THE TIMING AND DEFORMATION OF CASCADIA SUBDUCTION ZONE EARTHQUAKES FROM COASTAL MARSH RECORDS OF OREGON (Invited Presentation)

Determining the timing of Cascadia subduction zone earthquakes and their associated magnitude of coseismic deformation remains a challenge for paleoseismology. Here, we use (1) new coastal records; (2) microfossil-based paleogeodetic techniques (errors < 0.3m); and (3) high-resolution AMS radiocarbon ages (errors < ±20 years) to better understand past ruptures and, therefore, the seismic hazard posed by the Cascadia subduction zone.

We develop a new regional foraminiferal-based Bayesian Transfer Function (BTF) for inferring the magnitude of coseismic subsidence from these records. The BTF method has the advantage of incorporating information readily available from secondary proxies (e.g., sedimentology) and allows non-unimodal taxa-elevation relationships. The use of secondary priors compensates for the effect of mixing across coseismic contacts. Application of the validated BTF to records pertaining to the AD 1700 rupture suggests greater coseismic subsidence in northern Oregon and southern Washington than previously estimated. However, these results still indicate heterogeneous rupture in this event and a magnitude near 9 as inferred by Wang et al. (2013 JGR).

To extend this analysis to earlier earthquakes we must produce robust correlations between earthquakes at individual sites. To achieve this, we compile a new dataset of high-quality AMS radiocarbon dates on earthquakes recorded in sediment beneath Oregon coastal marshes. We employ two age models Bchron and Oxcal. Model results from both approaches are consistent when maximum and minimum ages are available and closely-spaced in age, and are proximal to the sedimentary contacts inferred to record coseismic subsidence and postseismic recovery. When these conditions are not met, the age estimated for an earthquake by Bchron and Oxcal may differ. This is most likely due to the additional sedimentation rate information used by Bchron to refine earthquake timings. These differences are most pronounced for an earthquake(s) occurring ~800 cal yrs BP with Bchron suggesting multiple earthquakes closely spaced in time in Oregon. In contrast, Oxcal results allow for both single and multiple-rupture scenarios. Both Oxcal and Bchron agree that more than one earthquake is recorded in Oregon between ~950 and 1300 cal yrs BP.