SCEC Award Number 21133 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Machine Learning-Based Tomography of Ridgecrest Region Near-surface S-wave Velocities
Investigator(s)
Name Organization
Peter Gerstoft University of California, San Diego Kim Olsen San Diego State University
Other Participants Michael Bianco and Zheng Zhu
SCEC Priorities 3d, 3e, 3f SCEC Groups Seismology, CXM, Geology
Report Due Date 03/15/2022 Date Report Submitted 03/22/2022
Project Abstract
 We perform ambient noise tomography (ANT) using data recorded on 342 seismographs within a 50\times50~km area inside which the July 2019 M7.1 and M6.4 Ridgecrest earthquakes occurred. We used the locally sparse tomography (LST) method, an unsupervised machine learning approach that learns to represent small-scale geophysical structures using only data from the immediate study.
The Rayleigh group speed obtained from LST better predicts travel times than conventional regularized least-squares inversion.
The 3D shear velocity model of the area obtained from the surface wave dispersion maps reveals a highly heterogeneous low-velocity zone (LVZ, with the primary velocity reduction in the upper 2-3 km) around the causative faults for the M7.1 and M6.4 events, with a 40\% reduction of the shear wave velocity. Further, correlation of other imaged LVZs in the model area with parts of the Little Lake Fault System without recent activity may indicate long-lasting damage zones.
Intellectual Merit We perform ambient noise tomography (ANT) using data recorded on 342 seismographs within a 50\times50~km area inside which the July 2019 M7.1 and M6.4 Ridgecrest earthquakes occurred. We used the locally sparse tomography (LST) method, an unsupervised machine learning approach that learns to represent small-scale geophysical structures using only data from the immediate study.
The Rayleigh group speed obtained from LST better predicts travel times than conventional regularized least-squares inversion.
The 3D shear velocity model of the area obtained from the surface wave dispersion maps reveals a highly heterogeneous low-velocity zone (LVZ, with the primary velocity reduction in the upper 2-3 km) around the causative faults for the M7.1 and M6.4 events, with a 40\% reduction of the shear wave velocity. Further, correlation of other imaged LVZs in the model area with parts of the Little Lake Fault System without recent activity may indicate long-lasting damage zones.
Broader Impacts The work has supported one graduate student and one Postdoctoral researcher.
Exemplary Figure Figure 1: (a) Vertical cross section of the shear wave velocities from the A1, A2, B1, B2, B3 and B4 station arrays. The intersection with the surface rupture of the M6.4 and M7.1 Ridgecrest events (green circles) and 2 km/s contours (white lines) are indicated. (b) Composite 3D image of shear wave velocities obtained from inversion of Rayleigh waves dispersion curves, delineating flower-shaped LVZs.
Linked Publications

Add missing publication or edit citation shown. Enter the SCEC project ID to link publication.