2008-02-05 DISC: TX / AR / MO / IL / KY / TN / MS

[Chris Hayes]

That is the longest track tornado since November 23, 1992.

 

[Jim Tang]

There were well over 1000 tornado warnings issued for this outbreak yet there were 70 reported tornadoes, many of these we are finding out were long tracked tornadoes and multiple reports for the same tornado. This means the NWS is getting 4-6% of the warnings right and this was on a historic outbreak were the storms were actually producing tornadoes.

There were nowhere near 1000 tornado warnings issued during the outbreak. The total was a little over 200 tornado warnings from 16 NWSFO's.

http://www.easternuswx.com/bb/index.php?s=&showtopic=155131&view=findpost&p=2507699

As for the central MS having a high FAR, well I think many tornado enthusiasts know that JAN issues far more tornado warnings for marginal/weak storms than almost all other WFO's in the country.

 

[Rich Thompson]

There were nowhere near 1000 tornado warnings issued during the outbreak. The total was a little over 200 tornado warnings from 16 NWSFO's.

http://www.easternuswx.com/bb/index.php?s=&showtopic=155131&view=findpost&p=2507699

As for the central MS having a high FAR, well I think many tornado enthusiasts know that JAN issues far more tornado warnings for marginal/weak storms than almost all other WFO's in the country.

The numbers I provided were from sometime around 1999. They have probably changed some since then (just a single year), and I know JAN has experienced more active years than the one quoted (all 11 tornadoes were F0-F1, I think). I don't know that I'd call JAN the most aggressive TOR warning office in the nation, however.

Rich T.

 

[Rich Thompson]

Hi Rich, calling it a false alarm rate (FAR) to me is inaccurate, at least with the respect to the Dopper radar. The reason, the Doppler detects mesocycones not tornadoes, that's always been its job and in that respect it has a very high success rate. Unfortunately when the original project was lobbied it was being sold for wind shear, correct. That was justified off Fujita's original work concerning aircraft safety, down bursts and commercial planes falling out of the sky. It wasn't just about saving lives, it was about general trust from the public about commercial flying. Of course now we have on board radar in the expensive aircraft. Voices started saying these Dopplers will protect us against tornadoes....I remember certain people trying to tell the truth, but they were drowned out. Thus the billions of dollars was spent to upgrade the most populated areas of the nation first, finally getting down to isolated areas like Goodland. In the rare cases when we do see a tornado on the end of the hook what we are actually seeing is the debris ball; again much larger than the typical tornado. Even with the new and more powerful radars coming on line how does one see a 100-150 meter wide event at 65 miles distant. That is pretty much what we are asking that piece of equipment to do. Frankly I don't think it's possible even before we consider (a) curvature of the earth and (b) the +.05 degree tilt for environmental safety. With respect to the Doppler seeing the mesocyclone, we know that's accurate and as chasers we can verify it's rotating. For example, how many times have you sat on a rotating mesocyclone for 1-2 hours before that supercell produced a tornado. Remember the Hill City event last season? The supercell sat in that area for how many hours before it finally produced a condensation funnel to ground, at least four, would that be a fair assessment? Meanwhile, the Goodland Doppler sees a potentially tornadic mesocyclone the whole time and warning after warning is issued, most counting as false alarms. I wish the public understood the relationship between what the radar is doing (correctly) vs true tornadogenesis. Armchair chasing on Feb 5th I watched a lone supercell take over from a cluster in far NE LA, it was the first supercell of the day. It move into SE AR and immediately spun up a strong meso. In fact it had meso signatures early on and base reflectivity showed a classic supercell configuration. Yet that storm as threatening as it was traveled for at least 100+ miles I bet, before starting to produce real tornadoes. As it moved north other cells formed and trained that supercell finally moving into the Memphis-western TN region. It was there when they really let loose with tornado after tornado. To not put the warning on such a cell early would be tantamount to duplicating the infamous Fort Smith event where hours of warnings were issued prior to the destruction in that city...with no warning. Some supercells produce tornadoes within 30-45 minutes, others wait hours, but in the end the tornadoes are just as strong.

Gene,

I can't really speak for the motivations of the folks that started the whole WSR-88D project. However, by the time I joined the NWS in '92, the primary focus was on "20 minutes of lead time" for tornadoes. That's where our problems started.

