Millsap TX Tornado and Supercell
Brock TX
15 May 13
SHORT: High-based junk storms watched near ABI. Headed E on I-20 toward confluence-line towers moving into high-SRH environment. Saw eventual Granbury supercell looking S and almost went after it. Instead took direct I-20 route to explosive, then-bigger storm to our NE. Watched excellent structure, brief/weak tornado(es), longer-lasting Millsap tube, and messy/multi-vortex circulation from Brock exit. Road-screwed for both storms after that. Tried to glimpse South Dallas storm after dark, too much rain.
LONG:
Pre-storm setting and early dryline convection
In a year that had been rather scant for observable tornado potential, we didn’t truly expect this day to yield much, either. Maybe it was pessimism. Maybe it was an unspoken wish not to jinx the chase, despite our disbelief in such paranormality. Maybe it was the peculiar pattern, with some split flow and the shrunken remains of a southwestern upper-level low slowly ejecting across Oklahoma. [For decades, didn’t Al Moller often extol the virtues of split flow aloft for outstanding Texas chasing?] Maybe it was the vibe that this was one of those down years (shown to be decidedly false by month’s end).
Reasonably large hodographs were reasonably forecast for part of north-central and NW TX, but generally displaced 100 miles E of the dryline, near the western lobe of a corridor of very mist return-flow air. Perhaps if an initially high-based storm firing off the NW TX dryline bulge can last long enough to get into that returning moisture and backed surface flow before dark, it could get happy. Perhaps this, if that…you know the saying: “If ifs and buts were candy and nuts, it would be Christmas every day!” Whatever the reason, we headed southwest on a classical spring chase trip to NW TX talking supercells but not tornadoes.
Rich and I decided to play the belt of enhanced midlevel winds and deep shear, S of the OK low aloft, W of the northward moisture bulge and E of the eastward dryline bulge. Initial target was the Haskell-ABI area, with (naturally) the season’s first Allsups burritos consumed along the way in Seymour. While at an abandoned farmstead N of Anson, the first deep towers erupted just to our WSW, practically in our laps and as if on queue.
However, more, even deeper towers could be seen in the hazy distance SW of ABI. Not seeing an appreciable difference in the environment of either, and both being in our target area, we decided to head toward the southern towers and keep the northern ones in view. The southern convection stayed more vigorous, and even garnered a severe warning–but turned to virga-blowing mush by the time we got S of ABI and in good intercept position to its E.
Meanwhile, the high base of the original convection, now two counties (or about 50 miles) to our N, could be seen clearly despite its distance, beyond the anvil material of the dying ABI storm. A quick check of the surface map and objectively analyzed moisture and CAPE fields revealed that the northern cell was a flimsy hope for making it far enough E to tap the reservoir of upper 60s surface dew points S and E of MWL, and (for now) shallow but building towers could be seen in the distance, corresponding to a confluence line located along the W side of the moist plume.
Maybe the northern storm somehow could get far enough east (unlikely). Maybe the cap could weaken enough along the confluence line to blow a storm or two and send it into the good juice (an intriguing but highly conditional possibility at that point). For either scenario, we would need to head back to I-20, then ENE at least 50-60 miles. If ifs and buts were candy and nuts…
As we scooted E of ABI and the northern storm shriveled, we were at our most pessimistic, as were many other storm observers–some of whom immediately abandoned the chase to start heading NW toward forecasts of central High Plains storm potential in the following days. We didn’t have that option, and besides, something very intriguing was happening across the eastern sky.
Target shift
We hung our hopes on those gradually deepening confluence-line towers–convection that got ever deeper and thicker as we rolled along the slab. The closer we got, the better it looked. The bigger towers began to sort themselves into young storms, and some of the development to our SE, ESE and ENE looked like it was going to erupt into big action very soon. The situation definitely was passing the eyeball test for those of us who have been doing this awhile.
Parameter check: effective shear was fine…SRH increasing E of the confluence line…CAPE was large, and we know what big CAPE does. Accordingly, an elongated bull’s-eye of significant tornado parameter (an index we invented, so we had better pay heed!) showed up and increased downshear from the building storms and W of I-35.
I could detect the scents of delectable convective comestibles cooking in the kitchen of the sky, soon to be placed on the smorgasbord of atmospheric violence for us to devour! The question was becoming: which entree shall we select?
