2016-05-08 EVENT: KS//OK/TX/MO

Apr 2, 2005
246
85
11
Norman, OK
www.chasetolive.com
This looks like a simple case where any consideration of the actual observations will suggest that the model moisture forecasts might be in question. As of 21z this afternoon, the NAM looks to be several degrees F too high across S TX with the dewpoints, as well as along the slow-moving front up toward nrn MO. I can't say for sure if it's the land surface model, the PBL scheme, or what's causing the excessive moistening, but it's been a problem since at least last year.

People tend to see what they want to see in the model forecasts, and something like the NAM now showing upper 60s can be enticing. Now, I'd like to know the last time anyone can recall upper 60s in wrn OK when the Tds were 4-8 F lower than that 24 hrs earlier across even the TX coast (aside from the lone 67 F at BRO)? That doesn't mean we won't see any tornado threat tomorrow, but I'll believe a 68 Td in OK when I actually see it.
 

Trey Thee

EF2
Mar 29, 2010
191
29
11
Tulsa metro
This looks like a simple case where any consideration of the actual observations will suggest that the model moisture forecasts might be in question. As of 21z this afternoon, the NAM looks to be several degrees F too high across S TX with the dewpoints, as well as along the slow-moving front up toward nrn MO. I can't say for sure if it's the land surface model, the PBL scheme, or what's causing the excessive moistening, but it's been a problem since at least last year.

People tend to see what they want to see in the model forecasts, and something like the NAM now showing upper 60s can be enticing. Now, I'd like to know the last time anyone can recall upper 60s in wrn OK when the Tds were 4-8 F lower than that 24 hrs earlier across even the TX coast (aside from the lone 67 F at BRO)? That doesn't mean we won't see any tornado threat tomorrow, but I'll believe a 68 Td in OK when I actually see it.
The NAM is consistent with its over forecast of dewpoints. My guess is we will see 62-65* dewpoints West of I35 tomorrow. That still results in a tor threat but I think it minimizes the risk of sig tor quite a bit.

I wouldn't be surprised to see the mdt peeled back to an enhanced unless the 00z runs continue to show substantial moisture. You would expect the NAM to start caving on moisture at 00z or 06z.

Sent from my SM-N920V using Tapatalk
 
Mar 2, 2009
124
52
11
Norman, OK
Agree with the above concerns on moisture. I'll add the WPC surface analysis from 00Z, which is ... not good if you'd like to see mid-60's F dewpoints in western OK. I'm thinking 60-63 F is most realistic, maybe with a few 64 F thrown in there.

upload_2016-5-7_23-45-50.png

However, I'll point out that even if you modify the NAM forecast soundings for low 60's F surface dewpoints, the LCL's are still within the inter-quartile range for tornadoes. The sounding below is the 08/00Z NAM for Hutchinson, KS modified for 61 F surface Td's, which gives an MLLCL of 1015 m, which is right about the median for significant tornadoes in that area.
upload_2016-5-7_23-51-41.png

The other thing you have to worry about is the lackluster moisture contributing to a cap bust in south central KS and OK, which becomes a very real possibility with the most recent NAM. Granted, Hutchinson is a little far east from where I'd like, but there's a hole in the Bufkit profiles in SHARPpy where I'd like to click. I'll be monitoring the moisture trends in the morning, and I'm considering heading out. Would probably target Alva, OK given the information right now.
 

Bryan Smith

Enthusiast
Sep 26, 2007
6
3
0
Norman/Witcher, OK
I'll be targeting Harper, Woodward, and Woods counties. Exit region of the H5 speed max, ~-13 @H5, and lower 60s dewpoints. For a good tornado, I would prefer a 13 g/kg lowest 100mb mean mixing ratio. This is based on sounding thermodynamic structures I looked at a few days ago for this event. I don't think we'll get a 13', maybe if we're lucky a 12' is potentially in the cards. The 12z FWD raob showed an 11.7 g/kg mean w.
 
May 6, 2005
240
104
11
Moore, OK
I saved a 3z surface obs last night and compared to 13z. Tds across central and northen texas increased by 5-6° F. 4km NAM is still showing mid 60s for oklahoma during afternoon. Per mesonet we are already seeing 60-62 so a 65 seems possible. No idea how deep the moisture will be though.

On an upbeat wish cast note... 4km NAM firing cells along the DL near Ok/Tx border and sending ENE.

Took this sounding on SE side of storm near Fort Cobb.

01d5cf03d6530ae94fd4d341540c90b7.png

5825993e40084b631588838e6198d234.png
 
  • Like
Reactions: James Gustina

James Gustina

Supporter
Mar 9, 2010
670
321
11
28
Dallas, TX
www.thunderingskies.blogspot.com
I think we might see scattered mid-60s by the evening but we're still right on track for the day. The FWD displayed a decently deep moist layer, larger than BRO or CRP this morning. Assuming this lead impulse leaves behind some residual cloud cover to start, our moisture might not crater out as badly with the higher SFC temps needed to coincide with our jet maxima to break the cap.

