SCEC Award Number 12123 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Development and validation of stochastic representation of small-scale velocity and attenuation structure in SCEC CVMs
Investigator(s)
Name Organization
Kim Olsen San Diego State University
Other Participants 1 graduate student researcher (TBN) (3 summer mo, 1 mo acad year). No charge to PI
SCEC Priorities 6c, 6a, 6e SCEC Groups USR, GMP, GMSV
Report Due Date 03/15/2013 Date Report Submitted N/A
Project Abstract
Small-scale, near-surface heterogeneities may affect ground motion estimates significantly, in particular as the highest frequency increases, and are currently not included in the SCEC Community Velocity Models (CVMs). We collected more than 700 Vs30 measurements in the greater Los Angeles basin. We then generated semi-variograms for all the collected Vs30 measurements from the greater Los Angeles basin. The semi-variogram for the Vs30 measurements suggests a Hurst number of 0.25-0.43 (fractal dimension 1.57-1.75). We then simulated 0-2.5 Hz ground motions for the 2008 Mw5.4 Chino Hills, CA earthquake in a subset of the SCEC CVM 4.0. We assume a minimum Vs of 200 m/s, and a fractal distribution of the near-surface heterogeneities with a Hurst exponent of 0.1 and standard deviation of 10% is added to the upper 800 m of the subset. We computed ground motions for a lossless model and for various linear relations between Qs and Vs. We find amplification/de-amplification of PGVs by up to a factor of 2, and that twice as large an area experiences de-amplified PGVs as compared to the area that experiences amplified PGVs. We find that the ’usual’ Qs-Vs relation Qs=50Vs (km/s) grossly underpredicts the PGV from the data at this and other basin sites. Results using other Qs-Vs relations (in particular, higher Qs values, such as Qs=250Vs) provide better match to the PGV and duration of the data. This is expected, as the the small-scale heterogeneities substitute some of the effects from the intrinsic attenuation specified by the Qs-Vs relation.
Intellectual Merit Developing statistical models of the near-surface heterogeneities of the southern California basins is a SCEC4 milestone. This project takes a first step toward developing such models, and understanding their effects on the ground motions. In particular, we should expect to modify the popular Qs-Vs relations when including the statistical models, due to the inclusion of scattering attenuation.
Broader Impacts Benefits to society includes more accurate ground motion prediction for seismic hazard analysis. The project has funded PhD student William Savran at SDSU, who will continue to work on the statistical models in the coming years. The results has been reported at internal workshop in Japan, as well as annual meetings of the SSA and SCEC.
Exemplary Figure Figure 4. 0-2.5 Hz Data (black) and synthetics for various simulations of the 2008 Chino Hills earthquake at the deep basin site LTP (white triangle in SW area of the model shown in Figure 3). Credit: Kim Olsen.