James Spann speaks to the media and the NWS concerning

[The Dude]

That's why we say in the TOR text "...NATIONAL WEATHER SERVICE DOPPLER RADAR INDICATED A SEVERE THUNDERSTORM CAPABLE OF PRODUCING A TORNADO..." or "...TORNADO EMERGENCY..." for highly populated areas. Other times we'll say "...LAW ENFORCEMENT OFFICIALS REPORTED A TORNADO ON THE GROUND...", etc. in the TOR or SVS.

It all depends on how much time you have and additional information you have as to how much you are able to change the default warnng text.

I think the problem with that approach is that the majority of the public likely don't see the text.

[isohume]

Well, there's a bit of a semantics issue there. I've heard both types referred to as QLCS tornadoes -- the weak spin-ups, and the legitimate circulations. It sounded to me like Spann was referring to the weaker variety, but I wish the nomenclature was a little bit clearer

(In truth, of course, it's more of a spectrum than a hard line between the two types.)

That's true. There is no hard line. I guess I was referring to your "this is something different entirely" statement. No wonder I couldn't figure it out. You are referring to bookend vortices and such associated with more prominent QLCS type systems. I get it now. Yes, those type of spin-ups are very hard if not impossible to warn for. In fact, they are usually missed. I don't see a large FAR with those types of tors.

[isohume]

I think the problem with that approach is that the majority of the public likely don't see the text.

Since the the majority of the public gets their weather/warning information from the media...the media needs to help make sure those messages are conveyed during live events. Not always plausible...but the information is there for the taking. It is always available on our warning products through the Internet and NWR.

[tornadotony]

I'm assuming that "20 tornadoes" was an exaggeration, but I have yet to see a situation in which a "shear axis" tornado warning has produced more than a few confirmed tornadoes. Of course, confirming tornadoes in such an environment is dubious -- in a fast-moving line, these weak circulations are entirely outbound-dominant, and these types of tornadoes are extremely small and brief. Sometimes, if you are fortunate to be within a county or two of a TDWR, a circulation might show up for one or two scans (1-2 minutes of real time). Usually, there's no evidence at all. I think that's the key -- is it a good practice to issue tornado warnings for large geographic areas when there is no actual direct evidence in support of them?

http://www.crh.noaa.gov/ilx/?n=02apr2006

http://www.crh.noaa.gov/news/display_cmsstory.php?wfo=lmk&storyid=67056&source=2

http://www.crh.noaa.gov/news/display_cmsstory.php?wfo=lmk&storyid=67020&source=2

April 19th of course was the "largest" tornado outbreak recorded in the state of Indiana, with the majority of the tornadoes, well, from shear zones. A few were from more solid mesovorticies. Similarly, April 2nd 2006 was one of Illinois's "largest" outbreaks, and the vast majority of tornadoes that day were from "shear axis" tornadoes. I just surveyed an outbreak of tornadoes from a "shear axis" where I was able to confirm 5 tornadoes, IWX confirmed another couple, and I know of at least a few other additional tornado damage points, adding up to likely a total of a dozen or so tornadoes. 20 is not an exaggeration. It's a great conundrum as to how to deal with these and I don't envy any warning forecaster who faces such a situation. Odds often end up being, as you mentioned, that there may only be a couple touchdowns. Sometimes, however, numerous tornadoes occur, and even with somewhat significant damage (I surveyed EF2 damage in the most recent shear zone event). The only real solution I see is a much more comprehensive radar system, ala CASA, which is many many years and a change in funding away.

[isohume]

April 19th of course was the "largest" tornado outbreak recorded in the state of Indiana, with the majority of the tornadoes, well, from shear zones. A few were from more solid mesovorticies. Similarly, April 2nd 2006 was one of Illinois's "largest" outbreaks, and the vast majority of tornadoes that day were from "shear axis" tornadoes. I just surveyed an outbreak of tornadoes from a "shear axis" where I was able to confirm 5 tornadoes, IWX confirmed another couple, and I know of at least a few other additional tornado damage points, adding up to likely a total of a dozen or so tornadoes. 20 is not an exaggeration. It's a great conundrum as to how to deal with these and I don't envy any warning forecaster who faces such a situation. Odds often end up being, as you mentioned, that there may only be a couple touchdowns. Sometimes, however, numerous tornadoes occur, and even with somewhat significant damage (I surveyed EF2 damage in the most recent shear zone event). The only real solution I see is a much more comprehensive radar system, ala CASA, which is many many years and a change in funding away.

