SCEC Award Number 17195 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Determination of Shallow Crustal Structure in Southern California and SCEC Community Model Validation Using Ambient-noise-derived Rayleigh Wave Ellipticity
Investigator(s)
Name Organization
Fan-Chi Lin University of Utah Amir Allam University of Utah
Other Participants Elizabeth Berg
SCEC Priorities 1d, 4a, 3a SCEC Groups Seismology, CXM, SDOT
Report Due Date 06/15/2018 Date Report Submitted 11/04/2018
Project Abstract
A self-consistent regional-scale seismic velocity model with resolution from seismogenic depth to the surface is crucial for seismic hazard assessment. Though Southern California is the most seismically imaged region in the world, techniques with high near-surface sensitivity have been applied only in disparate local areas and have not been incorporated into a unified model with deeper resolution. In the present work, we obtain isotropic values for Rayleigh wave phase velocity and ellipticity in Southern California by cross-correlating daily time-series from the year 2015 across 315 regional stations in period ranges 6 to 18 seconds. Leveraging the complementary sensitivity of the two Rayleigh wave datasets, we combine H/V and phase velocity measurements to determine a new 3D shear velocity model in a Bayesian joint inversion framework. The new model has greatly improved shallow resolution compared to the SCEC CVMS4.26 reference model. Well-known large-scale features common to previous studies are resolved, including velocity contrasts across the San Andreas, San Jacinto, Garlock, and Elsinore faults, mid-crustal high-velocity structure beneath the Mojave Desert, and shallow Moho beneath the Salton Trough. Other prominent features that have previously only been imaged in focused local studies include the correct sedimentary thickness of the southern Central Valley, fold structure of the Ventura and Oak Ridge Anticlines, and velocity contrast across the Newport-Inglewood fault. The new shallow structure will greatly impact simulation-based studies of seismic hazard, especially in the near-surface low-velocity zones beneath densely populated areas like the Los Angeles, San Bernardino, and Ventura Basins.
Intellectual Merit We combine Rayleigh-wave H/V ratios and phase velocity measurements in a joint Bayesian inversion to determine a regional shear velocity model for Southern California with improved resolution in the surface, shallow and upper crustal structure. Previous models such as the CVMS4.26 (Lee et al., 2014) have incorporated information from ambient noise and full waveforms but did not incorporate amplitude information and therefore have a relatively weak constraint on structure above 3km depth. By combining H/V ratios and phase velocity measurements, we gain sensitivity to shallow and mid-crustal shear velocity structure. The obtained large-scale mid-crustal features are similar to previous high-resolution models, lending confidence in the new model overall. The main improvement is the addition of new shallow features in the updated model, including more accurate basin depths and other near-surface low-velocity zones that have strong implications for studies of seismic hazard. The final model is a self-consistent regional-scale seismic velocity model with resolution from seismogenic depth to the surface.
In addition to resolving large-scale features of the crust, our shear velocity model includes small-scale shallow structure previously only seen by local imaging studies. In the north this includes the shallower-sediments in the southern tip of the Central Valley, high velocity of the Sierra Nevada Range, shallow slow velocities in the Coast and Transverse Ranges and evidence of fold and thrust faults. We resolve similar shallow structure in the LA basin to the CVMS4.26 while also imaging the Newport-Inglewood fault and Whittier faults. We also are able to see the San Bernardino basin and differing velocity structure across the Elsinore, San Jacinto and San Andreas faults. In the southern end of the region, we recover the Salton Trough and Peninsular Range with similar structure to active source studies. Our results demonstrate the considerable improvement to ambient noise imaging that can be gained from the incorporation of spatially dense Rayleigh wave H/V measurements to constrain shallow structure.
Broader Impacts The project supported the training of a PhD student, Elizabeth Berg. The student was also supported to present her work at the SCEC meeting and interact with other members of the SCEC community. The work results from this project has been accepted by JGR for publication. The 3D model constructed through this study will be share with the SCEC community through the IRIS Earth Model Collaboration (EMC) web portal.
Exemplary Figure Figure 2. Cross-Sections (Figure 1e) of (left) final inversion Vsv results and (right) difference between final and initial (CVMS) Vsv. (a) A-A’ cross section with the San Cayetano, San Gabriel, Clearwater, San Andreas, Lockhart and Garlock Fault surface traces marked. (b) B-B’ cross-sections with the Newport-Inglewood, Whittier, Sierra Madre and San Andreas fault surface traces marked. (c) C-C’ cross-sections with Elsinore, San Jacinto, Banning and Mill Creek fault surface traces marked. (d) D-D’ cross-section with Elsinore, San Jacinto and San Andreas fault surface traces marked. (e) E-E’ cross-section with Elsinore and Superstition Hills faults and Brawley seismic zone labeled. (f) F-F’ cross-section with San Cayetano and San Andreas faults marked. (g) G-G’ cross section with Newport-Inglewood, Sierra Madre, San Gabriel, San Andreas, and Garlock faults marked.

Berg, E., F-C. Lin, A.A. Allam, H. Qiu, W. Shen, and Y. Ben-Zion, Tomography of Southern California via Bayesian Joint Inversion of Rayleigh Wave Ellipticity and Phase Velocity from Ambient Noise Cross-Correlations, JGR, in press.