Observations and Simulations of Polarimetric Radar Signatures During the 24 - 25 May 2011 Tornado Outbreak

Abstract

From 2011 to 2013, dual-polarization was implemented in the Weather Surveillance Radar 1988 Doppler (WSR-88D) network, enabling the identification of hydrometeor type, orientation, and size from radar data. With dual-polarization data, several features have since been identified as relevant to supercell intensification and development. Such features include, but are not limited to, the differential reflectivity (ZDR) column, ZDR arc, spectral differential phase (KDP) column, KDP foot, and correlation coefficient (RHV) and ZDR rings. These dual-polarization radar signatures provide critical insight into the microphysical and dynamical evolution of supercells, yet few studies have examined the temporal evolution of such signatures across multiple storms within a single outbreak or compared observations directly with high-resolution simulations of polarimetric radar data. This dissertation addresses both of these gaps through three integrated studies of the 24--25 May 2011 tornado outbreak. In the observational component, polarimetric radar data from KOUN and KTLX were analyzed across six supercells — three tornadic and three nontornadic — during the outbreak. Tornadic supercells exhibited consistently larger mean areas across all polarimetric signatures (ZDR column, ZDR arc, and KDP foot), had decreases in ZDR column area after the genesis of an EF4+ tornado, and produced larger ZDR-KDP separation angles than their nontornadic counterparts. Differences between storm categories were maximized in the 15-minute window immediately preceding tornadogenesis, and a characteristic three-stage ZDR arc development was identified across all tornadic supercells. In the simulation component, tornadic supercells were modeled using Cloud Model 1 (CM1) with three bulk microphysics schemes — Morrison 2-moment (MORR), NSSL 2-moment (NSSL2), and NSSL 3-moment (NSSL3) — and a polarimetric forward operator. Three melting models were evaluated, and a novel Combined Melting Model (CMM) was developed that produces more physically realistic distributions of mixed-phase hydrometeors and removes artificial ZDR artifacts above the freezing level present in prior approaches. Microphysics scheme comparisons revealed that the NSSL3 scheme best reproduced polarimetric signatures below the freezing level and the observed vertical evolution of near-surface rotation, while the MORR scheme most accurately captured signatures above the freezing level. However, the NSSL schemes produced the largest updraft areas both at 1 km AGL and above the freezing level, with the NSSL3 scheme producing both the warmest cold pools and smallest KDP foot areas. Despite the differences across each microphysics schemes, all three schemes reproduced the orthogonal ZDR-KDP separation vector orientation characteristic of observed tornadic supercells. These results demonstrate that bulk microphysics schemes can recreate key observed polarimetric–dynamical relationships in tornadic supercells, while also highlighting persistent challenges in representing mixed-phase hydrometeors and phase change processes in simulated storms. The observed component of this dissertation highlights the need for additional observations of the full lifecycle of supercells. The CMM and modified parameterized forward operator from the simulated component provide a computationally efficient framework for future studies of simulated polarimetric signatures which can be expanded to a range of convective modes and radar wavelengths.