4/7/2006 DISC: Midwest and South

[Jeff Miller]

Amazing and unfortunately highly destructive, devestating day for many in the midwest and south. Major metropolitan areas such as Nashville, Tennessee area and counties such as Sumner County took very hard hits.

Just looked at some video and saw Sumner County took a particularly hard hit from a particularly long track tornado that looked to have strong F3 to low F4 damage in a subdivision.

What a day as all the conditions came together for strong supercells in the southeast. As many as 7 fatalities have been reported by media outlets.

What a day, my heart goes out for all those affected today by this situation. What are your thoughts on today's outbreak? And let me be the first to say, huge kudos to SPC for nailing this today.

 

[Jonathan Whitehead]

Ok I just got back to my apartment from cousins in Edmond. I had to go take care of some family business this afternoon. Anyways I got an email from stormspotterlive.com (of which i am a member of, those of you that are familiar with it) at 3:08pm reporting that

"Nashville tornado takes out NWS capability....the NWS office in Nashville at 3:08pm is no longer issuing warnings.....no word from NWS Nashville since tornado in area....NWS Morristown, TN now issuing for entire state....Debris seen on Nashville radar"

Im attaching a screenshot of the email just for proof.

Now I was tracking the tornadic supercells early this afternoon just before they were coming into the city. Im attching the last radar image I saw before I had to leave to take care of business.

But I am very interested to know what happened in Nashville today. Does anyone have any information as to what happened and where specifically b/c as we all remember it was just a few years ago downtown Nashville got hit dead on.

[attachmentid=104]

[attachmentid=105]

A side note, I saw on the local news at my cousins a short video clip of kid in Collierville, TN (which is a suburb of Memphis on the east side) holding baseball size in his hand there were several others littering the ground nearby where he was standing.

Also as im typing im watching The Weather Channel's coverage and they are reporting 8 deaths in Sumner Co. and 2 in Warren Co. all in Tennessee

Any further details would be appreciated

 

[Scott Olson]

Hi Chris, Nashville metro wasn't hit by a tornado. There wasn't a storm in the city when the communications went out but it's possible that one of the several supercells caused damage to the infrastructure which it relies on. In any event a tornado may have tracked as little as 8 miles north of downtown Nashville. "Debris seen on Nashville radar' this is most likely refering to the scan very near the outage where the storm (which isn't to far off) shows a sudden appearance of 55dbz in the hook echo/over couplet. This was a pretty low slice 1kft-2kft.
It appears the orignal supercell is responsible for most/all of the currently known fatalities, causing extensive damage NE of Nashville. Another tornado developed over SW Louisville, KY and appears to have tracked through the heart of southern Louisville and it's suburbs. I saw it on a webcam in Louisville at the time, with a beautifully sculpted updraft and rope like appearance.

Hi Chris, Nashville metro wasn't hit by a tornado. There wasn't a storm in the city when the communications went out but it's possible that one of the several supercells caused damage to the infrastructure which it relies on. In any event a tornado may have tracked as little as 8 miles north of downtown Nashville. "Debris seen on Nashville radar' this is most likely refering to the scan very near the outage where the storm (which isn't to far off) shows a sudden appearance of 55dbz in the hook echo/over couplet. This was a pretty low slice 1kft-2kft.
It appears the orignal supercell is responsible for most/all of the currently known fatalities, causing extensive damage NE of Nashville. Another tornado developed over SW Louisville, KY and appears to have tracked through the heart of southern Louisville and it's suburbs. I saw it on a webcam in Louisville at the time, with a beautifully sculpted updraft and rope like appearance. The Louisville office had to switch responsibilities as the tornado was close enough to force them to take shelter.

Also, there are some outstanding and widespread large hail reports coming in. Lots of soft-ball size hail reports and even larger, including one of 4.25" hail. Keeping in mind these are preliminary, it's pretty reasonable to assume some areas experienced some significant hail damage.

 

[Jeff Snyder]

"Debris seen on Nashville radar' this is most likely refering to the scan very near the outage where the storm (which isn't to far off) shows a sudden appearance of 55dbz in the hook echo/over couplet. This was a pretty low slice 1kft-2kft.[/b]

I think I stopped getting KOHX 88d data around 2:25pm. I have the reflectivity image from 2:13pm, but it isn't entirely too noteworth (well, not for that storm at least). I do remember seeing an area of enhanced reflectivity near the tip of the hook echo about that time (2:25pm). Given that KMRX took over for KOHX for the next couple of hours, I assume OHX experienced communications failure. From GR3, the lowest scan tilt should have been about 350 feet above the ground in the area of the mesocyclone of that storm; I think the hook approached to within 7-8 miles (to the north) of the OHX radar.