I'm having a hard time wrapping my brain around this "shear axis" tornado concept. If I understand correctly...these are tornadoes capable of producing EF2 damage with no discernible SRM couplets? Could you point me to a paper or training module?

[The Dude]

Since the the majority of the public gets their weather/warning information from the media...the media needs to help make sure those messages are conveyed during live events. Not always plausible...but the information is there for the taking. It is always available on our warning products through the Internet and NWR.

No doubt the information is there for anyone who actively seeks it out, and the media certainly bears responsibility for ensuring that details are communicated. That said, at the end of the day, the goal is to make sure the public actually gets the information they need, not just that they should have gotten the information they need. To that end, if the NWS can make the information more "fool-proof" it would be less reliant on the intermediary messenger understanding and providing the proper context.

[isohume]

No doubt the information is there for anyone who actively seeks it out, and the media certainly bears responsibility for ensuring that details are communicated. That said, at the end of the day, the goal is to make sure the public actually gets the information they need, not just that they should have gotten the information they need. To that end, if the NWS can make the information more "fool-proof" it would be less reliant on the intermediary messenger understanding and providing the proper context.

I'm not sure how we can be any more clear. What are your suggestions?

[The Great Zo]

I'm having a hard time wrapping my brain around this "shear axis" tornado concept. If I understand correctly...these are tornadoes capable of producing EF2 damage with no discernible SRM couplets?

Probably not. I would guess that most EF2 or greater tornadoes, even if they developed along a squall line, probably had a discernible couplet on the radar. They may still be brief, but we're probably talking 5-10 minutes instead of 1-2 minutes (like many of the EF0 spin-up squall line tornadoes I've seen).

Trying to issue a warning on a couplet that only lasts 5-10 minutes is hard to do. Trying to issue a warning on a couplet that only lasts 1-2 minutes is basically impossible. The only way to handle the weak spin-ups is a large-scale tornado warning on the shear axis where these circulations might at some point develop -- but that leads to a huge false alarm surface area (a "statistic" which isn't actually tracked) and a blog post from James Spann .

As Tony alluded to, developing a warning philosophy for this is a nightmare.

[tornadotony]

I'm having a hard time wrapping my brain around this "shear axis" tornado concept. If I understand correctly...these are tornadoes capable of producing EF2 damage with no discernible SRM couplets? Could you point me to a paper or training module?

No, I can't. And that's a problem. Because when there is a tornado capable of snapping 10 newer, very large wooden power poles, some at their base or shearing them off so that only a tiny toothpick-looking piece of wood is left and there is no real discernible radar signature from the nearest WSR-88D, that's not good.

[The Dude]

I'm not sure how we can be any more clear. What are your suggestions?

As I noted above, I think it would be helpful to have the ability to use a different headline to distinguish the severity of the threat.

This is a very good point. We need to have something to distinguish a warning based on a radar-indicated potential for small spin-ups, and warnings based on spotter-verified violent tornadoes on the ground. Perhaps we could use Tornado Advisory for the former and Tornado Warning for the latter, or simply expand the use of Tornado Emergency to cover the latter.

[isohume]

No, I can't. And that's a problem. Because when there is a tornado capable of snapping 10 newer, very large wooden power poles, some at their base or shearing them off so that only a tiny toothpick-looking piece of wood is left and there is no real discernible radar signature from the nearest WSR-88D, that's not good.

Wow...never heard of such a phenomenon. Phantom strong tornadoes. lol. I've seen tornadoes with weak/moderate rotational velocities, but I've never seen them with no discernible SRM signatures, producing that level of damage. You Midwesterners can keep all that.

[tornadotony]

Wow...never heard of such a phenomenon. Phantom strong tornadoes. lol. I've seen tornadoes with weak/moderate rotational velocities, but I've never seen them with no discernible SRM signatures, producing that level of damage. You Midwesterners can keep all that.

Haha yeah. Trust me, there are folks at WFO LOT, as well as myself, looking into what happened up here the morning of 5/25 and trying to discern if there was any conceivable way to detect these things (I say things because so far we have confirmed 5 tornadoes from that morning, have a very likely 6th, and possibly a few more, all but one mid-high EF1 or low-EF2, and with path lengths 2 miles to as much as 6.3 miles long, sans the EF0 which was quite short-lived).

