Tuesday, June 21, 2011

Unusually strong cold-core tornadoes in KS and NE - 20 June 2011

I had intended not to write about so-called "cold-core" tornado settings (see this paper) for quite awhile, as most tornadoes associated with such settings are weak. Given the terrible death tolls from violent tornadoes in 2011, I feel the main focus should be on settings that produce stronger tornadoes and how we might reduce the number of deaths. But the prolific cold-core tornado event on Monday 6/20/11 in northwest KS and south-central/central NE was atypical; some of the tornadoes were strong in intensity (preliminary rating EF3, rare for tornadoes in a cold-core setting), so I felt a short write up might be justified here.

Many cold-core events start relatively early (late morning or early afternoon) due to the strong dynamics and cold air aloft. The 20 June setting was no exception. The first tornado report came near Quinter KS at late morning just northeast of the surface low, and large tornadoes were in progress in the Hill City/Norton area of Kansas by early afternoon (see Mark "Storm" Farnik's impressive photos above near Long Island KS after 2:00 PM CDT, see also David Mayhew's awesome photo above from the same storm; go to this link for an incredible sequence of his images). Tornadoes continued along a corridor into south central NE with this and another supercell, and other tornadoes occurred at late afternoon with yet another supercell in central NE near York. All these tornadic storms were northeast of the surface low near the warm front/dryline interface, a typical focus area for cold core tornadoes with the closed mid-level low and associated cold air aloft not far to the west.

What made this event so prolific with strong rather than weak tornadoes? I modified the RUC analysis sounding at Hill City at 1900 UTC (2 PM CDT, 3nd image above) for surface inflow parcels coming into the tornadic cell north of Hill City at that time on north or northeasterly winds (see surface observations northeast of the surface low on the surface map above at 1930 UTC). This resulted in an unusual amount of CAPE for a cold-core event, on the order of 2500 J/kg (most cold-core events are associated with CAPE values less than 1000 J/kg). In addition, 0-1 km storm-relative helicity (SRH) was sizable, in the range from 200 to 300 m2/s2 within a focused area ahead of the storm. This resulted in large energy-helicity index values (0-1 km EHI greater than 3.0, see 5th image above) in a corridor ahead of the tornadic storm, also unusual for a cold-core event. The sharp trough and tight closed mid-level low moving into the KS/NE area were unusually strong for mid to late June (see NAM forecast for midday above), a time when, unlike early spring cold-core events, large CAPE values can be present with surface dew points in the 60s F or higher. This large CAPE and SRH certainly provided support for larger and longer-lived tornadoes on 6/20/11 than typically associated with most tornadic cold-core events.

NOTE: Cold-core events aren't defined by whether the storms are "mini-supercells"; yesterday's case featured larger storms that were essentially classic supercells because of the larger/deeper CAPE, contrasting with early spring or fall cold-core systems that may have surface dew pointsonly in the 50s F. The important issue is the _pattern_, with the closed low aloft that organizes intersecting boundaries in a fairly narrow corridor ahead of the surface low and cold air aloft as the midlevel system moves east or northeastward, setting the stage for potential rapid tornado formation. This is a common repeating pattern with cold-core systems that produce tornadoes, and can be very useful operationally in anticipating such events.

An additional factor on 6/20/11 was probably steep low-level lapse rates in the vicinity of the tornadic storm at early afternoon over northwest Kansas. On the last SPC mesoanalysis graphic above, notice the axis of steep lapse rates extending through the surface "dry slot" over western KS into the environment of the northwest KS tornadic storm at 1900 UTC, along with plentiful 0-3 km CAPE. This combination could enhance stretching in low-levels that would already be rather strong due to the cold air aloft from the nearby midlevel low, even as cloud bases might rise some with this hotter/steep lapse rate surface air mixing in from the east from the surface dry slot (again, see surface map above). Indeed, many of the tornadoes formed quickly and were "dusty" (see the Mayhew and Farnik photos above), suggesting non-mesocyclone stretching processes similar to "landspouts", but combining with supercell processes involving strong SRH (see SRH field, also in the last graphic above). According to both Storm Farnik's and David Mayhew's descriptions, multiple dusty tornadoes were sometimes in progress simultaneously in close proximity, suggesting that enhanced low-level stretching processes were ongoing in the northwest KS storm environment northeast of the surface low.

Thankfully, there have been no reports of injuries with any of Monday's tornadoes so far, even though damage as high as EF3 in intensity was observed with at least one of the tornadoes in northwest KS.

Thanks to "Storm" Farnik for sharing his excellent photos and descriptions, and to David Mayhew for allowing me to use one of his images. Check out David's awesome and beautiful sequence of shots from this tornadic storm at:

Jon Davies - 6/21/11


Evan Bookbinder said...

