NE Texas/SE Oklahoma Tornadoes & Hail: 4/24/2020 - Operations Proving Ground

A shortwave trough traversing the Central Plains set the stage for severe thunderstorms across southeast Oklahoma and northeast Texas (and the Ark-La-Tex region) on the afternoon of April 24, 2020. The SPC highlighted the threat area as an Enhanced Risk for all modes of severe weather, based on a 10% probability of tornadoes and a 30% probability of large hail and damaging wind (Figure 1). Surface dew points were primarily in the mid-50s (F) in northeast Texas as of 12Z 4/24; thus, the return of surface moisture was something to monitor throughout the day.

By 21Z, an area of surface convergence was evident along the dryline just to the east of the DFW metroplex, as indicated by a developing line of cumulus (Figure 2). With MLCAPE values of ~2000 J/kg and 70 kts of effective bulk shear, the environment was conducive for the development of severe thunderstorms, assuming deep convection could be initiated in this region. Note in the observations that surface dew points were able to reach the low 60s in advance of the developing cumulus (Figure 2). Indeed, NWS Fort Worth provided an excellent targeted graphic highlighting the areas most at risk for severe hazards in the upcoming 2-3 hours (Figure 3).

As seen in the Day Cloud Phase RGB from GOES-16 imagery with lightning overlay, convective initiation did occur along and just east of the dryline (Figure 2). This was a result of deep mixing and the removal of CIN in the vicinity of the boundary. However, due to the strong advective winds as illustrated by 700 mb-300 mb mean wind (black arrows, Figure 4), convective attempts quickly moved off the quasi-stationary dryline before maturing. While the region was characterized by synoptic-scale lift as evidenced by height falls and warm-air advection (Figure 5), such lifting was not on the order of the mesoscale lift associated with the shallow dryline circulation which initiated the convection. Thus, the immature thunderstorms were unable to sustain themselves once they moved off the boundary, likely owing to a combination of strong deep-layer shear and dry-air entrainment. Furthermore, because the convective attempts moved so quickly to the east, these initial updrafts were unable to condition the environment (via locally moistening) which would have aided in subsequent convective attempts along the dryline. Once the initial convective attempts occurred and ultimately failed as they moved off the boundary, it became clear that the threat for sustained updrafts in this region was mitigated. This would have allowed the mesoanalyst to shift their mental probability density function (PDF) regarding severe hazards to the left (i.e., lowering potential severe hazards) for that particular area.

Concern then shifted further up the boundary, to an area with stronger synoptic-scale forcing closer to the mid-level vorticity maxima. This region north-northeast of Dallas, up to the Oklahoma border was characterized by MLCAPE values of 1000+ J/kg and 45-50 kts EBWD, and was thus supportive of supercells. By 2240Z, convective initiation was beginning to occur along the front in southern Oklahoma. Meanwhile, in the region immediately downstream of this developing convection, surface winds backed substantially along and north of a mesoscale warm front located just south of the Red River (Figure 6). As a result, this was an area of enhanced low-level storm-relative helicity (SRH). The 22Z HRRR seemed to have a good handle on these backed surface winds which resulted in a 00Z forecast sounding for the Fort Towson, OK area that depicted an impressive 200+ m2/s2 of surface-1 km SRH swept out by the looping hodograph (Figure 7). Therefore, despite dew point temperatures that had only climbed into the low 60s, a mesoanalyst could have reasonably considered a localized tornado threat in this corridor due to the strong low-level SRH. Not to mention the threat for other severe hazards such as significant hail with the rotating updrafts. It is also worth noting that a modification of the model sounding based on the current surface conditions for Paris, TX would have resulted in greater 0-3 km CAPE and low-level lapse rates than otherwise indicated. This would lend additional support for tornado potential based on the possibility of stronger low-level vertical ascent and stretching.

Moreover, along the northern periphery of the dryline (now being taken over by the cold front) that extended north-south just east of Dallas (southern portions discussed earlier), red visible satellite imagery showed a gradual increase in growing cumulus (Figure 8). This occurred in advance of the approaching area of lift from the west-northwest as evidenced by the line of developing convection just east of Ardmore, OK. These observations could have been used as a sign that the towering cumulus west of Paris, TX would remain colocated with sufficient lifting, unlike previous convective attempts further south. Also, this area was within the region of enhanced low-level SRH, further indicating a localized threat of severe hazards associated with supercells, including a tornado potential. It is worth noting the singular NW-SE oriented line of cumulus near Paris, TX early in the animation of red visible imagery (Figure 8). Given the orientation of this line of cumulus, it is possible that this is further indication of the low-level backed surface winds and enhanced SRH.

Ultimately, multiple tornadoes occurred just west of Paris, TX, including one rated an EF-2. A mesoanalyst might have initially been skeptical of tornado potential in this area given ML LCL’s  at the high end of the tornado-producing parameter space. However, thinking in terms of sufficiency, the localized area of enhanced low-level SRH, attributed to the backed surface winds, was suggestive of rotating updrafts with the potential to produce tornadoes. With this understanding, a mesoanalyst could then properly anticipate a localized threat for a tornado or two, and therefore help inform warning operations in a region in which the lowest radar scan was around 10,000 ft AGL (making rotation detection difficult). Moreover, the tornado threat, along with additional severe threats such as large hail, could be communicated to partners with effective lead-time.

Special thanks to Ted Ryan, SOO at FWD, and Rich Thompson, SPC Lead Forecaster, for their review and suggestions to this analysis.