Spontaneous and triggered aseismic deformation transients in a subduction fault model

[1] Aseismic deformation transients can emerge as a natural outcome of the rate and state friction processes revealed in laboratory fault-sliding experiments. When that constitutive formulation is applied to model subduction earthquake sequences, transients can arise spontaneously for certain effective stress ( s) variations with depth. We show that if interstitial fluids are present and pore pressure is near-lithostatic around and downdip from the frictional stability transition, transients with recurrence intervals of 1 year are predicted on the basis of laboratory friction parameters and their temperature (hence depth) variations. The recurrence interval decreases with s and reaches 14 months when s is 2–3 MPa. Dimensional analysis and numerical studies show that the fault response primarily depends on a parameterW/h*. Here the high pore pressure zone extends distance W updip from the stability transition, and h* is the stable patch size for steady sliding. Evidence that such fluid conditions may actually be present is independently provided by the occurrence of nonvolcanic tremors as apparent responses to extremely small stress changes and by petrological constraints on expected regions of dehydration for the shallow dipping subduction zones where transients are observed. Transient sequences can also be triggered by a modest, one-time, step-like interseismic stress perturbation on the subduction fault, due to nearby earthquakes, or to pore pressure changes, e.g., during episodes of metamorphic fluid release. Properties of triggered transients and future thrust earthquakes depend on the interseismic time when the perturbation is introduced, its relative location along the subduction fault, and its magnitude.