Soundings don't tell the whole story (5/18/11 Oklahoma Bust)

[Jeff Snyder]

I just wanted to point out that the 00z OUN sounding appears to be another frustrating case of parcel theory failure. I was assuming that this was a true cap bust, and I was expecting to see something like -50 or -100 j/kg MLCIN. Instead, the Norman sounding has -10 j/kg MLCIN, -5 j/kg SBCIN, with 2700 j/kg MLCAPE. The low-level shear profile, as expected, is pretty fantastic, with ~300 0-1km SRH, ~30 kts 0-1 km "shear", ~650 m2/s2 0-3km SRH, and 46 kts 0-6 km "shear". There is a weakness near 550 mb (15 kts), but that was forecast. Considering what looked like rather appreciable surface convergence in the Hinton - Watonga area (and another area of convergence down near Lawton), there must have been strong capping just a couple of counties west of OUN. However, the sfc temps were also quite a bit higher out there (the OUN sounding has 77 F temp), with low-mid 80s in some locations E of the dryline. We even saw some clumping Cu and maybe some TCu around 3:30-4:00 pm SW of El Reno, but those dissipated by 5-5:30. Though the attempts were not as solid as those around the OKC area on 6/7/07 (another day with an uncapped sounding, very high CAPE, strong shear, but not a single sustained storm in central OK), it is frustrating to see that shear actually verified higher than forecast. Heck, any time I see these shear numbers (SRH, etc.) with 2500+ j/kg CAPE and 800 m MLLCL E of a Plains dryline, I can only think thoughts of tornadic supercells.

I still find it strange that synoptic scale subsidence on the order of centimeters per second can preclude intense convective updrafts (30-40 m/s) associated with 2500-3000 j/kg MLCAPE. Since we never saw a Cb S of the activity up near Fairview, it's evident that those surface parcels never really reached their LFCs. Perhaps it was more of a problem of the DEPTH of convergence today.

I don't have the calculated convergence fields from today, but I did make a surface map of observations at 2230 UTC based on METARS and mesonet observations. There certainly looks like there's decent convergence, though things aren't always like they appear.

 

[Robert Edmonds]

I still find it strange that synoptic scale subsidence on the order of centimeters per second can preclude intense convective updrafts (30-40 m/s) associated with 2500-3000 j/kg MLCAPE.

Until a storm gets going there isn't that 30-40 m/s vertical flow to get parcels past that warm air. There's was another day (last year, don't remember the date) where this played a large role too, where the cap was nearly gone.

 

[Chris Vagasky]

I was watching the radars all afternoon, and the VWP from TOKC had me drooling and envisioning those same tornadoes.

I'd agree with you, Jeff...I don't think it was the cirrus that killed it, more likely the ridge axis.

Taking a look at the 00z sounding, we came fractions of a degree from of hitting Tc, not that we would have automatically had development, but we were really close.

 

[Tim Supinie]

So I guess the question is which assumption of parcel theory was violated today? I don't see entrainment being much of an issue, as T/Td spreads were 6 C or less until well above the LFC. I also don't see exchange of heat between the parcel and its environment being much of an issue, since the parcel isn't that much warmer than its environment until above the LFC. Parcel pressure being the same as the environment? My gut says that's a pretty safe assumption, since the vertical velocities weren't great (obviously, or we would have had a thunderstorm). Am I missing one?

As for depth of convergence, KOUN is the only useful observation above the surface, and it was directly out of the southeast. This would suggest that we had reasonable convergence above the surface, but I don't have observations above the surface west of the dryline to support that.

 

[Jeff Snyder]

Tim -- I have no idea. I'm not sure parcel theory was violated, but it's a good reminder that parcel theory doesn't tell the whole story. A lot of forecasters and chasers (myself included) spend a lot of time looking at forecast soundings. If I had seen the 00z OUN sounding in a forecast sounding from any chase day, I would be extraordinarily "excited".

The better question may be one of representation. Certainly, if you have low-level convergence in an area with the environment defined by the 00z OUN sounding, parcel theory would suggest that you'd get an intense storm, and the various composite indices (SCP, STP, etc.) are on the high end of the scale in this case. The only thing I can really think of (or, rather, the most likely thing) is that the sounding was not representative of the environment where surface convergence was observed. Of course, if the OUN sounding is not representative of conditions near Anadarko, Hinton, or Watonga (where the sfc temps were actually higher than in OUN), only 50-60 miles from OUN, then there's certainly going to be issues with some of the studies that use "proximity soundings" as a sort of ground truth for the environment in which particular events occur.

