Mark D. Petersen

Seismicity and fault interaction, Southern San Jacinto Fault Zone and adjacent faults, southern California: Implications for seismic hazard

The southern San Jacinto fault zone is characterized by high seismicity and a complex fault pattern that offers an excellent setting for investigating interactions between distinct faults. This fault zone is roughly outlined by two subparallel master fault strands, the Coyote Creek and Clark-San Felipe Hills faults, that are located 2 to 10 km apart and are intersected by a series of secondary cross faults. Seismicity is intense on both master faults and secondary cross faults in the southern San Jacinto fault zone. The seismicity on the two master strands occurs primarily below 10 km; the upper 10 km of the master faults are now mostly quiescent and appear to rupture mainly or solely in large earthquakes. Our results also indicate that a considerable portion of recent background activity near the April 9, 1968, Borrego Mountain rupture zone (M_L=6.4) is located on secondary faults outside the fault zone. We name and describe the Palm Wash fault, a very active secondary structure located about 25 km northeast of Borrego Mountain that is oriented subparallel to the San Jacinto fault system, dips approximately 70° to the northeast, and accommodates right-lateral shear motion. The Vallecito Mountain cluster is another secondary feature delineated by the recent seismicity and is characterized by swarming activity prior to nearby large events on the master strand. The 1968 Borrego Mountain and the April 28, 1969, Coyote Mountain (M_L=5.8) events are examples of earthquakes with aftershocks and subevents on these secondary and master faults. Mechanisms from those earthquakes and recent seismic data for the period 1981 to 1986 are not simply restricted to strike-slip motion; dipslip motion is also indicated. Teleseismic body waves (long-period P and SH) of the 1968 and 1969 earthquakes were inverted simultaneously for source mechanism, seismic moment, rupture history, and centroid depth. The complicated waveforms of the 1968 event (M_o=1.2 × 10^(19) Nm) are interpreted in terms of two subevents; the first caused by right-lateral strike-slip motion in the mainshock along the Coyote Creek fault and the second by a rupture located about 25 km away from the master fault. Our waveform inversion of the 1969 event indicates that strike-slip motion predominated, releasing a seismic moment of 2.5 × 10^(17) Nm. Nevertheless, the right-lateral nodal plane of the focal mechanism is significantly misoriented (20°) with respect to the master fault, and hence the event is not likely to be associated with a rupture on that fault. From this and other examples in southern California, we conclude that cross faults may contribute significantly to seismic hazard and that interaction between faults has important implications for earthquake prediction.