7/13/04 REPORTS: Great Lakes/Ohio Valley

It wasn't 'obvious' that a 6hr discrete supercell with a violent tornado would occur synoptically, at least the night before. The first storm initiated at about 9 am. Sfc winds were progged to be almost WSWrly (they ended up being SWrly almost Srly) with a fairly unidirectional profile, though speed shear was progged to be extreme with extreme instability/steep lapse rates. Yes, I understand SR flow (hodos) and can adjust for NW flow days, but I don't recall the directional shear as transpired as being progged. Scott
Click to expand...

I'm not sure why you're saying the winds were progged to be unidirectional, since you already said the surface winds were progged to be "almost" wswesterly, and the 500mb flow was progged to be northwesterly. That's still more than 45 degrees of directional shear, and you've mentioned the speed shear was progged to be extreme.

How is this "unidirectional" shear, and why would this not be conducive for long-lived supercells, base on the 00Z or 12Z ETA forecast?

A closer inspection of hodographs and and low level helicities would have to be considered to determine "tornadic" supercell potential. But based off what has been mentioned here, conditions certainly aren't bad.

Jim
http://stormgasm.com
 
I'm not sure why you're saying the winds were progged to be unidirectional, since you already said the surface winds were progged to be "almost" wswesterly, and the 500mb flow was progged to be northwesterly. That's still more than 45 degrees of directional shear, and you've mentioned the speed shear was progged to be extreme.
Click to expand...

Winds were progged to be unidirectional, at least >6 hours in advance of the event. I think what Scott means by "fairly" unidirectional is that SFC winds were southerly, turning westerly at 850mb, northwesterly at 700mb and unidirectional from there. Between ~850-500mb (about 3-6km), the profile was unidirectional for the most part.

Take a look at the graphic I posted, the ETA shows ~45 degrees of directional shear (which I would consider unidirectional), and thats 9hrs before the event. Good tornado potential generally requires 90 degrees of directional shear, so I wouldn't consider 45 degrees good.

Speed shear and directional shear are two different components. You can have extreme unidrectional speed shear with relatively low helicities as well.

All in all, the threat for tornadoes didn't appear to be significant until about 3 hours before the event. SPC didn't even catch this, and had a mere 5% tornado probability. If the threat were more evident, it would have been higher than 5%. They did upgrade the probability in later outlooks, but that was only becase current obs dismissed what the models were showing at the SFC, which was WSW winds, compared to the SSW or S winds.
 
And the differences weren't all that bad between 12 Z eta and observations.
Click to expand...

Most of the region was fine, it was just those few stations in central IL, which were probably a result of mesoscale features. This locally enhanced the helicity/shear, which wasn't picked up by the models (most mesoscale events arn't picked up, as you said). My thinking is that if these winds would have been more westerly, as progged, then the tornado probably wouldn't have been as strong. While there was some directional shear progged, it was no where near what was actually observed, thus the tornado potential seemed small when looking at the models...

Of course this is all just a guess, and thats what makes it interesting...
 
It wasn't 'obvious' that a 6hr discrete supercell with a violent tornado would occur synoptically, at least the night before. The first storm initiated at about 9 am. Sfc winds were progged to be almost WSWrly (they ended up being SWrly almost Srly) with a fairly unidirectional profile, though speed shear was progged to be extreme with extreme instability/steep lapse rates. Yes, I understand SR flow (hodos) and can adjust for NW flow days, but I don't recall the directional shear as transpired as being progged. Any factor of boundaries is not synoptics. Was it obvious that it COULD happen? Yes, you *never* ignore (you treat it very seriously) that kind of instability with that kind of speed shear, but it was not obvious (at least the night before, I never had a chance to do a proper fcst the day of) that it would be a day with a discrete classic supercell lasting for many hours. Possible, sure, but how likely? If it was, where were all the forecasts of that preceding it? I saw nothing from anyone. Hindsight is convenient. There is some culpability in models and overreliance on them of course; and it also again illustrates how fairly subtle differences can be of paramount importance. Additionally, the ETA showed precip not firing till late afternoon/early evening, a far cry from mid-morning supercells with a peak in the early afternoon.
Scott
Click to expand...

I think you may want to review the radar data again - it was NOT the morning supercell that produced the tornado in Woodford county - the main tornadic cell was less than 45 minutes old! Morning RUC forecasts were excellent for a regional perspective - which of course is initialized with the 12Z ETA analysis. It showed > 200 helicity with > 4000 CAPE colocated along and east of I-39 from the I-88 to I-70 corridor. Why SPC wasn't impressed with this guidance I don't know - but is was evident this could be a significant tornado day - and clearly was going to be a supercell day - which they were on top of forecasting very large hail - rare for IL.

