Tornado near Wakeeney, KS, May 22, 2007
Three Tornadoes/Many Supercells
5721 Miles Total
TARGET: Northeast WyomingHere are Thom and me, the intrepid California chase team. We have decided to target northeastern Wyoming. The forecast soundings and hodographs for Gilette and Miles City have 1100-1400 J/kg sbCAPE, with nice anticyclonic loops to the hodographs. Remember that the "total" shear 0-6 km is the 6 km wind minus the sfc wind. That gives 37 knots of speed shear for Gilette and 43 knots of shear for MLS.
In addition, the low level shear is expected to be very strong this afternoon in that strip (you can see that strip on both the RUC and NAM forecasts) as winds come around to strong easterly and upslope. We expect a reasonable chance of rotating storms in that area today. And this keeps us close to tomorrow's provisional target of southwestern SD.BTW, every motel (with wireless) that we tried in Scotts Bluff/Gering was booked last night. So we ended up in Torrington (Holiday Inn Express)...which is already booked for tonight. High school graduations.
TARGET: +/-100km north/south Interstate 90 near South Dakota/Wyoming border
We've just had our morning map discussion...Thom and me in the motel room here in Rapid City, and Chuck on the road gesticulating wildly as he and Vickie make the transit north from Valentine to meet us. Here is Thom with Chuck and Vickie, and Chuck and me.
Anyway, NAM/WRF forecast hodographs for this afternoon for KGCC, KRAP, KCUT all look very good...for directional shear...but the best flow hanging back still gives us marginal/mediocre deep layer shear of around 35-40 knots in the I90 corridor.Then there is relentless moisture problem. Both the RUC and the NAM continue to overforecast dew points in the Dakotas (and, probably, elsewhere). In one of the verification time panels yesterday, for example, the NAM had +65F dew points to the Canadian border. I realize that much of this is evapotranspiration related...but there were no dew points close to that anywhere, even in TX, yesterday afternoon in the Plains.
On our morning run, Thom and I noticed a wind shift from ESE to SSE at around 9AM here and the low stratus cleared out rapidly..this was the synoptic-scale warm front passing through. We note that there are mid 50 degree dew points south of us, with a few high 50s.Both the NAM and RUC suggest that this makes into the southwestern quarter of SD by the afternoon, so our initial plan is to wait. We think, and agree with others, that the best combination of ingredients will be +/- 100 km or so of I90 near the Wyoming/SD border.
I wish I could see a lot better 500 mb flow, or a more significant push of moist air at the surface, for the afternoon, but I don't. The low level jet should create the interesting hodographs I mentioned above, though.
The first storm turned out to be the best. It developed on the Black Hills and quickly became a supercell. Our Mobile Threatnet is down for the count...so I have no screen captures to show. But the storm developed rotation quickly, but just as quickly became undercut by outflow. As that happened, and part sof the cloud mass evaporated, we could see funnels, none of which became a tornado. We estimate that we saw at least five brief well-developed funnel clouds.
We stayed with this storm as it moved southeastward. It was early in the day, so we expected more. Not to be the case. This was the show.
TARGET: Southwestern quarter of South Dakota
Given that we're starting the day in Rapid City, we see little reason to go much of anywhere right away. The RUC and the NAM/WRF seem to keep the surface low a short ways to our southwest by this evening, which is where it is at 8AM. Going northward very far doesn't seem like a good plan, since we expect to chase KS tomorrow. We have stronger winds aloft headed our way with surface moisture no worse than yesterday, so we'll hang around here today and go after what happens in sw South Dakota.
TARGET: Southwestern/south-central KS
No time for a significant discussion here. We needed to be on the road. All parameters, except for CIN, are favorable for tornadic supercells in Kansas, particularly the southwestern portion/south central. That's our target.
SPC's outlook agreed with our assessment. All the tornadoes eventually occurred within SPC's 10% risk area.
