Stress accumulation and release at complex transform plate boundaries

Abstract

Finite element methods are used to model the dynamics of deformation along complex transform plate boundaries, specifically the San Andreas fault system, California. Effects of mantle rheology and fault geometry on the stress buildup and release are investigated. No prior knowledge of the earthquake cycle time or amount of fault slip is assumed and the results suggest that the San Andreas fault slips at low shear stress (≈15 MPa). Although the maximum stress on the fault is 15 MPa, models with an upper mantle shear zone deforming entirely by dislocation creep accumulate stresses that exceed 100 MPa, a stress level high enough to drive localized dynamic recrystallization and a shift in dominant deformation mechanism to diffusion creep. Models in which the mantle shear zone deform locally by diffusion creep reach a dynamic steady state where lithospheric shear stresses never exceed the specified fault stress anywhere in the model and indicate that the strength of the upper mantle is an important parameter in the dynamics of plate boundary deformation.

The Loma Prieta, California, earthquake of October 17, 1989: Tectonic processes and models

“…In all the models, shear stresses on the peninsular section of the San Andreas fault are lower than on the east-bay faults. A first-order result of this and previous modeling (Verdonck and Furlong, 1992) is that the apparent role played by the peninsular section of the San Andreas fault as the major plate-bounding fault in the San Francisco Bay region implies that this fault section must somehow be weaker than the east-bay faults because it appears to fail at lower fault-parallel shear stresses.…”

Lower crustal structure in northern California: Implications from strain rate variations following the 1906 San Francisco earthquake

“…Temporal variations in shear strain rates in California have been modeled using both elastic and viscoelastic models [e.g., Rundle and Jackson , 1977; Thatcher , 1983; Rundle , 1986; Li and Rice , 1987; Pollitz and Sacks , 1992]. More theoretical studies of linear, time‐dependent deformation in California, though not directly compared to geodetic data, have been undertaken by Lehner et al [1981], Cohen [1982], Cohen and Kramer [1984], Verdonck and Furlong [1992], and Furlong and Verdonck [1994], among others. The effect of nonlinear rheologies during the earthquake cycle has been investigated by Lyzenga et al [1991] and Reches et al [1994].…”