Paleoseismic investigation of the San Gorgonio Pass fault zone near Cabazon, California

The discontinuous nature of the San Andreas fault at San Gorgonio Pass has led to competing models about whether or not large earthquakes can rupture through this structurally complex region. To test these models I excavated trenches across the San Gorgonio Pass fault zone (SGPFZ) near Cabazon as part of a larger effort. The trench site is located ~100 m to the east of a site excavated in 1997 by Yule where trenches uncovered evidence for at least one earthquake 500-700 yrs ago. At the 1997 site extensive bioturbation and low sedimentation rates hampered efforts to distinguish events. The new site, located where two alluvial fans merge, was selected for its higher sedimentation rates and excellent stratigraphy. Excavations at the new site contain incipient soil layers that subdivide the stratigraphy into six units (Units 100 to 600, from base to top of the section). The units form tabular to wedge-shaped layers up to ~1 m thick. In general, the base of each unit consists of yellowish-gray, thin- to medium-bedded sand and pebbly sand, with occasional gravel lenses and mud and silt interlayers. The tops of each unit are more massive; they locally contain numerous burrows and display a slight orange-red oxidized soil horizon. A distinct angular unconformity lies at the base of Unit 400. Units that underlie the unconformity (100-300) are cut by a fault that strikes N35-50° E and dips 15-30° NW. The fault shows 1.0 to 1.5 m of vertical separation, equivalent to 2.5 to 6 m of slip resolved parallel to a N45°W slip vector. Unit 400 is up to 1 m thick at the south end of the trench and pinches out against the fault scarp on the north. A channel fill sequence in Unit 500 caps the scarp. The stratigraphic (onlap) and structural (faulting) relations clearly restrict the most recent event (MRE) to have occurred post- and pre-deposition of Units 300 and 400, respectively. Radiocarbon ages from detrital charcoal fragments constrain the timing of this earthquake to between ~600-800 yrs BP. I suspect that the story here is incomplete because, due to a road that I could not excavate, I did not extend the trenches across the entire fault zone. However, results from this and a 1997 study led by D.Yule (which trenched the complete fault zone) suggest that the most recent earthquake occurred 600-700 yrs BP. Evidence from Cabazon therefore supports a model with infrequent ruptures (by San Andreas fault standards) in San Gorgonio Pass and rare through-going events.