By the time we approached exits leading to BWD and MWL, one healthy-looking storm already was beginning a right turn near BWD, and was warned. Two younger and closer ones were apparent and looking robust visually, but not yet as impressive on radar–approaching SEP (this became the eventual Granbury/Cleburne storm) and near MWL (the eventual Millsap storm). We had our pick of the litter, and ultimately wouldn’t have gone wrong with either of the two closer choices. Though the southern one of the closer pair initially was smaller and took longer to mature, it ultimately produced a violent tornado in deepening darkness. We briefly stopped briefly to look S at that young, future Granbury/Cleburne storm to our SSE, small but already showing some visual supercellular characteristics.
Meanwhile, the cell near MWL went berserk in ten minutes–two volume scans–explosive towers evident through intervening low clouds in the near NE sky, and remarkably rapid growth in areal coverage on the radar screen. Between that trend, the choppier terrain in the path of the much smaller SEP storm, the presence of an Interstate to take us straight to the northern storm in legal haste, and the poorer road options S of I-20, we quickly selected the MWL cell for close investigation. Along the way, and between hills and other obstructions, we saw a wall cloud quickly form and occlude (without tornado) under the base. This storm went from disorganized mess to serious supercell in less than half an hour, and as Jim Leonard might say, “This is serious business now”.
Smorgasbord delivered
Finding high yet accessible vantages in that part of north TX is a challenge; so we grabbed the easiest and first spot we could find in the storm’s inflow region — a staging area for construction work right off the Brock exit. Unfortunately, this means there is road-work material and/or the Interstate in the foregrounds of some of the images; but I guess they tell a story and add some foreground texture too. As for the storm, what a structural delight that greeted us! We remained at this spot, or just across the overpass (later stages) through the entire tornadic stage of this storm.
One fairly low wall cloud developed while the storm still had most of its spectacular deep-layer structure visible, with a turquoise tinge starting to show up in midlevels. A few minutes later, around 1841 CDT, a funnel (deep enhancement) developed on the near left edge of the wall cloud, with scud sometimes rising rapidly right off the ground and more slowly rotating into it as part of a distinct, helical column. It was about as weak and low-end of a supercellular tornado (deliberately underexposed zoom) as one can imagine, but nonetheless, there it was. This little vortex lasted just 2-3 minutes and whetted our appetite for more.
In the succeeding 20 minutes after that “wimpnado” ended, a big gob of rain slowly wrapped around what was left of the wall cloud — now becoming ill-defined — in the form of a precip-filled RFD. Then, at 1903, another funnel (deep enhancement) became apparent within the translucent bear’s cage, a little more distant from us (maybe by a couple miles) than the first vortex, and near Millsap. Though no power flashes were visible, faint and low-contrast ground contact of condensation in these early stages confirmed another tornado for us–albeit cheesy at this stage.
Quickly, the whole mesocyclone wrapped deeply in rain and completely obscured any remains or evidence of a tornadic vortex within. As such, we do not know if there is true physical continuity between that vortex and the next one in the same general area. However, this view at 1907, as the precip began abating and the mesocyclone got more deeply occluded, shows no obvious visible evidence of a continuous tornado.
Here’s the next visible vortex! This showed up suddenly at 1909 (deep enhancement), representing the start of the “Millsap tornado” that many observers watched. It may or may not have been continuous with the previous tornado. Some translucent curtains of rain still were orbiting this increasingly deeply occluded and quasistationary meso. The tornado became better visible as a classical, partly rain-wrapped cone beneath a scuddy and ragged storm-scale circulation.
Meanwhile, the parent supercell was moving slowly SE, away from the tornadic circulation–effectively kicking the old occlusion farther back through the rear of the storm. A new mesocyclone started stem-winding just ENE of us by just a mile or two, at most. We had to keep our heads on a swivel, watching the adjacent, tightening area of rotation while admiring the tornado from a greater distance.
Look here, look there…left, right…look here, look there! Someone watching a camera trained on me might have thought I was observing a slow-motion tennis match. Unfortunately, I didn’t shoot the closer meso yet because our ideal tornado vantage on the SW side of the interchange put the Brock Road bridge in the way of seeing much beneath the circulation that was practically in our laps. What a mesocyclonically bipolar conundrum! I was ready to run over or under the bridge, however, to shoot right down the Interstate in case the nearby circulation tightened to tornadic intensity.