Also, there are some outstanding and widespread large hail reports coming in. Lots of soft-ball size hail reports and even larger, including one of 4.25" hail. Keeping in mind these are preliminary, it's pretty reasonable to assume some areas experienced some significant hail damage. [/b]

Given the degree of organization of some of the supercells today (fantastic BWERs and intense and large mesocyclones present on/in some), the massive hail does not surprise me. Usually the gorilla hail days are those with stronger CAPE (>=3000), though I suppose that's only through personal observation. With the strong 0-1km SRH in parts of the area (helping to support the strong mesocyclones), I have little doubt that the updrafts were likely considerably stronger than one would otherwise expect in a thermodynamic environment similar to what was seen today over the area. I am VERY interested in seeing if we see any >4.5" hail reports from the supercell that was north of Pontotoc MS this afternoon. There was a nice area of >75dbz reflecitivities relatively vertically-stacked just north of that town (with VILs maxing out at 80 kg/m2). Given the velocity and reflectivity structure of that storm at that time, in addition to the mid-level thermdynamic profile, I wasn't surprised that various hail algorithms (GR2AE, multi-radar mesh estimates, etc) indicated hail in the 4.5-5+ inch range.

Interestingly enough, I think the highest NEXRAD-indicated (whatever algorithm is used) estimated max hail size on that storm was like 3.75"... I say that's interesting because it seems like that algorithm hasn't been too hot lately. I remember OUN saying that it was grossly overestimating hail size with the supercell south of Elk City on 4/1... In addition, we sampled the areas where the largest hail is typically found (in the hook, just inside the hook / between the RFD / FFD and the updraft, etc) on the supercell near Ada on 3/30, and the supercell that tracked through Pryor OK yesterday (where we felt the full force of the RFD), and the largest hail I encountered was dime size (MAYBE dime size, if not smaller). I believe that both of those storms were warned for hail larger than golfballs. Just an observation.

 

[j_scheerer]

This maybe off topic. If it is, my apologies. But could the tornado that hit the Volunteer State Community College have been an F-5?

From yahoo.com news;

"John Stevens, a taxi driver who was visiting Volunteer State, said the building where he waited out the storm shook as the tornado passed over.

After the storm, he found his minivan had been thrown about 150 yards from its parking space. The vehicle was twisted and smashed to pieces.

"It's like some giant sat on it," he said. "

 

[Joe Nield]

WFO Nashville lost all comms for a period today, at least according to what I heard. We at IND took over warning responsibility for ten minutes from WFO Louisville while I was on shift tonight, but didn't have to issue anything as they had all storms adequately covered prior to preparing to shelter.

The largest hail in the Indianapolis County Warning Area was 2 inches, in Bloomington and northeast Greene County. Several reports of golfballs were received as well.

On the subject of the 88D hail algorithms, I can't speak for other offices, but I rarely use them. They seem to be nearly useless on predicting size, and are only marginally better on presence of severe hail. Base data interpretation is always better.

 

[Greg Stumpf]

On the subject of the 88D hail algorithms, I can't speak for other offices, but I rarely use them. They seem to be nearly useless on predicting size, and are only marginally better on presence of severe hail. Base data interpretation is always better.[/b]

Joe,

Just curious - how do you use the base data alone to determine probability of severe hail or estimated size?

I'm not knocking your technique, and I'm not necessarily trying to "defend" the algorithm (which, BTW, is technology over 14 years old) - we (in the NWS) are using this information to help make warning guidance more useful.

greg

 

[Jonathan.Brazzell]

Joe,

Just curious - how do you use the base data alone to determine probability of severe hail or estimated size?

I'm not knocking your technique, and I'm not necessarily trying to "defend" the algorithm (which, BTW, is technology over 14 years old) - we (in the NWS) are using this information to help make warning guidance more useful.

greg
[/b]

Greg,

Just to respond to what Joe said, It is true that the hail algorithm does tend to over estimate the size a little. Mostly depends on what is entered into the rpg for the freezing level and -20C and if it is kept up to date. Still forecasters do tend to show some improvement over the algorithm in some cases by using the base data when combined with expierence. For instance, I look at the height of the 50dbz echo and the size of the core aloft and based on what the freezing level is, combined with expierence you have a pretty good feel for the hail size. Plus looking at the base data is faster than waiting on the algorithm to run, so you can get a jump on it.

There is a pretty good write up on this from Rodney Donovan out of the Des Moines office.
Here is a power point presentation that he put together when he worked in Grand Forks: Hail Size and Reflectivity and I know there have been other off shoot studies based on his study. Just shoot him and email and he will be glad to share his information with you.

Jonathan

 

[Jeff Snyder]

Well the first two ratings are in from Nashville, they sent three teams out today so more reports should filter in this afternoon.
[/b]

Alex,

Just as an FYI... Those two damage assessments are from the Memphis (MEG) NWSFO. I'll try to keep the following update --> 4/7 Damage Assessments. If you read some damage assessments that aren't in there yet just post them there... I'm going to update the first post in that thread as much as possible; I'll be busy tonight, so I may not get to update it much this evening should some PNSs come out.

