Role of fault gouge dilatancy on aseismic deformation transients

Author Posting. © American Geophysical Union, 2010. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research 115 (2010): B10414, doi:10.1029/2010JB007522.In the vicinity of episodic aseismic transients in several subduction zones, the presence of interstitial fluids and near-lithostatic pore pressure has been proposed to interpret seismic observations of high P to S wave speed ratio and high Poisson's ratio. Under such conditions, fault stabilization by dilatancy-induced suction during increased shear strain rates becomes very efficient. We analyze the frictional and hydraulic conditions for spontaneous transients on a fluid-infiltrated fault including dilatancy and pore compaction in the framework of rate and state friction with a “membrane diffusion” approximation. In both a simplified spectral model and a 2-D Cascadia-like subduction fault model, the fault response is mainly controlled by three nondimensional parameters: (1) W/h*, the along-dip width of the high pore pressure, velocity-weakening fault relative to a characteristic nucleation size, (2) a drainage parameter U, the relative time scales for fluid diffusion and friction evolution, and (3) a dilatancy parameter E, the relative contributions to stress drop from dilatancy and friction evolution. The incorporation of dilatancy enables aseismic transients at much larger values of W/h* than is possible under conditions of constant pore pressure. An analytic estimate of the maximum slip velocity as a function of W/h*, E, and U is derived and agrees reasonably well with the simulation results. The dependence of the properties of modeled transients on the drainage parameter U is similar to that on the dilatancy parameter E. For U (E) less than 1, maximum velocity decreases, while recurrence period remains relatively constant. For U (E) greater than 1, maximum velocity approaches the steady state velocity, and recurrence period approaches the period at neutral stability. In the subduction fault model using gabbro gouge friction properties, the slip per episode and the recurrence period increase with W/h*, generally following the trend defined without dilatancy. The maximum velocity with dilatancy can be several orders of magnitude smaller than that without, in particular for larger values of E and values of W/h* near the no-dilatancy stability limit.This study was supported by a Department of Geosciences Hess postdoctoral fellowship to Yajing Liu at Princeton, USGS-NEHRP award 08HQGR0047 to Allan Rubin that partly supported Y. Liu, and the J. Lamar Worzel Assistant Scientist Fund to Y. Liu at WHOI