"Slow slip events" have been discovered in subduction zones, including the Cascadia subduction zone in western North America, with precise Global Positioning System (GPS) networks. These slip events displace the earth?s surface in a similar fashion to earthquakes, however instead of lasting seconds to minutes, they last days to weeks, and even years. Because they occur so slowly, slow slip events do not radiate damaging seismic waves. They appear to be located beneath the megathrust faults that generate magnitude 8 to 9 earthquakes. Slow slip events in Cascadia occur fairly regularly every 10 to 16 months, and incrementally increase the stress on the locked fault zones. It seems likely that slow slip events occur in a frictional transition zone between the locked and steadily creeping faults. Therefore, understanding the occurrence of slow slip may lead to improved forecasting of damaging subduction zone earthquakes.
The physics of slow slip have remained poorly understood. Theh investigators suggest that rate-state friction nucleates slip under drained (constant pore-pressure) conditions, but as slip accelerates and becomes effectively undrained (no flow), dilatancy induced pore-pressure reductions quench the instability. Theystudy this, employing a simplified isothermal, membrane diffusion model and the Segall-Rice [1995] constitutive law for dilatancy. Numerical simulations exhibit either slow or fast (dynamic) slip depending on dilatancy and friction parameters, as well as effective normal stress. Stable slip is favored by low effective stress (high pore-pressure), consistent with seismically inferred vp/vs ratios in some slow-slip zones. They will extend this work by developing finite difference calculations of fluid transport that will be coupled to the friction/elasticity simulations. Slow slip modeling must ultimately consider thermal, as well as frictional, weakening. The researchers have shown that thermal pressurization becomes important at quasi-static slip speeds, before seismic radiation; the precise rates depend on permeability. This suggests that whether slip is fast or slow depends on whether or not dilatancy limits slip-rates below thermal pressurization limits. To investigate this requires coupling thermal and pore-fluid diffusion, with friction, elasticity, and dilatancy. For slip that becomes fast compared to the characteristic diffusion time across the actively shearing layer, they must explicitly consider the finite thickness of the shear zone in the computations. This work should shed significant light on the physics of slow slip, and the factors that ultimately control whether slip accelerates to inertially limited speeds characteristic of damaging earthquakes.
