May 4 2007 chase
click here for Minnesota Public Radio interview
click here for information on purchasing a book written by Greensburg survivors... all profits go to Christian Disaster Relief
My target on May 4, 2007 was the Highway 183 corridor east of Dodge City. I was scheduled to work a day shift, but anticipated a late show which afforded me an opportunity to chase. The 18Z DDC RAOB showed moisture still only 50 mb deep and a strong cap associated with a firm 9°C 700 mb thermal ridge. I assumed a deeper ribbon of moisture resided farther south and east that wasn’t sampled by RAOBs, and figured the cap might break locally by 00Z along the dryline as the better moisture retreated. Significant nocturnal tornadoes were a definite threat given the forecast accelerating/backing of the low-level jet within an environment of very strong CAPE. High-resolution 12Z GFS and NAM models also showed the 700 mb thermal ridge subsiding a bit by 00Z with the approach of a mid-level impulse (unrecognizable on water vapor by the time I left KC)…which ended up verifying to a “tee” per 00Z RAOBs and doubtlessly aided in initiation.
I departed at 250pm. One of my inverters failed en route, so I waited until I reached Kingman to fire up the laptop. Moisture advection into southwest KS was slower than progged per area METARs. Additionally, a tornado-warned supercell was moving toward Woodward OK and sending messy left splits up toward my target area. Thus, I had the typical “chase what’s there or wait for something that’s not there yet?” decision to make. I dropped south from Pratt to keep both options open. By the time I reached Alva OK, the supercell was surprisingly in full view to my distant southwest—a shriveling LP crawling nearly due east (still out of reach), as supported by radar. Stronger convective development was occurring amongst the left-split area in southwest KS as rich boundary layer moisture surged up from OK. I committed and raced west. Dick and Darin called, indicating that a supercell was ongoing near Coldwater and looking good. As I turned north, a new severe thunderstorm warning was issued for what was to be my target storm—a young updraft on the southern flank of the convection, near the state line.
A field of accas shrouded my view of the storm until about the time I crossed back into KS at 820pm. A nicely structured, laminar-looking supercell updraft appeared to my west, with a mid-level inflow fin streaming in from the east. The developing core to my north was already electrically active. The storm rapidly acquired low-level rotation, and NWSDDC concurrently issued a tornado warning. I moved west on Highway 160 and watched as the storm began to look solidly surface-based. At 834pm a low-LCL wall cloud and large conical funnel formed on the inside (north) edge of the rainfree base. Tornadogenesis appeared imminent if not ongoing already (trees blocked my view of the ground), but an RFD clear slot wrapped swiftly around the wall cloud and the funnel dissipated.
I passed through the hamlet of Protection KS and parked the car. Dusk was settling in, and contrast against the deep grey forward flank core was horrible. Still, occasional lightning flashes revealed a ropey funnel emanating from the occluded wall cloud by 843pm, which snaked its way fully to the ground soon after. The tornado lasted about 2 minutes. Inflow from the southeast was becoming strong, and the updraft overhead was gorgeously striated. A new wall cloud formed at the triple point to my due north. I moved back east of town and continued to monitor the new wall cloud via lightning flashes. It didn’t appear as explosive as the one prior, but certainly displayed occasional funnels. At 857pm I broke eastward back to Highway 183 in order to gain some ground on the storm. Lightning activity dropped off considerably, with most of it occurring in the vault region.
Darkness had fallen by the time I turned north on Highway 183 toward Coldwater KS. At 905pm, lightning suddenly illuminated a significant-looking cone tornado to my distant northwest. It had just touched down per other chasers… and little did I know would remain on the ground for the next 62 minutes. I passed through Coldwater. Backlighting of the rainfree base continued to be rare. At 911pm I finally caught another amazing glimpse of the tornado, a bit closer now—a strong-looking stovepipe that curved gracefully at its midsection. Very strong easterly inflow into the tornado ensued. I pulled over and attempted to get some decent video—in fact, the inflow winds ripped the car door out of my hands at that time. Lightning activity picked up some, and I watched the tornado double and triple in size about 6 miles to my northwest. By 917pm the tornado already appeared wedge-like beneath a classically sloped low-level mesocyclone. I moved a couple miles north again and pulled over behind a parked semi-truck and law enforcement vehicle. An occlusion downdraft began to wrap in, and the tornadocyclone strongly resembled the Gothenburg NE cyclic monster of 1974. Unfortunately, my video during this segment turned out very poorly.
It was about the time I crossed into Kiowa county at 925pm that the event became very surreal. The tornado was changing shapes with every lightning flash, with multiple vortices apparent on one occasion. Meanwhile, a rope tornado touched down a couple miles ahead of me just west of the highway, in the northeastern quadrant of the primary tornado. It orbited slowly northward and then westward around the far side of the tornado, never to be seen again. (In fact, it may have become absorbed). A third, more highly-based truncated cone tornado also appeared to be satelliting in the southeastern quadrant of the main tornado and was only a mile or so away from my location. I pulled over and stayed put until this nearest tornado definitively roped out. Soon after, the low-level mesocyclone became very ugly (upside-down bowl look) and began to descend—in a fashion not unlike the Blue Mound KS event of February 28. The rapidity with which the cloud structure changed shapes (per intermittent lightning flashes) also suggested it was rotating with extreme violence. In very short order, the entire low-level meso was in contact with the ground: a massive wedge tornado.
