Slow slip events as stick-slip precursors in laboratory friction experiments on simulated fault gouges

Since the relatively recent discovery of slow slip events (SSEs), the nature of the relationship between SSEs and ordinary earthquakes has become one of the most important questions in earthquake science.  Specifically, questions as to whether SSEs decrease or increase the likelihood of large-magnitude earthquakes, whether or how slow and fast earthquakes can occur on the same fault patch, and whether SSEs are potential earthquake precursors have important implications for earthquake hazards.

Here, laboratory friction experiments on simulated fault gouges are used to gain insight into the relationship between SSEs and ordinary earthquakes.  The experiments are conducted water-saturated, at room temperature and at low pressure (10 MPa effective normal stress) to simulate the shallow, near-surface portions of major fault zones.  A key feature of these experiments is employing driving velocities as low as 5 cm/yr (1.6 nm/s) to simulate natural far-field tectonic driving rates.  From a larger dataset which includes a wide range of simulated fault gouges, four gouge types exhibited consistent stick-slip and these are analyzed further.  These materials are pyrite, hematite, gypsum, and Carrara marble powders.

Preliminary results show that the pyrite and hematite gouges exhibit small stress drops and increases in sample sliding velocity, interpreted to be SSEs, prior to stick-slips.  The SSEs occur near the peak in friction before the large stick-slip stress drop, suggesting that they are precursors.  In hematite at 5 cm/yr, the precursory SSEs exhibit stress drops on the order of 10’s of kPa and peak slip velocities within an order of magnitude of the driving rate, whereas the stick-slips exhibit stress drops of about 1 MPa and peak slip velocities of up to ~1 mm/s.  The peak stress at which the SSEs occur is within 1% of the peak stress prior to the stick-slip events.  Gypsum and Carrara marble, however, did not exhibit SSEs prior to stick-slips.  The results suggest that both slow and fast slip can occur on the same fault patch under the same conditions, and indicate the possibility that SSEs can be used as earthquake precursors in some cases.  However, the lack of precursory SSEs in the gypsum and marble gouges suggests that precursory SSE behavior is not universal and requires further investigation.