The pattern expected for the afternoon over the western portion of Kansas had the following features/ingredients: (a) weak to moderate deep layer shear that we expected might be the biggest issue in the development of rotating storms; (b) strong southeasterly surface flow around a triple point intersection of dry line, cold and warm fronts somewhere near Dodge City or so; (c) moderate CAPE values associated with a tongue of high dew points extending northward from Oklahoma. Forecast hodographs for the area north of Dodge City had a very strong anticylonic loop, and low level shear. But the forecasts suggested that the best shear would await the latter part of the afternoon/evening. As it turned out, the deep layer shear was just adequate as the initial storms started to form near Wakeeney, KS.
Tornadic supercells...Kansas. Sorry...in very late. Lots of pictures. We caught a storm from initiation to tornadic development. We also tried to intercept the storm just west of the first...a wildly rotating storm. But it's 11PM now...and time to clal it a night.
Joining Chuck, Vickie, Thom and me were Scott Landolt, Dan Porter, Chris Juckins and Rob Handel. Here's a shot of Scott and Chuck discussing the upcoming tornado. Here I am with Scott Landolt and Chris Juckins. As the storms began to develop we were watching evolution on Mobile ThreatNet, as were Rob, Chris and Dan.
The first storm rapidly became a supercell, which split. The right split formed a rain free base, and then had some underhangs/forming wall clouds. As the right split crossed Interstate 70, it presented a striking image on Mobile ThreatNet (you can see the left mover to the north). Soon thereafter, the storm developed a bell-shaped updraft and had shear/rotational signatures on Mobile ThreatNet. You can also see the folloiwng storms that became tornadic near Oakley to the west.
The Wakeeney storm developed a crisp anvil that extended to the east, with underhanging mammatus. The storm's updraft base began to develop wall clouds and here's Chris pointing out small funnel shaped underhangs....we couldn't tell if the latter were rotating. The storm's bell-shaped lowering continued to evolve, with occasional rear flank downdraft cuts and wall clouds evident. Here's a picture of Thom and of Thom and me at atound this time.
Finally the storm began to develop a very ominous look, with curving inflow bands, horseshoe base, and initial formation of a funnel cloud. Here are a series of structure shots showing the structure of the base, and the developing tornado. Image A, Image B, Image C, Image D, Image E. And here are some closeups of the tornado. Image F, Image G.
After this storm began to dissipate, we headed off towards the Oakley storm, first passing through the impressive flanking line of the Wakeeney storm. As we approached the storm, at sunset, it was putting down tornadoes near Oakley and Angelus. The structure was magnificent, with a gigantic rotating wedding cake appearance, with inflow bands pulled in from the east (right) and at midlevels, with rotational collars.
TARGET: South-central KS/extreme northwestern Oklahoma--revised while on the road to ne TX Panhandle
The morning surface map had the slow moving cold front/stationary front draped southwest-to-northeast from the TX Panhandle through southwest Kansas. South of the boundary, a stream of true tropical air, with dewpoints in the middle to high 60s was moving northwest. The afternoon forecasts created a wave of on this front in the southwest TX Panhandle taking a warm front back west, and streaming the tropical air underneath a 500 mb jet axis (shown as broad black and white arrow on the surface chart). So we decided to target southwest KS around Protection, KS.
However, while we were in Dodge City, the boundary slipped back northwestward very early, as we were eating lunch. Strong northwest winds stilled....and then changed to southeasterly...a massive change in the atmosphere that brought the tropical air back over Dodge City as we ate. This was visibile in the flags...suddenly draping downward in the calm before the wind shift, and then pointing northwestward as the low stratocumulus in the deep moist tropical layer overspread us. This led us to reassess our original target to the area around Canadian TX in the Panhandle.
