U.S. Department of Commerce


Date of this Version



Published in National Weather Service Central Region Applied Research Papers, Volume 26 (July 2002).


Recent research into supercells has focused on narrowing the number of important storm processes that lead to these storms. For years it has been known that supercells generally deviate slightly to the left or right of the mean wind, but that in some rare cases, supercells move in a very atypical direction. For example, Corfidi (1998) examined the conditions present for the Jarrell, Texas tornado, which exhibited such an atypical motion, and found the environment to be extremely unstable with weak storm-relative helicity. He also found that the southwest movement of the Jarrell, TX storm was closely tied to the southwest movement of a strong convergence zone along a pre-existing boundary. This discrete propagation allowed the storm to move southwest despite mean cloud-layer winds indicating an east to northeast motion. An earlier study by Weaver (1979) found similar conditions for a tornadic supercell moving west, yet little has been studied concerning common parameter values among these events. The goal for this study will be to determine if any common characteristics exist among various parameters for these events.
In determining storm type and evolution, wind shear has been found to play a significant role. Studies by Klemp et al. (1981), Rotunno and Klemp (1982, 1985), and Klemp(1987) found that midlevel rotation or mesocyclones are created by the tilting and stretching of the horizontal vorticity found in the pre-storm vertical wind shear. This vertical vorticity produces a dynamically induced pressure deficit that is strongest in the midlevels of the atmosphere, effectively establishing a non-hydrostatic pressure gradient on the storm's flanks. These vertical pressure gradients force the updraft to move toward a particular flank, thus allowing it to become best correlated with the vertical vorticity on that flank (Weisman 1996). Davies-Jones (1984) found that when streamwise vorticity was ingested into the updraft of a storm, it was converted to vertical vorticity within the updraft. A later study by Davies-Jones et al. (1990) found that Storm-Relative Helicity (SRH) in the lowest two or three kilometers of the atmosphere is most relevant to the development of a midlevel mesocyclone. They went on further to state that differing values of SRH often can be associated with weak, strong, and violent mesocyclones. Weisman (1996), found that the 0-4 km or 0-6 km wind shear of the pre-storm environment is more beneficial to the operational forecaster in anticipating supercell rotation compared to SRH since an estimate of storm motion is not needed to make the calculation. Another similar study from Colquhoun and Riley (1996), found that the surface-600 mb shear magnitude to be best correlated with the intensity of a tornado.
Although ambient vertical shear and SRH have been found to be well correlated with midlevel rotation within supercells, recent tornadogenesis studies have pointed to the importance of midlevel winds and their role in redistributing precipitation away from the updraft. Brooks et al. (1994a,b) found that the midlevel storm-relative winds are important to the development of lowlevel mesocyclones, since their conceptual model indicates that the strength and lifetime of lowlevel mesocyclones is based on the balance between baroclinic generation of vorticity and storm outflow development. A later study by Rasmussen and Blanchard (1998) though showed that the same combination of parameters used by Brooks et al(1994a,b) demonstrated little skill in discriminating between tornadic and non-tornadic supercells or discriminating between tornadic supercells and ordinary thunderstorms. Thompson (1998) used Eta model initialized soundings to calculate a storm-relative wind in the midlevels (500 mb) and near surface (15 mb above the surface). This study did show some success at discriminating between tornadic and non-tornadic environments when Eta model storm-relative winds at 500 mb and the surface exceeded 15 knots. Rasmussen and Straka (1998) evaluated a data set of supercells and concluded that low precipitation (LP) supercells generally have anvil-level storm-relative wind speeds > 54 kts (~28 m/s), classic supercells have speeds between 35 and 54 kts (~18-28 m/s), and high precipitation (HP) supercells generally have speeds < 35 kts (~18m/s).
A very popular measure of instability defined by Moncrieff and Miller (1976) is Convective Available Potential Energy (CAPE). Rasmussen and Blanchard (1998) used CAPE as one of their investigated parameters and found it had some utility at forecasting tornadic environments. Combining shear parameters with CAPE, further enhanced forecasting skill. In addition, Rasmussen and Blanchard (1998) investigated Energy-Helicity Index (EHI), Vorticity Generation Potential (VGP), and Bulk Richardson Number (BRN). They found that EHI and VGP showed the highest skill at discriminating between tornadic and non-tornadic environments, while BRN showed significantly less skill.
Recently a study by Bunkers et al. (2000) described a method for forecasting supercell motion known as the Internal Dynamics (ID) method. The ID technique uses the 0-6 KM shear vector along with the mean wind to find a storm motion estimate. The advantage to this method is that it is Galilean invariant, meaning that the storm motion is the same, relative to the vertical wind shear, no matter where the vertical wind shear profile is positioned with respect to the origin of the hodograph. Because this technique has shown skill in predicting supercell motion compared to previous methods, this study will also test the Bunkers et al. (2000) scheme on storms exhibiting an atypical motion.