Greensburg/Trousdale tornado 4 miles wide?

[Warren Faidley]

This is an excellent report. I have always been torn between the Greensburg event and the one I ended up following, the Arnett tornado. I started out almost in the exact middle of where the two events began, and only headed towards the Arnett storm because I figured it would be easier to intercept with NE movement. I also considered the fact that southern storms often steal the energy out of northern storms, although the alignment and distance did not seem to trigger that concern. I also chased the southern Arnett storm because I figured the photographic contrast would be better -- and it was. The EMT side of me wanted to be in Greensburg, although I did not hear about it until later that night. Ironically, the Arnett supercell, had it been the (or another) dominate storm that evening, would have likely made a direct strike on Woodward, as I chased the diminished cell right over the city. It was still producing tennis ball (and larger) hail just west of the town, but thankfully it lost all visual signs of serious rotation before getting there.

Regardless, I am beginning to wonder if (on days like the Greensburg event) that the Arnett storm played some part in enhancing the Greensburg cell either by an outflow feature (wind/rh/thermal dynamics) interacting with the RFD, etc., or something else a little higher in the troposphere. Or, possibly the single Arnett storm served to limit additional atmospheric turnover to the south. It's obvious that outflows enhance tornadogenisis, but I'm wondering if there is not more to this. It would be interesting to do a study on how many violent supercells had a substantial southern cell(s) in similar proximately, configuration, movement and life cycle timing. Maybe the new Vortex project will shed some light on this.

W.

 

[Joel Wright]

If you review the radar from that evening, you'll notice as the Woodward County supercell dies off, the Greensburg storm becomes more organized looking. Shortly after the Woodward storm is gone, the Greensburg storm goes crazy. At that point it is the southernmost storm, and has full access to the rich inflow from the southeast. An interesting question to me is why did the Woodward storm die, and allow the Greensburg storm to dominate? Perhaps if the Woodward storm could have maintained a bit longer, the Greensburg storm may not have become as strong, or strengthened after leaving Kiowa County.

I think another interesting feature from the Greensburg storm is the big left split that traveled from Ford County north-northwest into southern Nebraska in just a few hours. I'm not sure that it's a left split from the actual Greensburg storm, but the storm(s) that moved north-northwest away from the GB storm sure acted like a typical left split. Whether this has any importance whatsoever I'm not so sure, but I thought it was sort of interesting nevertheless.

And by the way, EXCELLENT report! Well done guys.

 

[Mike Smith]

There was a weak east-west wind shift that passed through the Wichita area around noon to 1pm. It became more or less stationary along U.S. 54 or just to the south during the afternoon.

It might have provided a localized wind shift the helped to increase the low-level helicity that evening south of Greensburg. Unfortunately, there were few clouds along the wind shift line west of Pratt so it is difficult to say where it was late in the day.

 

[Brian Stertz]

That was probably the cap/subsidence moving in behind the weak wave that rippled northward from W.OK/E. TX Panhandle into S.Kansas during the evening. This what likely caused that storm to eventually vaporize and that new supercell to blossom. It was amazing to watch and would be consistent with that type of ebb/flow. I know the synoptic warm front was over N.Kansas, but there was some sort of boundary across S.Kansas. Similar to what happened on the Protection KS tornado event back in April '07. Just not nearly on such a grand scale/magnitude.

 

[Lanny Dean]

If you review the radar from that evening, you'll notice as the Woodward County supercell dies off, the Greensburg storm becomes more organized looking. Shortly after the Woodward storm is gone, the Greensburg storm goes crazy. At that point it is the southernmost storm, and has full access to the rich inflow from the southeast. An interesting question to me is why did the Woodward storm die, and allow the Greensburg storm to dominate? Perhaps if the Woodward storm could have maintained a bit longer, the Greensburg storm may not have become as strong, or strengthened after leaving Kiowa County.

I think another interesting feature from the Greensburg storm is the big left split that traveled from Ford County north-northwest into southern Nebraska in just a few hours. I'm not sure that it's a left split from the actual Greensburg storm, but the storm(s) that moved north-northwest away from the GB storm sure acted like a typical left split. Whether this has any importance whatsoever I'm not so sure, but I thought it was sort of interesting nevertheless.

And by the way, EXCELLENT report! Well done guys.

John Davies did an excelent write up on the event and described a basic "cold pool" from the Woodword cell which obviously helped the Greensburg cell. Johns write up can be found here: http://members.cox.net/jondavies6/050407greensburg/050407greensburg.htm

There were actually 3 cells that started right on the OK border, as they moved North and East you had "left splits" all along and then the Southern cell became dominate. The last " left split" occured in just North of Protection KS, this is where Dick M, Darrin B, Randy and myself filmed two tornadoes on the ground at the same time. Visually the "left split" tornado(the tornado Northwest of Protection) seemed to be the stronger of the two....it was on the ground for quite a while and then dissipated as the GT touched down not long after. This "left split" cell began to race off North and I believe made it into Nebraska before it died out.