 

[Aaron Kennedy]

Plus looking at the base data is faster than waiting on the algorithm to run, so you can get a jump on it. [/b]

If by base data you mean reflectivity at the lowest elevation tilt, you'll be in a world of hurt. I completed a small project during my undegrad career looking at hail reports compared to the highest reflectivity (on the base tilt) that passed over the location within a 15 minute window. Think of it as an expansion of:

"The Relationship Between Low-Elevation WSR-88D Reflectivity and Hail at the Ground Using Precipitation Observations from the VORTEX Project
Arthur Witt"
http://www.nssl.noaa.gov/swat/SLS18/slscon96.aw.html

The results were horrible; it was basically a crapshoot...which shouldn't be surprising. Most of our recent climatological hail studies are based off spotter reports which have issues, both in size and time. We need to bring back some of the hailpads like past field projects and CoCoRahs. Stick them near KOUN and compare the results to some dual-pol data. and see how it lines up with CHILL. Even better, we need a bunch of hail probe vehicles

Aaron

 

[Jonathan.Brazzell]

By base data we mean the raw reflectivity at all elevation angles. Yeah it would be crazy just to look at the lowest tilt.

 

[Tom Tackett]

For this event, statistically speaking, there were over 700 Warnings issued by local NWS offices. From 9:00 A.M., CDT Friday, April 7th, until 7:00 A.M., CDT Saturday, April 8th......There were 197 Tornado Warnings Issued and 519 Severe Thunderstorm Warnings. That is almost one warning every two minutes. Quite an event. A Huge THANK YOU to all our members and viewers who are with NOAA, the storm spotting community, local media, chasers, and emergency management officials for providing outstanding work with regards to ALL these recent significant severe weather events!!!!!!

Source of Data: http://kamala.cod.edu/svr/

Tom Tackett---
Fort Worth, Texas

 

[Joe Nield]

Joe,

Just curious - how do you use the base data alone to determine probability of severe hail or estimated size?

I'm not knocking your technique, and I'm not necessarily trying to "defend" the algorithm (which, BTW, is technology over 14 years old) - we (in the NWS) are using this information to help make warning guidance more useful.

greg
[/b]

The typical stuff...height, extent, and intensity of reflectivity cores, etc. We've been using a chart put together by another office (their exact identity escapes me at the moment), that relates current freezing levels to the height of 50-55 dbz necessary for severe hail, and it has performed admirably for discriminating strong from severe, but that's just one tool, of course.

For the most part, we "calibrate" our size expectations with the first several hail reports we receive. I don't remember ever seeing anyone here look at the size estimations in SCAN, but then again, I don't remember seeing many people use SCAN. I don't want to slam the work you guys have done with the algorithms, far from it. The TVS algorithm performed very admirably during our 3/31 event, noticeably better than usual, in fact.

 

[Kenny Drake]

Usually the gorilla hail days are those with stronger CAPE (>=3000), though I suppose that's only through personal observation.
[/b]

Mathematically, the updraft speed is directly proportional to CAPE. So, based on updraft speed being able to recirculate the hail: the higher the CAPE, the larger the hail. Of course, this is very generalized as it neglects many things, some previously mentioned such as the freezing level.

 

[Jeff Snyder]

Mathematically, the updraft speed is directly proportional to CAPE. So, based on updraft speed being able to recirculate the hail: the higher the CAPE, the larger the hail. Of course, this is very generalized as it neglects many things, some previously mentioned such as the freezing level. [/b]

Yup, but remember that CAPE (buoyancy) is only half the story. Due to the rotation of the updraft (mesocyclone), the vertical perturbation pressure gradients sometimes can contribute just as much to the intensity of the updraft as buoyancy can. There've been field studies in the past that have noted negative buoyancy at the cloud base of significant supercells (with theta-e at cloud base in the updraft less than the theta-e of the surrounding environment). The presence of vert. pert. PGF leads to the notion that supercells with 1500j/kg CAPE can have more intense updrafts than non-supercells (multicells, squall lines, etc) with 2500j/kg CAPE. This time of year, supercells can probably get away with less CAPE since the freezing level and wet-bulb zero level tend to be lower than later in the spring and summer. Of course, too low t=0c or Twb=0c works against hail production.

In addition, CAPE distribution plays a role in updraft intensity. Short, fat CAPE profiles tend to support stronger updrafts than long, slender CAPE profile, even if the given amoutn of CAPE is the same in both cases, due to negative contributions from water loading, cloud detrainment / dry-air cloud-edge entrainment, and other processes. Perhaps the time of year and cooler mid-level temps support steeper lapse rates, which tend to fatten the CAPE profile. In addition, later in the year, Tds in the 70s can present more significant water loading issues. So, a supercell in March with inflow layer characterized by (t/td) 75/62 and 2500j/kg CAPE may have a stronger updraft than a supercell in June with inflow characterized by 86/73 and 2500j/kg CAPE.