5/4/07 DISC: KS

[Johnathan Brouwer]

I was thru that town last year, it's very frightning to see that amount of damage. I was monitoring the spotternetwork yesterday and there were a lot of chasers all around that area. Maybe we as a community can converge there and lend a hand.

 

[jason kissam]

my gut told me to chase there yesterday thats where my original target was. my team and i headed to hill city area and had a neat storm. needles to say the time it took to go north instead of south i thank above for leading us in the right direction. Im sure we would have been in the middle of it if we went south. Our prayers are with greensburg and areas surrounding.

 

[Stan Rose]

Those aerial shots are incredible. Reminds me of the Jarrell damage. On first glance, you'd have to say 'if that's not EF-5, then there is no such thing'. Certainly some of the most widespread damage i've seen. On the other hand, most of the debris is wood planks, and it's hard to say how good that construction was, but i do see semi-intact houses next to wiped-clean foundations. Also, i noticed most of the cars seem pushed or damaged by flying debris, but not tossed or thrown. So, the damage may look bad, but not necessarily mean EF-5. The twisted metal is certainly a strong indicator. That's gonna be some extensive survey to conduct.

 

[Gabe Garfield]

Today's batch of tornadoes nearly paralleled the Friday night event, so it will be extremely difficult to differentiate tornadoes that occurred on Friday from tornadoes that occurred on Saturday. Will be interesting to see what DDC comes out with in the coming days.

Gabe

 

[Alexandre Aguiar]

 

[afischer]

Today's batch of tornadoes nearly paralleled the Friday night event, so it will be extremely difficult to differentiate tornadoes that occurred on Friday from tornadoes that occurred on Saturday. Will be interesting to see what DDC comes out with in the coming days.

Most definitely. I forgot to mention in my chase post that I encountered damage from 5/4's cyclic tornadic supercell on a few occasions while chasing on 5/5. The worst was east of Macksville in Stafford county. Another area was west of Hudson in Stafford county, and still another north of Ellinwood in Barton county.

 

[fplowman]

Reading Jason Kissams post. I too am thankful when I look back that I zigged instead of zagged. If I remember correctly that same storm fired down by Woodward Oklahoma. I think it was the same one I was monitoring on Baron all evening.

Me and Jim Cross chased the northern cell into the Osborne/Smith Center area before it was to dark to chase it. Then we headed back to Russel to get a room.

I recall telling Jim, There is a nasty cell with like 3 or 4 shear profilers on it East of Dodge. Im afraid it might come through Russell while were sleeping at 1 or 2am. Then we woke in the morning and it was all over the news.. Far out!

 

[George Tincher]

That radar loop from page 2 is unreal. Notice how the hook appears to turn into a "ball" just as it crosses over Greensburg. The reflectivity values also increase sharply as it passes over the town. Somehow I never noticed that the night of the event while watching the single scan updates on GR Level 3 but I am now convincved that was a debris ball showing up while the tornado was crossing the town.

 

[Kevin Bowman]

That radar loop from page 2 is unreal. Notice how the hook appears to turn into a "ball" just as it crosses over Greensburg. The reflectivity values also increase sharply as it passes over the town. Somehow I never noticed that the night of the event while watching the single scan updates on GR Level 3 but I am now convincved that was a debris ball showing up while the tornado was crossing the town.

the only thing more impressive are the pictures and videos, what a scary looking tornado

 

[David Yoho]

The NWS classified the Greensburg tornado as an EF-5.

http://www.foxnews.com/story/0,2933,270298,00.html

 

[Jeff Snyder]

For those w/ GRLevel3... I've uploaded the *rv3 files from 01:31:38 UTC to 03:27:12 UTC at http://www.tornadocentral.com/now/greensburg/ . The first file is when the supercell was NW of Coldwater (note the possible damage ball signature), and it ends as the 2nd huge tornado after the Greensburg tornado begins to develop.

Right-click and save-as, then drag those files into the GR3 window. All files should have, in the least, BR1, BR2, SRV1, and SRV2 data.

I have never seen an intense couplet as large as the one associated with the low-level mesocyclone / tornado that developed after the Greensburg tornado. There were many times during which the BR imagery showed a full "eye" feature at the center of the huge circulation. At the time, it reminded me a lot of the Hallam NE supercell and meso from 5-22-04. The inbound velocities associated with the inflow of that storm were absolutely remarkable.

The entire life-cycle of that storm was remarkable as well. I noticed very intense RFD surges occurred near the beginning of each circulation, when a notch of high reflectivity from the western side of the precip core would rotate extremely rapidly to the southeast then east. When these winds met up with the meso, I'd often see that circulation intensify and increase in size markedly.

