As a part of my third HokieStorm Chase excursion last month, I had the incredible opportunity to witness the Dodge City cyclic supercell that dropped a highly disputed number of tornadoes. It was by far the most spectacular tornadic display from a single cell of my career as a meteorologist, so I thought I would look a bit deeper into the setup that made the storm possible.
Morning Environmental Analysis
A surface low was positioned along the OK/TX/CO/NM intersecting borders, expected to migrate north and east through the early afternoon hours on May 24th. Its associated stationary frontal boundary weaved its way NE through eastern CO and western KS, while a dryline extended SW through the TX panhandle. Two overnight MCS clusters continued east through central Oklahoma and Kansas, providing a slow moving outflow boundary along the KS/OK state line just after daybreak. Taking a look at the Dodge City, 
KS 12z sounding, most of the ingredients for an explosive atmosphere were already in place. Surface CAPE > 2000 J/kg; MUCAPE > 3000 J/kg; 0-6km shear ~ 30kts (expected to increase); Mid-level lapse rates 8-9*C/KM; and the beginning stages of backing winds at the surface really set the stage. A bit of daytime heating + the influence of a lifting mechanism and the atmosphere would be loaded.
Overall, the SPC was generally in agreement for their morning update. An explosive day, outlooking two main areas with an “Enhanced Risk” of severe weather. The first near the KS/NE/CO border where backing winds around the Low would be the greatest…and a second along the dryline bulge around the KS/OK border.
As expected, the daily surge East of the late spring dryline was aided by the lows migration NE through the early afternoon hours. Initiation began in the early afternoon hours, with the first storms of the day located mainly in the north outlooked area. Around 3pm, a cell popped up near Scott City Kansas, and remained discrete for quite sometime producing the first sustained supercell (and tor warnings) of the day in central KS. Stretching to the south, the cumulus field began building along the stationary front and outflow boundary in southern KS.
In the meantime, the environmental recovery struggles in north-central OK became evident, with CAPE and shear values much lower than most models thought they would be.
5:33pm Appleton, KS.
By 4 – 4:30pm CDT, the first sustained cumulus towers began building SW of Dodge City, finally breaking the weak CAP. Initially, the storms were multicell messes, with cell mergers the theme of the hour, but it didn’t take long for things to get organized. The first Tornado warning on the cell that became the “cell of the day” was issued at (TIME), when the storm was still multicellular in nature. The warning was issued due to a “spotter confirmed funnel cloud.” While I can’t be certain that a cell merger or boundary interaction didn’t produce said funnel, I believe this report was a bit pre-mature by an overanxious spotter (which happens more than people realize). Either way, I certainly don’t fault the National Weather Service DDC for issuing when they did. If nothing else, it increased lead-time for the dangerous situation that was about to unfold.
Above illustrates the progression through the initial supercell stage of the storm until the first tornado touchdown. First image depicts our view at 5:37pm CDT where the last two were taken right at 6pm CDT as the first Tornado touched down. 6 o’clock magic anyone?
4 Panel KDDC Radar Loop (5:57pm – (TIME)) Z, BV, CC, ZDR
The radar loop above shows the incredible evolution of this storm, beginning just prior to the first Tornado touchdown through the majority of the event. A few interesting things of note:
The initial tornado rotational signature formed in an area of little to no reflectivity values. This likely occurred because of a few reasons, but the most likely dealing with the positioning of the WSR-88D system itself. By attempting to look through a massive hail/rain core into main circulation center, data can be obscured or lost, especially at the lower levels. It also could have to do with the initial lack of a full condensation funnel at the surface in the early stages of the first tornado.
The evolution of the velocity couplets in this storm was incredible. Anywhere from 3 – 5 (or more) separate couplets can be seen, including an incredible split as the first occlusion process takes place. I also found the couplet “wobble” that took place just as the storm passed west of Dodge City very interesting. It was almost as if the storm was attempting to occlude for a second time, which could have caused the new wall cloud to sputter further east over the city. Certainly an appropriate name for the city.
A debris signature showed up numerous times in the Z, CC, and ZDR data. It is most evident when the first tornado’s condensation funnel continued to grow to near “wedge tornado” size to the SW of Dodge City; and again as it impacted the western portions of the city during its second cycle.
One of the most interesting parts of the event for me, was the warning progression throughout the event. On one storm, we were able to witness every “stage” of the new Impact Based Tornado Warning process.
Base Tor Warning -> Confirmed Tor Warning -> Considerable Tor Warning -> Tor Emergency
Having the opportunity to witness the event on the ground, comparing it to radar data, and seeing the impact based tags issued by NWS DDC; it really shows the benefits of the new impact based warning methodology which is becoming widely implemented and accepted across the NWS.
The results of the Dodge City damage assessment is presented above. The office concluded that three different tracks resulted from this storm.
1.) The initial touchdown which continued onto becoming an EF-3 wedge tornado that occluded just to the SW of Dodge City and ENE of Ensign, KS. Stayed pretty rural.