I continued north. Two compact cars and a semi-truck crested the next hill, southbound, barely beating the tornado. I became shocked by the sheer size of the wedge and found myself uttering the words “this is not good” about a hundred times in a row. I got some dramatic video between 931-935pm as the mile-and-a-half wide monster lumbered across the highway, moving from about 210 degrees. I continued north, and only moments later curtains of precipitation wrapped entirely around the tornado and obscured it. I called NWSDDC to give them a report and let them know the tornado had become rain-wrapped. I then began fumbling with my camcorder to make sure I’d been recording the previous experience, when I glanced back up and noticed that the tornado was visible once again to my north-northeast. I got some additional footage from 945-950pm as the tornado—unbeknownst to me—curved leftward around the updraft and headed directly toward the town of Greensburg. A powerflash illuminated a well-defined and strong-looking stovepipe tornado to the right of the wedge; I am not sure whether this tornado was a true satellite or not.
I got a sick feeling in my stomach as I began to encounter destroyed structures at roadside and injured cattle standing dazed on the highway. The devastation was maximized along Highway 183 due to the small angle with which the huge tornado had crossed the road. A natural gas plant on the west side of the highway—which I had parked next to while chasing in southwest KS only 6 weeks prior—had been destroyed and was hissing ferociously as it spewed smelly white fumes. I nervously drove through its RFD-driven gas clouds, as did several other vehicles. I had circumnavigated 1.8 miles of damage when I came upon a toppled powerpole in the road. A state trooper, another chaser, and a truck driver were stopped there. I got out and asked if everyone was okay, as there appeared to be a damaged building just up the hill. After waiting about 10 minutes, I turned around and headed back south.
I drove to Medicine Lodge KS and pulled up to a Budget Inn to get a room. “Is it going to storm here?” the couple checking in asked the clerk. I told them most of the activity would stay to the north, and let them know about the unfortunate damage that had occurred in Kiowa county. They were surprised to hear it. I checked into my room at 1130pm and, exhausted, dozed off on the bed. I had a restless night of sleep...occasionally checking the web for updates on the damage in Greensburg KS, and watching the supercell on radar as it continued to produce tornadoes through central KS.
My heart goes out to the folks in Greensburg, and everyone affected by the tornadoes that night.
|All images (sorry for the poor quality):|
812pm - crepuscular rays atop accas as I approach the OK-KS state line
|833pm - updraft of classic supercell that would go on to produce the Greensburg tornado. The updraft is about 10 miles to my due west. Note young wall cloud adjacent to FFD. The wall cloud produced a probable truncated cone tornado ~836pm.|
|841pm - rapid clear-slotting and occlusion of wall cloud, looking northwest from near Protection|
|844pm - lightning backlights a rope tornado, located about 5 miles northwest of Protection; note clear slot at picture top|
|905pm - a brilliant bolt of lightning reveals the birth of the "Greensburg tornado" to my distant northwest, as well as the comparatively enormous parent supercell updraft. The tornado is about 12 miles away from me at this point.|
|917pm - tornado becoming wedge-like beneath descending wall cloud/low-level meso... about 5 miles to my northwest|
|918pm - another bolt of lightning silhouettes the updraft base, inflow fin (upper right), and most of the wedge tornado (lower left)...view looking due north|
|925pm - two satellite tornadoes apparent; primary large tornado ongoing in lower left, with debris... view looking north-northwest|
|925pm - nearest satellite ropes out one mile down the road; primary tornado to left "wedging out" again|
|927pm - other satellite tornado wraps westward around or into the primary tornado|
|928pm - tornado has a violent look to it with multiple vortices and debris cloud; my elbow visible in far left-hand portion of image as I check for cell phone reception|
|928pm - low-level mesocyclone "bowl" rapidly descending|
|929pm - i head north again as the tornado fully transitions into a solid wedge shape. Note light shining through portions of the base of the tornado as it continues to "fill in."|
|931pm - i pull over on the left side of the road and set up to video the tornado, which is centered 2.5 miles to my due north|
|933pm - tornado begins to cross Highway 183. Shiny surface in foreground is my car roof, with which I anchored the camcorder.|
|934pm - tornado has reached its maximum width of 1.7 miles, as surveyed by NWSDDC. Tornadoes this large are extremely rare.|
946pm - tornado on Greensburg's doorstep, about 5 miles to my north-northeast. Note rain curtains on the tornado's east flank. Car headlights up the highway mark approximate southern fringe of damage path.
|948pm - transformer powerflash, possibly caused by a separate tornado adjacent to wedge|
|950pm - the tornado entered Greensburg KS at about this time|
|950pm - one last look, with lightning illuminating an enormous occlusion downdraft above and to the left of the tornado|
|Maps, weather data|
observed tornadoes (paths estimated)
NWS-surveyed tornado paths in Kiowa county; note cyclic tornadogenesis following occlusion of Greensburg tornado (NWSDDC)
TVS signature of nearly unprecedented strength as tornado crossed Highway 183 at 938pm (NWSDDC)
03Z modified RUC sounding at Greensburg KS
(RUC had issues with BL moisture depth during this event, with a dry adiabatic dewpoint lapse rate ahead of the retreating dryline;
thus, carefully built this sounding via 02-04Z METARs and 00Z/06Z/12Z interpolated DDC/AMA/LMN RAOBs, with some input from RUC model trends above the BL)
Sfc T/Td: 77/68 F
MLCAPE: 4562 J/kg
MLCINH: 27 J/kg
0-3 km MLCAPE: 167 J/kg
MLLCL: 1085 m
MLLFC: 1586 m
representative hodograph via DDC VWP/RAOB and Haviland profiler
storm motion: 218 deg @ 21 kts
0-1 km SRH: 417 m2/s2
0-3 km SRH: 611 m2/s2
0-1 km bulk shear: 44 kts
0-6 km bulk shear: 53 kts
7-10 km SR flow: 35 kts
Click here for radar imagery from the event (831pm-1006pm CDT)