257607d668a2f7fb9fe35cad86f7c31f.jpg

Sent from my XT1080 using Stormtrack mobile app
 
  • Like
Reactions: Robert Forry

Zach

Enthusiast
May 4, 2012
9
3
0
Not chasing but looks like Altus/Lawton area might produce. HRRR been somewhat consistent with this supercell complex. Some agreement with the HRRR and NSSLWRF. Dews are near 65F and I know there was some worry they would get into the mid 60s. Decent clearing too with HRRR suggesting 3000 j/kg CAPE. Veering profiles look decent. Good luck to those that are chasing down in SW OK.
 
  • Like
Reactions: Chris Frost
May 18, 2012
292
151
11
Gaines, MI
I'd be curious to see you guys feedback on this. I know there were some tornadic storms around in NC KS and SW OK, but after taking a look at the things prior to initiation (18Z soundings at OUN and DDC) it looks like there were a few things acting against the atmosphere to allow for a greater production of tornadoes. Initially, in reviewing the surface data, it looked to me like the surface flow, while backed, was mostly anemic. Additionally, I thought the winds and temps at H850 were a little on the weak side (20 kts). But after looking at those specially launched soundings, it looks like this was more of a VBV issue (S Shaped hodos) again, acting against tornadogenesis.

I ended up foregoing the 14 plus hour drive from Michigan initially because of my concerns with deep layer moisture, but after watching the SPC mesoanalysis page on and off throughout the day, I quickly realized that that wasn't going to be an issue, in turn causing me to question my decision.

Below are the 18Z soundings from OUN and DDC:
OUN:
OUN.PNG

DDC:
DDC.PNG

Both soundings show a a pretty solid cap, we these were at 1PM locally.
What are your thoughts?
 
Dec 9, 2003
4,839
121
11
Oklahoma
... I ended up foregoing the 14 plus hour drive from Michigan initially because of my concerns with deep layer moisture, but after watching the SPC mesoanalysis page on and off throughout the day, I quickly realized that that wasn't going to be an issue, in turn causing me to question my decision.

Both soundings show a a pretty solid cap, we these were at 1PM locally.
What are your thoughts?
I can't chase as much as I used to owing for family reasons, and I opted to sit this one out despite being in the Oklahoma City area. By yesterday morning, model forecast wind profiles were going down the tubes (no pun intended), which played heavily in my decision to "save up" my chase days (hey, I'd have to use a lot of familial political capital to ditch my wife and son on mother's day after being out of town for a week ;P ).

The shear profiles were unremarkable despite, as has been the case several times so far this spring, earlier model forecasts of strong low-level shear. A weakness in the flow around 700-750 mb and, in perhaps a few small area, a veer-back-veer vertical wind profile yielded S-shaped hodographs that looked unsupportive of significant tornadic supercells. [As Jeff notes below, you don't necessarily need veer-back-veer to get an pseudo-S-shaped hodograph. In this case, the weakening of the flow from ~850 to ~700 mb was the nasty part, IMO). Indeed, in the entire risk area in Oklahoma, there was only a single, briefly tornadic supercell. For the most part, left-moving, anticyclonic supercells were preferred over the right-moving, cyclonic supercells; there looked to have been more negative SRH for left movers than there was positive SRH for right movers. Looking at the hodographs across central and western Oklahoma, this is not surprising. The zone of strong destabilization and weakened CINH was also not particularly large owing to the cloud cover that persisted across much of central Oklahoma.

@Jeff Duda posted the following on Facebook (I don't think he'll mind me posting it here since it's weather related), which is the hodograph based on the FDR VWP/VAD valid a little after 22z. "Blech" in terms of a significant tornado threat... (Click any of these thumbnails to enlarge)
13178701_10103762017324069_6490090126955183529_n.jpg

Below is the 00z OUN sounding, which wasn't that far from the briefly tornadic supercell east of Lawton. Note that there is only cross-wise vorticity above 1 km AGL. As was the case a couple of weeks ago, the 0-1 km SRH wasn't too bad, but the 0-3 km SRH was on the low end for supercells. Not too that the cap was going to quickly devour any surface-based storm that approached the I35 corridor, which is pretty much what happened (note convective temperature of 89 F where surface observations showed 2 m AGL temperatures in the low 70s).
20160509_00z_OUN_sounding.PNG
If we go into model analyses to look a little west of the I35 corridor to get closer to the storms, we see hodographs that are no better... Example below from the 00z NAM analysis for a location near Lawton (first image) and somewhere between Alva and Coldwater just north of the OK/KS border (second image):
nam_2016050900_000_34.84--98.05.png

nam_2016050900_000_36.28--97.92.png
 
Last edited:
  • Like
Reactions: Robert Forry

Jeff Duda

EF6+, PhD
Staff member
Supporter
Oct 7, 2008
3,535
2,585
21
Denver, CO
www.meteor.iastate.edu
Robert,
IMO, the failure mode of the day was actually early initiation of storms coupled with 'S'-shaped hodographs. There wasn't much actual veer-backing going on, as 500 mb winds were past 225 degrees. However, you can still get the "S" shape without veer-backing.