The earlier Doppler radars were not collecting data unless a "big" day was expected, thus the early storm samples may have been biased toward high-end mesocyclones and tornadoes. Hence the old "50% of mesocyclones produce tornadoes" statistic. We've come to realize the fraction that produce tornadoes is much smaller, likely somewhere in the 10-20% range (depends on how you define a meso).

Gene is correct - the 88D almost never resolves an actual tornado, thus it's not fair to "blame" the radar for everything. In reality, the decision thresholds and situational awareness of the warning meteorologists play the biggest role in POD and FAR.

Rich T.

 

[Alex Lamers]

Does anyone have a listing of the longest tornado tracks on record...and where this one would fall?

AJL

 

[Karen Politte]

This thread is a great discussion of Tuesday's events. Thanks, everybody, for making it so interesting and informative to read!

Jason and I have returned from our damage survey of the single, long-tracked tornado that affected Pope County through Sharp County, Arkansas on Feb. 5th, 2008. We surveyed the damage path from Atkins, AR in Pope County, through Shirley, AR in Van Buren Co.

Unfortunately, daylight and roads available, coupled with extreme traffic congestion in Clinton did not allow us to follow the damage path of this tornado any further than approximately >10 miles northeast of Shirley, AR in Van Buren County and consequently this is where our own damage survey had to cease. We did not get to view the damage in the Mountain View/Zion/Highland areas.

Our preliminary impression of the damage by this tornado is somewhat lower than we had expected given the news reports etc. in the days preceding this one. There is, of course, a large amount of tree damage - however, about 85% of the tree damage we encountered was simply a case of the “blow-down”, there was a complete lack of denuded and debarked trees in all but a very select few cases. The majority of the trees we saw were shallow-rooted evergreen trees that had simply been blown over in their original place, with turf-ball still in place around the roots at base.

There were two or three instances along our survey journey that unearthed questionable high-end EF-3/low-end EF-4 damage to brick homes. One was in a rural setting southwest of Cleveland, AR, and the other was a foundation that was missing a home in Clinton, AR. We will have to review the photographic data that we collected prior to forming a firm opinion on the ratings these structures displayed, however. It does, however, seem very likely to us that this was one single tornado - given the frequency of tree damage encountered along many single track back roads that we ventured into. At various points in time road passage was difficult.

More later, including images.

KL

 

[Brian Stertz]

Pretty remarkable tornado for sure

I went back on list of the long tracked tornadoes since 1950 and found one that may get the asterisk. According to Storm Data (March 1990), the C. Nebraska long tracked F4 tornado on March 13, 1990 had a 124 mile damage track. In contrast, Tom Grazulis' green book mentions this was 124 miles consisted of different "tornado family" members.

At any rate, its remarkable how a tremendous violent tornado can remain in such balance to create such a long damage track. This especially in a colder winter type severe wx setup with the main 850mb jet displaced over towards the Mississippi River and points eastward (where 0-3km SRH was shown to be well over 500 ms/s2 by early evening). The benefits and support were certainly there with the main 700/500mb wave pivoting the NW 1/3 of Arkansas and the 90-100 kts. of wind at 500mb.

 

[afischer]

The only other recent extremely impressive tornado (with regard to path length) that comes to my mind is the Stockton MO tornado of 5-4-03, with a path length of ~83 miles. 120+ miles is unbelievable. The 00Z Little Rock sounding showed ~1800j/kg MLCAPE which was obviously more than sufficient amidst the strong environmental wind fields. The temp profile was also "cool" with a 3km deep moist layer and a relatively impressive 125j/kg 0-3km MLCAPE... not terribly dissimilar to the thermo environment on 5-04-03. Seems to me a lot of the nastiest tornadoes vastly prefer a cool profile rather than having a significant elevated mixed layer. (e.g. Greensburg)

 

[Karen Politte]

Hey Brian,

That's interesting about that central NE storm with the 124 mile long track that may be one single tornado. I hadn't heard of that one - maybe I need to do some more research.

I'd edge away carefully from thinking that Tuesday's AR tornado maintained it's intensity for any significant length of time, however. It was certainly impressive to see a damage path continue uninterrupted for so far - but from our survey of the path today it was apparent that the vast majority of damage was in the EF-0 - EF-2 range. The track varied in width at several points - it was wide at Atkins, narrower at points northeast of there. Most notably however was the fact that - at points northeast of Atkins and southwest of Clinton, it was sometimes very hard to determine the damage path - the circulation appears to have got very diffuse and weak. It was still fairly wide - but the level of damage it was inflicting was greatly reduced compared to east of Atkins. Sometimes all we saw were some snapped twigs here and there. So - I would assert that this tornado went through several stages and morphs along it's path - and maybe have been multi-vortex in structure northeast of Atkins for a time. EF-4 damage is extremely hard to come by - as is EF-3. Mostly the tremendously long path consisted of nothing more than downed trees and sheet metal wrapped around fenceposts.

