Evidence for Seismic and Aseismic Vertical Slip Along the Hayward-Rodgers Creek Fault in Northern San Pablo Bay, California

Distinguishing between seismic and aseismic fault slip in the recent (last few hundreds of years) geologic record is difficult, yet fundamental to estimating the seismic potential of faults. We integrated ultra-high-resolution chirp sub-bottom imaging with targeted coring and multi-disciplinary core analyses to determine whether the across-fault vertical deformation visible in chirp data along the Hayward-Rodgers Creek fault was the result of aseismic creep, seismic slip, or both. During the fall of 2015 we collected a series of vibracores targeting a fault bend basin previously recognized in the chirp seismic data.

Through integrated seismic stratigraphic and core analysis, we identified and traced four key horizons (R1-R4), all younger than approximately 1500AD, that cross the fault and extend throughout the fault bend basin. Stratigraphic age models were developed for two cores on either side of the fault zone using detailed down-core radiocarbon and radioisotope dating combined with records of anthropogenic metals and stable isotopes (δ¹³C, δ¹⁵N). We calculated vertical slip rates across the key horizons and developed a tectonic model to explain recent deformation along the fault. Vertical fault offset of strata younger than the most recent surface-rupturing earthquake on the Hayward fault in 1868 suggest near-surface aseismic creep is occurring along the fault in northern San Pablo Bay. Vertical separation of the two oldest horizons, R1 and R2, suggests vertical slip-rates of 1.7 to 2.1 mm/yr. These vertical slip-rates are likely too large to be attributed to creep alone. Together with stratigraphic evidence of widespread folding and subsidence of pre-R1 strata, a combination of seismic and aseismic deformation since approximately 1500AD is suggested. Additional radiocarbon ages from terrestrial organic material will be used to calibrate the local marine reservoir correction and refine our age models.