An Analysis of Associations Between Polarimetric Supercell Signatures

Authors

Erik Green, University of Nebraska-LincolnFollow

First Advisor

Matthew Van Den Broeke

Second Advisor

Adam Houston

Third Advisor

Clinton Rowe

Date of this Version

1-2022

Comments

A THESIS Presented to the Faculty of The Graduate College at the University of Nebraska In Partial Fulfillment of Requirements For the Degree of Master of Science, Major: Earth and Atmospheric Sciences, Under the Supervision of Professor Matthew S. Van Den Broeke. Lincoln, Nebraska January, 2022

Copyright © 2022 Erik R. Green

Abstract

Supercell thunderstorms produce unique polarimetric radar signatures that are not often observed in unorganized deep convection. Repetitive signatures include deep and persistent differential reflectivity (Z_DR) columns and the Z_DR arc signature, which are both indicative of thermodynamic and microphysical processes intrinsic to supercells. Prior investigations of supercell polarimetric signatures, both those observed by operational and research radars, and those simulated numerically, reveal positive correlations between the Z_DR column depth and cross-sectional area and quantitative characteristics of the radar reflectivity field. This study expands upon prior work by incorporating a dataset of discrete, right moving supercells from across the continental United States, as observed by the operational, Weather Surveillance Radar 1988-Doppler (WSR-88D) network. Several quantitative metrics from Z_DR and Z_HH signatures are compared against characteristics of Z_DR columns, including the depth of the column, and the cross-sectional area of the column within ~1 km of the environmental freezing level. Sample statistics including median, mean, and maximum metric values were compared and tested using non-parametric similarity tests, including a Mann Whitney U-test and the two-sample Kolmogorov–Smirnov test. Cross-correlation coefficients were calculated between Z_DR column metrics and the remaining polarimetric signature metrics with increasing positive and negative lag of up to 45 minutes, for both individual storm observation periods, and as whole tornadic and nontornadic samples. A bootstrapping method (i = 5000) was conducted on the observed data, where bootstrapped distributions of metric median, mean, and maximum values were obtained, and the tornadic – nontornadic difference in the 95^th percentile median values were compared against the respective observed statistic value differences. Paired metric comparison datapoints were also bootstrapped over all offset values, and the cross-correlation coefficients were compared against the observed values. After completing the analysis, the results reveal: 1) Significant (95% confidence level) differences exist between most of the tornadic and nontornadic sample metrics including larger max Z_HH storm-core and mean Z_DR arc values and larger inferred hail areal extent among the nontornadic sample, and deeper and broader Z_DR columns within the tornadic sample; 2) Significant correlation values between metric comparisons from the tornadic sample involving Z_DR arc characteristics indicative of polarimetric associations unique to pretornadic and tornadic supercells; 3) Significant correlation values between Z_DR column metrics and inferred hail radar metrics supportive of prior observations indicative of cyclical processes in both tornadic and nontornadic supercells.

Advisor: Matthew Van Den Broeke