8/19/09 DISC: IL/MN/OK/KS

[Lanny Dean]

EDIT: sorry was posting the sane time Jon was., didnt know he and Mike had posted.

Granted, there were storms 60-100 miles away, but this is not that unusual. It happens in the middle of the traditional storm season too. The question should become: Why did the storms initiate in SE Kansas and not NW Oklahoma or northcentral Oklahoma?

Jeff, Mike....first of all, a storm did fire in that area near Enid. This WAS a supercell and actually had some decent look to it. That said, the further it moved east, the more it died. Subsidence probably is somewhat of an issue here as it crossed "over" into the "warm side" of the OFB??? I did note that there was some "decent" warming that was clear in the 12z OUN but I personally do not think this was an issue in the area from Stillwater to basically Enid/Alva, at least I did not see that in the LMN ROAB.
I am curious if the PV had any sort of roll in this "waste of a day".
I first noticed the PV shortly before 4:00 CDT...hmmmmm???

It is interesting that Mike mentioned Greensburg because believe it or not, yesterdays parameters greatly exceeded 5-4-07!

8/19/09
5700j/kg, 350+ m2/s2, LIs -13 and an EHI value of 10 with STP of 8! those parmeters were insane! Just to put that into perspective Greensburg was (respectivly)
5/4/07
ML CAPE 3830j/kg
0-1km EHI 4.9
205 m2/s2
LIs -6 -8
STP 5.9 - 6.2 (adjusted/modified)

The big difference IMO is the weakining near 700mb not to mention the temp. I believe we were looking at nearly 13 degrees C. I would be lying if I told you I knew totally what was going on and why things did not happen. I don't have a clue.
IMO, things should have gone nuts and with parameters like that, we could have seen a violent/long track tornado!

 

[mike scantlin]

After watching/editing the timelapse of the "death" of the Enid cell, some interesting observations occurred. Even after the updraft 'detached' from the downdraft region, it continued to efficiently tower and spin for a good 30-45 minutes before completely petering out. I will post a link to the timelapse, but here's a terrible MSpaint pano showing how far the updraft is from the core. ***EDIT*** also, this is looking EAST! not north. This is the awkward orientation I am talking about. As the storm matured, the updraft got further and further behind the downdraft.

 

[Chad Cowan]

Did anyone else notice the analogs with this sounding?? Wow. They're all significant, but one analog stands out more than the others: May 3, 1999 at 23z in Oklahoma City.

That was quite an environment yesterday, especially for August! My best guess is that the late shortwave arrival was the limiting factor. That combined with some subsidence in the wake of the initial shortwave is what probably kept widespread, deep convection from developing, even with very little CINH.

 

[Benjamin Rock]

The NWS/Chicago confirmed an EF-1 in Elburn, Illinois Kane County. I saw the tornado as it became rain wrapped and moved into the corn fields east of me.

http://www.crh.noaa.gov/news/display_cmsstory.php?wfo=lot&storyid=30579&source=0

 

[Steve Worthington]

I`m not going to beat a dead horse or myself up any longer over yesterdays chase but from what we saw it looks as if nothing in Okla couldnt get above 16000' with out the top being pushed off. We started chasing at the KS, Okla state line on Hwy 8 running south down to Hwy 64, chasing a little cell that only topped at 16400 from what we could see on GR3. This cell looked like it was going to take off but as it reached the Biiling Okla area it was dieing at a fast rate. The cell did have some rotation but never looked like it was going to do anything. It was a very small "single cell" ((The above cell the Mike Scantlin posted))

After reaching I-35 we headed north to chase the Kansas cell where all the action was. As bad as I hate chasing Estern Kansas the towers were pushing up and the cell was producing. We saw a few funnels in Kansas but nothing to report.
My question is WHY could the cells be so close together but the south cells just couldnt get any higher? Every thing was in place in western Okla thus the reason we changed our target more west of I-35 then we had planed early in the day.

 

[Lanny Dean]

I`m not going to beat a dead horse or myself up any longer over yesterdays chase but from what we saw it looks as if nothing in Okla couldnt get above 16000' with out the top being pushed off. We started chasing at the KS, Okla state line on Hwy 8 running south down to Hwy 64, chasing a little cell that only topped at 16400 from what we could see on GR3. This cell looked like it was going to take off but as it reached the Biiling Okla area it was dieing at a fast rate. The cell did have some rotation but never looked like it was going to do anything. It was a very small "single cell" and not real sure I`d even call it a "super cell" ((The above cell the Mike Scantlin posted))
My question is WHY could the cells be so close together but the south cells just couldnt get any higher? Every thing was in place in western Okla thus the reason we changed our target more west of I-35 then we had planed early in the day.

Steve, to be honest, the cell that went up near Enid was well above 16,000kft. I do not believe that is an issue as to why things didn't happen....Maybe your Level 3 didn't update? Or maybe you didnt see it when it was closer to Enid?

