Poster #032, San Andreas Fault System (SAFS)

Observation-constrained multicycle dynamic models of southern San Andreas fault and the San Jacinto fault: the effect of the Big Bend and Cajon Pass on rupture dynamics

Dunyu Liu, & Benchun Duan
The poster PDF is private. For more information, please contact the author(s).

Poster Presentation

2020 SCEC Annual Meeting, Poster #032, SCEC Contribution #10717
Macroscopic fault geometrical complexities such as restraining bends and stepovers, i.e., earthquake gates, show impediments to earthquake dynamically propagating ruptures. Earthquake gates are common in natural fault systems and they occasionally pass dynamic propagating ruptures, leading to multi-fault and multi-segment ruptures. Assessing the likelihood of multi-fault and multi-segment ruptures is in urgent need, given the 1992 Landers, the recent 2016 Kaikoura and the 2019 Ridgecrest earthquakes. In this study, we simulate earthquake cycles on southern San Andreas fault and the San Jacinto fault, with realistic fault geometries and an emphasize on physics-based dynamic ruptures. We try t...o understand how the Big Bend and Cajon Pass, two major earthquake gates in the system, affect rupture extents, slip distributions, and statistics of recurrence intervals. We use straining rate from GPS measurements to load the fault system. Simulated results are compared to long-term slip rate, rupture extents and statistics of recurrence intervals from paleoseismological records to gain insights of mechanical factors that may explain past rupture patterns. The 2D simulation method has a finite element model for co-seismic dynamic ruptures and an analytic viscoelastic model for interseismic fault stress evolutions. In one end-member models with uniform shear loading, the results indicate that maximum shearing at 10-15˚ in compression relative to the SAF south of the Big Bend yields some ruptures resembling historical events. However, the significant change in strike around the Big Bend on the San Andreas fault and GPS measurements indicates nonuniform maximum shear loading directions. In the other end-member cases with maximum shearing loading parallel to local fault strikes, supported by some seismological evidences, the system yields mainly ruptures of the entire SAF in the model, ruptures of the whole system, and ruptures of individual SJF segments. This loading scheme significantly reduces the impediments of geometrical complexities to ruptures and the results cannot reflect rupture extents implied by paleoseismological records. A maximum shear loading condition that is parallel to the SAF north of the Big Bend but is aligned at 10-15˚ in compression relative to the SAF south of the Big Bend could yield rupture patterns implied by past earthquakes.