The pattern expected for the afternoon our target area had the following features/ingredients: (a) moderate to strong deep layer shear overtopping strong southeasterly flow--this was an ideal setup for rotating storms; (b) moderate to high CAPE values associated with a tongue of high dew points extending northwestward from the Gulf of Mexico. In the middle and upper troposphere, a strong short wave trough was on its way towards the area...creating layer lifting and destabilization. The only factor that did not support tornadic storms was the relatively weak flow at 300 mb, not much stronger than the middle tropospheric flow....this would create weak anvil level storm relative flow and might (and did) imply that storms might precess from classic to HP supercells very quickly. HP supercells might have strong tornadoes associated with them, but have wrapping rain curtains that makes it difficult, if not impossible, to chase such storms. We did concur with SPC's moderate risk.
The early storms that initiated became supercells quickly. Tornado warnings were quick to appear. The issue for us was that most of these storms were moving through areas with a dearth of roads. The most significant storm initiallywas between Canadian and Stinnet, TX. It was cycling between classic and HP mode...and had 4" diamater hail wrapping around the mesocyclone. We took Highway 83 towards the lowered base and mesocyclone. It appeared to develop stronger rotation...with the base associated with cloud bands wrapping cyclonically around the mesocyclone...all the while we heard tornado warnings associated with the storm. But we were trapped by the road system...the tornado would be out of sight from us...and inaccessible from the road system, failing a dangerous core punch, which we will never do. Before trying to move north to get a view of the inflow area, we took some spectacular shorts of the anvil mammatus. We never could get out of the rain and large hail, some one inch in diameter, centrifuged out of the rapidly rotating core in order to see the updraft area....so we just decided to drive east and escape the storm.
It turned out that there was indeed a tornado that moved along highway 281 to its intersection with highway 83. The next day we drove that road and found the evidence that a moderate tornado at least passed along highway 281. According to the storm motion, that tornado would have moved nearly parallel to west-east highway 281 and with a slight angle across it. We found a half mile long swath of downed power poles, snapped like twigs, grass and vegetation laid down flat either indicating north to south or south to north flow depending upon where the tornado was located...and lots of incidental damage to radio transmission lines/towers (note the sheet metal wrapped around one of the guy wires). Chuck has made a report of this damage to the Amarillo NWS WFO.
After this storm weakened, yet another storm formed on its southwest flank, just north of Lake Meredith, with the mesocyclone near Boreger, TX. This, again, was in a nearly roadless area. The one road north would take us right into the teeth of the mesocyclone, with diminishing light. The updraft base lay to our northwest, with an obviously roating lowering. We thought we could see a tornado enveloped by darkness and occaisonally backlit by lightning, perhaps 15 miles northwest of us in open country. We could also see evidence of a wall cloud. However, eventually, it was too dark to see much of anything, and we decided to overnight in Amarillo.
TARGET: Repositioning north
Time to get towards northwest Kansas/northeast Colorado.
The models suggest that, in the wake of the cold front, a new short wave will bring falling pressures east of the Rocky Mountains. That suggests moisture return over the next several days, underneath moderate 500 mb flow.
We trekked northward from Amarillo to Liberal, KS. Along the way, there were spectacular displays of wild flowers. Also, the low level cumulus were developing in a sheared environment, and we can see them rolling over to form horseshoe vortices.
TARGET: Northwest Kansas
We targeted the northwest portion of KS and were rewarded with an extremely photogenic, nearly tornadic, supercell.
No maps to show for this day. Return flow finally brought 50+F dew points to northwest KS, underneath 30 mph 500 mb flow. 2000 J/kg CAPE with a strongly curved hodograph of little length meant that right moving supercells would be favored, but that their opportunity for tornadic development would be brief before they "morphed" into HP mode (as shown above...note the dark rain area, which represents forward flank downdraft air that wrapped completely around the updraft.) The storm was a hard right mover and, peculiarly, had a twin just to its west...neither Chuck Doswell nor I could explain how that could happen. Here are some radar plots showint the storm and our GPS position (yellow circles around a position white square): Radar Image 1, Radar Image 2, Radar Image 3, Radar Image 4, Radar Image 5, Radar Image 6.
