29 April 2009 Cedar Hill, TX Tornadoes

[Matthew Grzych]

I've started to put together a small case study for this event on my blog:

http://wxforecastnow.com/wxblog/2009/10/21/29-april-2009-cedar-hill-tx-tornadoes-case-study/

This was quite an interesting event and tornadogenesis appeared to unfold in a not-so-classic manner. We were tracking an isolated supercell from Plainview that went up on the dryline. Initially it was high based and outflow dominant. We observed at least 2 cycles that had strong, cold RFD surges and blowing dust. I believe it was the 3rd cycle that produced a strong and persistent anticyclonic flare on the south side of the RFD boundary. We rushed north and east to get in position anticipating a new developement when, what seemed to be very rapdily, the storm's base lowered dramatically and a clear slot developed way off to our southwest (maybe 3-4 miles). In the matter of minutes a large cone funnel developed. I haven't yet done a time series radar analysis, but this just seemed unconventional. Visually it almost appeared that the remnant RFD boundary surged and interacted with the pre-existing outflow boundary causing new rapid development well south of where we anticipated. The strange thing is that we were tracking that area of anticyclonic rotation in the vicinity of where the tornado developed. Any thoughts? Did anyone else see what I believe I saw?

 

[Adam Lucio]

I didn't chase that event but I was glued to the radar much of the day. I think intersecting boundaries played a key role in tornadogenesis this day. I saved an image here that clearly shows the boundaries intersecting and the locations of the tornado reports.

My non-expert guess would be the low level helicity was greatly enhanced by the interaction of the 2 boundaries.

 

[Matthew Grzych]

I definitely believe that the boundaries played a huge role in tornadogenesis on this day. The dryline / outflow boundary triple point sure does show up nicely on radar. The actual intersection of the boundaries may not play a big role in increasing low-level helicity, but helicity was surely greater on the north side of the OFB with greater backed low level flow.

I guess what I need to do to help answer my questions is to plot a time series of radar images with our position plotted to get a better feel. I think what threw me off was that the 4th (I believe) RFD cycle induced tornadogenesis in a not-so-classic position relative to where one would expect. Typically in cyclic supercells, after a mesocyclone occludes, the new meso/wall cloud forms east or northeast of the old meso where the previous RFD boundary pushes ahead. In this case the new meso and wall cloud formed (what appeared to be) way off to the south or southwest. Maybe this was a totally new updraft that formed as the old RFDB interacted with the OFB. But we did race east to get ahead of surging RFD, so maybe my orientation got screwed up.

 

[David Drummond]

Upslope flow of the caprock no doubt had a role as well. The edge of it was about 15 miles or so from the initial tornadogenesis areas. I've seen this many times out here where a storm would spin up tornadoes as it approaches and moves across the caprock escarpment.

 

[Scott Weberpal]

Upslope flow in itself does not induce vorticity, its not different than winds over a flat surface, except they are going over and significant vertical obstruction. I imagine tornadogenesis near the edge of the caprock has more to do with the storm being able to feed off deeper moisture east of the caprock, and potential mesovorticity from canyon interactions.

Upslope flow can aid in storm initiation, but I've never heard of it aiding in tornadogenesis - but I could be wrong.

 

[Matt Hughes]

We were on the western edge of the Caprock when the first tornado formed. "Caprock Magic" was definitely in play on this day. The tornadoes were boundry-related as well. In fact, if I had one definition for the 2009 tornado season, it would be "The Year of the Boundaries" (or "Needles in Haystacks" lol). Most of the tornadoes in the alley this year were in direct relation to crossing boundaries.

 

[David Drummond]

I don't have deep technical knowledge of the process Scott, but after a couple of decades of seeing this happen, I do have a theory. Part of it has to do with what you mentioned, access to deeper moisture. More to the point, that deeper moisture hitting an almost vertical wall and being forced upward thereby helping to force-feed any storms.

If you note some of the canyons there form nice V shapes as they terminate at the edge of the caprock, thereby acting like a funnel if you will. I wouldn't be surprised to learn if some sort of turning in induced by this as well, but my theory in that regard is that the caprock almost acts like a boundary of it's own, and thereby inducing some of the vorticity we more commonly see on a traditional "air boundary".

That's my theory anyway. I don't know if anyone has done any in depth studies on it, but I've seen it happen way too many times for it to just be a coincidence.

 

[Scott Weberpal]

If you note some of the canyons there form nice V shapes as they terminate at the edge of the caprock, thereby acting like a funnel if you will. I wouldn't be surprised to learn if some sort of turning in induced by this as well, but my theory in that regard is that the caprock almost acts like a boundary of it's own, and thereby inducing some of the vorticity we more commonly see on a traditional "air boundary".

I certainly don't disagree that there is something happening with the caprock, and I trust your observations as you've probably seen as many storms on the caprock as any other active chaser (except maybe an old fart or two such as Bary Nusz).

