Tushka Culmination
Southeastern Oklahoma, 14 April 11
SHORT: 3 supercells intercepted in SE OK, likely end of Tushka tornado witnessed after navigating damage.
LONG:
Ryan Jewell, Jaret Rogers and I left Norman after working the day shift, headed SE toward Ada to get ahead of the supercell that already had crossed I-35 near Davis. Vigorous convection had gone up just E of Norman too. We eschewed that option completely while E of Lexington, given that it was heading into an environment of weaker shear and buoyancy than farther S, and had messy radar structures and mergers happening by then.
Alas, so did the old Arbuckles storm, by the time we saw its inflow region, looking NW from near Stonewall. We had beaten the hail core through Ada by just a few minutes, and took the time to marvel at the turquoise color of the wrapping precip and the long tail cloud. Motion into the mesocyclone region was really fast, inflow screaming from the SE; but mergers with other cells to its S had rendered the supercell, still with an intense mesocyclone, a garishly greenish HP stormzilla.
Two more supercells beckoned to our SW, the first headed for the area between Tupelo and Clarita, and easy to intercept. The storm had a great reflectivity hook, and was TOR-warned; but the low-level velocity structure didn’t look particularly vigorous on radar. Visually, as seen from Clarita (looking WNW), the convective structure was very robust, the base far less precip-infested than the Ada/Stonewall storm. Still, it had a long, scuddy rear-flank gust front, with similar scud drifting about under the putative mesocyclone region. A broad, ragged, very low-hanging, but only weakly rotating wall cloud formed, then fell apart. Not too impressed, we didn’t longer longer, and made haste down OK-48 through Coleman, to OK-22 and westward after the southernmost (anchor) storm near Tishomingo.
We arrived within sight of the storm’s base just a tad late to see the Tishomingo tornado, having this NW-looking view of two wall-cloud regions as the storm moved close to Milburn. The nearer (dark) one fell mercy to the rear-flank downdraft wrapping around the more distant, northern, older, and occluded meso near Milburn (brightened by clear-slot light in this later shot). Here we met and chatted for a spell with Ashton, who later ended up in the damage track of the Tushka tornado very near us as well.
The near (southern) meso thus undercut, we headed N to get a closer look at the circulation near Milburn, but it too was weakening. A wrong turn in Milburn led us to some knobby and often partly broken hail up to 2-1/4 inches in diameter, which we photographed before heading the correct way out of town. We zigzagged from Milburn to Coleman, getting close to the storm’s back edge of the re-intensifying supercell’s hook region. We considered an unmarked road to the east, through Boggy Depot, but rejected it as unsafe, given its unsavory look on the road map, and its precarious position slightly N of astern with respect to the new mesocyclone area. Take that road, we knew, and huge hail wrapping around the backside of a hook echo would be a sure fate.
Instead we busted S back to OK-22 and E to US-75, then N past Caney, arriving to see a mesocyclone with dark, amorphous murk beneath moving past the road in the distance.
Not yet knowing that a tornado was somewhere within that murk, we proceeded NNE up US-75 into Tushka, unaware at first of the intensity of the damage until we had gotten irreversibly into it, embedded in a stream of slow-moving traffic. Broken and uprooted trees appeared on both sides of the big roadway, becoming more numerous in the first mile or two, along with assorted sheet metal, pieces of wood, broken road signs and other variegated flotsam. The blended scent of shredded vegetation, split green wood and rain deeply saturated the air, a unique aggregate aroma I’ve experienced only amid fresh tornado damage.
Right after we drove several feet safely beneath some low-hanging power lines, an adjoining tractor-trailer rig hit them…and kept going full speed! It busted at least three power lines, yanked down a power pole that still had been standing, and may have dragged another already downed pole a short distance. I suspect one of my most abiding memories of this chase will be the furious, staccato “THWAP THWAP THWAP” noises as the cables snapped just behind and next to us. We were aghast that the driver would just plow right through low-hanging power cables, snap them mercilessly, and keep on truckin! I guess he had places to be.
