Quantifying Coseismic Subsidence during the Giant AD 1700 Earthquake along the Oregon Coast

The elastic dislocation technique, employed to define seismic hazards for the Pacific Northwest, uses the amount of coseimic subsidence, geodetic strain measurements and thermal data to constrain estimates of coseismic slip and model earthquake magnitude and rupture area. To date, coseismic subsidence has predominantly been estimated using qualitative or semi-quantitative techniques hampered by large unquantified error estimates. Therefore, quantitative estimates of coseimic subsidence in the coastal zone are essential to our understanding of seismic hazard in the Pacific Northwest.

Changes in intertidal foraminiferal assemblages are useful for reconstructing Holocene relative sea-level change and are employed at the Cascadia subduction zone to estimate earthquake-related subsidence. However, the accuracy of foraminifera in quantitative reconstructions hinges on the presence of appropriate modern assemblages that reliably reflect past faunal-environment relations We developed a modern training set of 83 samples and 18 foraminiferal species from five Oregon intertidal zones and statistically identified two elevation-dependent ecological zones: Faunal Zone I dominated by Balticammina pseudomacrescens, Haplophragmoides wilberti and Trochammina inflata in upland, high and middle marsh; and Faunal Zone II dominated by Miliammina fusca occupied the low marsh and tidal flat.

We developed a foraminiferal-based transfer function using this modern training set, which produced accurate (r2jack = 0.82) and precise (RMSEP = ~0.21 m) reconstructions of paleomarsh elevation of fossil sediment sequences from the most recent megathrust earthquake in Cascadia (AD 1700). The quantitative coseismic subsidence estimates varied between 0.42 m and 0.90 m (± 0.27 m to 0.32 m) along coastal Oregon. These estimates were further supported by a multiproxy approach of lithostratigraphy, environmental analyses (carbon isotope analysis) and faunal and floral zonation. Importantly, these results further refined coseismic estimated subsidence used to validate elastic dislocation models, and also they suggested spatial variations in plate interface character (slip, strain and area).