with these new guidelines won't come up with some new formula for issuing a tornado warning vs a severe thunderstorm warning - I am saying that the PDF does not say that. According to some within the NWS - that is not the case.

I am sure there are meteorologists that work for the NWS reading this thread. Perhaps they could shed more light on the topic for their partners. This seems to be a point of concern with many - the lack of communication on this subject with partners. I would love to hear clarification from someone within those offices participating.

I would agree that the goal of this experiment should be to reduce the number of tornado warnings (especially for some of these marginal squall line events that produce brief tornadoes).

I wish the NWS would stop issuing so many tornado warnings for squall line events and go with severe thunderstorm warnings that include the "severe thunderstorms can and occasionally do produce tornadoes" - you and I agree on that topic. I also wish they would increase the wind speed criteria for severe thunderstorm warnings - cut down on the number of severe thunderstorm warnings. I don't know exactly how many severe thunderstorm warnings produce little or no damage - but I know it is MANY.

The disagreement on this subject is what the initiative actually states vs what will happen. I hope you are right. It concerns me that NWS meteorologists don't seem to know exactly what it means. I am sure that concerns you, as well.

I agree that the interpretation of this initiative may be in the eye of the beholder. Which - is a concern, as well. There needs to be as much uniformity as possible - from office to office (also within the media). I am sure we agree on that subject. Realistically I know that is impossible - at least complete uniformity. This is because (obviously) each warning coordinator makes their own decision. Many of us have watched tornado warnings exit one NWS office (WFO) only to become severe thunderstorm warnings in the next forecast area (WFO). When this happens repeatedly and a pattern is establish - well we know that one office is more liberal with their tornado warnings vs another office.

Not arguing with you - I am just pointing out there seems to be a conflict of how meteorologists are interpreting all of this (even within the NWS). Hopefully all this will be ironed out soon.

 

[Elaine Spencer]

This story from STL gives a different description of the new warning system than what I've heard so far:

http://www.stltoday.com/news/local/...cle_f2696646-42d0-5275-ba96-65832c936220.html

According to this story, there will be four tiers of warning: radar indicated tornado (rotation), "observed" tornado (by spotters and/or debris detected by new dual-pol radar), "significant" damage potential, and "catastrophic" damage potential. No mention of PDS or "tornado emergency" wording.

 

[Mike Smith]

Elaine, please read the PDD which I link to in posting #1 in this thread. You will find the "tornado emergency" language.

 

[CGardner]

After reviewing what has been proposed by the NWS and reading the comments, I agree with changing the issuance. If this dual-pol radar can discern a strong rotating meso from a major tornado causing a large debris ball. I am all for this, I just see a few issues with this that will not change the outcomes of the past.

1. A more clarified, accurate, and advanced radar will probably take more time to update from scanning. The average NEXRAD update is usually 4-6 minutes, so even then we could easily say that the new "dual-pol" radar should update in 6-7 minutes, worst 10. Since NWS is wanting to run this, and the budget is being reduced for IT, then guess what people. IT won't be able to put in or afford the required upgrades in processors to analyze the data faster.

Historical problem: The Joplin tornado went from being non-existent on radar to the next update showing the mother of all hook echos and debris balls in 4 minutes. How can we expect dual-polarization to warn us in time? "Oh, a very severe thunderstorm is approaching with an increasing updraft." Six minutes later, "Well, the radar shows a clearly defined massive tornado right overhead doing dam..." There have been many other pop-up supercells that just appear out of nowhere and mature within 15 minutes. We need faster processors for the data coming in, yes accuracy helps, but in these situations. TIME IS THE CRITICAL FACTOR!

2. Common folk, ones that are not familiar with severe forecasting or terminology in general will still do what they have always done. A social habit for Midwesterners is to go down in the basement or the interior hallway and plop the mattress over themselves. Yes, for maybe an EF-0 to EF-1 this will fly. However, last year should of been a wake-up call for most everyone in Tornado & Dixie Alley that things need to change. We have to go back and look at how our grandparent's houses were built, with a basement or cellar. Now, I'm not saying every house should have a basement, but if it is in a tornado prone area, like eastern Tornado Alley, Dixie Alley, Hoosier Alley, Corn Alley, and the never-mentioned Missouri interchange. Yes, maybe insurance underwriters and homebuyer's should look at mandating storm shelters or re-inforced basements. People will continue to seek shelter in their new, suburban, homes with crappy, cheap construction and end up with worse injuries than the 78 year old man in a 1920's farmhouse reinforced with stone.

There needs to be a change to how the protection issues will be addressed for very apparent storms that are producing severe or catastrophic damage.

(Missouri interchange)- The interchange for Tornado, Hoosier, Corn (MN, IA, WI, and Northern IL), and Dixie alleys. Basically, for multi-day events, Missouri either doesn't get it or gets what is coming.

 

[Elaine Spencer]

Elaine, please read the PDD which I link to in posting #1 in this thread. You will find the "tornado emergency" language.

I know that the "tornado emergency" language was in the PDD -- I was pointing out that it isn't mentioned in this particular news story, however. Whether that is due to oversight/misinterpretation by the reporter, or because someone at NWS decided to modify the language a bit more, I don't know.