Abstract

Slip on faults is often modeled or simulated assuming that the surrounding host rock is purely elastic. However, it is generally understood that the crust is a porous, fluid-filled medium and thus could be better described as poroelastic. Recently there has been considerable interest in the relationship between fault slip and changes in pore-pressure through various processes such as dilatancy and fluid-injection. Such investigations require treating the crust as poroelastic if the goal is to achieve a self-consistent framework for simulating the complex coupling of slip and pore-pressure changes with the surrounding host rock.

The talk will have two main parts. First, I will review the fundamental concepts and theories that enable us to simulate the entire spectrum of slip behavior on faults during the seismic cycle. Second, I will apply these concepts to analyze and simulate slip on a fault embedded in a poroelastic medium, where I account for the fully coupled dilatancy of the fault gouge. I start by using a linearized stability analysis to gain insight into this coupled and non-linear system. I identify a previously unknown stabilizing mechanism associated with the expansion of the fault through dilatancy. Further, I generalize the stability results of Segall and Rice (1995) to a poroelastic continuum. Finally, I show simulations using a novel spectral boundary integral formulation. I solve the coupled problem of a fault undergoing simultaneous pore-pressure changes from dilatancy and fluid injection. I systematically vary the injection rate but maintain a constant injection volume to investigate how different injection strategies affect the nucleation time of events. The results indicate that the nucleation time depends strongly on the injection rate, but the ultimate size of the earthquake only modestly depends on the injection rate for the simple injection strategy explored.