Earthquake nucleation by stick-slip ruptures

Abstract:

Earthquake nucleation processes are difficult to decipher in natural fault systems. Earthquake dynamics requires weakening of the host fault, however, the nucleation processes and their associated weakening intensity remain enigmatic. We analyze here the nucleation processes in short-lived, fast-propagating ruptures of stick-slips on a circular experimental fault (10 cm diameter) that has no free edges, and thus imitating an infinite fault. The experiments consistently generated spontaneous, bilateral ruptures that propagated from a recognizable nucleation site. The rupture fronts were monitored by ten rosette strain-gauges at 1 MHz sampling rate, in addition to 5 kHz monitoring of macroscopic stresses and velocities.

The majority of the rupture fronts (~ 70%) propagated at supershear velocities. The determined strain values indicated that the rupture fronts intensely reduced fault strength by 8-58% during slip-displacements of 3.7-43 microns, and that this weakening intensity was proportional to the front propagation velocity. These results and supporting stick-slip observations along granite faults (Brace & Byerlee, 1966; Lockner et al., 2017; Passelègue et al., 2013, 2016), revealed that fault strength drops as high as 90% can occur during slip-displacements shorter than 1 mm. This stick-slip weakening is more efficient, by orders of magnitude, than friction weakening in steady-state, high-velocity experiments.

We propose that natural, undetectable, tiny earthquakes (M < -5), which are similar in magnitude to experimental stick-slips, can intensely weaken the host fault and thus serve as ideal earthquake nuclei. However, these tiny earthquakes could grow into large earthquakes only if sufficient elastic energy is available in the surrounding crust. Finally, Brace & Byerlee proposed that earthquakes are natural stick-slip events, and this concept fits well the generation of tiny earthquakes, but does not address the large slip-displacements associated with moderate and large earthquakes. We argue here that tiny earthquakes provide ample fault weakening to nucleate and drive large earthquakes when sufficient elastic energy is available in the surrounding crust, but otherwise will remain just tiny, failure events.