SCEC Award Number 17071 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Earthquake Ruptures in Damaged Fault Zones with Along-Strike Segmentation
Investigator(s)
Name Organization
Yihe Huang University of Michigan
Other Participants One graduate student
SCEC Priorities 3d, 2e, 3c SCEC Groups FARM, Seismology, SDOT
Report Due Date 06/15/2018 Date Report Submitted 06/07/2018
Project Abstract
 Major strike-slip faults are usually surrounded by several-hundred-meter-wide damaged fault zones (DFZs). However, how DFZs contribute to the nucleation and arrest of large earthquakes is not well understood. In particular, DFZs show along-strike variations of damages that may be correlated with historical earthquake ruptures. These segmented structures may either act as barriers or assist rupture propagation, which can significantly affect the final sizes of earthquakes. We use dynamic rupture simulations to elucidate the effects of along-strike variation of fault zone damage on earthquake rupture propagation and termination. Our results show that intact rocks adjacent to a preexisting DFZ act as material barriers for earthquakes nucleated inside the DFZ. The barrier effects also vary with the earthquake nucleation size, hypocenter location in the DFZ, DFZ width and velocity, DFZ structure smoothness, DFZ asymmetry, and stress heterogeneity along the fault. A break-through rupture is more favorable when DFZs become longer, wider, sharper or more damaged, indicating a higher likelihood of having larger earthquakes in more established fault zones. A break-through rupture is also possible when a high-stress asperity is located at the ends of the more damaged fault zone, but the asperity needs to have a sufficiently large size, which requires a long period of interseismic stress accumulation. Our results suggest that a priori knowledge of the along-strike variations in near-fault rock properties, which may be related to cumulative displacement, fault lithology, tectonic setting and historic earthquake rupture, is critical in predicting the possible size of future large earthquakes.
Intellectual Merit The size of an earthquake largely depends on where rupture stops. Our results show that besides stress and friction heterogeneity, along-strike variations in fault zone structure that surround major strike-slip faults can contribute significantly to rupture termination, which has not been well understood before. Our models predict that small earthquakes tend to concentrate in the highly damaged fault zone segment and occasionally break through the intact rocks outside. It provides a consistent explanation for the seismicity distribution in the Anza section of the San Jacinto fault. Our models also predict a higher likelihood of having a break-through rupture when DFZs become more mature and well-established and when stress accumulation in the intact rocks is sufficiently high.
Broader Impacts This project has supported an early-career scientist when she started her tenure-track position. It has also fostered collaborations between two early-career scientists in two institutions within the SCEC scientific community. The goal of the collaboration is to build more realistic 3D rupture scenarios for earthquakes that break through the Anza gap along the San Jacinto fault, which has been one of the most seismically active fault zones and may host the next large earthquake in Southern California. The results from this project also motivate the development of earthquake cycle simulations with fault zone structure variation by a first-year graduate student at University of Michigan.
Exemplary Figure Figure 2. Top: Slip rate function for every 1 t_c of ruptures nucleated in a 1 L_c wide sharp fault zone with 50% velocity reduction that end at a distance of (a) 4 L_c and (b) 5 L_c from the hypocenter when the half nucleation size is 0.6 L_c. The bottom panel shows rupture speeds normalized by the shear wave velocity of the intact rock. It shows whether rupture stops as it reaches intact rocks depends on the relative distance from the hypocenter to the zone of intact rocks. Rupture that propagates in the damaged fault zone for a sufficiently long distance may penetrate into intact rocks and continue propagating afterwards, as it tends to have higher slip rates and thus carries larger dynamic stresses before reaching intact rocks.