After we noticed the storm becoming a hard right mover, we jumped in the cars and drove south. As we did, we noticed a rain free base to our west. As we neared the storm, the striated updraft became visible, and several times it had blocky lowerings. In the last picture, you can see the stair stepped appearance of the updraft area, each stair step representing a counterclockwise helical rotation. The main updraft had flanking line towers building into it, with midlevel bands tailing into the updraft cyclonically. Wrapping rain curtains and hail shafts circled south of the updraft core, and the storm began to develop a wedding cake appearance. The main lowering had an impressive look, but began to develop a shelfish/outflow look as the storm started to cycle to HP mode. Here's Thom taking an observation with our Kestrel4500 or Nimbus 3000.
Here are some good closeups starting from southwest to northeast of the storm as it entered its HP mode. First the rainfree base and main updraft area. Second the main rain core having encircled the updraft area, with scud-like lowerings, and two pictures (A and B) of the beaver tail extending northeastward.
TARGET: Northeast Colorado
We targeted the Palmer Divide.
SPC's Day I outlook had a slight risk over the eastern third of Colorado. Although flow at 500 mb was mediocre, surface flow was expected to be upslope by afternoon in that area, with moderate buoyancy. There was an outside chance of a rotating storm. Fortunately, this was close to Denver's airport, and Thom was schedule to leave on May 27.
As it turned out, the forecasted moisture never really developed, and only a few multicell storms developed.
TARGET: Nebraska Panhandle
We targeted the northwest portion of KS and were rewarded with an extremely photogenic, nearly tornadic, supercell.
After dropping Thom off at the airport in Denver this morning, I picked up Scott Landolt, who will be my chase partner for the remainder of this year's trip. The morning surface chart suggested a deep moist air mass moving northwestward through the Nebraska Panhandle. The forecast for the afternoon had mediocre shear moving over the area, but enough shear to get storms to rotate. Sure enough, we encountered two storms that were supercells.
We headed northeast up I-76 and then north towards Alliance. Storms were visibly going up to our east and as we approached Sidney, NE, we noticed a nice updraft to the northeast. The lower part of the updraft appeared to rotate with a nice lowered base. After watching for a while, it appeared to visibly split so we stopped to take a picture. As we got out of the car, a thin needle-like funnel appeared to the right of the lowered base and a debris cloud became visible from the ground, halfway up to the base.
The tornado lasted about five minutes with the debris cloud still visibly rotating after the funnel had disappeared. We decided to leave the storm and continue north as another multicellular storm formed in the north of the first storm in NW Garden County. As we drove, Scott noticed a lowered base associated with one of the updrafts and what appeared to be a landspout tornado underneath it. Scott managed to capture a quick picture of it before it went behind a hill and when we finally emerged from the hill, it had dissipated. The contrast on the tornado was poor, but I'll enhance it when I have time and post the picture then. Interestingly, the storm on Threatnet never showed a reflectivity value higher than 35 DBz. The storm fell apart soon after.
We continued north towards Alliance eyeing the storm the had formed to the northeast of the city. After a call from Chuck Doswell that he was on it, we changed course to intercept. We met up with Chuck and Vickie in time to watch a beautifully striated LP that had taken a hard right according to threatnet. The storm produced an abundance of funnel clouds, most of them at the mid-levels and we watched it for at least a half hour before it visually started to shrink. We continued on north, passing the storm to the east and driving through a small section of rain.
We continued on to Chadron where we had dinner and then to Rapid City where we are staying the night tonight. All in all, a great day on a day we weren't expecting to chase much, let alone see tornadoes.
TARGET: North Central South Dakota, near North Dakota border
Unfortunately, the upper trough gained amplitude at the expense of its progression eastward. This kept our target area under a strong ridge, and a cap. Adding to the cap problems was a thick layer of cirrus that overspread the area and prevented surface heating.
