SCEC Award Number 18190 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Characterizing Source Descriptions That Account for Nonlinear Effects in Linear Simulations
Investigator(s)
Name Organization
Kim Olsen San Diego State University
Other Participants Zhifeng Hu, PhD student
SCEC Priorities 2b, 2c, 4c SCEC Groups GM, CME, Seismology
Report Due Date 03/15/2019 Date Report Submitted 05/02/2019
Project Abstract
 Recent large-scale simulations (e.g., Roten et al., 2017; 2014) have shown that near-fault nonlinear (plastic) effects can reduce the ground motions by 50% or more, even at longer periods and at appreciable distance from the causative fault (such as in LA due to large events on the San Andreas fault, SAF). A main reason for this reduction is that Love waves are prevented from developing to otherwise large-amplitude phases by near-fault plastic effects. We have implemented and tested a novel and efficient approach to account for nonlinear/plastic effects in CyberShake with its current adjoint, linear framework. We first characterize first-order differences in slip, peak sliprate, and shape of the source time function due to near-fault plastic effects. Then, we implement the source description into AWP and verify that the ground motions produced by the linear simulations are sufficiently close to those produced by the equivalent nonlinear simulations.
Intellectual Merit The approach will possibly allow inclusion of near-fault nonlinear/plastic effects in CyberShake with its current adjoint, linear framework.
Broader Impacts Including nonlinear effects into CyberShake using this approach will likely improve the resulting hazard curves and hazard maps significantly, by removing unrealistically large ground motions due to its current linear approximations.
Exemplary Figure Figure 4. PGV distribution for the southern San Andreas Fault region, obtained for (a) linear, (b) sandstone and (c) the EKS model.

Credit: Z. Hu.