The Interaction Between Frictional Slip and Viscous Fault Root Produces Slow Slip Events

We consider a model where an unstable frictional region, governed by rate and state friction, interacts with a viscous zone with Newtonian rheology. The system is loaded at distance with a constant velocity. Pore pressure variations are considered and we show that the model of Segall and Rice (1995, https://doi.org/10.1029/95jb02403) relating porosity changes to variations of the state variable could be derived considering viscoplastic deformation of a population of identical asperities. We perform a linear stability analysis in the case of a constant pore pressure in agreement with the full numerical results. For a given value of the viscosity of the viscous region, stable slip is promoted at low normal stress and unstable slip at high normal stress. Near the transition from stable to unstable slip, modest acceleration of slip, resembling slow slip events (SSE) are observed. We show that our model can reproduce real SSE sequences in the Guerrero subduction zone which are the largest worldwide. The best fit parameters suggest that SSEs happen in areas of low effective normal stress (for the frictional region) and low viscosity (for the viscous region). In our model, SSEs happen in a regime where the viscous region is able to counteract the instability of the frictional one. We show that considering a rate strengthening rheology or a non newtonian one leads to the same linear stability results. Our work shows that a simple model with homogeneous spatial properties can lead to complex dynamics, covering a wide range of observed sliding modes, from steady‐state creep to seismic slip and SSEs. Plain Language Summary Our model studies the interaction between a frictional transition zone (FTZ) and a viscous zone. The seismogenic fault zone (SFZ) that hosts large earthquakes is considered locked and does not play any role in our model. Switching from unstable slip in the SFZ to stable slip at greater depths requires the existence of a FTZ that we model considering a weakly unstable rate and state friction fault zone. The viscous fault zone is modeled using a viscous newtonian rheology. We discuss the possibility of considering a non‐newtonian rheology and even a rate strengthening response and show that our stability analysis remains valid to first order. When asked to meet some observables of the long‐term SSEs of the Guerrero subduction zone, our model implies the existence of near lithostatic fluid pressures in the FTZ. The temperature range consistent with th