[isohume]

As I noted above, I think it would be helpful to have the ability to use a different headline to distinguish the severity of the threat.

Tornado warnings should always be taken seriously. Once you get cute with the wordage...that will be the time when one blows up on you. We can't say..."...RADAR BASED TORNADO WARNING..." in the headline. That would imply that there may be no real threat or really a vague need for the warning in the first place. The overwhelming majority of tornadoes are detected and issued based on radar. We already have a general statement in the text that says why the warning was issued. The main thing you want is a reaction to the phrase "TORNADO WARNING"...you don't want folks trying to figure out or interpreting the headline.

[isohume]

Haha yeah. Trust me, there are folks at WFO LOT, as well as myself, looking into what happened up here the morning of 5/25 and trying to discern if there was any conceivable way to detect these things (I say things because so far we have confirmed 5 tornadoes from that morning, have a very likely 6th, and possibly a few more, all but one mid-high EF1 or low-EF2, and with path lengths 2 miles to as much as 6.3 miles long, sans the EF0 which was quite short-lived).

So all of those tornadoes had no warning? Wow.

[tornadotony]

So all of those tornadoes had no warning? Wow.

Correct. A couple were covered by an SVR. One of the potential tornadoes that I didn't get to survey was actually covered by a TOR that was issued based off a spotter report of it on the ground, so it came after it formed. Understandably, with that being the only report and the crappy (and I do mean CRAPPY) radar display, the warning forecasters were extremely hesitant to issue a TOR, especially a giant blanket TOR which would have been the only way to do it with any sense of legitimacy and would have had to have covered parts of six or seven counties.

[The Dude]

Tornado warnings should always be taken seriously. Once you get cute with the wordage...that will be the time when one blows up on you. We can't say..."...RADAR BASED TORNADO WARNING..." in the headline. That would imply that there may be no real threat or really a vague need for the warning in the first place. The overwhelming majority of tornadoes are detected and issued based on radar. We already have a general statement in the text that says why the warning was issued. The main thing you want is a reaction to the phrase "TORNADO WARNING"...you don't want folks trying to figure out or interpreting the headline.

I agree that tornado warnings should always be taken seriously. The problem is that we have at least anecdotal evidence that they weren't being taken seriously in Joplin (and elsewhere) because people were assuming it was either a false alarm, or a small, short-lived tornado.

By introducing differing degrees of tornado warnings, I absolutely agree that there is a risk that people will begin disregarding the lesser of the two. I think that risk is outweighed by learning to treat the stronger one as an imminently life-threatening situation.

[tornadotony]

http://www.crh.noaa.gov/lot/?n=20110525storms

[isohume]

Correct. A couple were covered by an SVR. One of the potential tornadoes that I didn't get to survey was actually covered by a TOR that was issued based off a spotter report of it on the ground, so it came after it formed. Understandably, with that being the only report and the crappy (and I do mean CRAPPY) radar display, the warning forecasters were extremely hesitant to issue a TOR, especially a giant blanket TOR which would have been the only way to do it with any sense of legitimacy and would have had to have covered parts of six or seven counties.

Wow. Well I imagine the MIC has instigated a root cause analysis on those events...or at least should. That's pretty bad missing 5+ tors that did significant damage across your CWFA.

[tornadotony]

Wow. Well I imagine the MIC has instigated a root cause analysis on those events...or at least should. That's pretty bad missing 5+ tors that did significant damage across your CWFA.

I don't know if it's an official analysis being done, but I can assure you quite a bit of investigation is ongoing. I don't want to speak to much of what has been found so far (not that a whole lot has other than what was rather obvious during the event) but it's an interesting case for sure that will require a ton of further study.

[isohume]

I agree that tornado warnings should always be taken seriously. The problem is that we have at least anecdotal evidence that they weren't being taken seriously in Joplin (and elsewhere) because people were assuming it was either a false alarm, or a small, short-lived tornado.

By introducing differing degrees of tornado warnings, I absolutely agree that there is a risk that people will begin disregarding the lesser of the two. I think that risk is outweighed by learning to treat the stronger one as an imminently life-threatening situation.