Even within a few minutes after the tornado, plenty of police and other first-responders already were arriving or there. We saw several vehicles flipped asunder, as we navigated around whole trees and large debris on the road; and every overturned vehicle I could see either had no one inside, or at least one person standing next to it. Under those circumstances, and mired the plodding traffic, the best move was to get through the mess safely and expeditiously as possible, and get out of the way of those who needed to be there. It reminded me eerily of coming into the damage swath along I-44 right behind the (also rain-wrapped) Catoosa F4 of 24 April 93. As we entered Atoka, the damage lightened; though we could see more off to the east.
With several miles of damage along the highway, the tornado obviously was fairly wide, even if not exceptionally violent, and took a very oblique path with respect to the road. Even in the deepening twilight, now quite dim under all that storm-cloud cover, we still could see rapid scud motions from left to right above the road (NNE of us), wrapping back around an unseen mesocirculation center to our NE. From Ryan’s I-Phone radar, we figured the mesocyclone (and temporarily, the storm as a whole) had taken a left hook toward the NE, with a new occlusion forming farther downshear. As it turns out, Rich confirmed this for us while we were still in the damage, calling him to report it just in case none of the other hundreds of people had. He also let us know that the mesocyclone was narrowing and perhaps even crossing back over the road as it veered leftward.
Lo and behold, right then we saw a power flash to our NNE, slightly leftward of the road, near Stringtown. Within a minute or two, a ghostly, dim but still very apparent cloud cylinder emerged from the wrapping precip, right in the direction of that power flash. The cloud column, whose bottom was obscured by a terrain ridge, contorted and gyrated wildly, with a ragged coil or helix of scud pulling around the near side shortly before it fell apart. The previous volume scan showed the narrowing couplet in that spot, and the next had it dissipated.
I am now convinced that was the deeply-occluded demise of the Tushka tornado near Stringtown, kicking back out of the rear of the precip-filled area (as I’ve seen a few times before). This is based on Rich’s real-time descriptions of the meso occlusion process on radar, my own post-event radar review, the path/timing of the old occlusion relative to our position N of Atoka, what has been documented of the path so far, and of course that wildly morphing cloud cylinder near where we just had seen a power flash. All things considered, not very satisfying. It’s ironic that the best view of the tornado (Gabe’s, while in some gigantic hail) was from the WNW–the angle we chose not to approach when we were just behind the storm in Coleman.
After the meso died, it was dark; and we found ourselves in a quandary. Dense cores kept crossing US-69 to the NNE, the NW route to Ada (US-75/OK-3) was blocked by a supercell, another supercell was back near Wapanucka headed for Atoka, and miles of tornado damage lined US-75/69 to our S. Solution? Eat. Also, sit down and plan some escape strategy.
At the Subway in north Atoka (an undamaged area), we heard of a triage setup elsewhere in town–the first sign that some casualties had happened. This was bad news indeed. [We didn’t hear of the two fatalities (elderly sisters) until after returning to Norman.] The next TOR polygon for the supercell to our W included Atoka; so the already frazzled employees started locking up to take shelter.
We finished eating out in the vehicle, as the supercell to our W weakened and passed just to the NW. The door out of the area now opened, we then headed home, listening to a combination of Metallica, Slayer, Iron Maiden and Pantera on Ryan’s phone to just wind down a little.
Meteorologically, our forecasting and pre-storm conceptualization was reasonably good. The only part that didn’t work out as well as I like was the large number of left splits and close-proximity cells that made storm modes messier than the orthogonal deep -shear vectors would indicate. Smaller early hodographs with some negative SRH probably contributed to that, along with a nearly simultaneous dispersion of CINH all up and down the area just E of the dryline, from the Red River to the KS border.
Daytime Mild, Evening Wild
Hennessey to Anadarko and Norman
13 May 9
SHORT: Observed multicells and supercells before dark from W of Hennessey to SW of Calumet. Witnessed Gracemont-Anadarko night supercell and power flashes from tornado and RFD winds. Observed part of separate supercell E of Norman after earlier tornado.
