Electronic Theses and Dissertations
Date of Award
Doctor of Philosophy
Charles Adam Langston
Christine A Powell
Mitchell M Withers
We develop and apply the curvelet transform to remove background noise and to classify waveform features for dense spatial seismic array data. The thresholds for noise reduction are determined by analyzing the characteristic statistics of noise coefficient magnitudes through the empirical cumulative distribution function (ECDF) method within a pre-event time window. Results using a dataset from an experimental array deployed in southeastern Ohio show that background noise can be significantly removed by soft thresholding in the spatial domain. This works well for high frequency signals and complements time domain wavelet methods. The application of the curvelet transform to the Long Beach array dataset yields a successful wave partitioning of teleseismic and local scattered wavefields in the curvelet domain with improved SNR. Results show large scale occurrence of P-to-Rayleigh wave conversions in the basin. We observe a radially propagating Rayleigh wave arising from the Signal Hill popup structure at a velocity of 1km/s. Another set of P-to-Rayleigh wave conversions is generated along the Newport-Inglewood fault zone (NIFZ), manifest by a group of planar Rayleigh waves propagating away from the NIFZ at a phase velocity of 0.7-0.9km/s. The mechanism of P-to-Rayleigh wave conversions is investigated using 3D finite difference solutions for the wave equation to infer the effect of topography of Signal Hill on the wave conversions and derive high-resolution information on fault zone (FZ) structures along the NIFZ. We find low velocity material above sea-level, with 60% reduction in Vp and 60% reduction in Vs in the popup structure. The NIFZ is best modeled by low velocity zone (LVZ) above a depth of 500m, with a width of ~100-120m, with a ~15% reduction in Vp and ~15% reduction in Vs compared to the surrounding sedimentary rocks. The investigation of body-to-surface wave conversions in Long Beach provides new insights into the structural complexity in the LA basin and helps reveal geological structure of the NIFZ.
Dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.
Zhang, Jia, "USE OF CURVELETS IN SPATIAL ARRAY PROCESSING: METHOD AND WAVEFIELD ANALYSIS" (2021). Electronic Theses and Dissertations. 2183.
Data is provided by the student.