All tornadoes can kill you, not just EF4/5s. If someone doesn't heed a tornado warning that is their personal business and their right. Of course Joplin, it probably didn't matter what action you took. The problem with tornadoes is they affect a very small area...and normally away from people. I can see why there is sometimes a complacent reaction toward TORs. They think..."well it's never affected me before...why would it now?" But as more and more people become affected and towns destroyed...people will begin to take them more seriously.

[isohume]

I don't know if it's an official analysis being done, but I can assure you quite a bit of investigation is ongoing. I don't want to speak to much of what has been found so far (not that a whole lot has other than what was rather obvious during the event) but it's an interesting case for sure that will require a ton of further study.

Yeah, I've never heard of that many damaging tornadoes being missed by one WFO. Interesting case indeed.

[Hoosier]

I don't know if it's an official analysis being done, but I can assure you quite a bit of investigation is ongoing. I don't want to speak to much of what has been found so far (not that a whole lot has other than what was rather obvious during the event) but it's an interesting case for sure that will require a ton of further study.

I wonder if part of the problem with 5/25 has to do with the distance from the radar. Romeoville is pretty far away from Newton/Jasper counties and would be scanning pretty high even on the lowest tilt. I'm not sure if the folks at LOT were utilizing other radar sites, but in any case, these things can be really difficult to pick up.

Check your PM's in a minute...

[tornadotony]

I wonder if part of the problem with 5/25 has to do with the distance from the radar. Romeoville is pretty far away from Newton/Jasper counties and would be scanning pretty high even on the lowest tilt. I'm not sure if the folks at LOT were utilizing other radar sites, but in any case, these things can be really difficult to pick up.

Check your PM's in a minute...

ILX and I believe IWX were also being used, that I can assure you. Problem is that they are even further from Newton/Jasper Counties than LOT.

[Hoosier]

I believe the majority of PAH tornadoes are from squall lines - brief and short lived. PAH issues a LOT of tornado warnings but they also like to go out and verify those warnings.

They do an excellent job at verification.

There is a belief here at the local office that there are more of these weak EF0-EF1 tornadoes than meteorologists realize and that quite a few are missed in storm surveys. Mainly because the damage occurs in fields/tree areas - other.

I have had a few conversations with several of the local mets about this subject.

That is an interesting subject. Personally I think QLCS tornadoes are particularly vulnerable to getting missed given their tendency to be brief and not always very visible. Think about our big nocturnal squall line events. You get a tornado that lasts a minute or two (if that) in darkness in a rural area and possibly wrapped in rain...yeah, good luck documenting it.

[radarman]

The only real solution I see is a much more comprehensive radar system, ala CASA, which is many many years and a change in funding away.

With regards to QLCS micro-vortices spinning up, probably the best case I've seen with CASA radars is from Apr. 2nd, 2010 in Rush Springs OK.

These things happen fast and lead time is very hard to come by. They certainly do have the potential to be dangerous and should be warned, as the damage from this one proved.

This wasn't warned and really wasn't visible on KTLX. I've written an algorithm to automatically detect high winds in these cases, and send out emails, tweets, and messages on NWS chat, and these are monitored by our local EMs.

http://www.kfor.com/...0,3648706.story

http://newsok.com/br...lery/3451169/1/

You're absolutely right about the funding and the work it would take to install our system all over. Frankly we've had trouble making the business case in the private market. And despite lots of support from folks in govt. agencies and civilians we all know the state of the federal budget. Nexrad cost billions...

I think you may see some of us attempt to privatize once the NSF funding comes to a halt, which will make it easier to hammer on cost reduction. Maybe in a few years we'll be closer to a larger deployment? A lot hinges on the success of CASA in Dallas, which we're going to begin in earnest very soon.

[Ian]

Plenty of people don't know the difference between a watch and a warning. Throwing a whole bunch of different types of advisories or warnings out will probably confuse more. It would be nice if there was a way to easily discern the difference between a radar indicated and spotted tornado but even that probably doesn't help as much as needed since a large tornado can drop pretty fast. We may just have to accept more people are going to die in big outbreaks from time to time.

