Determining the depositional age of peat: implications for dating paleoearthquakes at the Burro Flats Paleoseismic site near Banning, CA

Determining the age of paleoearthquakes hinges upon acquiring age data that accurately reflect the age of the ground surface at the time of the rupture. About a dozen paleoseismic studies along a 500 km stretch of the southern San Andreas fault provide rupture chronologies and provide the possibility of correlating earthquake ruptures between sites, thereby constraining the timing and relative sizes of San Andreas paleoearthquakes. However, sites with well established paleoearthquake chronologies, like those at Wrightwood and Pallett Creek, show that about half of the paleoearthquakes recognized at one site are not evident at the other. This apparent non-correlation is somewhat surprising considering that these two sites are 30 km apart. This project explores one possible explanation that the organic material commonly used to constrain San Andreas (and other) paleoearthquakes may be inaccurate. This study examines peat deposited shortly before and shortly after ground rupture at the Burro Flats paleoseismic site near Banning, CA. The peat layers consist of a variety of organic material that both predate and post-date the age of deposition. Components that predate the depositional age include detrital charcoal, wood fragments, seeds, and silt- and clay-sized organic soil components. Components that post-date the depositional age primarily consist of roots that cut across stratigraphic layering. Radiocarbon dating of a bulk peat sample will therefore yield a mixing age that may not reflect the depositional age of the deposit. In addition to obviously older and younger materials, the peat layers at Burro Flats contain layer-parallel fibers interpreted to represent annual reeds and/or grasses that die and are buried parallel to the ground surface. These layer-parallel fibers perhaps provide the best estimate of the depositional age of a given stratigraphic layer. At the Burro Flats site, each peat sample was examined under a microscope, and the various components were hand-separated. Radiocarbon ages were obtained for all but the crosscutting roots. The individual and bulk samples yield ages that vary in age by as much as 300 years for a given stratigraphic layer. The detrital components, bulk samples, and layer-parallel fibers consistently yielded the oldest, intermediate, and youngest ages, respectively. Relying upon radiocarbon data from bulk peat and/or detrital charcoal to constrain the timing of paleoearthquakes can therefore result in ages that are too old. The radiocarbon ages of layer-parallel fibers appear to best represent the depositional ages of the layers. Layer-parallel fiber ages can provide the most accurate constraints on the timing of paleoearthquakes, and can provide reliable chronologies useful in attempts to correlate paleoearthquakes between sites. Providing accurate paleoearthquake chronologies will lead to a better understanding of the past rupture behavior of the southern San Andreas fault and will lead to a more reliable forecast of the seismic hazard facing the southern California region.