It’s a good thing that didn’t happen yet, because the meso was smack-dab over the road to our immediate ENE, with a good deal of truck traffic zooming this way and that.
With the tornado still in progress to the NW, a brief, small funnel appeared to the N, but with no ground circulation evident beneath. The funnel sprouted from the cyclonic-shear zone (and likely ribbon of low-level vorticity) in between but connecting with the two mesocyclones. Clearly, this was a much different environment than our moisture-starved convection out by ABI.
Back on the occluded, certifiably tornadic area NW of us: the rain gradually fell away to reveal a classical tornado specimen in its full splendor, going through assorted conical forms from nearly symmetric to tapered and curved for several minutes, finally roping out with a twist. A detached condensation puff marked this tornado’s dissipating gasp at 1919, and its parent mesocyclone very quickly followed suit.
No time existed for basking in the joy of a harmless and beautiful tornado as just seen. The nearer mesocyclone actually retreated N of the Interstate, exhibiting a similar storm-relative backtracking as the precious attempt–but without the long-lasting, high-contrast tornado. Instead, it offered a short-lived, fuzzy, scuddy, ragged, multivortex circulation that lasted 2-3 minutes starting around 1922. We promptly drove the hundred yards or so back onto the eastbound service road and over Brock Road to get a better view, by which time the multivortex action to our NNE had consolidated beneath a fat little tube that was coiling ragged scud off the ground.
Fortunately, the brief tornado continued to back away from the Interstate, soon dissipating. Yet another meso was developing downshear, a few miles to our SE, as the supercell jumped flanks.
Decisions in Dallas darkness
Unfortunately, this southeastward, effectively discrete propagation led us right to the W edge of a road void ENE of Dennis, looking E at an increasingly rain-wrapped meso. We had no viable option back around to the inflow side, and no view within. A very well-defined radar hook, a tight velocity couplet, and rapid N-S precip motion on the back side left us wondering just what was happening “in there”. The precip was just too dense to see through, and the viewing angle was bad (except for the rainbow across a field of flowers).
Refreshingly rain-cooled air, tinged with a blended floral and earthy scent, carrying bird songs, was a consolation for an inability to ascertain what the later storm reports indicated was nothing of importance happening inside the mesocyclone. We surprisingly encountered Jack Beven and Margie Kieper on this remote stretch of unmarked back road NNE of Dennis; they had seen the main Millsap tornado from the distant E after a late start from visiting friends in the DFW Metroplex.
We all could have made the loopy plunge southward to intercept the now tornado-warned Granbury-Cleburne storm (whose flanking towers were visible and not far away) at or just after dark. Common sense prevailed. Core-punching a tornadic supercell, from the N or NW, on twisty and hilly roads with few escape options, in deepening darkness, did not appeal. Instead, Rich and I grabbed dinner at Whataburger in Weatherford, then headed back toward FTW for the turn northward and homeward.
Along the way, a small supercell with a nice hook erupted over Arlington and cruised E across the Oak Cliff section of SW Dallas. meanwhile, the Granbury storm was headed ESE toward I-35 and could have been targeted (again, we decided against that). Though it was dark, the projected path of the smaller Dallas storm took it toward south Mesquite; so we jumped down to I-2o to get into its inflow region. Very shortly after doing so, we encountered a huge traffic jam…going the other way! A wreck had the westbound Interstate backed up for miles; but our side was smooth sailing. Before we got into decent position SE of the storm, though, a dense cluster of heavy showers formed and moved right over us; their outflow weakened and ultimately killed the south Dallas supercell. We turned N through its feeble remnants, wheeled around LBJ freeway and drove on back to Norman through occasional elevated storms.
A north Texas chase day finally bore fruit for us with a classically structured supercell and a few low-impact tornadoes casually observed from one spot. The bad news was that the storm we didn’t chase (and wisely so) produced a terrible killer tornado near Granbury, and a big, destructive wedge after dark, around Cleburne. This marked the sad day that the switch had flipped on from a shockingly inactive tornado season to a dreadfully busy and tragic two-week stretch, at least for the southern Plains.