This talk was originally given at the AMS Radar Conference at Austin, Texas in September, 1997 during a special evening session on the Jarrell, Texas F5 tornado. The presentation slides have been redone for a special session on Jarrell to be given at the SOO COMET workshop on February 10, 1998
The model was initialized with a composite sounding. The composite is defined below 450 mb with the 1945Z M-CLASS sounding with BL mixing ratio and theta adjusted up (consistent with ACT/GRK/HLR observations) just until the capping inversion was removed. The M-CLASS sounding appeared convectively contaminated at pressures lower than 450mb and winds are missing above 300mb so the 3 hour RUC forecast for Temple, TX (valid 21Z) was used in the upper half of the composite. The composite sounding yields CAPE=5800 J/kg, LI=-11, and BRN=1100 m2/s2 (using our in-house sounding analysis package, "skewplot"). Updraft theta-w (for a 1km mixed BL parcel) is 25.8¡C which is actually slightly lower than pre-storm observations at ACT (27.7¡C at 17Z) and GRK (26.5¡C at 19Z) so the composite may underestimate actual pre-storm CAPE values. We ran two simulations. The first simulation used a homogeneous environment initialized with a standard 4 deg "bubble". A single cell storm is produced with initial updrafts briefly exceeding 100 m/s, however it "gusts out" by 1 hour.
For the second simulation, we defined an idealized horizontal "windshift line" (of 1 km depth) oriented NNE-SSW with cyclonic shear (constant SSE winds to the east and NNW winds to the west of the line). Winds are light at 4 m/s either side of the line consistent with the 19Z observations at CLL/AUS/HLR/ILE. The environment is kept horizontally homogeneous in terms of T/Td profiles and winds above 1km.
In the second run, new updrafts continually form to the SW along a flanking line and merge with the main storm complex. Intense circulations are periodically generated at low-levels. The initial max updraft exceeds 100 m/s while the average max updraft is about 70 m/s.
This study demonstrates the importance of vertical vorticity along a windshift line to storm morphology and evolution. No horizontal vorticity was initialized along the boundary. It appears that the vertical vorticity continually available along the windshift line as the storm moves SSW makes low-level rotation possible and aids storm sustainance. We also speculate that weak midlevel winds may have provided favorable conditions for the storm to remain "anchored" to the boundary.