LONG: Given the very strong capping that was forecast, and the increasing uncertainty of diurnal storm initiation with southward extent, Elke and I targeted a conceptual zone just NE of the projected “anchor area” of a front-dryline intersection zone. The plan was to grab storms discretely backbuilding SW down the front and watch them in stair-step fashion, until either they lined out or became less interesting than newer development. Once, again, confidence in cap strength gave us the luxury of a late lunch in Norman, before we headed up to Kingfisher. Storms fired to our NNE near the KS border (one of them being the briefly tornadic event that Joel and Blufie captured on the way W from TUL…great work, dudes!). We briefly considered that activity, but turned our attention to newer cells visibly erupting into the sky over Major County, NW of town.
When we arrived at a decent viewing spot between Hennessey and Okeene, the bases looked rather flat and linear, as if frontally forced; but the then-anchor storm got better organized, with a nice anvil push with mammatus, and some tail cloud development off the NE side. ESE storm movement and CGs chased us from the spot. We headed S — coincidentally on the same paved back road W and SW of Hennessey from which we photographed a spectacularly unplanned “planned burn” earlier this season. Other smaller cells started to develop farther SW, merging into the complex, and contributing to a hard net right turn and SSE motion that had us zig-zagging on some paved, but at times quite rough, unmarked roads from Kingfisher SW to Calumet, keeping within viewing distance of the storms now pounding areas around Watonga and Geary.
The daylight was getting short and the low-level jet began to crank up, so we had some hope for more convincing supercellular structure. After exiting Calumet, those hopes came to fruition. On the NE side of the newer cluster of storms to our W (the same one under which Kiel O and KMan observed the landspout up close), a bonafide supercell formed and quickly developed a very low-hanging, slowly rotating wall cloud (wide angle and zoom).
We headed over I-40 S of Calumet, faced with either a 32-mile detour E, S then W on good roads, but with little daylight left, or 4 miles of thinly graveled dirt road leading to to an unmarked paved byway headed S toward Cogar with a bridge over the Canadian River. I chose the latter; and the “crappy” road actually was friendlier to my car than the many more miles of paved but horrendously maintained trash we had been traversing between Loyal and Calumet.
Meanwhile, the wall cloud had gotten undercut, but the broader storm began to acquire striations, and assumed a more circular appearance. In the fading twilight, we could make out a more sharply banded, “stacked plates” appearance to the storm while rounding the corner from Cogar to Gracemont, staying just ahead of its hard southward charge.
Here comes some meteorological discussion for the unitiated to skip, if compelled. I figured this laminarity was related to the balance between the storm’s improving organization and the strengthening of both capping and environmental low-level shear. A mixed-layer lifted sounding curve was “capped” in a pure parcel theory sense, and growing more so all the time through gradual diabatic cooling of the near-surface layer. Despite that apparent handicap, the supercell’s strengthening deep mesocyclone and vertical pressure gradient force caused deep ascent of parcels from the boundary layer through the environmental capping inversion. Effective lifted parcels still were bringing near-70 degree dew point air from the surface into this storm, while the LCL was lowering due to cooling temperatures and loss of deep boundary layer mixing. The strong inversion also was keeping the storm cluster rather isolated from a few others farther NE along the front, so it had no “outside competition” or impediment whatsoever for high theta-e source parcels, other than being able to maintain its own inflow-outflow balance. Storm-relative inflow was quite favorable, thanks to the supercell’s deviant SSE to S motion right into the intensifying low-level jet, which also was enlarging the 0-1 KM AGL hodograph and storm-relative helicity quite a bit. As long as the storm could forcibly inhale surface-based air in this window of opportunity (before either losing access to the boundary layer or evolving into a bow echo), it could survive, thrive and perhaps get really dangerous.
OK, that was the end of most of the jargon…
I described all that because it seems to fit a pattern we often see precede tornadoes with very late afternoon supercells that seem rather unproductive by day, then go berserk at or just after dusk. This one did!
I tried to find a vantage between Gracemont and Anadarko to watch the storm coming in from the N. It’s a good thing we didn’t succeed. Instead we turned E out of Anadarko toward a favorite viewing spot of mine — a service driveway for a hilltop broadcast mast located exactly 6 miles E of town, off the N side of Highway 9.