What you mentioned about V-shapes of the canyons that terminate at the edge of the caprock may have a similar effect as what the DCVZ does along the front range, only on a smaller scale. I don't know a ton about the DCVZ, but in a paper by Rasmussen and Pietrycha, the DCVZ form from a process that may be similar to what happens at the edge of the caprock.

"[SIZE=-1]The Denver Convergence-Vorticity Zone (DCVZ) develops as southeast low-level flow interacts with an east-west ridge known as the Palmer Divide (located between Denver and Colorado Springs), and the north-south axis of the Rocky Mountains Front Range."[/SIZE]

Someone needs to do some mesonet transects of the caprock during a svr day. Perhaps it's already been done?

 

[Jeff Duda]

Numerous studies of the dryline in TX, OK, and KS have been done involving mobile mesonet observations, but I don't think any were in the Caprock.

 

[Brian Stertz]

Agree with David's force feeding theory the deep moisture up the Caprock when those winds are east or southeast. It seems to give them that extra degree of intensity as the supercells move off the dryline and towards the edge of the escarpment. The TX Panhandle and South Plains area is a case study always in progress. Just when you think you've seen it all...go there. You get the best of all worlds seemingly colliding there. Steep lapse rates...returning warm fronts, strong pressure falls, strong shear, surging Gulf Moisture, upslope, dryline...and so much more. There is truly something to be said about Texas Panhandle Magic.

 

[Wesley Luginbyhl]

The tornado did seem to form in a sort-of uncommon area to the SW of the old circulation as Matthew said. I was on the storm when it first formed as a cluster west of Plainview and stayed with it through each cycle waiting for it to hit the boundary. I do remember dew points were up to 67-68 range just off the caprock behind the boundary, where at Plainview they were only about 60. The cloud base heights before and after the boundary were obvious.

After the second cycle, I had lost faith the storm was produce even after I thought it crossed the boundary. I started to race south on 207 after turning off of 2286 hoping to catch up to the eastern storms. Just before I got to highway 97 (East to Cedar Hill) I noticed a small funnel right above me along what I though was the RFD from the 2nd cycle. The funnel caught my attention enough to drive down 97 and stop 2 1/2 miles west of Cedar Hill.

I recorded a zoomed out view of the storm for about 5 minutes leading up the Cedar Hill tornado coming down. You can see a large funnel coming from way above the cloud base. The updraft gets wrapped around the funnel and the tornado dropped immeadiately afterwards. You can see this process in my video from about 1:00 to 1:40 - Video .

I have been trying to figure what happen with that storm on the April 29th ever since then. If I didn't notice that funnel as I was heading south I would have likely missed the whole thing. That was definitely the quickest I have seen a storm pick up intensity visually and produce from where it was at a few minutes before.

Below is a photo a few minutes before the tornado came down. You can see the funnel from the top right of the photo down to the center. I am sure it is directly link the the boundary interaction, exactly how, I have no idea.

 

[Jay McCoy]

I was well ahead of the storm for a while and up until about 5-10 minutes before it produced the tornado it was very outflow dominate with numerous gustnados, lost of dirt being thrown out ahead of it, and almost a shelfcloud appearance and cold winds. I almost gave up on it a couple times but since it was still the best storm around I stayed with it hoping it would get near the caprock. I didnt feel any inflow winds until I was WNW of Cedar Hill. My guess is thats where I crossed that boundary.

 

[Greg Blumberg]

Just from first glance, the radar image posted strongly reminds me of the Pampa, TX storm of this year. I believe it happened on May 14th, if I remember correctly. The same type of interaction of the two boundaries did produce a tornado on that storm too. I wonder if the cells on that day had similar cycling patterns to this scenario due to the boundary interactions.

I imagine that up-slope flow can help in tornadogenesis. First of all, the ground would be higher up, allowing the tornado a shorter distance to reach the ground. Secondly, it would help in the drawing of surface parcels if they were much more easier to draw up into the storm.

The question with a theory of an enhanced updraft/rotation by upslope flow is figuring out whether or not the speed of the upslope flow would actually have an affect on the speed of the updraft. In short, what is the ratio of the updraft's speed to the upslope flow's speed?

 

[Wesley Luginbyhl]

I was on the Pampa storm from when it first went up as well. That storm on the 15th was a different type of setup. That storm formed north of Pampa at the intersection of the cold front and dryline. The 500mb winds were about 25kts if I remember right, so the storm ended up tracking almost due south once it finally started to move off the boundary. I dont remember the Pampa storm having more than one cycle, more of just one big cycle with multiple tornadoes. The storm that day was being force feed a storng low level jet, which definitely help it produce 4 tornadoes.

The storm on the April 29th was helped by moving east across crossing the boundary, which happen to lay near the Caprock. Panhandle magic at its best.