06/06/08 tornado environment

Summary:

A mini-supercell produced long-tracked EF-3 and EF-2 tornadoes near Park Rapids MN between 900-1000am local time on 06/06/08. The synoptic weather pattern strongly resembled typical "cold core" tornado events of the past (as studied in depth by Jon Davies)... with the time of day of the event being one of its most bizarre and tough-to-anticipate attributes. A seasonably strong closed mid-level low was lifting NNE through eastern ND, with its attendant surface low about 100 miles to the east near Fargo ND. The tornadic storm initiated in the northeast quadrant of the mid-level low on the "fringe" of the vorticity, which is very classic in these events. A pronounced (if dirty/stratocu-ish) dryslot was also nosing into the area, which is a hallmark of many cold core tornado cases as a means to shut down precip and heat/destabilize the environment ahead of the mid-level low. As for the surface pattern... given the intensity/duration of this event, my assumption based on hourly METARs is that the tornado occurred primarily to the cool side of the lifting warm front. However, the data would have to be carefully scrutinized in order to be prove this for absolute certain. Low-level shear magnitude in the open warm sector south of the front dropped off quite dramatically, but was still generally sufficient for tornadic development given the strong LLJ observed on the 12Z MPX sounding. Regardless, my assumption in this study again is that the mini-supercell updraft tornado'd primarily north of the front... and weakened with time as it moved into low stratus and cooler surface temperatures.

To create a representative proximity sounding, I used the 14Z RUC 00-hour forecast at BRD... modified just slightly at the surface with the 14Z PKD METAR (immediately before the storm and the warm front lifted through PKD). Based on the 12Z ABR and MPX RAOBs (superimposed below), the RUC T/Td profile at BRD looked very reasonable... correctly showing the mid-level cold pool (3C and -16C at 7H and 5H respectively) overspreading a boundary layer characterized by 15C 100mb ML dewpoints. A weak/shallow frontal inversion was present on the proximity sounding (which seems feasible, though none was observed at INL); note that the inversion had little net effect on MLCINH. As for the other "numbers" cranked out by the sounding, they are pretty typical of a cold core mini-supercell event. The ML Lifted Index of -6C is a good showcase of the "fat" mid-level CAPE profile, suggesting stronger updraft accelerations were possible than one might expect based on modest total MLCAPE (622 J/kg). Meanwhile, 0-3km MLCAPE was moderate in magnitude at 87 J/kg (though strong relative to the total CAPE), given the thermal profile didn't become particularly cool until above 700mb. This is still pretty impressive low-level instability considering the time of day and that low-level lapse rates had just begun to steepen via diabatic heating. As for sufficiently low LCL and LFC heights... they are rarely difficult to achieve in cold core tornado events, and this one was no exception.

The 14Z RUC BRD hodograph likewise looked quite reasonable based on the 12Z hodographs observed at INL and MPX. Its surface wind was 125 degrees at 11 kts, which surprisingly lines up quite well with the 14Z PKD ob... important since low-level shear magnitude is often strongly dependent upon the surface wind vector. The 0-1 km SRH computed for the RUC hodograph was very impressive, and really sets this event apart from others. Strongly backed surface through 200 m winds just north of the warm front, with flow increasing to 40+kts by 500 m, resulted in 0-1 km SRH of 462 m2/s2!!! (In studying the most prolific cold core tornadic minisupercell events, I found their environments most commonly possessed 0-1 km SRH in the 100-150 m2/s2 range.) The hodograph also shows a strong sickle/hooked shape, a shear profile that is often observed in violent tornado environments as recently shown by Giuliano/Esterheld and Miller. The 0-6 km bulk shear is 39 kts--supportive of supercell structures. In my experience, cold core tornadic cells commonly originate in the spiral band of DPVA beneath the 35-50 kt 5H isotachs... as did this event. This seems to be a "sweet spot" with regard to balancing the bulk shear and relatively small/shallow CAPE... specifically because the deep layer shear here is sufficiently strong away from the mid-level low center to support supercellular processes ... yet not too strong (as beneath the mid-level jet axis farther east--commonly 70 kts+ in these setups) so as to overwhelm or rip convection apart. As a final note, it's possible 0-6km shear isn't a particularly good gauge of "bulk shear" for these cold core events, given the tropopause/EL is much lower than in more typical tornado setups... in fact, 6 km is sometimes right near the EL in minisupercell environments. The RUC BRD sounding showed the EL/storm top in this case to be around 23,000 feet (7km).

Figures:

12Z NAM-WRF 03-hour forecast of geopotential height/wind at 500mb & 850mb