Optimizing the WSR-88D Mesocyclone/Tornadic Vortex Signature Algorithm Using WATADS—A Case Study

Abstract

A tornado outbreak occurred in east-central Minnesota around 0000 UTC 22 July 1995, within 80 km of the KMPX WSR-88D, at the Minneapolis?Chanhassen Weather Service Forecast Office (WSFO). The WSR-88D Mesocyclone/Tornadic Vortex Signature algorithm (MTA) identified only two Tornadic Vortex Signatures (TVSs) out of 14 confirmed tornadoes. A detailed examination of Doppler velocity products revealed operator-defined TVSs, intense gate-to-gate azimuthal shear, in 10 of the 14 storms. This outbreak presented an opportunity to optimize MTA performance because the algorithm performed poorly and the KMPX radar sampled many weak tornadoes at close range. Numerous combinations of two adaptable parameters, threshold pattern vector (TPV) and threshold TVS shear (TTS), were systematically tested using the WSR-88D Algorithm Testing and Display System (WATADS) and archive level II data. MTA?s performance improved significantly when default values were optimized. The number of tornadoes detected increased from two to seven and average TVS alert lead time, within 60 km of the radar, increased from 10 to 15 min, but the number of TVSs falsely identified increased from zero to two. In 1995 and 1996, the NEXRAD Operational Support Facility (OSF) authorized radar offices to lower TPV and TTS from their default values under the Unit Radar Committee level of change authority. Recommended values for TPV and TTS, based on research from data collected from many different sites across the United States, were provided for field office use. The values of TPV and TTS used to optimize MTA performance during the Minnesota tornado outbreak described in this paper independently agreed with OSF recommended values of TPV and TTS.