 

[Lanny Dean]

Maybe John could hit some highlights from his paper for us if he is not to busy?

 

[Mike Umscheid]

Greensburg initiation

There were actually 3 cells that started right on the OK border, as they moved North and East you had "left splits" all along and then the Southern cell became dominate. The last " left split" occured in just North of Protection KS, this is where Dick M, Darrin B, Randy and myself filmed two tornadoes on the ground at the same time. Visually the "left split" tornado(the tornado Northwest of Protection) seemed to be the stronger of the two....it was on the ground for quite a while and then dissipated as the GT touched down not long after. This "left split" cell began to race off North and I believe made it into Nebraska before it died out.

I thought I would comment a bit on the genesis of the Greensburg storm. Lanny is correct in noting the more or less 3 "congealed" cells centered along the KS-OK border shortly after 0030 UTC. As you step forward in time to 0045 UTC, these 3 cells are still loosely connected and are largely disorganized "garbage" storms across eastern/southeastern Clark County. Look to the southwest of this cluster at 0046 between Rosston and Englewood and just barely to the east of that line. There is a 40-45 dbz echo aloft around ~ 25 kft. This is the incipient updraft of the Greensburg storm and is trailing the "cluster of garbage" about 15 miles to the Southwest. This updraft appears on radar separated from the small left-moving garbage. It then "explodes" at 0100 just as it crosses into KS just east of Englewood. We (Les and I) don't believe these initial tornadoes were in any way associated with left-split storms. These tornadoes west and northwest of Protection were part of the east-west elongated zone of cyclonic shear (parent "right-member" cyclonic mesocyclone{s}). At 0115 to 0130, the supercell does begin to shed its own left-splits, racing northward toward Bucklin by 0145-0150. From that first Greensburg left-split, other splits are born and this cluster moves due north toward Hill City area by late in the evening. That initial 3-cluster garbage that developed first at 0030 died northeast of Russel, KS around 0345.

Mike U

 

[Jon Davies]

Maybe John could hit some highlights from his paper for us if he is not to busy?

I hadn't been following this thread, but someone called my attention to it this morning, and that Lanny had asked for some input.

The main elements I remember with 5/4/07 and the Greensburg storm were the very strong warm-moist advection that was taking place during the evening in the P28-DDC area in response to the intense upper system swinging out into the plains, in addition to normal diurnal backing of the dryline. The RUC did a terrible job picking up on this, and was more than 50-60 miles too far east with the dry line (see my case study write up on my web site). The "moist" front moving back northwest combined with the upper system had much to do with firing up the Greensburg storm and other cells in the same area, not unlike the Lubbock tornado setting back in 1970.

The net result was tremendous moisture influx into the Greensburg-Protection area in the 01z-03z time frame, along with strong backing winds and increase of 0-1 km SRH. This setup generated CAPE of 3500-4000 J/kg (near dark!), and 0-1 SRH that went from < 200 m2/s2 at 01z to around 400 m2/s2 by around 04z. Those are big _combinations_ that one doesn't see very often where convection develops. (The Parkersburg IA EF5 tornado on 5/25/08 appeared to have around 4000 J/kg CAPE and roughly 200 m2/s2 in its environment). An additional factor was the strong deep layer shear through midlevels, > 50 kts, helping to really intensify and support the Greenburg supercell, given the already great evening CAPE-SRH setting (not forecast that well by the models).

I don't remember writing about a "cold pool" from the Woodward cell on 5/4/07 that had any big effect on the formation of the Greensburg storm. The Woodward storm formed from strong surface heating at the dryline over ne TX/nw OK at the edge of the afternoon "cap", but seemed to get snuffed out during the evening as the cap built back in a little to the north, and the larger scale forcing and moisture transport shifted to the north-northwest.

I don't know alot about the radar evolution Mike talked about... the left splits probably had to do with the rightward stretched out hodograph above 1-2 km AGL. The lead "garbage cells" that Mike mentioned may have been associated with the rapid moisture return to the northwest... it could be that the initial Greensburg cell echo slightly southwest of these was in deeper moisture behind the northwestward push, and was able to take root better at the edge of the cap. But that's all speculation. What a storm!

Great write up, Mike and Les...

