Thursday, February 12, 2009

Feb 10 review: significant tornadoes strike Edmond and Lone Grove, OK

February 10 looked to be the first potential severe weather outbreak of the slow-to-start 2009 severe weather season. I had planned to chase the event from a week out as long-range models developed a strong lock on the general pattern, but was later scheduled to work a day shift the following morning (February 11). Given the tornado threat initially looked somewhat marginal before dark and probably would be solidly on the TX side of the red river (a 7-hour+ drive back home after the chase), I decided I should sit it out. Guidance the morning of suggested severe storm potential might shift a bit more to the NW, but I still wasn't sure how significant the daylight tornado threat would be given initially semi-modest low-level shear just E of the dryline and the rather expansive nature of convection expected to develop by mid-afternoon. My virtual target was Ardmore (just E of Lone Grove), but we all know how targets change and how one can be suckered and swayed through the course of a chase day. I definitely wouldn't have thought to target the I-44 corridor from OKC to Tulsa, that's for sure, where the "daytime storm-of-the-day" ultimately trekked.

Below is the modified 18Z OUN RAOB and 21Z Purcell profiler to represent the environment of this storm as it produced an EF-2 tornado that struck Edmond, OK around 300PM. This storm fired on the northern flank of the returning deep BL moisture (another reason I'd probably have never targeted this region, all things considered), and had no trouble becoming a strong supercell producing baseball hail for a few hours. The rapidity with which the storm became such a nasty little beast is slightly surprising given very strong deep layer shear (the Purcell profiler sampled 6 km flow as high as 88 kts!!), as I usually like to see stronger low-level shear for such set-ups than was in place for this one. However, moderate CAPE was present (approaching 1500 J/kg...certainly stronger than forecast), and low-level CAPE was very strong (150 J/kg below 3 km) given seasonably cold temps above the BL... these factors probably played a role in allowing for the long-lived supercell.

Sadly, after dark, the second and final significant tornado of the day struck portions of Lone Grove, OK and caused 9 fatalities. By around sundown, the synoptic cold front was overtaking the dryline across west OK and northwest TX, encouraging upscale convective growth to a solid squall line. Convective initiation concurrently took place farther east in the warm sector, with a supercell emerging south through east of Wichita Falls, TX. Pretty much everything was in place for this storm to become a violent tornado-producer as it approached the western edge of the nocturnally-accelerating LLJ core. Even if surface temperatures had fallen toward the mid 60s after sundown in advance of this storm, the quality of the boundary layer moisture (surface dewpoints near 64°F) and absence of a significant EML would have allowed for a sufficiently strong surface-based environment to persist. As it was, surface temperatures managed to hold in the 70-72°F range in the storm's inflow environment through at least 8 P.M... resulting in the maintenance of near-zero CINH and very strong low-level instability (150-200 J/kg 0-3 km MLCAPE). In addition to the absolutely ideal low-level thermodynamic environment after dark, MLCAPE was > 2000 J/kg, with quite strong deep layer shear and 0-1 km SRH exceeding 300 m2/s2 (see modified 00Z DFW sounding and its hodograph blended with 01-02Z VWP data from the FWS 88D). The tornado was on the ground for over an hour, and preliminarily has received a rating of high-end EF-4. This case continues the unnerving trend for nighttime killer tornadoes in the Plains the past few years... more often seen in the deep South. Luckily the storm was swallowed shortly thereafter from behind by the squall line... and despite the prefrontal warm sector remaining mild and weakly capped all night, additional convective initiation did not take place there.

Sunday, February 8, 2009

Feb 9 forecast

Just a brief post for tomorrow's system. Looks like it could be an interesting day for the middle/lower MO river valley, at least by almost-still-the-middle-of-winter standards. Strong upper level trough crossing northern Baja this morning will eject negative tilt through the central Plains tomorrow. The associated surface low will seemingly have no trouble undergoing bombogenesis as it lifts NNE overnight tonight through tomorrow, potentially making a run at 982 mb by peak heating tomorrow. This isn't a classic Davies cold core set-up by any means, as the mid-level system (a negative-tilt shear axis) is not closed off... and will be moving at a high rate of speed. Showery precipitation, driven by strong high-level difluence/divergence preceding the upper trough, will spread through the middle and lower MO river valleys tomorrow morning. The potential for mini-supercells will follow in the warm sector of the surface cyclone (potentially initiating in the thermal ridge near a prefrontal confluence band), as the dry slot attempts to wrap northward and northeastward into the base of the negative-tilt trough. The farther north the dry slot can advance, the better, as my sense is that cells just to the cyconlic side of the mid-level jet will struggle to organize amidst 700-500mb flow of 50-80 kts, given SBCAPE will likely be AOB 500 J/kg. Respectable low-level CAPE (700 temps around -5C!) and healthy low-level shear (0-1km SRH potentially as high as 200-300 m2/s2) will aid in organization and the potential for mini-supercells and tornadoes. We've certainly seen past low-CAPE/super high-shear settings produce tornadoes (e.g. 11/12/05), but those cases seem pretty rare. Tomorrow's system has a huge bust potential, particularly if it races through the area more quickly and/or farther north like the GFS has been hinting at.

EDIT @ 930 P.M. : am liking the looks of the latest models even better. They continue the trend of mid-level dryslotting into the mid/lower MO river valley by early afternoon, heating the surface well into the 50s. Also, models actually bring dewpoints in the low 50s into the area--definitely looks at least possible given the deep (2 km), saturated boundary layer observed on the DDC and OUN RAOBs with ML dewpoints of 7-11C. Finally, the wind gradient in the cyclonic side of the mid-level jet is forecast to be far more diffuse, with non-outrageous 700-500mb flow (40-55 kts)... I'm actually seeing some semblences of a "warm sector" cold core pattern emerging.