[Fred Gossage]

After watching, analyzing, and forecasting events in this part of the country for over 17 years... and running radar and working in-studio severe weather operations for one of the most aggressive severe weather market stations in Dixie Alley (FOX 6 WBRC is just as equipped, dedicated, and aggressive as Spann and his clan at ABC 33/40).... I think the main impact that can be made with getting the public the information they need to assess a threat in a "QLCS tornado warning event" is how we handle the situation on television. There are going to be anomalies in nature, such as Tony's situation on May 25th... but the majority of the time, shear-axis type squall line tornadoes are not going to do damage that will injure or kill you if you're treating it as what I call a "hallway storm" (lowest floor, center of the house, away from windows.... needs to be done for straight-line winds too). There was an instance on May 11th, with a severe squall line moving across Alabama... where we had multiple tornado warnings issued for "signatures" that did not even have -10kt / +10kt couplets. If we're on the air, talking about the overall straight-line wind threat... and how 70-80 mph winds that can and will knock down widespread tree and power lines over a large area.... are moving through... and doing a squall line type storm track with times of arrival, instead of a fan-type track and hyper-local street level mapping.... we are still covering the area that may produce a spin-up and telling people to take action..... but we're also covering the areas that will receive straight-line wind damage just as adequately, without trying to focus on one trouble area... that may not even be a trouble area for more than 2 or 3 minutes, out of a 30-45 minute tornado warning issuance... and causing everybody else around them to ignore the wind threat, when trees and lines may be coming down around them.

In these cases.... I think the two biggest things for broadcast weather people to remember are... 1.) 80 mph winds will do the same damage, whether they are spinning or blowing from one direction.... 2.) The radar systems like FasTrac, VIPIR, Titan, and ESP have squall-line storm tracking features for a reason. USE THEM!

[isohume]

Plenty of people don't know the difference between a watch and a warning. Throwing a whole bunch of different types of advisories or warnings out will probably confuse more. It would be nice if there was a way to easily discern the difference between a radar indicated and spotted tornado but even that probably doesn't help as much as needed since a large tornado can drop pretty fast. We may just have to accept more people are going to die in big outbreaks from time to time.

True. The more complicated you get with a TOR message the more it will be misinterpreted by the public. We are told to write to a 9th grade level...meaning less is more and no ambiguity.

The vast majority of TORs are based on radar and when that is the case, the phrase "CAPABLE OF PRODUCING A TORNADO" is included in the warning text. This phrase is often overlooked I've noticed and it needs to be better understood by the public. Certain radar signatures on one storm may produce a tornado...while the same features on another storm will not. This is one of the limitations of radar science right now and it needs to be more in the forefront of our communication with the media and the public.

In any case...people should take cover when they are under a TOR warning. It's better to waste 30 mins and be sure than to take the possible chance you will bet hurt or killed.

[gymengineer]

Plenty of people don't know the difference between a watch and a warning. Throwing a whole bunch of different types of advisories or warnings out will probably confuse more. It would be nice if there was a way to easily discern the difference between a radar indicated and spotted tornado but even that probably doesn't help as much as needed since a large tornado can drop pretty fast. We may just have to accept more people are going to die in big outbreaks from time to time.

...like 2/1998, 5/1999 or 2/2008 large, not like 4/27/11 large. Sure, 60 people still may die in violent outbreaks once every few years, but 330 is just not a number that "needs" to happen. Even if you disagree with every one of Spann's points (I don't think he's right on everything), there will be another set of identified shortcomings and recommendations upcoming in the NWS's own Service Assessment. I do think they are going to say something about both the weather radios and the standardization of tornado emergencies.

[Ian]

True. The more complicated you get with a TOR message the more it will be misinterpreted by the public. We are told to write to a 9th grade level...meaning less is more and no ambiguity.

The vast majority of TORs are based on radar and when that is the case, the phrase "CAPABLE OF PRODUCING A TORNADO" is included in the warning text. This phrase is often overlooked I've noticed and it needs to be better understood by the public. Certain radar signatures on one storm may produce a tornado...while the same features on another storm will not. This is one of the limitations of radar science right now and it needs to be more in the forefront of our communication with the media and the public.

In any case...people should take cover when they are under a TOR warning. It's better to waste 30 mins and be sure than to take the possible chance you will bet hurt or killed.

Yeah I guess I wasn't clear.. I meant it would be nice to easily see that it is radar/spotter/etc up front rather than buried in the warning following a lot of standard text. The main way to get the public to listen is to scare them IMO. Maybe we need color coded tornado warnings. Green = grab your camera and stand in the driveway like a goon. Red = pray you aren't impaled by 4000 toothpicks before your skin is ripped off.