KL

 

[Karen Politte]

NWS LZK has just issued another Public Info. Statement confirming that the AR tornado's path from Feb. 5th was 123 miles long - making it the longest confirmed tornado path since 1950. I imagine to top it one would have to look towards the Tri-State event and other historical markers.

http://www.srh.noaa.gov/lzk/cgi-bin/wxs1.php?pil=PNSMAIN&max=13

KL

 

[Perry Williams]

Unfortunately, daylight and roads available, coupled with extreme traffic congestion in Clinton did not allow us to follow the damage path of this tornado any further than approximately >10 miles northeast of Shirley, AR in Van Buren County and consequently this is where our own damage survey had to cease. We did not get to view the damage in the Mountain View/Zion/Highland areas.

KL

Karen, from what I understand, some of the most intense damage occurred in the Mountain View and Zion areas. In that area, several vehicles were reportedly thrown several hundred yards; others are reportedly still missing.

 

[Danny Neal]

Not directly related to this specific event but along the same topic, didn't the June 2nd, 1990 outbreak produce some long tracked single tornadoes? I seem to remember an F4 from IL into IN on the ground for almost 100 miles. I'll try to look it up on Wiki or one of those pages.

Is there a timeframe on which this single tornado occured? I haven't looked at the NWS Little Rock page since I am not home and am on a dial up connection and it would take 2 hours for it to load.

 

[Brandon Clement]

There were nowhere near 1000 tornado warnings issued during the outbreak. The total was a little over 200 tornado warnings from 16 NWSFO's.

http://www.easternuswx.com/bb/index.php?s=&showtopic=155131&view=findpost&p=2507699

As for the central MS having a high FAR, well I think many tornado enthusiasts know that JAN issues far more tornado warnings for marginal/weak storms than almost all other WFO's in the country.

Thanks for the info. I read earlier that a NWS employer said over 1000 tornado warnings were issued. I think the reporter meant over 1000 warnings now that you have provided the correct info.

I did not realize Jan was viewed by most this way.

 

[Rob Satkus]

NWS LZK has just issued another Public Info. Statement confirming that the AR tornado's path from Feb. 5th was 123 miles long - making it the longest confirmed tornado path since 1950. I imagine to top it one would have to look towards the Tri-State event and other historical markers.

http://www.srh.noaa.gov/lzk/cgi-bin/wxs1.php?pil=PNSMAIN&max=13

KL

I believe they are talking only about tornadoes that have affected Arkansas...there have been at least 3 with reported longer paths since 1990 elsewhere, including the previously mentioned 3-13-90 NE tornado at 124.0 miles, the Brandon MS F4 tornado of 11-21/22-92 at 128.0 miles and the already noted 11-23-92 NC F3 at 160.0 miles...whether these were adequately surveyed I have no clue. I would expect there were others from 1950 - 1979 with longer paths as well due to lack of surveys/knowledge of cyclical supercells.

RS

 

[Matthew Crowther]

I believe they are talking only about tornadoes that have affected Arkansas...there have been at least 3 with reported longer paths since 1990 elsewhere, including the previously mentioned 3-13-90 NE tornado at 124.0 miles, the Brandon MS F4 tornado of 11-21/22-92 at 128.0 miles and the already noted 11-23-92 NC F3 at 160.0 miles...whether these were adequately surveyed I have no clue. I would expect there were others from 1950 - 1979 with longer paths as well due to lack of surveys/knowledge of cyclical supercells.

RS

Grazulis in Significant Tornadoes is of the opinion that the 1990 NE tornado event was a family and that the NC event was as well.

Also, if compare that NWS list to the Grazulis book, there are some startling discrepancies:

2/20/1951: longest path length in AR...7 miles.

4/30/1954: 3 separate tornadoes in AR with lengths respectively...35, 23 and 35 miles.

3/20/1955: 40 miles.

12/24/1982: one 55 miles, one 48 miles.

I wonder where the NWS office got their data?