Late morning obs should have been the guidance that sfc winds were not WSW but more SSW across this region. The 12Z sounding at ILX was amazing for this part of the country. A ETA forecast of less than 12 hours is of little value for short-term convective forecasts - the physics take too long to get spun up to be of value for convective forecasting. I'd use the 00z and 06Z and compare to the current ob trends for quality, then morph to RUC for < 9 hours + current obs trends. I didn't check older ETA runs closely for this event - maybe the guidance was not very good - but I recall that they were consistent with a morning squall line that didn't materialize - so forecasts needed significant adjustment from model guidance. While I didn't post my forecasts here as I was out of town that day - I did do so within my local group of chasers late morning of that day, and had this event pretty well forecast. The real wildcard, IMO was that the significant tornado came from such a young cell. Prediciting even an hour in advance where the most probable location for a tornado would be at a county level would have been very challenging.

Glen
 
Winds were progged to be unidirectional, at least >6 hours in advance of the event. I think what Scott means by "fairly" unidirectional is that SFC winds were southerly, turning westerly at 850mb, northwesterly at 700mb and unidirectional from there. Between ~850-500mb (about 3-6km), the profile was unidirectional for the most part.

Take a look at the graphic I posted, the ETA shows ~45 degrees of directional shear (which I would consider unidirectional), and thats 9hrs before the event. Good tornado potential generally requires 90 degrees of directional shear, so I wouldn't consider 45 degrees good.

Speed shear and directional shear are two different components. You can have extreme unidrectional speed shear with relatively low helicities as well.

All in all, the threat for tornadoes didn't appear to be significant until about 3 hours before the event. SPC didn't even catch this, and had a mere 5% tornado probability. If the threat were more evident, it would have been higher than 5%. They did upgrade the probability in later outlooks, but that was only becase current obs dismissed what the models were showing at the SFC, which was WSW winds, compared to the SSW or S winds.
Click to expand...

IMHO, you simply don't much directional shear between 850mb and 500mb if there exists 45 degrees of directional shear between the surface and 850 (including speed shear), and strong speed shear from 850mb to 500mb. But, I remember the 850mb winds were not forecast by the ETA to be northwesterly, they were progged to be almost due westerly, which would have yielded deent diretional shear between 850 and 500mb anyways. And given the storm motion, the SR surface winds would have been much more backed, anyways.

Whatever the case, neither of the profiles mentioned is unidirectional, IMO. And good tornado potential requires a good analysis of a hodograph, not necessarily directional shear...and 90 degress of directional shear is absolutely significant.

I thought the SPC underplayed the tornado potential since the 00Z ETA run. The forecasted shear profiles were fine for supercells. The forcing was going to be weak, so no worries about immediate linear development, very high cape, and the forecasted hodographs looked very favorable for tornadic supercells once they moved into central Illinois. Plus, the storm motion was going to be fast (35 knots), so moving into the better shear environment was not a problem. I'm not saying I pinpointed the area exactly, but there was certainly a greater threat for tornadoes than forecasted over the central/north central Illinois area. Actually, the ETA underforecasted the speed of the surface trough/cold fronts progression east, which is one reason why, I think, the tornadoes were where they were.

I think the real issue at hand here is the northwesterly flow. This catches many forecasters, including the SPC (at times), off guard. I'm still not sure why there wasn't at least a hatched area for tornadoes, given the extremem CAPE, coupled with great shear.

Does anyone know why the other sounding posted, besides the one I posted from NCAR, shows 6000 CAPE at 18z, rather than the NACAR 4000 CAPE observation? Strikes me as odd.

This was a very interesting setup, and thus far has been an interesting discussion about it.