Jon Davies

http://www.jondavies.net

 

[Lanny Dean]

I thought I would comment a bit on the genesis of the Greensburg storm. Lanny is correct in noting the more or less 3 "congealed" cells centered along the KS-OK border shortly after 0030 UTC. As you step forward in time to 0045 UTC, these 3 cells are still loosely connected and are largely disorganized "garbage" storms across eastern/southeastern Clark County. Look to the southwest of this cluster at 0046 between Rosston and Englewood and just barely to the east of that line. There is a 40-45 dbz echo aloft around ~ 25 kft. This is the incipient updraft of the Greensburg storm and is trailing the "cluster of garbage" about 15 miles to the Southwest. This updraft appears on radar separated from the small left-moving garbage. It then "explodes" at 0100 just as it crosses into KS just east of Englewood. We (Les and I) don't believe these initial tornadoes were in any way associated with left-split storms. These tornadoes west and northwest of Protection were part of the east-west elongated zone of cyclonic shear (parent "right-member" cyclonic mesocyclone{s}). At 0115 to 0130, the supercell does begin to shed its own left-splits, racing northward toward Bucklin by 0145-0150. From that first Greensburg left-split, other splits are born and this cluster moves due north toward Hill City area by late in the evening. That initial 3-cluster garbage that developed first at 0030 died northeast of Russel, KS around 0345.

Mike U

Mike, here is a shot of the "garbage" south west of Protection sometime near 0045 I believe???. If I remember correctly, you issued the first tor warning not long after. Photo is looking Southwest from just south of Coldwater. KAKE called me right as you issued the warning and Jay wanted a phoner so I missed recording the first warning.
As far as the left splits and the first tornado(s) along and North/South of Hwy 160....are you saying they are from the elongated "garbage" or from the "parent" cell i.e. the Greensburg cell or what would become the Greensburg cell?

 

[Joel Wright]

The interaction of a supercell just west of Greensburg is quite interesting too. As the city of Greensburg was being impacted from the tornado, a second (separate) supercell over northeast Ford/southeast Hodgeman counties was merging with the GB supercell. It almost looks like the GB supercell "sucked" the other storm into it from the west. Right as the two supercells were merging the next big tornado touches down northeast of Greensburg.

It would almost appear there was sort of a fujiwara effect taking place. Maybe that's why there was such an extreme leftward turn to the GB tornado, which did almost a complete circle. Perhaps if the second (western) supercell hadn't have been there, or been further away the GB tornado would have missed the town to the east?

 

[Lanny Dean]

An Interesting artical with regards to this case study and paper has been written by the Wichita Eagle can be found here: http://www.kansas.com/news/story/638855.html

Might be of good reading in addition to the paper... of particular interest is Mike's comment regarding the comparison of the 5-3-99 event to the 5-4-07 event.

 

[Paul Knightley]

This is a very interesting thread IMO. I have been very interestest in the nomenclature of tornadoes for some time now. In fact some time ago I asked on this forum whether people thought there was some kind of intrinsic difference between what one might call a 'normal' looking tornado (e.g. some kind of elephant trunk type thing), and a big 'wedge' tornado.

I wonder whether many tornadoes form in the same way (i.e. stretching of vertical vorticity, irrespective of where that vorticity comes from), but then there future life and strength is then determined by rather subtle things, such as updraught strength/longevity, etc etc.

Perhaps in the case of a tornado on the scale of the May 4 2007 events, the whole mesocyclone effectively builds down to the surface over time. The vertical velocities within this 'ground mesocyclone' could be very high - high enough in fact to stretch the ambient high low-level vorticity into an extremely violent event. So does this then mean that this is still a tornado? Well, it is still a violently rotating column of air pendant from a cumuliform cloud, etc etc, but does seem to be in a completely different class to the 99% of other tornadoes.

 

[Shane Adams]

The definition of a tornado doesn't account for size, just the process, so of course the May 4 event consisted of "tornadoes" and not "something else".

There's already enough 'hybrid' tornado events out there that are cast aside by classification because "we don't know exactly what it is/how it happened"....there's no need to try and create new definitions for things.

May 4 was just an extreme case of the processes that come together to create tornadoes. I only wish the event had been diurnal, we'd have learned much more.

 

[Paul Knightley]

That's a fair point, Shane. There are, of course, various magnitudes of many things in nature. There must still be something rather special about the processes involved in situations like this, however.

Merry Xmas by the way!

 

[Shane Adams]

There are, of course, various magnitudes of many things in nature. There must still be something rather special about the processes involved in situations like this, however.

Agreed. I've never believed that tornadogensis is the same thing every time. I've seen too many weird and varying ways in which tornadoes develop to believe the process is set in stone. And like you mention, I also believe these varying processes have varying degrees of intensity or severity, depending on factors like the local environment and the upper air system itself.

Like I said, had 5-4-07 been a diurnal event, we could've gleaned so much more by comparing the radar information with visual observations. Not to mention the fact many lives would've been saved no doubt.