Implication of seismicity for failure of a section of the San Andreas Fault

Abstract

abstract On January 15, 1973, a magnitude ML 4.1 earthquake occurred near Cienega Road on the San Andreas Fault about 20 km south of Hollister, California. A 3-km-long segment of the fault southeast of the earthquake was aseismic for the 7 weeks preceding the event, although microearthquakes occurred at both its ends. The first day's aftershocks occurred at the northwest end of the aseismic segment; later aftershock activity migrated to the southeast, filling the remainder of the segment. If the discontinuous surface trace of the fault can be extrapolated to the focal region of the earthquakes to define fault geometry at depth, then aftershocks occurred primarily on one continuous segment of the fault and epicenter locations and direction of rupture propagation (inferred from the azimuthal pattern of P-wave radiation) of the precursory shocks correlate with the discontinuities in the trace that terminate the segment. The 1970 to 1976 deficit in seismic slip within the segment suggests that fault creep accounts for a significant part of cumulative slip within the segment. The pattern of seismicity is consistent with the hypothesis that creep on the segment before the main shock caused a buildup of stress at the ends of the segment or at the ends of adjacent offset segments. Correlation of seismicity and discontinuities or bends in the mapped fault trace are the basis for an extension and refinement of the “stuck” and “creeping” patch model of the San Andreas Fault in central California. Patch boundaries extend from the free surface down through the seismogenic zone. Creeping patches lie beneath smooth continuous segments of the fault trace. Stuck patches lie beneath discontinuities or bends in the fault trace.