Seismic and aseismic slip on the San Andreas Fault

Abstract

Field and experimental evidence are combined to deduce the mechanism of slip on shallow continental transcurrent faults, such as the San Andreas in California. Several lines of evidence portray the central section of the San Andreas fault as a very smooth and flat surface, with a very low frictional strength in comparison to the breaking strength of intact rock. The Parkfield earthquake of June 27, 1966, and its aftershock and creep sequences are examined as a detailed example of fault slippage that includes both types, seismic and aseismic. It is shown from considerable number of field data that during the main shock a region from about 4 to 10 km in depth slipped approximately 30 cm. In response to this slippage, creep and aftershocks were generated. The creep and aftershocks are not directly interrelated, but they are microscopically identical processes of time‐dependent brittle friction occurring in parallel in different regions. The creep occurred by time‐dependent stable frictional sliding in the 4‐km‐thick surface layer; the aftershocks, by time‐dependent stick‐slip at the ends of the initial slipped zone. This model is in good agreement with laboratory results which show that slippage should occur by stable (aseismic) friction in the upper 4 km, by stick‐slip accompanied by earthquakes from about 4 to 12 km, and by stable sliding or plastic friction below 12 km on the fault. One feature not observed in the laboratory is the episodic nature of creep. These episodes can be predicted with an accuracy of about 1 week.