Cheers,

Jim
 
It wasn't 'obvious' that a 6hr discrete supercell with a violent tornado would occur synoptically, at least the night before. The first storm initiated at about 9 am. Sfc winds were progged to be almost WSWrly (they ended up being SWrly almost Srly) with a fairly unidirectional profile, though speed shear was progged to be extreme with extreme instability/steep lapse rates. Yes, I understand SR flow (hodos) and can adjust for NW flow days, but I don't recall the directional shear as transpired as being progged. Any factor of boundaries is not synoptics. Was it obvious that it COULD happen? Yes, you *never* ignore (you treat it very seriously) that kind of instability with that kind of speed shear, but it was not obvious (at least the night before, I never had a chance to do a proper fcst the day of) that it would be a day with a discrete classic supercell lasting for many hours. Possible, sure, but how likely? If it was, where were all the forecasts of that preceding it? I saw nothing from anyone. Hindsight is convenient. There is some culpability in models and overreliance on them of course; and it also again illustrates how fairly subtle differences can be of paramount importance. Additionally, the ETA showed precip not firing till late afternoon/early evening, a far cry from mid-morning supercells with a peak in the early afternoon.
Scott
Click to expand...

I think you may want to review the radar data again - it was NOT the morning supercell that produced the tornado in Woodford county - the main tornadic cell was less than 45 minutes old! Morning RUC forecasts were excellent for a regional perspective - which of course is initialized with the 12Z ETA analysis. It showed > 200 helicity with > 4000 CAPE colocated along and east of I-39 from the I-88 to I-70 corridor. Why SPC wasn't impressed with this guidance I don't know - but is was evident this could be a significant tornado day - and clearly was going to be a supercell day - which they were on top of forecasting very large hail - rare for IL.

Late morning obs should have been the guidance that sfc winds were not WSW but more SSW across this region. The 12Z sounding at ILX was amazing for this part of the country. A ETA forecast of less than 12 hours is of little value for short-term convective forecasts - the physics take too long to get spun up to be of value for convective forecasting. I'd use the 00z and 06Z and compare to the current ob trends for quality, then morph to RUC for < 9 hours + current obs trends. I didn't check older ETA runs closely for this event - maybe the guidance was not very good - but I recall that they were consistent with a morning squall line that didn't materialize - so forecasts needed significant adjustment from model guidance. While I didn't post my forecasts here as I was out of town that day - I did do so within my local group of chasers late morning of that day, and had this event pretty well forecast. The real wildcard, IMO was that the significant tornado came from such a young cell. Prediciting even an hour in advance where the most probable location for a tornado would be at a county level would have been very challenging.

Glen
Click to expand...

I have saved all grib data/sfc/upper air/etc (don't have radar saved though)...

Since I am not sure exactly what is up to debate here, these are my thoughts:

1) Did the synoptic setup appear favorable for large tornadoes, at least on the ETA, >12 hours before the event? No
2) Did the synoptic setup appear favorable for supercells >12 hours before the event? Yes
3) Was the 18Z KILX sounding favorable for tornadoes: Yes
4) Did the RUC do an excellent job at forecasting this event < 6 hours prior to the event? Yes (see image below)...
5) Was this a mesoscale or synoptic scale event? Mesoscale

I think the confusion may be coming from the fact that helicity was relatively low for a large tornado (typically >300m2/s2), but the fact that CAPE was >5000J/KG compensated for this. As for the threat of large tornadoes being seen >12hrs prior, it wasn't. Climatology also suggests that most NW flow events are associated with damaging winds, large hail, and isolated weak tornadoes.

The supercells also fired along a confluent/outflow boundary, which the RUC apparently picks up very well. Without this boundary, the SVR development probably would have occured later in the day, and since the CAP would have been totally gone (as opposed to "weak"). Given very high instability and no CAP, even weak linear forcing would generate a line of severe convection...

Below is the RUC 17Z F003 forecast... It puts the QPF in the right spot, has good helicity, with very high CAPE values... This would be indicative of a tornado-producing environment, and was forecast 3 hours in advance. Now, if I were to look at the 12Z ETA, the threat of tornadoes isn't evident (high CAPE, very little helicity, etc.). Also, it says "Lifted Index", but thats because I manually put in the CAPE parameter, and the icon was clicked on "Lifted Index".


[Broken External Image]:http://www.waveformpc.com/roanoke/ruc_17Z_F003.gif
 
Does anyone know why the other sounding posted, besides the one I posted from NCAR, shows 6000 CAPE at 18z, rather than the NACAR 4000 CAPE observation? Strikes me as odd.
Click to expand...

I believe the 4000J/KG CAPE is MLCAPE (Mixed Layer CAPE), and >6000J/KG is either SBCAPE (Surface Based CAPE) or MUCAPE (Most Unstable CAPE)

When I plug in the 18Z KILX RAOB into nsharp, I get (all un-modified) SBCAPE/MUCAPE of 6699J/KG with 0C CAP and a LI of -19C!! MLCAPE of 4536J/KG with a 1C CAP and a LI of -15C!..