As far as nothing going yesterday, my vote is the the issue with subsidence after the "first" shortwave....either way, what a wasted day IMO!

 

[Steve Worthington]

We were having trouble getting up-dates from KICT so we changed to an Okla site that showed 24000` but I just wasn't buying it. We ran that cell from Nash to near I-35.

Either way, it was really a let down but our play on Kansas made up for it
Video and pics posted on TVN, Could have been better but It was better then nothing....lol

 

[Dustin Wilcox]

I gave this a little more thought overnight. Take a look at the tornado parameters on the LMN sounding (above). They are higher than the respective values for Greensburg. These values were consistent with what the 18Z RUC from yesterday were predicting for 23Z, so they were not out of line. Perhaps there was some mid-level warming in the area the balloon did not sample, but it is not like a cap killed convection. There were severe thunderstorms 60 miles northeast of the sounding in the same air mass.

We are, understandably, picking apart surface obs and layers that dried (for example) but the fact is that these values of "off the chart" -- meaning if the sounding did not perfectly capture the air mass, the lowering of the SIG TOR to, say, a value of 6 would still be around the value of Greensburg.

So, given LMN sounding was consistent with that the RUC was forecasting for northern Oklahoma / southern Kansas, I think it would be more worthwhile to ask "what is wrong with the index" (and what is wrong with our understanding of the tornadic environment) more than "what is wrong with the sounding." Clearly, there is something eluding us.

From many personal experiences, I put little to no value in Sig Tornado, (and many other composite indices for that matter) come the summer months. It's nothing to see "end of the world" parameters across NE & IA what seems like almost every summer day, with the result a nice setting sun beneath clear blue skies. It took learning the hard way, but now a Sig Tornado of 10+ and EHI's of 15+ mean very little to me in July and August; in fact anymore it seems quite the contrary, outrageous parameters seem to indicate an environment which favors hot summer sun. I'm not sure what specifically many of those parameters define, for example if the Sig Tornado is simply indicating the chance at tornadoes given a supercell, it might not need all that much modification; however if it is defined as indicating the overall potential of tornadoes, then agreed, it's got the summer time blues and needs modified.

We were having trouble getting up-dates from KICT so we changed to an Okla site that showed 24000` but I just wasn't buying it.

FWIW, Vance, OUN and ICT all had it up near 40K and VIL at 35+ for at least a shortime...

 

[Mike Smith]

This is an excellent discussion everyone, thanks for the comments. I'd like to make a few more.

OK, so yesterday, from an "index" POV, was more ominous than Greensburg and an "significant" analogue to 5/3/99 (a prolific tornado day over the same region) and still nothing happened. If no storms had developed, we could write it off as a cap bust, etc. But not only did storms develop, they were severe and tornado watch/warned. This is, to me, what makes this non-event so important. I think our field is doing a great job warning of tornadoes, but the FAR continues stubbornly too high. What can we learn?

I made my earlier comments from an a priori perspective meaning "before the event occurred." Yes, we can go back now and point to parameters and observations that looked "off" afterward and those are very useful (I have a couple of new insights -- thanks) and how science works. But, by any reasonable a priori perspective, yesterday should have been a "tornado day."

It is fine to say I/we "don't have faith in the indices" but they are what they are and failed miserably as they seem to quite often (even in spring) by my unofficial tally. So, is there something wrong with the input (i.e., consistently bad input from the RUC) or the index may not be a reliable proxy for tornado likelihood? If the latter, perhaps it should be eliminated.

It wasn't a single sounding that indicated "tornado day" yesterday. The 18Z LMN sounding looked very good and the forecast plots of the RUC model output via TwisterData looked great: high EHI, high CAPE, high helicity, etc. over northern OK and southern KS. I don't believe it is "fair" to focus "blame" on the LMN sounding. The 00Z LMN sounding looked like one would have expected based on the 23Z forecast from the 18Z RUC.

With regard to convergence, there wasn't a great deal of surface convergence on 5/3/99 in central Oklahoma or south central Kansas. So, the fact that there wasn't strong convergence yesterday shouldn't have been a deal-breaker.

Jon, I agree completely with your comment about "not knowing how to measure" inhibition and subsidence. This better states what I said in my earlier post about something "eluding" us. That something or (more likely) somethings seem to have bitten us numerous times in the 2009 storm season. Perhaps we (and I certainly include myself) don't understand the pre-tornado environment as well as we think we do. And, if that is the case, what implications does it have for "warn on forecast?"

 

[Dustin Wilcox]

Perhaps the forecasters at the SPC for yesterday's event should chime in. While the indices and it appears many of the chasers forecast was for a worthwhile tornado event, keep in mind despite being aggressive with the red box, they for the most part ignored (or as is required took other things into consideration) the composite indices and only had the area under a relatively low 5% tornado risk...

