EARTHQUAKE BEHAVIOR ACROSS A FAULT SEGMENT BOUNDARY: A HIGH RESOLUTION PALEOSEISMIC STUDY IN THE SOUTHERN SAN JACINTO FAULT ZONE, IMPERIAL VALLEY, SOUTHERN CALIFORNIA

We excavated trenches at several sites along strands of the southern San Jacinto fault to determine the rupture history across a 3 km step separating the Coyote Creek fault (CCF) and the Superstition Mountain fault (SMF), Imperial Valley, southern California. All of the trenches were excavated below the shoreline of Lake Cahuilla, and there are coeval lacustrine deposits preserved at each site that we use to determine the relative and absolute timing of events. At a new site along the SMF at Carrizo Wash south of the step over, we found evidence for 6-7 surface rupturing events, with the past 3 events between A.D. 1085 and 1680. The most recent event (MRE) produced only ~15 cm at this site and is not recognized only 3 km to the SE at the shoreline site of Gurrola and Rockwell (1996). The two earlier events together produced about 6 m of slip, with the latter event correlating to the MRE at the shoreline site where 2.2 m of slip was determined. At another site, located within the southern break of the 1968 Coyote Creek rupture but still south of the step over as defined by microseismicity, we found evidence for up to 11 events. In addition to 1968, only one other small (1-2 cm) event ruptured the site after A.D. 1600. Two earlier events were considerably larger than the two most recent events but slip was not resolvable. Based on their stratigraphic position, one of these is almost certainly the same event as recorded at Carrizo Wash farther south. North of the step over, two sites are located within the middle break of the 1968 rupture, one at the northern shoreline and one in the area of maximum slip. We found evidence of repeated slips after the last lake with 1.4 m of cumulative slip after A.D. 1680. We also found evidence for a very large surface rupture that probably correlates with the last large event on the SMF about A.D. 1600, that produced a much broader rupture zone than did 1968 or any other event after the last lake. We take these data to suggest that: 1) the southern third of the 1968 rupture is actually part of the SMF; 2) the CCF produces frequent small slips whereas the SMF produces less frequent but much larger earthquakes; 3) CCF ruptures rarely propagate onto the SMF (as in 1968) but SMF ruptures apparently do rupture onto the CCF. Therefore, the 3 km wide step is sufficient to stop slips of 30-50 cm but insufficient to stop displacements of several meters.