SCEC Award Number 09194 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Realistic Initial Conditions for Large-scale Earthquake Rupture Simulations
Investigator(s)
Name Organization
Guanshui Xu University of California, Riverside James Dieterich University of California, Riverside David Oglesby University of California, Riverside
Other Participants Fang, Zijun
SCEC Priorities A10, A9, A6 SCEC Groups FARM, FARM, SoSAFE
Report Due Date 02/28/2010 Date Report Submitted N/A
Project Abstract
 Large-scale physics-based computational simulations of earthquake ruptures have been undertaken by SCEC to assess potential strong earthquake ground motions in southern California. The simulations generally do not include the spontaneous earthquake initiation process for complexities in both computation and modeling. Our studies of earthquake nucleation on complex 3D fault systems indicate that fault geometry and properties together with initial stress conditions strongly influence where and when earthquake nucleation occurs (Zhang, et al., 2004, 2006, Fang, et al., 2010, 2011, 2012). For a rupture simulation along a pre-defined fault, the point of rupture initiation strongly affects ground motions. For this reason, we proposed to investigate how earthquake earthquakes are nucleated in geometrically and frictionally complex fault systems. In particular, we intended to develop a comprehensive mechanical model that allows us to study earthquake nucleation on general non-planar faults and then implement 3D simulations of the San Andreas fault at the resolution used for the large-scale rupture simulations. We hoped the results can then be directly used as to set initial conditions for the large-scale rupture simulations.
Intellectual Merit This work has produced an improved method for determination of hypocenter location for geometrically complex faults under realistic loading conditions. It has strong implications for SCEC research objectives in Fault and Rock Mechanics as well as ground motion prediction.
Broader Impacts The method developed in this project may be of great use in producing realistic earthquake nucleation locations for scenario earthquakes in Southern California and beyond. In turn, better estimates of nucleation locations may help reduce uncertainty in ground motion models via a better estimate of effects such as directivity. In addition to its scientific results, the project also helped train an engineering graduate student (Zijun Fang) in fault dynamics; this student went on to a seismology post-doc at Stanford, and now has a promising career in industry.
Exemplary Figure Fig 1. a) A schematic configuration of a strike slip fault with a ramp between two fault segments. (b) Earthquake nucleation velocity field when the bend angle  is 2.9 degrees. (c) Earthquake nucleation velocity field when the bend angle  is 8.6 degrees.