5/22/08: FCST

The supercell/tornado threat looks pretty decent across the western half of Kansas on Thursday (5/22/08)...in fact, the pattern looks like a fairly classic Great Plains tornadic supercell setup (with the exception of a few minor deviations). A dryline will arc south across far western Kansas...with a warm front extending east across northern KS from a sfc low over northeast CO. The only caveat appears to be weak convg along the dryline as well as somewhat unidirectional mid/upr-lvl flow. In most cases, this might favor a rapid transition to a squall line...however, the weak convg along the dryline may off-set that tendency and still yield discrete cells. Storm motion will be north/northeasterly...which may quickly carry those cells forming closer to the warm front across the low-lvl thermal gradient and into a much cooler/stable boundary layer airmass. That may limit the tornado potential along the warm front, though extreme low-lvl shear along the boundary could still result in brief but intense low-lvl rotation. The better option appears to be waiting on storms to form south/southwest of I-70...allow one of those to mature, and then pick it up as it moves northeast across the very unstable and highly sheared warm sector east of the dryline (evident by the NAM fcst sounding for KHYS). I plan on targeting Hoxie, KS initially...hoping to pick up any cells forming directly south/southwest by mid to late afternoon...and then follow north through the evening.

DRC Formation and Vertical Wind Shear

I've conducted a set of sensitivity tests in order to determine which environmental predictors might play a role in DRC formation. The simulated supercells showed a strong association between the 0-1 km shear vector and the occurrence of the DRC. As the magnitude of the 0-1 km shear increases, the DRC becomes more prominent. In addition, when the vertical wind shear profile is linear (e.g., straight hodograph), the DRC does not appear at all. The simulated supercells develop a vortex couplet on the rear side of the supercell (behind the rear flank gust front), and this vortex couplet is colocated with the formation of the DRC. I would speculate that horizontal vorticity in the lowest 1 km is tilted upwards at the rear-flank gust front (creating positive vertical vorticity), and is then tilted back downwards behind the gust front (creating negative vertical vorticity), which creates the pair of counterrotating vortices. In addition, just prior to the onset of strong low-level cyclonic updraft rotation, an area of relatively strong rear-to-front flow (i.e., rear inflow jet) develops at the rear of the updraft, straddled by the two counterrotating vortices (as documented by Rasmussen et al. 2006). All of these features have been documented one way or another throughout the past literature which highlight fine-scale radar observations of tornadic supercells, so it appears that the simulations may be pointing toward one *possible* method of DRC formation.

Owl Horns and DRCs

Several recent papers have highlighted reflectivity features present on the rear-flank of supercell thunderstorms. Kramer et al. (2005) have documented a pair of counter-rotating vortices which appear to be a short-term indication of updraft splitting. They have coined this feature the "Owl Horn Signature," since it resembles the great horned owl when observed using high resolution mobile radar. Through a set of idealized numerical simulations, they showed that this storm feature develops through the tilting of horizontal vorticity (which is produced by strong vertical wind shear) as a storms cold pool expanded outward. In addition to storm splitting, the authors noted that tornadogenesis often takes place after the owl horn signature has been observed.

Another rear-flank reflectivity feature is the descending reflectivity core (DRC). Rasmussen et al. (2006) define the DRC as a "protuberence of reflectivity that descends from the echo overhang in the right-rear flank of a supercell." Aaron Kennedy has studied this reflectivity feature for several years now, and has put together a great source of information at this link here. The DRC may be another indicator of increasing low-level updraft rotation, possibly through a mechanism of arching vortex lines.

During the last few months, I have performed a set of idealized numerical simulations in order to produce storms which display DRCs. For the most part, the DRC has been present in a majority of the storms simulated (which I'll highlight in a later post). An interesting evolution in the rain water fields of these simulated storms seems to occur on a regular basis. 1) 30-45 minutes after storm initiation, the storm goes through a splitting process, and the owl horn signature appears, 2) then the anticyclonic member of the owl horn couplet migrates toward the south end of the rear flank of the right mover, 3) mini owl horn signatures develop every 10-20 minutes, but the cyclonic member of the couplet is much weaker than the anticyclonic member, 4) an area of locally enhanced rain water curls up into a ball around the anticyclonic vortex on the back side of the udpraft, 5) this blob of rain water descends toward the surface, 6) strong cyclonic rotation within the low-level updraft eventually develops.

Based on these simulations, I believe the dynamic processes leading to the "owl horn signature" and the DRC are one and the same, and I plan to investigate this relationship in the near future.

8/6/06

I'm adjusting to life in the "mid-south" (working at the NWS in Paducah, KY)...it looks like the Plains are still dead with regards to svr-wx, so I'm not missing much. It will be fun to experience a few cool season severe weather events here in KY during October and November...as well as cooler and less humid conditions.