We got to our target area to find a band of startocumulus along each of the boundaries forming the triple point intersection. However, these clouds never developed and there was no initiation. We overnighted in Pierre, South Dakota.
TARGET: SW Nebraska/nw KS, ne CO
Trough is lagging back over Great Basin and should pull out into Rockies this afternoon...decent shear, backed surface winds in nw KS to e CO. We're trying to make a run from South Dakota.
TARGET: Oklahoma Panhandle
We are also taking a down day on May 30.
If the RUC is basically correct, return flow will spread high dew point air into ne New Mexico and the Panhandles overnight tonight into tomorrow, continuing Friday, as the synoptic scale boundary comes to halt and returns as a warm front tonight. The NAM progs 30-40 knots of 500 mb westerly or west-southwesterly flow over those regions on Thursday and ~30 knots of west northwesterly flow on Friday (Note: if the models continue fast with the current system, then it could be that (a) the westerly flow will persist until Friday...and (b) the jet streak coming down the backside of the low could be stronger than the models suggest, meaning that the 500 mb flow will stay westerly). With the 10 knots or so of southeasterly flow progged on both days for that region (including also, of course, some very backed east-southeasterly flow right near the boundary), then we could have around 40 knots of deep layer shear both days, with favorable directional shear. Since the NAM has been systematically too high with its dew point forecasts, it could be that the progged 2000-3000 J/kg sbCAPEs will not verify. But it is, at least, better than a poke in the eye.
We picked Guyman last night and are still here this morning. At the moment, there's nothing that compels us to run off anywhere - we can adjust as things progress during the day. Basically, our reasoning is comparable to Don B.'s: the best combination of deep layer shear, moisture, and instability is progged more or less in this vicinity. Both AMA and DDC soundings this morning will be decent with some more moisture. AMA's hodograph is already decent, and DDC's may improve if the forecast surface low develops to the west.
From last night's NAM/WRF, we produced a pretty great sounding for GUY this evening - naturally, this is contingent on the quality of that model forecast! We have excellent wireless connections here at the Day's Inn.
While we were on this storm, we did notice the RFD outlined in the cloud base. At the occlusion point, a lowering did form, and finally a funnel. We did not see any evidence of ground level rotation. However, while driving we did notice some debris/dust and occasional laminar features that suggested that there may have been brief tornadoes during this phase. But these were gone each time we attempted to stop and get pictures.
MTN showed, basically, an HP storm, with mesocyclones in the proper position, but often wrapped in rain visually. An isolated storm in Dallam County Texas began to approach the dew point tongue. We dropped south to intercept that, only to realize our mistake, too late to return to the first storm, which had intercepted higher dew points and briefly became a tornadic classic supercell. As we drove south, a spectacular double rainbow greeted us (in the process of creating a higher quality image of this).
TARGET: Northeastern New Mexico and adjacent portions of Texas Panhandle
The pesky closed low in the middle and upper troposphere continued to have moderate flow focussed over the northeastern New Mexico and Texas Panhandle areas. At the surface, a synoptic scale boundary, and outflow boundaries were interacting south of Amarillo. Storms were firing on the boundary back into New Mexico, so we targeted the area south and west of Amarillo.
Further south we could see a distant storm develop between Midland and Lubbock, which was probably the storm of the day, but was 150 miles distant. We could see overshooting top, backbuilt anvil, rotational features in the updraft area. We later found out that it came close to producing a tornado, but never did.
As Scott and I left for our journey north to an overnight at Raton NM, MTN guffawed and produced an outrageous surface wind observation. The storms we had left behind had organized into a squall line.
Thanks from myself, Thom Trimble and Scott Landolt to Chuck and Vickie Doswell for being terrific chase partners. We were happy to see Matt Crowther, the original "Twister Sisters", John Moore and his daughter Lisa, Brian Curran and Ed Calienese. Also, it was a treat to run into Chris Juckins, Rob Handel and Dan Porter.
See y'all again next year.
John Monteverdi, June 5, 2007