S11C-4363:
New Insights About Pre-Stress and Fault Interaction at Fault Step-Overs from 3D Numerical Simulations

Monday, 15 December 2014
Kayla Kroll1, David Douglas Oglesby1, Keith B Richards-Dinger2 and James H Dieterich3, (1)University of California Riverside, Chino, CA, United States, (2)UCR, Riverside, CA, United States, (3)UC Riverside, Riverside, CA, United States
Abstract:
The stressing conditions on faults continuously evolve due to several factors, such as tectonic stressing, fault interactions, pore-fluid perturbations, and viscoelastic relaxation. Initial stressing conditions have a profound effect on the rupture process, including rupture propagation through stepover regions. Knowledge of rupture behavior near fault stepovers is critically important to properly quantify and mitigate seismic hazards. Previous studies have shown that stepovers along faults can influence both the earthquake magnitude and recurrence interval when rupture either jumps across or is arrested at a stepover. Landmark investigations of en echelon fault interaction with uniform initial stresses that promote super-shear rupture reported successful jumps occur at step-over widths ≤ 2.5 and 5 km for compressional and extensional stepovers, respectively (Harris and Day, 1993).

We use the 3D quasi-dynamic, physics-based simulator RSQSim to investigate how stress evolution effects rupture propagation at fault stepovers over multiple earthquake cycles. Comparisons of single-event ruptures at fault stepovers between RSQSim and the dynamic finite element code FaultMod demonstrate nucleation locations on the receiver fault similar to those of Harris and Day (1993). These simulations use uniform initial stresses with rate- and state- and slip-weakening dependent friction for RSQSim and FaultMod, respectively. Here, we present results from multi-cycle event simulations on en echelon faults using evolved stress states that arise due to fault interaction and tectonic loading. Results indicate that successful rupture jumps only occur at stepover widths of 1-1.5 km for both fault step types. The spatial pattern of rupture re-nucleation locations is strongly influenced by the evolved stress state and is dissimilar to the pattern predicted by studies. Finally, initial rupture nucleation always occurs before the magnitude of the pre-stress reaches values high enough to cause super-shear rupture, due to the heterogeneity of stress and the rate- and state- frictional properties. These results suggest that mechanisms such as extreme weakening may play a larger role than initial stress on supershear ruptures, and that observations of rupture jumps > 1 km may be explained by fault connection at depth.