How can tidally modulated low-frequency earthquakes in the mid crust along the San Andreas Fault be reconciled with its shallow apparent locking depth?

Earthquake-cycle models for major strike-slip faults suggest that stationary, localized interseismic deformation and large-scale, rapidly decaying postseismic transients may be reconciled if a strong lithosphere, a low-viscosity mantle asthenosphere and a largely aseismic shear zone extending from the brittle-ductile transition to the mantle asthenosphere are incorporated. Localized late interseismic deformation around major strike-slip faults seems to call for a “stiff” shear zone (maintaining shear stresses of at least several MPa) in the lower crust and mantle lithosphere, and rapid postseismic deformation requires that part of the shear zone (likely in the mid crust) should have a low effective viscosity per unit width, at least during the postseismic interval (Yamasaki et al., 2014, Hearn and Thatcher, 2014).

Recent paleoseismological studies suggest a preponderance of M 7.3 to 7.7 earthquakes during the past 650 years along the 1857 rupture segment of the San Andreas Fault (SAF), with just two events comparable in magnitude to the 1857 earthquake (Scharer et al., 2014). Very low shear stresses (less than 0.001 MPa) and a frictional rheology have also been inferred along the northern part of this segment between depths of 16 and 29 km, based on tidal triggering of low-frequency earthquakes (Beeler et al., 2013). I have refined my 3D viscoelastic finite-element earthquake-cycle models to address whether extremely low SAF-parallel shear stresses in the mid crust can be reconciled with localized and invariant interseismic deformation (and an apparent locking depth of 16-18 km) if earthquakes are smaller and more frequent than assumed by Hearn and Thatcher (2014). Results from models incorporating frictional afterslip and viscous shear zone creep will be presented for a small set of plausible rupture histories that take smaller earthquakes into account.