2.) A brief (EF-1) rope that formed and tracked through a few fields SW of Dodge City (Likely the one pictured above acting as the “Twin.”
3.) A third (EF-2) that formed during the occlusion process, and from our point of view, gained multi-vortex qualities as it began to impact the western suburbs of Dodge City.
4.) A fourth (EF-3) touchdown occurred just north of Dodge City, again, likely prior to the dissipation of the third Tornado described above.
Somehow, injuries remained minimal during this event and no recorded deaths resulted from this series of tornadoes. Luckily, the cyclic supercell heading right toward Dodge City somehow just skirted it’s western suburbs, but not without putting on one heck of a show for seemed like thousands of storm spotters/chasers. Some impressive video of the event from multiple sources is provided below.
Taylor Kanost (WFXR Roanoke, VA) Timelapse
The biggest draw of a storm chasing experience for me isn’t the pictures, the people, or data gathering. The largest benefit is the experience gained forecasting. The ability to make a forecast in the field, make hard (and costly) decisions on that forecast, and to view the results first hand is an incredible opportunity. This gives you an environment to learn from your mistakes and successes, while also giving you the opportunity to share and learn the tools, tricks, and general experiences to/from others.
In our case, the student forecasting was incredible. Each active chase day, we found ourselves in the correct area synoptically. In fact, we found ourselves nestled into a Tornado Watch Box prior to issuance 5 days in a row before ending the trip. Not surprisingly, things weren’t perfect when it came to finer scale forecasting (where will the first storm go up…etc), however, we did land on 2 of the 3 best storms of the trip. In a normal year, catching even 1 of the 3 best storms would have been a rare and major success. The only storm we missed out on was the long track 
brilliant views on D4, to multiple Tornadoes and beautiful Supercell structure on D6 + D9; it is hard to deny how well the atmosphere treated us.
Taking a look at the visible/IR satellite this AM. An outflow boundary from an overnight MCS in north-central KS is visible from the OK panhandle/KS border stretching ENE through southern KS. The progression of this boundary actually seems a bit further than the SPC originally expected, which certainly could come into play a bit later today. Otherwise, heating is underway for about half of the eastern TX Panhandle + western OK.
A Tor watch was issued as we continued NW in Kansas. After a brief lunch/restroom stop, we decided to shoot North in central Kansas, in an attempt to give a bit of distance between us and the dryline/sfc low bulge. We quickly backtracked once a healthy looking tower shot up over our lunch stop area, and decided to setup shop just to the southeast of Dodge City, KS…where we saw this monster mature through every stage of a thunderstorm lifecycle:
Visible Satellite imagery this AM show the remnants of an MCS continuing east through the Red River Valley region of TX/OK. A few boundaries are clearly evident on the loop, the first being an outflow feature stretching nearly W to E across the TX Panhandle just south of Lubbock turning north into far W OK. A few other minor boundaries are also in play, and will likely be a focus point for initial convection later today due differential heating and a steep moisture ridge.
12Z Surface Observations show multiple surface lows beginning to form along the TX/NM/OK borders. An associated cold front extends SW through east-central NM lagging behind a dry line feature surging E currently oriented along and west of the TX/NM border. SFC winds are generally south to southeasterly just east of the dry line with the exception of the eastern red river valley currently being influenced by the remnants of an MCS. Other outflow boundaries left by this system also have some micro scale effects on the screen winds and moisture transport this AM.
Overall, a good chase day! It was a bit slow to get going, but luckily we made just enough “good decisions” to end up on the only Tornado producing storm in TX during the daylight hours. While we didn’t see the tornado itself, we did end up coming across its damage path which included a tractor trailer blown into the median of I-40 a good bit east of Amarillo. Chasers actually got footage of the incident, and I was able to contact NWS Amarillo on their behalf to get the report submission process going. No report on the SPC storm reports as of this morning, which likely just indicates a delayed report and/or potentially deeming the incident to be caused by straight-line winds. After viewing the video (bad lightning to watch it), it could have gone either way, however, timing-wise and witness accounts makes me lean toward a weak and brief Tornado Touchdown. A tornado warning was issued less than 30 seconds after the incident.
Day 4 started out very similar to Day 3, with not much in the way to look forward to chase-wise. Ridging continues in the southern plains as a late season cold front continued to wipe the atmosphere clean of moisture throughout all of TX. Waking up in the panhandle of Oklahoma, we were only a short ride from favorable upslope convection type terrain in NM/CO. Lucky to see CAPE values > 1000; and non-existent forcing/shear, expectations weren’t high. A picturesque thunderstorm would be considered a success… and boy did we exceed those expectations.
Not much to add from yesterday. As mentioned prior, bust potential was rather high due to the lack of lift, and pretty stout cap in place. The closest storm reports to us were well to the North, in East-Central WY and the Dakotas. Luckily, we knew this going in, and didn’t dilly dally too long in the north, potentially putting us out of play today. After an extended lunch overlooking the Rockies and Denver International Airport, we hauled South to Dumas Texas for the night.
Decided to punt the Rio Grande Slight Risk for many factors, but mainly poor chase