Upper mantle heterogeneities contribute to the synchronization of earthquake cycles on oceanic transform faults

Abstract:

Earthquake synchronization behavior has been reported in many fault systems. The general idea of such behavior is that faults in the system have similar repeating intervals and are positively coupled through stress interaction. Even though static stress perturbation has been shown to be important for the synchronization of earthquakes in nearby adjacent segments, viscoelastic stress should also be important, especially for segments that are relatively far away. Here we investigate how viscoelastic stress transfer can contribute to earthquake synchronization on ocean transform faults, and in particular, how heterogeneity of upper mantle might be important in affecting the dynamics of fault processes. We take the Gofar transform fault in the East Pacific Rise as the study fault, where M6 earthquakes repeatedly occur on several segments in a synchronized manner. We use the integral method to couple the dynamics of fault slip with viscoelastic flow governed by dislocation creep where lab-derived rheology parameters are applied. The model consists of two seismic patches separated by an aseismic barrier. The mantle is represented by a 2D cross section, due to the instability issue in the 3D inelastic Green’s function. We find that the viscoelastic relaxation cannot alternate the cyclic pattern of M6 events. Varying rheological parameters shows that a uniform effective viscosity of 1e17 Pa s would lead to synchronization when adjacent fault segments are no more than 1 fault segment length (20 km) from each other. A ten-times increase of dislocation creep pre-constant does so as well when the distance is no more than 7 km. Such an amount of increase corresponds to a water content 8 times larger under the same water fugacity exponent, or 5% melting under a melting constant of 45. Part of this might be due to the relatively small magnitude of earthquakes on Gofar (M6.0). Therefore, in order for viscoelastic stress transfer to effectively modulate synchronization behavior here, the effective viscosity would be much smaller than expected from lab-derived rheology. Upper mantle heterogeneity, such as localized shear zone or low effective viscosity resulting from temperature, water content, melting, composition, etc., could perhaps resolve the discrepancy between modeling and lab experiments, explain the observed synchronization, and provide an observational constraint on upper mantle rheology.