Role of advection, vertical stresses, and strain accumulation in the partitioning of deformation along the San Andreas Fault, Southern California

Thermal advection, accumulation of strain, and vertical displacement can cause relatively small changes to the strength of the crust, which can have profound effects on the partitioning of strain and resultant seismicity within an active plate boundary. Here we present three-dimensional visco-plastic numerical models of the southern San Andreas Fault system using a finite difference code that simulates the deformation of Earth material using mechanical and thermal constitutive relationships. These models are driven by a basal shear consistent with geodetic velocities and are defined with a temperature-dependent viscous model and a strain-dependent plastic model (Mohr-Coulomb with strain weakening). Model temperature controls the overall strength of the model by altering the plastic and/or viscous failure criteria. Deformation, including the formation of large-scale fault systems, is driven by the partitioning of strain within the system and is not dictated by pre-existing mechanical or thermal heterogeneities. The model reproduces the first-order characteristic horizontal velocities, strain rates, and stress rates associated with the San Andreas Fault that compare favorably to reported measurements and previous modeling results. Additionally, vertical displacement and velocity results reproduce first-order vertical offset marker and tide gauge datasets and agree well with interseismic vertical deformation produced by a 3D semi-analytical model that applies geologic strike-slip rates along active faults. Both numerical and analytical models suggest four primary areas of the San Andreas Fault System where vertical motion is significant: the Salton Trough, Death Valley, and the Big Bend. The Salton Trough and Death Valley are areas of subsidence controlled by crustal transtension. These zones of subsidence result in thinned and weakened crust resulting in a lower potential to accumulate large stresses. The Big Bend region has uplifted due to a transpressional restraining bend. The substantial uplift and evolving crustal stress conditions within this area would have thickened and strengthened the crust resulting in relatively larger stress accumulations. The long-term effect of the non-linear feedbacks between vertical deformation, thermal advection, and crustal rheology causes a dynamic link between geomorphic signal, strain accumulation and long-term seismicity.