 

[Chip Redmond]

What about the nearly moist-adiabatic temperatures from 600-200mb? Could this stable layer effect the below atmosphere and prevent it from reaching LFC? Also, backing winds behind the dryline could have weakened the convergence perhaps?

Chip

 

[Jeff Snyder]

The lapse rate in the 200 mb to 600 mb layer isn't very good, but it's not abysmal either. Notice that the mixing ratio in the middle and upper troposphere is typically quite small (in other words, there's not really much water vapor available), so latent heating is usually also quite small. As such, the moist adiabatic lapse rate tends to approach the dry adiabatic lapse rate. In this case, there's not much in the middle and upper tropospheric temperature trace that looks that bad to me -- no big CAPE robber or midlevel inversion. The mid-level temps could have been cooler, that's for sure. I wouldn't have minded -12 or -13 C at 500 mb and 7.5-8 C/km 500-700 mb lapse rates.

The winds behind the dryline did back slowly, and I remember seeing that there was questionable convergence along the dryline S of I40. Near Hinton/Watonga, however, there certainly looks like there was relatively persistent convergence. Obviously, since no sustained convection developed, the capping was too strong for the given convergence. So, if the capping wasn't very strong, then neither could have the convergence.

 

[Jeff Duda]

In my research, I have seen that convective initiation is a very sensitive entity to forecast, especially at high resolutions. When looking at the 00Z OUN sounding, I do see a very small pocket of stability in the 750-800 mb layer. I realize it's not much and it could be a factor of the thickness of the lines, but it looks clear to me that the virtual parcel path does sit to the left of the virtual temperature profile in that layer. Given that this layer was 50 mb or more above the top of the PBL, perhaps even the convergence along the dryline (assuming this sounding was representative of that at the dryline) was simply not enough to overcome what little CIN there was. Given the potential measurement/instrumentation error, it's possible that the CIN was actually a tad higher, too.

 

[Tim Vasquez]

That, and really there is a lot of small-scale noise in the fields which really isn't resolved well, especially when you're talking about one little balloon that has to give a representative sample of 1.6 million km^3 of troposphere. There's only so much that the first-guess fields can contribute, and basically when all is said and done you still have only an approximation of the storm environment. The marginal performance of forecast soundings kind of bears this out.

Tim

 

[Zac Flamig]

I think the theta-e vs. pressure plot there provides some clues. With the exception of the absolute surface, the atmosphere between the surface and 800mb is neutrally buoyant suggesting that parcels being forced upwards are not gaining any vertical velocity due to thermodynamically driven accelerations. Then you'll notice the subtle inversion at about 800mb suggesting there is sinking motion taking place there, not to mention that this is the same location at which the CINH is present so the parcel after struggling to make it to the 800mb level would quickly be forced back down by both the dynamic subsidence and thermodynamic buoyancy effects.

This seems like a case where the numbers can be slightly misleading.

 

[Jeff Snyder]

Zach -- that is possible. However the constant theta-e below 800 mb is just the result of a well-mixed boundary layer. With the same mixing ratio and potential temperature near the surface, you'd expect theta-e to be essentially constant. In order to see a decrease in theta-e, you'd need to either have superadiabatic lapse rates (so potential temperature / theta decreases with height) and/or a moisture profile characterized by decreasing mixing ratio with height. In my experience, the latter is not terribly good, though neither of these are likely through most convective boundary layers (except very near the surface). So, I wouldn't read into the observation that theta-e remains the same below 800 mb too much.

There is a slight inversion (perhaps it's just an isothermal layer -- it's hard to tell since it's so shallow) at the top of the CBL, but that too is relatively common, particularly considering that 23 z (presumed launch time) is after peak heating and the typical warmest temperatures of the afternoon (21-22 z). I'm not sure that's the result of subsidence aloft, however. If you think of an unsaturated, sinking parcel, mixing ratio will be conserved as temperature incrases (assuming there's not precip falling through the layer, etc.), and you'll typically see relatively dry air as a result. It certainly is possible that there was subsidence occurring, but that cap is very, very, very weak -- -10 j/kg MLCINH, and the the absolute difference between the parcel temp and environment there couldn't have been more than 1 C, and that's for a 77 F surface parcel (temperatures just W of OUN were 85-89 a couple hours earlier). The TCu that occurred didn't look terribly vigorous, which leads me to believe that the actual parcels characterizing the updrafts of the TCu were quite a bit different from the actual surface parcels (e.g. there was mixing between post-dryline and pre-dryline air). Of course, it's also possible that there was much warmer air at 800-775 mb just west of OUN that actual as a stronger cap.