Helicity is:

0-2km: 101m2/s2
0-3km : 106m2/s2

0-2km shear: 50.49knts
0-3km shear: 46.25knts
0-6km shear: 28.48knts

I will try to get a graphic up later...
 
I have saved all grib data/sfc/upper air/etc (don't have radar saved though)...
Click to expand...

Last I checked the NCAR site still has the NIDS data available, and should for another day or two. Level II data is not yet available through HDSS.

Glen
 
The issues were how apparent was the setup synoptically and how early preceding the event, as well as for me most importantly: the longevity of one of the cells (which was bothering me). I haven't seen a radar summary of the day yet, but was told (by more than one person, some of them knowledgable) that it was the cell that developed on the WI border that continued until evolving into a derecho near BMI. I should have been more questioning of that itself instead of trying to fit that cell to the meteorological setup right away. Did I think tornadic supercells were likely even the night before? Yes. I honed in them being possible even before that. My forecast to friends would show that. It's the longevity (as a discrete CS in a very weak cap environment and as I recalled less than ideal progged directional shear) that was baffling me, and I'm glad that's cleared up.

I look forward to doing a little case study when I find the time.

Scott
 
I had a pretty decent chase on July 13th... There wasn't anyway for me to get down to central IL, which is where the best potential for any tornadic supercells was. Yet, MI still looked pretty decent, once the cold front lit up by the late evening... A line of severe thunderstorms developed over lake MI by around 00z, with several discrete supercells forming southeast of the main line. I targetted LAN/Ionia, hoping to get to something quick, when it was still discrete.

I did catch the first LAN-area (near Clinton) supercell -- which picked up some decent cloud-base rotation fairly quickly (was SVR-warned). I witnessed at least one funnel... With a decent RFD kickup south of the funnel. All in all, I got some pretty decent lightning (these things were major lightning producers) and the funnel pretty much made it a good chase.

Plenty o' CAPE in the atmosphere at the time as well... More than >4000j/kg SBCAPE near the supercell, accompanied by strong, unidirectional flow.

I'll have the account and pics on my website soon.
 
I went on a local chase, to catch the wind/lightning, got hit with 50-60mph wind, very heavy rain, and very small hail (< pea sized)...

I have some nexrad images of the MI stuff that I chased, with some comments added: http://www.waveformpc.com/july13_dtx/

If you look at the DTX image from 0250Z (I posted if, cause its interesting...), there appears to be a supercell with a hook echo, which I do not remember seeing before. Storm relative velocity indicated weak rotation, were you on this cell Nick?

My house also got blasted by 70MPH winds, guess I should have just stayed home for this one...

[Broken External Image]:http://www.waveformpc.com/july13_dtx/dtx_nids_0250.gif
 
Here is a picture I took of my laptop, of the storm when it was in Eaton and Ingham County; I talk to the NWS, they mention there was some mid-level roation along the leading edge of the storm and it was anitcyclonic. The NWS mention no reports of funnel clouds or anything, when I talked to them. concerning Eaton, Clinton, Ingham counties. There was some low scud hanging from the shelf cloud, as it approached my location, in Eaton County, I was near the Eaton and Ingham County line, quickly followed by 60 mph winds and pea size hail.

Radar Image I captured showing Eaton and Ingham Counties.
[Broken External Image]:http://mgweather.com/xm1.gif

Another view on the storm:
http://mgweather.com/xm3.gif

Mike
 
"That sure looks like a hook echo to me! Where in the world is the TOR??"

It's where it should be - not issued unless rotation is strong & tight or if mesoanalysis shows something that would help the rotation.

- Rob
 
You must log in or register to reply here.

Similar threads

Dan Robinson
2024-05-08 REPORTS: MO/IN/IL/KY/TN/OK/TX
Replies
0
Views
399
Dan Robinson
Dan Robinson
Dan Robinson
2024-05-26 REPORTS: AR/MO/IL/TN/IN/KY
Replies
5
Views
1K
Bill Hark
Bill Hark
Dan Robinson
2024-04-25 REPORTS: TX/KS/NE/CO
Replies
8
Views
844
John Farley
John Farley
Dan Robinson
2024-07-29 REPORTS: IL/IN
Replies
0
Views
407
Dan Robinson
Dan Robinson
Dan Robinson
2024-04-02 REPORTS: IN/OH/KY/WV/TN
Replies
0
Views
583
Dan Robinson
Dan Robinson
Back
Top