Use of WSR-88D Spectrum Width as a Diagnostic and Prognostic Tool in the Tornado Warning Decision-Making Process

Tornado warning decision-making continues to be one of the most challenging tasks faced by National Weather Service forecasters, evidenced by the relatively low probability of detection and high false alarm ratio for warnings.  In the past few years, two new tools were made available that aid the tornado warning process.  The advent of dual polarization data, specifically the tornado debris signature, is helping improve detection, at least for tornadoes near the radar, but it provides no lead time (at least initially).  The Multi-Radar Multi-Sensor time trends of azimuthal shear, known as rotation tracks, permit straightforward temporal analysis of the circulation strength of potentially tornadic mesocyclones or mesovorticies.  But despite past efforts to demonstrate the utility of spectrum width as a tornado warning decision-making tool, it remains underutilized in this process (while these other tools are coming into common usage).

This study assesses the utility of spectrum width as both a diagnostic and prognostic tool using a database of over 200 tornadoes from eastern Iowa and northern Illinois between 2005 and 2015.  Spectrum width values observed by the lowest two elevation angles of 0.5° and 0.9° prior to and during tornadogenesis are analyzed to determine a threshold value, lead time, and POD.  Data are stratified by storm mode to evaluate any potential performance differences between supercells and quasi-linear convective systems.  Results are summarized and recommendations are made for how spectrum width data should be worked into the tornado warning decision-making process.