15 April 2006: Southeast Nebraska

I had a pretty decent chase on Saturday (4/15/06)...I made a few mistakes early, but still managed to observe a tornado at the end of the day. I left Lincoln, NE sometime around 4pm after deciding that a developing storm southwest of Beatrice, NE looked promising. As I entered Beatrice at around 4:40 pm, I was greeted with nickle size hail. I then went east on highway 136, at which point it became apparent that I was within the RFD of a strong supercell. This storm was moving northeast fairly rapidly, and had already produced a tornado to the south of Beatrice just a few minutes earlier. If I would have left Lincoln 10 minutes sooner, I probably would have been in position to observe this tornado.

It was not a fun experience driving through the RFD, with 60 mph winds surging eastward toward the rain free base, which was still displaying strong rotation (evident by the broadly rotating rain curtains on the outer edge of the updraft base). I eventually managed to drive east of the updraft; it displayed some very nice structure but I didn't have time to stop and take pictures. I continued east on 136 and intercepted a second supercell in Tecumseh, NE at 5:30 pm. This storm had a very long RFB with a clear slot notching into its southwest flank. After shooting a little video, I kept on with my eastward trek, eventually reaching Auburn, NE. I decided to drive a bit south and investigate an LP supercell north of Dawson, NE. This cell died, and the dryline was now just to my west.

I decided it was time to give up this chase, and started the northward drive up highway 75. As I was approaching Auburn, NE, I noticed a large cylindrical cloud feature sticking out of the back side of a supercell that was between Auburn and Nebraska City (~6:35-6:40 pm). It soon became obvious that this was a tornado, although I was still about 10 miles south of the vortex so I didn't bother to shoot video or take any pictures...the contrast was not very favorable. This tornado eventually contracted into a thin needle funnel and dissipated soon after. I then let these storms move east into Iowa and called it a day.

FCST: 15 APRIL 2006

It looks like I'll get another shot at chasing a closed mid-level low pattern on Saturday. Looking at the 00 UTC 4/15/06 NAM and GFS forecast cycles, significant differences between the two model solutions are noticable. The NAM is slower and further to the south and west with the placement of the surface and mid-level low, while the GFS is a bit quicker and further to the northeast. The GFS advects the dryline eastward into eastern NEB by 00 UTC, while the NAM appears to occlude the system early causing the dryline to hang back to the west. Both models deepen the surface low to around 988 mb by 00 UTC tomorrow, and both models develop a tounge of mid 60s F dewpoints advecting northwestward toward their respective surface lows by mid afternoon.

The basic chase options appear to remain the same for either model (although the geographical targets are different)...a mid-level dryslot should work its way cyclonically around the mid-level low, which will allow surface heating and subsequent destabilization to take place. Strong large scale lift should further act to erode a strong cap. Low 60s F dewpoints seem possible, and steep low and mid-level lapse rates should reside underneath a cold mid-level airmass to the east/southeast of the closed low and within the clear slot. I believe this will lead to thunderstorm initiation by early afternoon beneath the upper-level low as well as in the vicinity of a strong dryline surge bulging northward around the KS/NEB border...points further south into KS may be too capped for initiation, evident by the warm temperatures at 850 and 700 mb. At this point, it seems as if the best option is to allow storms to move north off of the dryline surge and intercept just to the east of the warm front/dryline intersection. NAM forecast soundings indicate that low-level shear profiles will be favorable for updraft rotation, with strongly backed low-level flow north and northeast of the dryline veering rapidly in the lowest 3 km's. Deep layer shear appears marginal for supercells in NEB, but that doesn't seem to be an issue with these closed mid-level low events.

April 6, 2006: Northeast KS

After today's chase, I hope I have learned to target the surface low when a closed 500 mb low is sitting immediately to the west. Brian Thalken and myself drove down to Topeka, KS, and sat waiting for storms forming near Wichita to move northeast into our area. Cells did make it up to Topeka, but they immediately weakened as they approached. After reviewing the evening soundings, it appears that a warm layer of air, seen here on the 00 UTC 4/7/06 KTOP sounding, likely inhibited sustained updrafts as they moved east off of the dryline (note the subtle inversion just above 850 mb).

We did follow a reintensifying cell north of Topeka, which developed into a small-compact HP supercell. It developed a very low base which quickly rotated around the back side of the storms wet RFD. Observing this storm saved us from experiencing a complete bust. Meanwhile, several tornadic supercells developed west of Topeka. We considered racing west to intercept them, but finally concluded that we couldn't catch up due to their rapid northerly motion. Apparently, these tornadic storms developed in a region which closely fit Jon Davies mid-level closed core low pattern, which can be found here. I hope I will recognize this pattern the next time it emerges.