January 21 brought an interesting severe weather scenario to the Tennessee Valley region, with a number of low-topped severe storms--including at least two tornadic mini-supercells. The tail-end charlie storm in particular (see 22Z visible satellite/radar overlay below) produced an EF-2 tornado in the city of Huntsville around 2325Z. Several videos on youtube show a remarkable structure to the storm, with relatively high-contrast features and pretty coloring with sundown.
A stacked low pressure system (see 00Z 500mb RAOB plot and 22Z subjective surface analysis below) moved ENE through the region and, in conjunction with diurnal heating beneath a large scale dry slot, was responsible for the severe storms. The regime had similarities to a "warm sector variety" Davies cold core set-up, although the tornadic storms occurred south of the path of the mid-level low--not necessarily tied directly to PVA ahead of the closed low. As with typical cold core setups, the cold temperatures accompanying the mid-level system contributed to ideal low-level thermodynamics for tornadoes.
Fortuitously, the 00Z RAOBs at BNA and BMX were launched just ahead of the occluded/cold front. With Hunstville sitting exactly between each, a representative sounding was easily created by compositing these two soundings (i.e. an average of the two) and then appropriately modifying the near-surface layer (63/53°F inflow environment). This composite sounding appears below (substituted on the BNA sounding). With -2°C 700mb temperatures overspreading ML dewpoints in the mid-50s°F, 0-3km MLCAPE is very large (175 J/kg); and total MLCAPE is respectable (1051 J/kg). MLLCL heights are also quite low (708 m agl). Again, as per Davies research, strong low-level stretching likely played a substantial role in allowing for the Huntsville tornado. Actually, given that the BMX RAOB used to make the composite actually showed a warm-ish mid-level subsidence inversion "hanging on" beneath or immediately south of the mid-level jet axis (which was no doubt not present to any extent at Huntsville's latitude), it's possible the low-level and total CAPE were even slightly more favorable than I've shown in this sounding.
The shear profiles observed on the BNA and BMX RAOBs were largely straight-line, with only ~60 m2/s2 of 0-1 km SRH using the observed motion of the Huntsville cell. It's possible these observations were a bit too close to the occluded front, which would explain the veering immediately above the surface. The local 88D at Hytop AL (HTX), however, sampled a somewhat more favorable low-level shear environment immediately downstream of the tornadic storm in the warm sector. The VWP data was quite consistent during the course of the event, and looked reasonable when eyeballing the size/orientation of the BNA/BMX hodographs against the FFC hodograph (N of the "wedge front"). The mean HTX VWP hodograph from 2303-2322Z is shown below. This hodograph was characterized by 0-1 km SRH of 155 m2/s2... which is more than sufficient for tornadic development when coupled with the ideal low-level thermodynamics. Deep layer shear and mid- through high-level storm-relative flow were also favorably strong... although with the low ELs/storm tops in this case, it is tough to know what sort of storm depths/heights are appropriate for gauging a parameter such as deep-layer shear.