April 26 Southern Kansas Chase

by John Farley

Forecast elements:

Starting five days or so ahead, the attention of the storm chasing world was on this system. Lots of chasers were excited about the possibilities, and I admit I was, too. However, a lot can and almost always does change from weather forecasts that far ahead, and by two or three days ahead, it was clear that this system was not the slam-dunk it looked like earlier. While there were consistent forecasts of very high instability (CAPE in the 3,000-4,000 range), there were still a lot of questions. There were questions about the timing of the jet stream relative to the daytime hours when the best stability would be in place - would the upper winds get there in time? There were questions about whether storms would initiate during the day, or wait until after dark, as some models were suggesting. And there was a persistent signal of a veer-back-veer (VBV) wind profile. For the non meteorologists reading this, the most conducive wind profile for tornadoes occurs when wind veers with height - in other words, it is southerly or southeasterly at the surface, then gradually comes from a more westerly direction as you go higher in the atmosphere, so it ends up in the upper atmosphere at something like a 90 degree angle to the surface winds. If the winds gradually turn this way as you go from the ground up, and also strengthen the higher you go, you have a wind profile that is favorable for updrafts to rotate consistently as they rise from near the ground to the upper parts of the storm. This kind of wind profile is very favorable for tornadoes. But with the VBV profile, the wind turns with height in a less consistent way - perhaps starting southeasterly at the surface, then as you go up shifting to south-southwesterly, but then at some point higher up backing again so that it is coming from the south or southeast, before finally veering back to southwesterly in the upper part of the storm. This pattern tends to produce a messier pattern in the updraft than the consistent ground to cloud-top rotation you get with more consistent veering with height. This can make the storms rain on their updrafts, or rain to wrap around the updraft, or lead to more complex multi-cellular storms rather than an isolated supercell storm. If the air is unstable enough you can still get big hail and perhaps high winds with the VBV profile, but strong, persistent tornadoes are less likely.

Still, mostly because of the very high instability forecast, a lot of forecasters and chasers felt that the possibility existed for quite a few tornadoes and some strong, long-track tornadoes. But it was clear that this was a possibility, far from a certainty. Still enough of a possibility to get a lot of chasers, including me, out to the plains to see what would happen. From the standpoint of avoiding huge chaser crowds, one positive point was the large area with severe weather potential. Tornadoes and other modes of severe weather would be possible anywhere from the NE-KS state line all the way down through KS and OK and well into TX. At least with so many possible targets over such a large area, the chasers would not all be in one place. Still, there would be a lot of chasers out.

I was interested for a couple days in the southern KS area. I considered northern KS closer to the triple point surface low pressure center, but was concerned that storms there could be messy and low-based, so even though there was somewhat better potential for tornadoes there, they might be hard to see. And then the predicted position of the triple point shifted north, making that a longer drive. The TX target was an even longer drive, and I was determined to avoid central OK because I do not like chasing within 50 miles of OKC due to the crowds. So Monday I drove to Dodge City to pre-position and then decide on a more precise target, probably somewhere south or west of Wichita. Another consideration in picking a KS target was that there were some signals for the possiblity of a secondary low forming around the TX panhandle and moving northeast into KS, which would strengthen the wind profile, back the surface winds to SE which would add more low-level wind shear, and perhaps act as a trigger for storms in southern KS.

Forecast issues the morning of the chase:

All over the eastern half to one-third of KS, the models were predicting high instability with CAPE in the range of 3,000 to 4,000 j/kg and in some cases higher. The question was whether there would be enough trigger to get storms - even very unstable air can end up having no storms if there is not a sufficient forcing mechanism to generate strong updrafts, or if a cap (layer of warm air aloft) prevents sustained updrafts from forming and rising into the upper atmosphere. And the VBV issue discussed above was still there. I was concerned that morning models were predicting a lot of convection in the woods east of I-35 rather than on the dryline, meaning less in the way of isolated daytime storms in good chase terrain (west of I-35). However, I did see signals from several models that somewhere near the OK/KS border west of I-35 might have potential. They were breaking out more isolated convection in that area, and also showing signs of some backing of the surface wind there. Hence, my morning target was around the Attica-Harper area.

The chase

I drove to Harper, had lunch, and checked data. It was clear that the dryline was a little farther east than I had figured, so I decided to head a little farther east. I soon could see towers going up to my south, so I went far enough to get east of them (about midway between Harper and Wellington) and then headed south. This was around 1:50 p.m. - I would be in active chase mode for more than 6 hours from this point. Looking at radar, I did not at first see much more than a narrow band of light precipitation. It seemed that this precipitation was "training" along the dryline, which was oriented NE-SW in southern KS, roughly extending southwestward from Wichita. In other words, the showers were moving laterally along the dryline, without it either advancing toward the east or retreating toward the West. Visually, I could see updraft base pretty much everywhere along this dryline, with a few updraft towers reaching the point where the water droplets turn to ice crystals and beginning to spread anvil clouds northeast of the updrafts. After a while, the radar showed that an isolated strong storm had formed just SW of Wakita, OK. That's right, the town featured in the movie, Twister, and the home of the Twister Museum. I was ultimately very pleased with my forecast and chase strategy, because it allowed me to get on the best storm in the northern OK/southern KS area almost from the time it initiated near Wakita, OK, and then to stairstep the back roads and stay with the storm from the stateline northeastward for better than two hours until it moved into the Wichita metro area, at which point I let it go. I did not see anything that I could confirm as a tornado with this storm, but did get a couple likely funnel clouds as well as a couple impressive wall clouds.

