SCEC Award Number 17058 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Nonstationary ground motion spatial correlations in CyberShake simulations, and implications for regional risk analysis
Investigator(s)
Name Organization
Jack Baker Stanford University
Other Participants 1 PhD student (tentatively, Yilin Chen)
SCEC Priorities 4d, 4c, 4b SCEC Groups CS, GM, EEII
Report Due Date 06/15/2018 Date Report Submitted 07/19/2018
Project Abstract
 Spatial variations in strong ground motion have a significant impact on performance of distributed infrastructure in earthquakes. These variations are also of great importance to insurance companies that have earthquake insurance policies for many buildings in a region. Currently, these spatial variations are measured using ground motion data from densely recorded earthquakes. While useful, this measurement process requires strong assumptions about stationarity and anisotropy of correlations.

This project has performed analogous spatial variations calculations on simulations from the CyberShake platform. The richness of that simulation set has allowed significant relaxation of these assumptions, and offered insights regarding the role of source and path heterogeneity on the spatial correlation of ground motion amplitudes. This work indicates an opportunity for a new dimension of ground motion simulation validation, as the estimated correlations can be compared to results from past earthquakes. Finally, this work helps make the case for the value of using physical simulations when evaluating risk to distributed infrastructure systems.
Intellectual Merit This proposal is to perform analogous spatial variations calculations on simulations from the CyberShake platform. The richness of that simulation set will allow significant relaxation of these assumptions, and potentially provide new insights regarding the role of source and path heterogeneity on the spatial correlation of ground motion amplitudes. Additionally, this work will serve as a new dimension of ground motion simulation validation, as the estimated correlations can be compared to results from past earthquakes. Finally, this work will help make the case for the value of using physical simulations when evaluating risk to distributed infrastructure systems.
Broader Impacts Spatial variations in strong ground motion have a significant impact on performance of distributed infrastructure in earthquakes. These variations are also of great importance to insurance companies that have earthquake insurance policies for many buildings in a region. Currently, these spatial variations are measured using ground motion data from densely recorded earthquakes. While useful, this measurement process requires strong assumptions about stationarity and anisotropy of correlations. On the other hand, the approach used in this proposal avoids those assumptions and uses insights from ground motion simulations to address the problem
Exemplary Figure Figure 1: Heat map of correlation coefficients for spectral acceleration at T=3s caused by San Andreas rupture. CyberShake Simulation (left); Empirical Model (right). A reference site is indicated by a triangle, and correlation coefficients between this site and all other sites in the region are indicated by colored shading.