We parked on the big signal hill watching and photographing that (by then) spectacularly sculpted supercell move in, illuminated by lightning. While we were still parked but packing up, and right before the power flashes started, Elke saw some sort of conical downward protuberance to the near right (N) of the eventual location of the flashes, to our W, silhouetted by faint lightning flashes.
Unfortunately, my still camera had been pointed 90 degrees the other way, northward toward that great structure on the storm’s E side and occasional CGs blasting through the vault region. The last shot was three minutes before the Gracemont-Anadarko tornado became indirectly apparent to the W, off the left side of the image.
We had to bail due to encroaching CG activity (that was my widest wide-angle!); otherwise I might have had a photo of a night tornado — or at least, wheeled the tripod head around for to capture its effect of snapping utility lines. As we were pulling out of our parking place, we saw around a dozen bright power flashes in several different spots ~6 miles to the W, within a 3 minute span…in or very near the N side of Anadarko. The flashes appeared to be buried in thick precip, but were quite vivid, displaying a variety of coloration — some blue, some green, some in between, even a couple with reddish and yellow tinges.
Unfortunately, damage obviously was occurring, but we failed in our attempts to report it in real time. Elke wasn’t getting any traffic on her pre-programmed HAM frequencies. [Turns out the net control operator was sustaining damage to his own business there in Anadarko.] I tried to call the 1-800 numbers I had for the WFO but instead got either a recorded message asking me to call a different 1-800 number to “chat with friends nationwide” (for one number) or rapid busy signals (for the other). [I’ve since gotten updated contact info…thanks Rick for responding so fast!] We then lost all cell signals until we got to the S side of Chickasha, which is very unusual. Finally, somewhere S of CHK on US-81, I was able to call work and ask them to relay the delayed report of the power flashes to the WFO.
We then heard the tornado warnings close to home in Cleveland County, abandoned our (by now) outflow-dominant HP monster SW of Chickasha, and made a beeline back to Norman. We arrived after its weak tornado near Stanley Draper Lake. From Highway 9 east, not far from our house (!), we saw a ragged wall cloud moving SE across the Lake Thunderbird area before it got lost in precip and distance. The lightning show on the back side of the supercell’s HP hook was continuous and dazzling, but mainly in-core. I did capture a few CGs (like this one) looking out my east facing, second-floor window at home.
Based on WFO Norman’s damage survey the next day, an RFD caused most of the damage in Anadarko; but a significant (EF-2) tornado did track from Gracemont southward into parts of Anadarko. This probably was the conical lowering Elke saw right before the power flashes. Here was that part of the WFO report:
-
TORNADO – CADDO COUNTY – NEAR GRACEMONT TO ANADARKO
922PM-940PM
THIS TORNADO LIKELY DEVELOPED NEAR GRACEMONT AND MOVED SOUTH ALONG
AND JUST EAST OF HIGHWAY 8. THE MOST SIGNIFICANT DAMAGE WAS NOTED
ALONG COUNTRY CLUB ROAD IN ANADARKO WHERE SEVERAL HOMES HAD
SIGNIFICANT DAMAGE. THE PRELIMINARY RATING FOR THIS TORNADO IS
EF2…ALTHOUGH THIS IS SUBJECT TO CHANGE PENDING FURTHER REVIEW OF
DATA. MOST OF THE DAMAGE IN ANADARKO WAS LIKELY THE RESULT OF
STRAIGHT LINE WINDS ASSOCIATED WITH A LARGE REAR FLANK DOWNDRAFT.
Since we were perpendicular to (east of) both the tornado and RFD tracks, I don’t know how many of the flashes were tornadic in origin, and how many were caused by RFD wind. The power flashes illuminated dense precip, and were rain-wrapped. It was sobering and sad to know that people were in danger, unable to communicate about it, and only hoping nobody would get killed or hurt seriously. I’m glad and pleasantly surprised that the Anadarko tornado/RFD damage event didn’t.