 

[Mike Hollingshead]

All I know is it never seems you see anything good under that region of the upper jet. Seemed like that area was under the left entrance region of that lead jet streak. I remembered looking at the day and being annoyed by where the plains targets fell in relation to that jet. The further west one got there the more they seemed to be in that crappy area. But I'm no expert on jet dynamics, etc. Perhaps the intense part diving south behind the trough was already effecting that area in a good way by then(doesn't seem so though).

 

[Rich Thompson]

Ok, I'll jump in the fray. I worked the day shift at SPC, was responsible for the 1630z convective outlook, as well as the Tornado Watch for OK/KS.

The parameter discussion highlights some misconceptions about the intent of the parameters. STP is simply meant to highlight where there is sufficient "overlap" of known supercell tornado ingredients (e.g., 0-6 km bulk shear, 0-1 km SRH, LCL height, MLCAPE, etc.). The "effective layer" version uses the fancy shear layers that depend on the thermodynamic profile (plus CIN), while the "fixed layer" version is a little simpler. Both composite parameters have shown to be better discriminators between significant tornadic and nontornadic supercells than any of the individual ingredients. STP is essentially the EHI with the inclusion of deeper layer shear, LCL height, and CIN, hence the similar values when the latter three "ingredients" are not limiting factors.

It's easy to find cases where STP was large (and there were no storms), or the CIN was large and STP (with CIN) was small, but tornadoes occurred. These parameters are not magic bullets, and they can't possibly be expected to summarize the complexity of the storm environment in one number. However, large parameter values *do* suggest volatile ingredient combinations that warrant a close look. If you choose not to "like" STP, that's your choice. Still, don't act as if you can reach a different conclusion looking at the same information that's included in STP!

To answer another question posed earlier in the thread, the STP represents the *conditional* threat for significant tornadoes, *given* the presence of a supercell. It is not meant to explicitly forecast storm initiation - any expectation along those lines is asking for trouble.

Regarding yesterday, we did go with lower probabilities than the parameter values themselves might suggest. Part of the reason was uncertainty regarding storm initiation and what form of the outflow boundary would be left by late afternoon. I fully realize that storm motion with respect to boundary orientation is critical, especially early on when the storms are just forming. Will the early updrafts along the boundary "detrain" from the zone of ascent before reaching an LFC? It's awfully hard to know parcel trajectories along a mesoscale boundary, so I don't pretend to have 100% confidence in any forecast involving storm initiation and a boundary.

In the end, I was too concerned with storm potential to sit around and be completely sure that storms would form before issuing a watch for OK/KS. It's fun to speculate about what went wrong. However, until I see something I can use that will give me supreme confidence to lay off that environment, or I get a call from somebody with a better forecast *and* an explanation in real time, I'll continue to have to issue the bust watch you saw yesterday.

 

[Chad Cowan]

All I know is it never seems you see anything good under that region of the upper jet. Seemed like that area was under the left entrance region of that lead jet streak. I remembered looking at the day and being annoyed by where the plains targets fell in relation to that jet. The further west one got there the more they seemed to be in that crappy area. But I'm no expert on jet dynamics, etc. Perhaps the intense part diving south behind the trough was already effecting that area in a good way by then(doesn't seem so though).

Exactly. Tuesday's 18z NAM would have been the ideal outcome, keeping the early morning MCS north of I-70 and laying down that OFB right in the path of the left-front quadrant of that powerful jet streak. Late Tuesday night I thought there was a decent chance the OFB would make it this far north because even though the MCS extended into OK, it was moving east much quicker than I expected. But the swrly LLJ ramped up and around 6z an E-W band of training elevated storms fired in sw KS, behind the MCS, and I knew we were in trouble. That stuff crept east while backbuilding and kept the best thermodynamics far south of the best kinematics.

EDIT: I mixed up entrance and exit terminology, Mike H has it right. I'm a noob.

From: http://www.theweatherprediction.com/charts/300/

 

[Jeff Duda]

No offense intended, Chad, but the 300 mb height chart was rather unexemplary to show your case. Looking at it closely, I agree with Mike: with the exception of far eastern/southeastern OK, the entire state was under BOTH the left entrance and right exit regions of the jet streaks around, thus implying quite a bit of downward motion under the kissing jets there.

 

[Mike Smith]

It's easy to find cases where STP was large (and there were no storms), or the CIN was large and STP (with CIN) was small, but tornadoes occurred.

To answer another question posed earlier in the thread, the STP represents the *conditional* threat for significant tornadoes, *given* the presence of a supercell. It is not meant to explicitly forecast storm initiation - any expectation along those lines is asking for trouble.

But that is the point. There were supercells in Kansas. Yet, no tornadoes with a sky-high index value. I can't remember a day without tornadoes when things looked so good and supercells formed.

None of this should be taken personally or as a criticism of SPC.