Abstract, submitted for Fall 1999 AGU

Recent rupture history of the San Andreas Fault southeast of Cholame in the northern Carrizo Plain, California

We conducted a paleoseismic study on the San Andreas fault southeast of Cholame, CA to investigate the record of earthquakes within an 80 km data gap between Parkfield and the Carrizo Plain. The only data previously available for this area were ambiguous geomorphic offsets. The site excavated is the LY4 site, located approximately 37.5 km southeast of Cholame. This site is on an alluvial fan that emanates from the Temblor Mountains and crosses the fault trace at the distal end of the fan. We excavated a 20-m-long and 3-m-deep trench perpendicular to the fault. We found evidence of four ruptures within the past 931-715 yrs. The oldest event, C4, caused apparent normal offset that is only present in the lower half of the trench. Event C3 is based on fracturing which is truncated above event C4 and had no apparent vertical displacement. The units ruptured by events C3 and C4 are not continuous across the fault zone. Event C2 is inferred because lower units that cross the fault are offset with an apparent normal sense of slip while those above are not. The event horizon C1 does not have an apparent vertical sense of offset and is defined by fractures that terminate above C2. Because of limitations in the production and preservation of datable material, the only age constraint is on a paleosol two units below the fourth event horizon. Two samples from this layer have a summary 2 sigma age range of 881-665 cal BP.

Paleoseismic sites in the Carrizo Plain contain five ruptures during the last 900 years. The geomorphic offsets in the Carrizo Plain were larger than the offsets southeast of Cholame during the 1857 Fort Tejon earthquake. The characteristic earthquake model, a model often used to consider earthquake recurrence, states that ruptures along a particular fault segment recur at similar magnitudes and with similar offsets. If there is a uniform slip rate on these segments, then this model suggests that there should be more frequent earthquakes on the Cholame segment with smaller offsets. However, our findings imply that the San Andreas fault in the Carrizo may have actually ruptured more frequently than along the Cholame. Rupture scenarios that link the Cholame ruptures with Carrizo earthquakes fail to maintain a uniform slip rate if the offsets are characteristic, given this result. Scenarios that link Cholame ruptures with Parkfield earthquakes do not conform to a characteristic rupture length of the Carrizo events such as experienced in 1857.