 

[jladue]

Gene,

I can't really speak for the motivations of the folks that started the whole WSR-88D project. However, by the time I joined the NWS in '92, the primary focus was on "20 minutes of lead time" for tornadoes. That's where our problems started.

The earlier Doppler radars were not collecting data unless a "big" day was expected, thus the early storm samples may have been biased toward high-end mesocyclones and tornadoes. Hence the old "50% of mesocyclones produce tornadoes" statistic. We've come to realize the fraction that produce tornadoes is much smaller, likely somewhere in the 10-20% range (depends on how you define a meso).

Gene is correct - the 88D almost never resolves an actual tornado, thus it's not fair to "blame" the radar for everything. In reality, the decision thresholds and situational awareness of the warning meteorologists play the biggest role in POD and FAR.

Rich T.

Check out Trapp et al. 2005

Latest stats for a algorithm identified mesocyclone (somewhat similar to a typical user defined mesocyclone) extending to within 1000m ARL is somewhere between 30 and 40%. These are the best stats out there with the widest representation. The 10-20% number considers all circulations that have 3km or more vertical continuity, and are at least rank 5 (see this online session about the mesocyclone detection algorithm). I'm sure the lower and stronger the circulation, the odds are even greater.

Another interesting project to consider is a project that Adam Houston at U of NE is undergoing. They're attempting to create a climatology of supercells, not just mesocyclones. That is, any mesocyclone algorithm cannot associate mesocyclones to its parent storm and so we don't have a good view of supercell climatology. Hopefully Adam's project can have some success in defining a percentage of tornadic supercells and relate the percentages to their physical structural details.

 

[Aaron Kennedy]

Hopefully Adam's project can have some success in defining a percentage of tornadic supercells and relate the percentages to their physical structural details.

I'm not familiar with the project, but it seems it would be subject to the caveat the study only includes significantly precipitating supercells... IE for classic and HP archetypes. Many LPs can be pretty darn anemic on radar.

 

[Brian Stertz]

I have gathered info from all the various damage surveys and have complied a list of the EF3 & EF4 tornadoes from the Feb. 5th/6th outbreak. Below is the link to the page on Vortex Times. There are still some details that have not been fully presented, and will update those as soon as they become available.

http://vortex-times.com/252608Outbreak

 

[Jack Beven]

Karen and all,

This description reminds me of something.

I was flying into Oklahoma City on May 5, 1999 when due to some snafu in air traffic control my plane did a 360 turn while on final approach. As a result, I wound up doing an inadvertent aerial survey of the Moore-Bridge Creek damage track.

One thing I was struck by was that it was a bit difficult to tell where the tornado crossed the Canadian River. There were well-defined tracks on either side of the river, but the damage swath was less well defined near the river itself. How much of this was due to changes in the tornado and how much due to changes in what the tornado hit, I don't know. I would say based on the changes in the track and what I've seen in chase videos that the tornado did undergo some changes while near the rvier, then regained its wedge form before reaching Moore. Maybe the chasers who saw it could comment further?

Jack Beven

Hey Brian,

That's interesting about that central NE storm with the 124 mile long track that may be one single tornado. I hadn't heard of that one - maybe I need to do some more research.

I'd edge away carefully from thinking that Tuesday's AR tornado maintained it's intensity for any significant length of time, however. It was certainly impressive to see a damage path continue uninterrupted for so far - but from our survey of the path today it was apparent that the vast majority of damage was in the EF-0 - EF-2 range. The track varied in width at several points - it was wide at Atkins, narrower at points northeast of there. Most notably however was the fact that - at points northeast of Atkins and southwest of Clinton, it was sometimes very hard to determine the damage path - the circulation appears to have got very diffuse and weak. It was still fairly wide - but the level of damage it was inflicting was greatly reduced compared to east of Atkins. Sometimes all we saw were some snapped twigs here and there. So - I would assert that this tornado went through several stages and morphs along it's path - and maybe have been multi-vortex in structure northeast of Atkins for a time. EF-4 damage is extremely hard to come by - as is EF-3. Mostly the tremendously long path consisted of nothing more than downed trees and sheet metal wrapped around fenceposts.

