A bimaterial interface along the northern San Jacinto fault through Cajon Pass

We investigate the existence of bimaterial interfaces along the San Jacinto fault zone (SJFZ) and the San Andreas fault (SAF) around their intersection at Cajon Pass (CP) to clarify if large earthquakes can rupture simultaneously parts of both faults. Analysis of fault zone head waves reveals a bimaterial interface along the northern SJFZ with a >7.5% velocity contrast (slower northeast side) that extends continuously through and perhaps slightly beyond CP. Other seismological studies show an opposite sense of velocity contrast across the nearby SAF. These observations, combined with expected properties of bimaterial ruptures, suggest that large earthquakes on the northern SJFZ are likely to propagate to the southeast and hence unlikely to trigger large SAF events. However, large earthquakes propagating on the SAF near CP can trigger continuing rupture on the SJFZ to the southeast. This inference is consistent with ruptures of large paleoseismic earthquakes and directivities of smaller events. Plain language summary Large faults typically separate rocks with different seismic velocities. A sharp velocity contrast across a fault (bimaterial interface) produces refracted phases (fault zone head waves) that travel exclusively along the interface and can have significant effects on the behavior of earthquake ruptures. Using fault zone head waves we detect the existence of a bimaterial interface in the northern San Jacinto fault (southern California) up to its intersection with the San Andreas fault at Cajon Pass and estimate the velocity contrast across the interface. Rocks northeast of the San Jacinto are shown to have velocities that are over 7.5% slower than rocks on the southwest side. Previous studies indicate that the San Andreas fault near Cajon Pass also has a sharp bimaterial interface but with slower rocks on the southwest side of that fault. These observational results, combined with theoretical expectations for dynamic ruptures on bimaterial interfaces, suggest that a large earthquake on the San Andreas can relatively easily develop into a continuing earthquake on the San Jacinto fault but not the other way around. The results can be used to refine the seismic shaking hazard in the area.