Intermittent possible damage-ball signatures, "hammed-head" hook-echo signatures (with anticyclonic circulation on the south side of the hook / to the right of the RFD), incredible "eye" reflectivity features, intense circulation measuring more than a few miles wide, etc. I have never seen a supercell that looked like that one did on radar.

Below is SRV2 from 0327 UTC. Note the anticyclonic couplet on the south side of the RFD surge. Of course, it's worth noting the size of the cyclonic mesocyclone as well.

 

[Kevin Bowman]

i was unable to get screenshots when it was over greensburg but this was just to the northeast.

 

[rdale]

Kevin - use SRV when you are looking for rotations within storms...

 

[Andy Wehrle]

Good advice in general, but the Greensburg couplet was so well defined base velocity resolved it just fine. I usually keep GR 3 running for the duration of severe weather events on several different radar site and save hundreds of images in several different modes including base reflectivity, base reflectivity with storm attribute overlays, base reflectivity smoothed, base velocity, and storm relative velocity tilt 1 and 2.

I'd be curious as to what closer meterological study of this event reveals as what made conditions so perfect for such a violent tornado. After all there have been plenty of very destructive tornadoes between May 4, 1999 and May 3, 2007, but none that quite reached this level of violence. I used to remark after seeing damage from some of the most significant tornadoes of 2005/2006 that if that's not F5 damage, nothing is anymore, but after seeing what's left of Greensburg I can see that there is something uniquely devastating about this tornado. However, on the face of it there didn't seem to be anything different about Friday than other classic severe weather setups in recent years that "just" produced (E)F3-4 tornadoes.

 

[Jeff Snyder]

Good advice in general, but the Greensburg couplet was so well defined base velocity resolved it just fine. I usually keep GR 3 running for the duration of severe weather events on several different radar site and save hundreds of images in several different modes including base reflectivity, base reflectivity with storm attribute overlays, base reflectivity smoothed, base velocity, and storm relative velocity tilt 1 and 2.

I'd be curious as to what closer meterological study of this event reveals as what made conditions so perfect for such a violent tornado. After all there have been plenty of very destructive tornadoes between May 4, 1999 and May 3, 2007, but none that quite reached this level of violence. I used to remark after seeing damage from some of the most significant tornadoes of 2005/2006 that if that's not F5 damage, nothing is anymore, but after seeing what's left of Greensburg I can see that there is something uniquely devastating about this tornado. However, on the face of it there didn't seem to be anything different about Friday than other classic severe weather setups in recent years that "just" produced (E)F3-4 tornadoes.

I think there were some storm-scale processes that we can't really forecast or even understand at this time. The forecast hodographs from the 12z NAM and RUC for areas east of DDC to Great Bend showed incredible low-level curvature, with >400 m2/s2 0-1km SRH. Strong instability for a surface / low-level parcel + very intense low-level shear + low LCL + well-organized, discrete supercell + factors we may not know about or be able to quantify (I'll call them "intangibles") = possibility of violent tornado. Of course, we can't really quantify all of those factors too well (though their have been attempts in the form of "composite" indices like the Significant Tornado Parameter that show some skill).

Looking at the loop of that supercell over a 3-5 hour period, it was evident that there were some very interesting processes occurring within the storm. The cycle of RFD surge -> tornadogenesis -> intensification -> curve in path to the left with decrease in size of low-level meso -> new meso to the east of the old tornado/meso -> new intense RFD surge -> tornadogenesis ... Rinse and repeat.

It's just too bad we didn't have any mobile radars on that supercell for a longer time. The OU/Umass group had the Xband dual-pol scanning east of Protection, but couldn't continue due to battery issues (I'm excited to see that data!). That said, data were collected from the Greensburg tornado before it hit Greensburg. The DOWs were farther north and weren't able to scan that supercell either. At least it spent some of its lifetime near the DDC radar, so level-II radar data should be interesting (I haven't had a chance to download the L2 data yet).

FWIW, it's tough to say that we've seen conditions like this since 5-3-99... The environment in which the storm developed and progressed can be characterized by such parameters as CAPE, SRH, and so forth, but those are based on objective analysis methods that depend upon discrete observations. So, even with a RUC mesoanalysis, it's still tough to perfectly quantify the near-storm environment. In addition, I think that you can put two storms in the exact same environment and still get two different results. Violent tornadoes are rare, and their seems to be an extremely delicate balance in the troposphere (and heck, maybe even in the very lower stratosphere) that support such intense supercells and tornadoes. As we all know, the results of a "setup" may depend strongly on the nature and structure of storms that occur in that environment. We've all chased in favorable environments that we thought would support tornadic supercells, only to end up with multicell convection or a squall line.