I guess I disagree with assigning these tornadoes a cold-core tag. The two primary storms responsible for vast number of tornados from Quinter to the N/NE occurred along an intersection of 3 sharp boundaries (including the first storms own gust front...whose NE surge led to the demise and rapid development of the storm to the immediate N).

The primary inflow trajectories came from parcels arriving from the northeast, utilizing the westward extension of the warm sector being drawn into the sfc low. The ambient airmass in this region was in the mid 80s with dewpoints in the 65-70F range, not something I would consider typical of a cold-core setup.

I am not disagreeing that the presence of multiple tornado simultaneously was certainly due to the huge volume of low level stretching present along this incredible boundary interaction, nor am I disagreeing that the two primary storms moved into (storm 1 ) or develop in (storm 2) a region of very steep low level lapse rates.

Instead, I saw a decent elevated storm over NW KS that crossed a boundary allowing it to tap the warm sector reservoir and extreme low level helicity. The radar structure change was nearly immediate (see my posts on Umscheid's FB prior to tornado touch down).

I think the RUC soundings and sfc environment were typical of classic supercells, just with an unusual storm motion/orientation, developing earlier in the day simply due to the dynamics and forcing present.

I *would* concur (although not addressed in your write-up) that the supercells and tornadoes that developing in a similar region late in the afternoon WERE definitely cold core driven storms. Although no chasers were on them, that environment certainly met the criteria you've set forth over the years.

Evan Bookbinder

Jon Davies said...

Hi Evan:

I appreciate your comments.

The bottom line is: without the presence of the cold-core low system aloft nearby and the resulting spiral organization of surface boundaries and air masses close to the upper low (a consistent repeating pattern with cold core systems that produce tornadoes), this event would _not_ have occurred. This is particularly true with the tornadoes occurring so early in the day and forming so quickly and prolifically.

Much of what you describe with the surface pattern ("westward extension of the warm sector into the surface low" from the northeast in a narrow band around the advancing dry sector; the presence of the intersecting boundaries you mention; the unusual boundary orientation and storm motions) is rather unique to the surface organization of cold core systems that produce tornadoes (go back and read my paper which focuses on a _repeating pattern_, which matched yesterday very well). Without the cold-core system and its orientation, those features would not be there in the first place!

Core-core cases aren't defined by whether the storms are "mini-supercells"; the more rare late spring/summer cold-core cases usually involve larger/deeper CAPE and resulting more classic supercells. It is the pattern and organization that are important for recognizing and forecasting them. Making more technical distinctions regarding categorization in these cases simply is not useful operationally.


Adam L said...

Awesome write up Jon as always. When I was looking at this system I made several comments and observations that this system appeared to have characteristics of a cold core setup yet I wasn't convinced we could have a true cold core setup this far south this late in the year. I definitely appreciate your analysis to confirm my thoughts and help me learn. Thanks!

Andy Fischer said...

The insane 06-06-99 cold core event in northeastern ND was thermodynamically similar, in that it was a "high-CAPE" event with upper 70s over upper 60s F at the sfc. Then again, that event also had virtually every possible "classic" hallmark of a Davies cc set-up in place as well!

Despite the proximity of the 7H-5H lows yesterday, temps above the BL were warm-ish... that's maybe the biggest difference I see between yesterday and other cc events, though with the very deep and rather rich moist layer, CIN was obviously eliminated anyway, w/ moderate-stg low-level CAPE in place.

Exactly where yesterday's storm traveled relative to the sfc boundary(/ies) is somewhat subjective IMO, but I've documented cold core tornadoes in past set-ups that were definitively located in the open warm sector... not traveling to the cool side of a boundary or tied to a boundary intersection.

Evan Bookbinder said...


I certainly see where you're coming from, and think this case opens up for further discussion.

What say you and Shawna, casino tonight? :)


Jon Davies said...


Good points... you're right. The temperatures above the boundary layer were quite warmish for a CC event, particularly at 700mb to the east of the 700mb and 500mb lows. However, the original tornadic storm was associated with a sharp spoke of energy coming around the midlevel low (quite visible on satellite), and the "capping" to the east seemed to really help the original storm by keeping the warm sector energy focused on one storm for quite awhile, without other storms popping to compete.

Yeah, there are some CC events where tornadic storms occur in the broader warm sector farther to the southeast... it's hard to figure when that will happen, and even harder to define what storm areas might be called "cold-core" or are "too far" away from the midlevel low (probably impossible to categorize). Lower LCL heights along the warm front boundary to the east or northeast of the low seem to help with tornado production, as well as increased SRH in that area; if LCL heights are significantly higher in the open warm sector, and SRH is less there due to winds that are not as backed, that may hinder things tornado-wise in the broader warm sector to the southeast... that seemed to be the case yesterday.

