OBSERVATIONS OF SUPERCELL TORNADO EVOLUTION USING A MOBILE, RAPID-SCAN, X-BAND RADAR

Abstract

On 24 May 2011, a series of supercell thunderstorms and violent tornadoes tore through central Oklahoma. A mobile, rapid-scan, X-band, polarimetric, Doppler radar (RaXPol), collected data from one of those storms as it produced two tornadoes west and northwest of Oklahoma City. Volume scans of 360° PPIs at nine elevation angles were collected every ~17 seconds for nearly 30 minutes, and single elevation angle (1°) PPI's were collected every ~ 2 seconds for 6 minutes. The first tornado, rated an EF3, was documented from intensification to decay, and the second tornado, reaching EF5 strength, was documented from genesis through mature phase. Maximum Doppler velocities in the second tornado were observed to reach 124 m s-1. The life cycles of the tornadoes and their parent supercell are examined herein, with particular emphasis on how their structures and features evolved over the short time scales observable by RaXPol.

The roles of storm-scale features in the formation, maintenance, and decay of the tornadoes are examined and placed into context with previous studies. In an effort to determine the chronology of how and when tornadic rotation evolves on short time scales, several analysis methods are employed to examine the time-height relationship of the circulation associated with the tornadoes. These methods include quantifying the difference between the maximum and minimum inbound and outbound velocities (ΔVmax) associated with the Doppler radar detected vortex, and three dimensionally calculating and analyzing an estimate of vorticity based on the radial velocity field. Other rapidly evolving aspects of the tornadoes' life-cycles are examined as well. Polarimetric observations are used to enhance the kinematic and storm-scale analyses.

An attempt was made to retrieve the two and three-dimensional wind fields from single-Doppler data using the Tracking Radar Echoes by Correlation technique, with the desired end result of calculating trajectories. Unfortunately, this method did not prove to be accurate enough to determine confidently the wind field, which would have allowed for a more detailed examination of quantitative storm-scale properties. Even its qualitative accuracy was questionable. Therefore, this method is not used for analyzing the storm.

The most important conclusions from the results of this study include: 1) Prior to the formation of tornado 2, rotation initially is present only at the lowest analysis level, but tornadic-strength rotation develops nearly simultaneously (within ~30 s) over a depth of several km. No evidence of the dynamic pipe effect is observed. 2) The parent supercell exhibits an atypical mesocyclone and tornado cycling process prior to the formation of tornado 2. 3) A rear flank gust front surge acts detrimentally to tornado 1 but beneficially to tornado 2. 4) A horizontal vortex just ahead of the rear flank gust front coincident with a weak reflectivity band and a narrow channel of inbound velocities in the radar data appears to contribute to the intensification of the tornado. All conclusions significantly benefited from the rapid temporal observations available for this dataset since they all involved processes that evolved over periods of less than four minutes; (1) and (4) occurred in less than two minutes.