Loading...
Thumbnail Image
Publication

Automating Seismic Waveform Analysis For Full 3-D Waveform Inversions

Lee, En Jui
Chen, Po
Abstract
Description
We present a semi-automatic seismic waveform selection algorithm that can be used in full 3-D waveform inversions for earthquake source parameters and/or earth structure models. The algorithm is applied on pairs of observed and synthetic seismograms. A pair of observed and synthetic seismograms are first segmented in the wavelet domain into a number of wave packets using a topological watershed algorithm. A set of user-adjustable criteria based on waveform similarities is then applied to match each wave packet obtained from the observed seismogram with the corresponding wave packet obtained from the synthetic seismogram. The selected wave packet pairs are then used for extracting frequency-dependent phase and amplitude misfit measurements, which can be used in seismic source and/or structural inversions. The algorithm takes advantage of time-frequency representations of seismograms and is able to separate seismic phases in both time and frequency domains. We demonstrate the flexibility of this algorithm using examples of full 3-D waveform inversions for earthquake centroid moment tensors and earth structure models at different geographic scales.
Date
2013-07-13
Journal Title
Journal ISSN
Volume Title
Publisher
University of Wyoming. Libraries
Research Projects
Organizational Units
Journal Issue
Keywords
Computational seismology,Earthquake source observations,Seismic tomography,Time-series analysis,Wavelet transform
Citation
Embedded videos