Shear heating and the brittle-ductile transition: thermomechanical earthquake cycle simulations on continental strike-slip faults

Kali L. Allison, & Eric M. Dunham

Submitted August 8, 2018, SCEC Contribution #8345, 2018 SCEC Annual Meeting Poster #161

We investigate interactions between coseismic slip in the seismogenic zone, deeper interseismic fault creep, and off-fault viscous flow in the context of earthquake cycle simulations on a strike-slip fault in continental crust, like the San Andreas. We focus on the effects of shear heating, which creates a thermal anomaly (i.e., temperature difference from a 1D background geotherm) around the fault and its viscous root. This reduces the effective viscosity near the fault, and therefore changes the depths of the brittle-ductile transition (BDT), earthquake nucleation, and the down-dip limit of coseismic slip. To investigate these issues, we have developed a finite-difference code for simulating earthquake cycles with rate-and-state friction and power-law viscoelasticity coupled to the heat equation.

We use flow laws for dislocation creep for feldspar in the crust and olivine in the mantle. We consider a range of background geotherms, parameterized by the lithosphere-asthenosphere boundary depth, degree of fluid overpressure, and frictional shear zone width w. We find that frictional and viscous shear heating contribute roughly equally to the thermal anomaly. Also, frictional shear heating produces a transient temperature rise, which for small w can reach hundreds of degrees, but its limited lifespan and spatial extent means that most cycle results are insensitive to this transient thermal anomaly. Instead, features like the cycle-averaged temperature distribution and the BDT are well-characterized by a steady-state approximation to the system with constant slip velocity and viscous strain rates (in which, as in our cycle simulations, the transition from frictional sliding to viscous flow is not imposed a priori but determined as part of the solution). We also find that the models without shear heating predict a BDT near or below the Moho, permitting interseismic fault creep below the Moho. In contrast, the thermomechanical models predict a BDT in the mid-crust, and many predict a 1-2 km wide zone in which coseismic slip and viscous flow both occur. Thus, we find that the simulations with shear heating are consistent with observations of faults which root in viscous shear zones in the lower crust, and observations of zones in which both viscous flow and coseismic slip occur, unlike simulations without shear heating.

Key Words
earthquake cycle, brittle-ductile transition, viscoelastic

Citation
Allison, K. L., & Dunham, E. M. (2018, 08). Shear heating and the brittle-ductile transition: thermomechanical earthquake cycle simulations on continental strike-slip faults. Poster Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Stress and Deformation Over Time (SDOT)