Optimization of Data Functionals for Full-3D Tomography

Alan Juarez, & Thomas H. Jordan

Submitted August 7, 2018, SCEC Contribution #8310, 2018 SCEC Annual Meeting Poster #147

Full three-dimensional tomography (F3DT) is an imaging technique for refining estimates of Earth structure by iteratively assimilating waveform data into 3D models of seismic wave propagation. We present a technique for systematically separating seismic phases using time-frequency spectra computed by the S-transform, which is based on Gaussian wavelets. Iterative waveform stripping decomposes seismograms into a finite set of waveforms, and subsets are then recomposed in an optimization process that localizes the seismic energy in the temporal domain and the structural sensitivity kernel in the spatial domain. The localization process reduces the high-wavenumber components of the structural kernel and improves the resolution to specific geologic structures. By increasing the coherent structural information that is derived from a single seismogram, this algorithm allows the identification of seismic phases that are not predicted by 1D structural models, such as basin-edge conversions. We show examples of seismogram decomposition and tomographic kernels computed for both synthetic models and recorded earthquakes in Southern California.

Key Words
Seismic tomography, data functionals, structural kernels

Citation
Juarez, A., & Jordan, T. H. (2018, 08). Optimization of Data Functionals for Full-3D Tomography. Poster Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
SCEC Community Models (CXM)