Jon Davies said...


Thanks for the invite. Shawna and I are pretty busy the next few days trying catch up on things post chase season. The casino hasn't been a healthy or productive outlet for us this year, either :-).


Jon Davies said...

Thanks for your comments, Adam. I approach CC events from a pattern recognition perspective, and yesterday fit that nicely.

Some would argue that when the 700mb and 500mb temps over the target area ahead of the midlevel low aren't that cold, like yesterday, it's not a "true" cold core event. But, even in those cases, the rapid erosion of temperatures aloft near the midlevel low can make things happen early and quickly (like yesterday). And the CC low aloft can tightly focus spokes of energy and the surface boundaries ahead of the low, setting the stage for things to crank rapidly.

I challenge anyone to find such a whacky surface pattern (tornadic storm _west_ of the surface dry intrusion with northerly inflow winds) that produces a long track tornadic supercell starting around noon in mid to late June, without it being associated with a cold core type system aloft.

Walker Ashley said...


Thanks for the interesting write-up. I don't have a bone to pick regarding the cold-core tag, I guess. As with everything else in meteorology/nature, there is a wide spectrum of these type of events. One could argue this is on the "warmer" side of the cold-core spectrum with the mentioned moderate CAPE, high Tds, and warmer 700 Ts.

Though I'm no expert on landspouts and their genesis, I'm unsure if I'd classify the many dusty tornadoes we saw as the landspout variety. These were all under areas of intense rotation (i.e., small, but extremely intense mesocyclones) along the S-N running FFD that continually were ingested into the main core. It was one of the most interesting things I've ever seen.

I do have a few intriguing animated gif/lapses illustrating the processes here: http://chubasco.niu.edu/filebox/events/11.06.20/ The most interesting is the first one which was looking west "ahead" of the northward moving Hill City beast, which had turned into a rain-wrapped barrel cone at this point (not pictured, but to the left of the frame). The first animation was taken between Almena and Edmond. The latter four animated gifs were taken north of the Stamford, NE tornado, which is not visible due to the wide-angle lens used. The tornado was embedded in the rain found to the left, and would make an appearance only briefly from our perspective.

I'm not sure how I would describe the tornado-producing processes within these storms. Maybe something like this ... you'd have a more classical (whatever that is) low-level mesocyclogenesis below the mid-level meso, with standard handoffs, while at the same time a continual pulse of smallish, intense, low-level mesos ahead of the larger mesocyclone to the south that would be slowly ingested or rotate around the "main" meso. I only wish this storm would have moved a bit slower and to the east to allow for more visual investigation -- it was a tough chase with its motion, lackluster dirt road network, and circulations forming all around us.

I'm interested in your interpretation.


Adam L said...
This comment has been removed by the author.
Adam L said...

Exactly Jon, I was in the hotel that morning going over the forecast with Skip. He kept making comments about how wacky the pattern was and I told him "I know, this looks like a cold core setup only its June and its warm."

What you are saying is exactly correct. I too was stuck on the notion that to be a true cold core needing that -20+ C low but you are right, its all about the pattern. I recognized the pattern, just didn't believe it to be true.

I wish I would have though, I would have planned my chase strategy much different to be on that storm at birth, instead we sat in Kearney watching it on radar too long and by the time we made our moved we got there for its last tornado.

I remember at one point on the meso analysis page you could see the theta-E feeding directly in that storm, you could see it riding up the boundary on radar. Just amazing to watch that all work.

I will [hopefully] not make that mistake again! Especially knowing what I know now after reading your blog.


Jon Davies said...

Great animations, Walker!

I agree... the primary processes with the 6/20 nw KS tornadoes were definitely supercellular. That's very evident from the structure and motions in your images, as well as in the Mayhew images.

What I meant to suggest was that some of the rapid tornado formation (other dusty tornadoes formed quickly north of the one Farnik and Mayhew photographed; two were even rotating around each other according to one description), particularly east of Norton, may have been assisted by steep low-level lapse rate air (enhanced stretching) being pulled into the storms on their east side along the edge of the dry slot. That steep low-level lapse rate air is something found in many landspout and gustnado cases (without the supercell environment ingredients), as well as some cold-core cases. Here it may have been a part of the mix, but that's speculative and I should not over-emphasize that point. Certainly, the primary processes here were supercellular.

Thanks for sharing your excellent animations... Very enlightening!