As I approached the intensifying storm I decided to position myself along the OK-KS stateline in a location that would be just east of this storm as it tracked northeastward into KS. Quickly, it got the first SVR warning of the day at 2:18 p.m. as it began to pound Wakita with heavy rain and hail. As I sat under the anvil as the core of the storm approached the stateline, there was continuous rumbling of thunder up in the anvil. This would have been around 2:20, close to or just after the time when the first SVR warning was issued. At first, there was no visible lightning, but there had to be a lot up in the clouds to get such a steady rumbling. Then cloud-to-ground (CG) bolts began to zap down in the precipitation core to my west. I stairstepped the back roads alternating between going north and going east to try to stay on the southeast side of the storm, as multiple reports of 1-1.25 inch hail came in with this storm. It had some slightly lowered areas, but nothing too dramatic at first. Then, around 3:05, it formed an elongated lowering that looked like something between a wall cloud and a shelf cloud. At the north end of this feature an inflow tail formed, and at the south end a lowered area with some weak rotation. Then, quite suddenly, another lowering formed closer to the inflow tail, which quickly developed into an impressive wall cloud:

This wall cloud formed around 3:10 p.m., west/NW of Caldwell, just before the first of the two TOR warnings this cell would receive was issued at 3:13.

Within a few minutes, the wall cloud wrapped up in rain. It really looked like there could have been a rain-wrapped tornado in there, but there were no reports of tornadoes at this time. (video capture, around 3:16).

After this wrapped up in rain so completely that I could see no features, I had to race to get back far enough northeast to be able to see the storm's features. And the storm began to cycle, i.e. one area of rotation would wrap up in rain (occlude) and a new one would form. For the rest of the time I was on this storm, this occurred repeatedly, with each area of rotation seemingly wrapping up and occluding before it could produce a tornado.

Likely funnel cloud about around 3:36, and farther northeast (video capture, taken from north of Caldwell). I was driving when this formed so could only video out the window. By the time I could find a safe place to pull over amid the chaser crowds, it was gone. I thought for a while this could have been what led to the one tornado report that was associated with this storm, and indeed it looks a lot like the tornado image on the NWS-ICT event summary page, but the time is about 20 minutes off, as I observed this before the time of the reported tornado, and to the southwest of its reported location. However, a severe weather statement from the Wichita NWS did indicate that spotters were reporting a funnel cloud at this time, and I have since seen other pictures of this from different angles that pretty much confirm that it was a funnel cloud.

Around 3:52 p.m., a few minutes before the time of the one tornado that was reported with this storm, I observed two lowerings. One looked to me like a skinny scud feature extending most of the way to the ground. To its left (south) was another lowering that had much more the look of a funnel, but only went perhaps 1/3 of the way down at most. Unfortunately, I messed up the video of this by having auto focus off when I thought it was on, so I have no decent pictures of these features. However, neither of these features looked much like the picture of the tornado on the NWS ICT web page (also on chaser Jason Caster's Facebook page), so I doubt the reported tornado was associated with either of these features. Also around this time, golfball hail was reported a little north of where the tornado was reported. At 3:54, the second of the two TOR warnings this storm would receive was issued, with rotation indicated 9 miles east of Argonia. The warning also mentioned that spotters had reported several funnel clouds and a rotating wall cloud.

A little later and right around the time of the tornado report, I did see a lowering associated with what visually appeared to be a mesocyclone. From the time and location, this would appear to be the same feature associated with the tornado report. If it was, however, I certainly was not looking at it from the same angle as other pictures I have seen. The time (3:57) is about right, and so is the location, since a close zoom of my video reveals that this was taken at Chicasa (route 49) and 40th St., with the lowering to my NNW. That would put it 3 or 4 miles WSW of Mayfield, near where the tornado was reported. But I did not think "tornado" at the time, and since I was still having the aforementioned focus problem, the video is not clear enough to ascertain rotation. Here is the best (sharpened and contrast slightly enhanced) video capture I could get:

Later and farther north, around 4:20 p.m., I observed a rapidly rotating wall cloud, looking north along route 81 north of Wellington. Then, a little northeast of that location, a little after 5:00, intense precip shafts to my NNW, as 2-inch hail was reported west of Mulvane with multiple reports of golfball hail at other locations in Sedwick Co. including east of Clearwater. Here is a picture of that, taken at what would be the northernmost point of my chase:

Shortly after this, I decided to break off this storm to try an intercept of another SVR-warned supercell coming out of OK. The storm I had been chasing for more than 3 hours had by now moved over the Wichita area, where it, along with other storms training along the same track ahead of it and behind it, produced torrential rain, flooding, and large hail up to tennis ball size. However, I had no interest in chasing in a metropolitan area, especially with flooded roads and big hail, and besides, the storm coming up from OK was now looking better on radar and seemed to be in an environment where it could ingest juicy air not comtaminated by outflow from other storms.

