Plane-stress finite-element models of tectonic flow in southern California

Abstract

The continuing, distributed crustal deformation in southern California was modelled as creeping flow in a nonlinear continuum. The crust was assumed to be in a membrane state of stress, except for the weight of topography, and subjected to plate-tectonic velocity boundary conditions. The flow law of this membrane contained a rigid-plastic term to represent the frictional faulting in the upper crust and a power-law term to represent dislocation creep in the lower crust. In addition to regional strength variations based on heat flow, some models had a weak belt representing the San Andreas fault. When calculated strain-rates were tested by cross-correlation with actual seismic moment rates, directions of shortening, and rates of vertical movement, the best model was found to be one in which the San Andreas fault is eight times weaker than the surrounding crust, perhaps because of clay gouges or fine-grained mylonites within it. This model is consistent with the absence of a detectable heat-flow anomaly at the San Andreas, and also with the difference between geologic and paleomagnetic plate velocities. The rest of the crust must be at least half as strong as is indicated by laboratory deformation, or the crust would slump southwest toward the continental shelf. However, the failure of all the models to match many details of local tectonics suggests that other faults share the anomalous weakness of the San Andreas. This is expected to limit the success of any future continuum models of the region.