KL

 

[Perry Williams]

I have gathered info from all the various damage surveys and have complied a list of the EF3 & EF4 tornadoes from the Feb. 5th/6th outbreak. Below is the link to the page on Vortex Times. There are still some details that have not been fully presented, and will update those as soon as they become available.

http://vortex-times.com/252608Outbreak

Hi Brian,
Just a FYI: per an article in yesterday's Huntsville, Alabama Times newspaper (link below), the death toll from the Lawrence/ Morgan county EF-4 has risen to 4......a critically injured 86 year old woman succumbed to her injuries at a Huntsville area hospital

http://www.al.com/news/huntsvilletimes/index.ssf?/base/news/1202638523211330.xml&coll=1

Take care,
Perry

 

[Karen Politte]

Per NWS LZK, they should make available later today a track map with map coordinates from ground/aerial survey data for the Arkansas EF-4 tornado from Tuesday last week. I will post a link when it is online.

It will be good to see this - because I am curious as to where the EF-4 damage was found along the track. I am thinking that this type of damage may have been located up in the Zion/Highland areas, which is the exact area we didn't make it to on Saturday.

KL

 

[Chad Cowan]

Per NWS LZK, they should make available later today a track map with map coordinates from ground/aerial survey data for the Arkansas EF-4 tornado from Tuesday last week. I will post a link when it is online.

It will be good to see this - because I am curious as to where the EF-4 damage was found along the track. I am thinking that this type of damage may have been located up in the Zion/Highland areas, which is the exact area we didn't make it to on Saturday.

Based on some LZK data I have looked at, I would have to guess the EF-4 damage is located ~ 5-10mi sw of Clinton. The radar presentation at this time was remarkable and if I calcluated with the correct storm motion, the G2G shear at this time was over 150kts. I will be posting some of these GR2AE images in the next few days.

 

[Danny Neal]

Was there really a "tornado" over Nashville? http://youtube.com/watch?v=l5u-kqBjWj8 I among others were watching the webcam live over Nashville and saw that amazing CG along with the wall cloud passing right over downtown....here is a view that maybe one didn't see. Another "chaser" decides to go out when storms are around.

 

[Heidi Farrar]

As just about everyone knows, the monster tornado of February 5th missed our house by 2.5 miles. However, the debris from the tornado has been showing up in areas more than 10 miles from the path. This afternoon I decided to go out and see what i could find in our neighbor's 200+ acres of fields. I rode around on their Kawasaki Mule and filled the Mule's bed with debris: huge pieces of sheet metal, so much insulation that I stopped picking it up, boards with nails sticking out, a broken broom ... but I was after more indentifiable stuff. Earlier I had found a piece of paper from a doctor's office. It was obviously from their filing cabinet, because the date on it was Nov. 2001. I thought that it had to be from Mt. View, about 35 miles to our southwest, since I know their small hospital got hit pretty hard. However, just as I was finishing up my loop over their property, I found something that literally blows my mind.

I happened to notice a rather clean-looking piece of paper on the edge of one of their fields. As I got closer, I could tell that it was an envelope, with some strange items lying next to and underneath it. I knelt down and saw that there was a return address on the envelope. It said Clinton, Arkansas. The envelope contained dental X-rays. And for the record, Clinton is 75 miles from here, "as the crow flies."

Here's the envelope, X-rays and the sign-in sheet. Notice insulation, shingles and boards in the background.

There's a close-up of the envelope. The blackish rectangles are the X-rays. ... One thing I've noticed about the paper debris I have found is that it is all folded very strangely, almost like a paper airplane, with multiple folds in interesting directions. It's not crumpled at all.

Finding the dental records has to be the most mind-blowing thing I have ever experienced.

 

[Karen Politte]

I have managed to get our informal thoughts and pictures from our damage survey of the Feb. 5th record-long-track AR tornado online. You can view it here:

http://stormskies.com/OpenDoor/?p=204

Or of course just click on the link to my Blog in my sig.

It's a pretty casual damage survey and in no way detailed - and of course we failed to make it to the end of the path in Sharp County - but it was still a very interesting exercise and still provided me with yet another great learning experience w/r/t tornadoes. Unfortunately, we were not able to do a walk-through inspection of any of the damage property we saw - mostly due to the fact that all of the structural damage was on private property in remote areas. We also wouldn't have felt good about walking through the damage and ooohing and aaahing about this being unanchored or that being poorly built when those poor families are trying to pick up the remaining pieces of their lives.

Anyway - hope it provides food for thought for some.

As I sit in bed and write this, we have 39oF temperatures outside, thunder, lightning and hail. I love weather diversity.

KL