Digging For Worms: Characterizing Near-Ground Vorticity During Supercell Tornadogenesis with Doppler Radar and Lidar

Abstract

The question of why some supercells produce tornadoes and some do not has motivated decades of research by the severe storms community. Advancing the understanding of the dynamical processes behind tornado formation is thought to be beneficial for improving tornado forecasts and warnings. The current understanding of tornadogenesis involves four stages (Fischer et al., 2024), the first being the formation of a rotating updraft called a mesocyclone. The second is the generation of rotation near the ground, which can be converged and amplified by the vertical motion beneath the mesocyclone. A recent numerical modeling study by Markowski (2024) proposes the idea that the inflow, not the outflow, can act as the origin of near-ground vertical vorticity for supercell tornadogenesis. This inflow vertical vorticity is associated with turbulent coherent structures in the surface layer of the planetary boundary layer (colloquially referred to as “vorticity worms”), which generate organized streaks of vertical vorticity often greater than 0.05 /s in highly sheared environments. This finding is contrary to nearly half a century of tornado research, which has pointed to the tilting of horizontal vorticity by a supercell’s downdrafts as the origin of surface rotation for tornado formation. Because vorticity worms exist in the typically clear-air inflow of supercells, they are difficult to capture in most radar observations. Additionally, vorticity worms are confined to the surface layer, only spanning a vertical depth of a few hundred meters, adding additional observational challenges. In the current study, we turn to observations of the 10 May 2010 Central Oklahoma tornado outbreak by the OU-PRIME radar, which was a polarimetric C-band system and was located in the inflow of the Moore-Choctaw EF4 tornado. The narrow beamwidth, high sensitivity, and use of a 0.25 degree elevation angle provide a rare opportunity to characterize near-ground (< 100 m AGL) vertical vorticity in all regions of the Moore-Choctaw supercell during tornadogenesis. We use the 2D multi-level wavelet transform to estimate the wavelength of inferred vertical vorticity from the azimuthal shear of radial velocity. Dividing the near-field of the supercell into inflow and outflow partitions, we find that wavelengths 200-800 m are the most common at ~ 100 m AGL in both regions. The wavelet analysis also confirms that the near-field inflow is predominantly composed of vorticity worms while the outflow is made up of vortex patch-like features (e.g., lacking preferred orientations). Partitioned time-height series of circulation from a dual-Doppler analysis (DDA) with OU-PRIME and KTLX show similar values between the inflow and outflow. Additionally, we inspect vertical vorticity flux into the inflow and outflow partitions, finding positive flux into the near inflow and negative flux into the outflow partition during the entire period. We speculate that this is due to the orientation of the storm-relative winds near the mesocyclone, which are southeasterly in the inflow and exceed 50 m/s during tornadogenesis. In contrast, the storm-relative winds in the RFD are very weak and have an easterly component in some regions. This orientation could indicate the inflow winds are better oriented to advect vertical vorticity towards the mesocyclone in this case, though the 2+ minute volumetric updates of this dataset prevent us from directly tracking this process. Vertical profiles of wind speed are constructed in the inflow, revealing a shallow layer (~ 50 m deep) of substantial streamwise horizontal vorticity. Lastly, we subjectively identify vorticity worms in the inflow and discuss their stretching potential in the context of the Fischer et al. (2024) tornadogenesis model. Though our analysis is insufficient to answer if vorticity worms directly facilitated tornadogenesis in this case, our findings highlight the inflow region as a potentially large source of vertical vorticity for tornado formation. Discussion of other supercell cases confirms that vorticity worms are not unique to our case or the Markowski (2024) simulations. Lastly, we discuss environmental parameters that may be important in the context of vorticity worms and present ideas to further investigate the influence of turbulent coherent structures on supercell tornadogenesis.