Comparison of algorithm performance assessing size sorting signatures in polarimetric radar data

Several polarimetric radar signatures have been repeatedly observed in supercell thunderstorms. Two such signatures are an enhanced region of specific differential phase (KDP) values in the forward flank and an enhanced region of differential reflectivity (ZDR) values located along the inflow edge of the forward flank. These regions and the horizontal separation between them are due to hydrometeor size sorting. Recent studies have analyzed this separation signature and connected it to the storm-relative winds and assessed differences in this signature between tornadic and nontornadic supercells. Two methodologies for quantitatively analyzing this signature in supercells are the “dynamic thresholding” technique from Loeffler et al. (2020) and the Supercell Polarimetric Observation Research Kit (SPORK) from Wilson and Van Den Broeke (2022). A novel algorithm being developed for potential real-time use within the Multi-Radar Multi-Sensor (MRMS) and Warning Decision Support System - Integrated Information (WDSS-II) frameworks, “w2sepvec” utilizes reflectivity (ZH), ZDR, KDP, correlation coefficient (CC), and azimuthal shear (AzShear), in addition to data from the 13km Rapid Refresh (RAP), to identify and track regions of enhanced ZDR and KDP. Due to the increasing number of methodologies being utilized to analyze this separation signature, a direct comparison of their performances is warranted. This study will present key similarities and differences between the three algorithms and their outputs from analyzing identical sets of tornadic and nontornadic supercells.