I backtracked south to South Haven, then west from there to Caldwell once again. From there, I followed the paved road that goes a little north and then west, until I was on the storm. It did have a nice lowered area for a while under the updraft, which I watched along with a good number of other chasers who had moved down to this storm.

Lowered area on second storm, west of Caldwell, around 6:15 p.m.

However, after this picture was taken, this storm was looking gradually weaker and more outflow-dominant, so I decided to move on once it had passed to my northwest, and try to intercept the next storm that was coming up from OK. That one weakened, too, and I began to think about going somewhere to find lodging for the night. I also nearly got stuck on a side road when I turned off to look at the last storm, and gradually realized I had gotten west into the area the first big storm had moved over, so some of the back roads were nearly undriveable.

Around that time, I was noticing on radar a kind of a large, whirlie meso-low feature to my SW in OK. Storms seemed to be whirling around its center, and it was actually pushing the NE-SW line of storms extending from its north side back to the northwest. Here is a radar image of that feature, which I captured off GRL-3 at 2355z:

I figured this was a big massive mess of wind, torrential rain, and hail, pretty much unchaseable. However, a strorm that appeared to take on supercell characteristics formed along that northern line, to the SW of Anthony, KS. That looked interceptable, and perhaps after that I could find lodging in Anthony or Medicine lodge, as it was now starting to get dark. So I continued westward on the road I had been on most of the time since coming through Caldwell. This road eventually changed from paved to gravel, but it was harder and better maintained than the unpaved back roads, so it was driveable even when the back roads were not, though there were a few places where an inch or two of water was running across the road in the low spots.

The storm coming up from OK toward or just west of Anthony got a SVR warning, and looked supercellular. My plan was to go into Anthony, perhaps grab a quick bite to eat, and meet the storm at the west edge of town. However, new convection began to go up just south of Anthony, and the supercell storm to the southwest now also looked like it would hit Anthony. I hit a drive-through, and decided to head east 5-10 miles and watch the storm as it moved over town. However, when I stopped about 5 miles out of town, I could see an oncoming gust front just to my south so I sat tight there, observing 50-60 mph wind and torrential rain, but no hail. A new SVR warning was issued for this line of storms, which had formed over southeastern Harper Co. and surged north. This seems to have been evolution of storms on the east/NE of the meso low; it developed rapidly and was quickly surging NE. Once this passed over, that was the end of the chase and I was fortunate to find a room in Anthony, as occasional positive bolts came out of the anvil in the stratoform precipitation on the back side of the storm complex for another hour or two.

Event page from ICT NWS:


This page has pictures of the large hail, of wind damage, and of the reported tornado if you click on different tabs, a well as a radar animation.

Chase route map (approximate)

As noted, this is very approximate - I am really not sure of the exact route I took stairstepping NE with the storm, but this is the general idea. The numbers correspond approximately to these stops/highlights:

1 - I am heading south, crossing under the NW-SE line of updrafts, watching as the storm over Wakita comes into view.

3 - Storm goes TOR-warned, wall cloud forms near NE end of elongated lowered area to my WNW, lowers, and becomes rain-wrapped. Around 3:10-3:20 p.m.

4 - Likely funnel cloud visible to my NW. Around 3:36 p.m.

5 - Two lowerings: what appeared to be a narrow scuddy lowering to near the ground to the right and a likely funnel 1/3 to the ground to the left. Features are to my WNW/NW, around 3:52 p.m.

6 - Location where I viewed what apparently was the lowering associated with the tornado report WSW of Mayfield.

7 - Rapidly rotating wall cloud just to my north along route 81, 4:20 p.m.

8 - Precip shafts as large hail falls around Clearwater and Mulvane, just after 5:00 p.m.

9 - Lowering on trailing storm, 6:15 p.m.

10 - Location where I almost got stuck on side road

11 - Gust front moves over me east of Anthony, 8:10 p.m.

Video Highlights:

This video clip has 3 short segments. The first shows the partly rain-wrapped wall cloud W/NW of Caldwell. The second shows the funnel cloud as I drove north of Caldwell on route 49, looking NW. The third segment shows the rotating wall cloud north of Wellington. Unfortunately, this video is a little fuzzy because of the focus problem mentioned above, but you can clearly see the rapid rotation.

Local Storm Reports:

A list of the day's local storm reports from NWS Wichita can be seen here.

Chase miles:

Day 1, travel to the chase area, 452 miles

Day 2, the actual chase, 341 miles

Day 3, returning from the chase area, 545 miles

Total: 1338 miles.

Return to 2016 Severe Weather Observation page