Event magnitude and timing from paleoseismic and geomorphic data along the central and southern San Andreas Fault, California

George E. Hilley, & Jeri Young

In Preparation 2003, SCEC Contribution #733

Determination of the timing and surface offset of paleoearthquakes is one of the central goals in paleoseismology and an important input to seismic hazard evaluations. We expanded on existing and developed new statistical methods for inferring the continuous surface-offset distribution from point measurements of geomorphic offsets along the central and southern San Andreas Fault (SAF). We found that reconstruction of surface-offset distributions from geomorphic data could be reconstructed with some degree of confidence for the last surface-rupturing event. However, events prevent reconstruction of the offset distribution earlier than the last event. Therefore, offset distributions for events older than the last earthquake are likely model driven and can not be used as a rigorous test of fault segmentation models. Using data from the central SAF, we analyzed earthquake timing between Parkfield and the Mission Creek segment of the SAF, California. We determined earthquake timing probabilities based on data from paleoseismic excavations using a Bayesian approach that uses stratigraphic sequence constraints, detrital ages, and peat accumulation rates to refine radiocarbon ages. Additionally, we implemented ordering constraints and accumulation rates to determine and refine event probabilities and the resulting recurrence interval probability density functions. We report event probability distributions and determine the mean recurrence interval and 95% confidence intervals for recurrence for each event and all events at each site. We found acceptable correlation of events between some sites that may image continuous, throughgoing ruptures. Our analysis of those sites shows complex ruptures whose ends do not always start or terminate in a constant location. Therefore, rupture segmentation models of the central and southern SAF that require thorough going ruptures to begin and end at segment boundaries and occur regularly in time may not be resolved in the available offset and timing data. Based on the current paleoseimic data, it is unclear if earthquakes occur regularly in time along this portion of the SAF. Therefore, it may be necessary to develop new methods that integrate paleoseismic, geomorphic, geologic, and geophysical data in a quantitative way to determine the appropriate model of rupture behavior for this portion of the SAF.

Citation
Hilley, G. E., & Young, J. (2003). Event magnitude and timing from paleoseismic and geomorphic data along the central and southern San Andreas Fault, California. Bulletin of the Seismological Society of America, (in preparation).