The Role of Fault Heterogeneity on the Diversity of Slip Styles in Megathrust Earthquakes

Some recent subduction earthquakes (2011 Tohoku, Japan and 2014 Pisagua, Chile) presented large nucleation processes punctuated by foreshock swarms, including repeating earthquakes, that can be interpreted as asperities driven by aseismic slip or cascading via intervening aseismic transients. The mainshock slip expanded into these nucleation regions without particularly strong high frequency radiation. Other regions of the megathrust produce smaller magnitude seismicity and participate in large earthquakes with remarkably strong high-frequency radiation. Hence, asperities seem to be capable of sliding slow or fast, with high or low frequency slip.

One of the frameworks to interpret the variety of fault slip styles (seismic vs aseismic, low vs high frequency, etc) is the spatial heterogeneity of mechanical properties of the fault zone, including frictional, thermal and fluid transport parameters. Field observations on exhumed faults support the view that fault zones are often mixtures of materials with different strength and frictional stability. The overall stability of a mixture is controlled by structural parameters, such as the ratio of stable to unstable materials, but also by time-dependent variables, such as pore pressure and loading rate.

The goal of this presentation is to review our current theoretical understanding of the mechanical behavior of heterogeneous faults, and to compare some model predictions to seismological observations of the nucleation and rupture processes of recent megathrust earthquakes. In particular we focus on rate-and-state models with velocity-weakening asperities embedded in a matrix of velocity-strengthening materials or with a transition from weakening to strengthening at high speed. Earthquake cycle simulations on heterogeneous faults provide insight into the conditions that allow asperities to generate high-frequency radiation during large events.