SCEC Award Number 19017 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Toward Full Waveform Tomography Across California (Year 2)
Investigator(s)
Name Organization
Carl Tape University of Alaska, Fairbanks Clifford Thurber University of Wisconsin, Madison
Other Participants Ryan Modtak, UAF postdoc
Avinash Nayak, UW postdoc
SCEC Priorities 4a, 3a, 3b SCEC Groups CXM, Seismology, CS
Report Due Date 04/30/2020 Date Report Submitted 04/30/2020
Project Abstract
We seek to establish a complete, open-source software workfow for simulation-based seismic inversion in southern California. Our goal is to establish seismic imaging tools that are documented and usable by a broader group of seismologists. This would provide a sustainable pathway to validating and updating the SCEC Community Velocity Models (CVMs), such as CVM-H15.1 and CVM-S4.26, which are used for probabilistic ground motion simulation projects such as CyberShake. Earthquakes and ambient noise cross-correlations can both be used for seismic imaging. To carry out this effort, in Year 1, we focus on extracting the CVMs using UCVM software, choosing a small target set of earthquakes and ambient noise “master stations,” performing moment tensor inversions, and developing a semi-automated 3D synthetic inversion based on a software package called Seisflows. Also, using standard tomography approaches, we improve upon the tomographic models in central California and aim to use these models within future simulation-based inversions. Our efforts in Year 2 fall into three main categories: (1) development of a full-waveform tomography workflow, (2) improved extraction of Rayleigh waves from ambient noise cross-correlations with an initial application to central California, and (3) improving the three-dimensional velocity model for the central California area.
Intellectual Merit We are promoting the use of highly accurate seismic wavefield simulations for the purposes of ground motion predictions and for improving 3D tomographic models of southern California. Simulation- and adjoint-based seismic inversion remains a formidable challenge, and we are seeking to reduce the effort by establishing a new workflow. Preliminary tests of the workflow have been successful.
Broader Impacts Our emphasis on open-source software is a key component that will allow the broadest range of seismologists to benefit from our work. In addition to using open-source software, we aim to provide suitable documentation, with the hope of attracting and sustaining users.

The project initiated a fruitful collaboration between UAF and U.~Wisconsin and also fostered the development of several postdocs new to the SCEC community: Avinash Nayak, Ryan Modrak, Ben Heath, and Vipul Silwal. Team members (Tape, Thurber, Nayak) participated in the SCEC CVM workshop in Pomona in October 2018, and also interacted with UCVM software developers over the course of the project.
Exemplary Figure Figure 3. Depth slices through the joint inversion model from body waves, Rayleigh waves, and Love waves.
(a-c) Depths of 1.5, 18, and 31.5 km (bsl) for Vp. (d-f) Depths of 1.5, 18, and 31.5 km (bsl) for Vs.