What I mean by early initiation is that models (especially the flavors of the NAM) suggested it wouldn't be until close to 00Z that the wind profile would become very favorable for bigtime storms and tornadoes. A lot of 21Z soundings across the region had 'S' shapes. See the below VAD derived from KFDR just after initiation:

KFDR_VWP_2102.jpg

Note the '+' indicating right-mover storm motion at the time - it's on the wrong side of the hodograph if you're looking for cyclonic rotation! The left-mover storm motion was around 205 deg. at 45 kts. Anyway, this hodograph was much more friendly for left-movers, and those are the ones that tended to survive. If storms had fired a few hours later and the wind profile had matured into the classic hooked shape (it did appear to get better towards 00Z), I bet this event would've resulted in quite a few more tornadoes. A look at mesoanalysis yesterday showed that 700 mb flow was less veered once you got to the TX-OK border and westward. And the dryline had begun to retreat as early as 21Z from what I judged by watching a radar fine line from KFDR. So later initiation would've also meant storms would have developed a little further west, closer to the region with a better wind profile (and with overall better wind profiles warm-sector wide). The Lawton storm was able to produce a tornado because it turned hard enough to the right to make the storm motion vector on the correct side of the hodograph (it was moving basically due east once it got organized).

Regarding the 18Z soundings you posted: Norman was generally too far east to sample the true environment these storms were living on, so I wouldn't put much further consideration into the OUN sounding. Also, there was a strong cap on the 00Z OUN sounding, stronger than forecast. However, widespread clouds during the day across C OK also limited destabilization and narrowed the warm sector, which is why the Lawton storm died as it moved into Stephens County. The veer-back in the 18Z DDC sounding is not all that bad even though it was low in the profile. However, the hodograph on the 00Z DDC sounding was much improved and would've supported intense storms had the thermodynamics worked out.

Overall the moisture ended up not being an issue at all, which surprised me. Surface flow was also not a problem: winds of 10-15 kts out of the SSE and even SE/ESE were common throughout Oklahoma during the afternoon.
 
Last edited:
Jan 7, 2006
596
828
21
USA
www.skyinmotion.com
Excellent discussion, Jeff. I agree completely. I was relatively confident we'd get a few nice tornadoes on the dryline yesterday up until mid-afternoon or so, because as you mentioned, earlier guidance had suggested initiation would be well timed around 22-23z. Instead, the main round of initiation occurred between 20-21z, something that CAMs had already begun to suggest by the night prior. This timing in relation to SRH ramping up was a subtle detail probably overlooked by many at the time, including myself.

Had these initial storms held off until even 90 minutes later, it might have been a much different chase day. It's hard to say, because soundings/VWP and storm behavior yesterday suggested that the wind profile in W OK may not have been particularly favorable up until right at (or even a tad after) 00z. As the first storms were initiating around 3:30pm, I was looking at 850 mb wind forecasts from the HRRR and noticed that it consistently failed to ramp up the LLJ beyond 15-20 kt until right at 00z, which was immediately discouraging. Even at 23z, the HRRR showed very anemic 0-1 km AGL shear.

Due to the capping issues over much of OK and lingering cloud cover that held temperatures E of US-281 in the 70s, there was a relatively narrow east-west corridor for storms to intensify and mature. The strong midlevel flow and associated fast easterly storm motions (for right movers) compounded this problem. However, that corridor from Frederick to Cordell to Seiling eastward to US-281 might have been able to produce some nice tornadoes had mature storms entered there at 00z, instead of 21-22z. And not to be too much of a downer, but this more favorable timing theoretically could've had a positive feedback, where one or two right-moving supercells established a very strong internal VPPGF to sustain themselves further E into the more capped environment. The fact that a few early afternoon runs of the HRRRX depicted a dominant supercell surviving past I-35 in the evening with high UH is some evidence for this.
 
Jul 5, 2009
1,367
1,433
21
Newtown, Pennsylvania
I was surprised that there were more TOR warnings and reports in northwestern KS. IIRC, surface temps and dewpoints were much lower there. What was more favorable in that region?
 

Jeff Duda

EF6+, PhD
Staff member
Supporter
Oct 7, 2008
3,535
2,585
21
Denver, CO
www.meteor.iastate.edu
I was surprised that there were more TOR warnings and reports in northwestern KS. IIRC, surface temps and dewpoints were much lower there. What was more favorable in that region?
Hodographs lacked the S-shape up there. Thermodynamics were the 'scaled' same (may have been somewhat cooler and drier at the surface, but it was also cooler aloft, so lapse rates were still pretty good and MLCAPE was also as high as it was in OK).