Computational Electromagnetics Applied to Scattering Observed by Polarimetric Weather Radar

Abstract

The primary topics of this dissertation are issues existing in the current ensemble scattering procedures. These procedures are failing to quantitatively reproduce polarimetric signatures from resolution volumes filled with ensembles of resonant size precipitation, biota, and anthropogenic scatterers. Sources of these failures are traced to the constraints on the topology that is admissible to the different modeling procedures. The dissertation evaluates in a systematic manner the current modeling procedures focusing on limitation sources and their effects on the overall process of polarimetric variable simulation. It re-evaluates limitations of the widely used T-Matrix approach and discusses sources of instability. Based on the identified limitations, a novel computational electromagnetics (CEM) approach to scatterer modeling and polarimetric variable calculation is introduced to mitigate the current limitations. Detailed overview of the process as well as guidance on applying the CEM to the polarimetric variable calculation is presented. This is the first systematic exploration of a specific CEM solver to modeling of polarimetric radar signatures from precipitation and biota. Finally, to demonstrate meteorological application the CEM approach is evaluated by comparison with some polarimetric radar observations of hail. Of main significance is modeling of large and giant hail having surface protuberances, or rough, irregular shape. Additionally, radar observations of biota and radar cross section (RCS) measurements are considered for aeroecology applications. As an example, the precise size and shape model of Brazilian Free-tailed bat (Tadarida brasiliensis) is created and compared to the RCS measurements, as well as to radar observations of bat emergence in Texas plains.