Abstract
The recent magnitude-9 (M9) earthquakes at subduction zones in Sumatra and Japan are focusing efforts to learn the history of infrequent earthquakes that are commonly accompanied by devastating tsunamis. The Pacific Northwest coastline along the 2 Cascadia subduction zone contains a sedimentary archive of past megathrust earthquakes. Where coastal stratigraphy records sudden subsidence due to relaxation of the overriding plate during rupture of the subduction fault, transfer functions can convert microfossil data into estimates of the amount of coseismic subsidence. Subsidence is then used in 3D elastic models of subduction-zone deformation to calculate earthquake magnitude. We tested this method with a simulated earthquake by “transplanting” a bed of modern high salt-marsh sediment into the low marsh, a vertical elevation change equivalent to coseismic subsidence of 0.62m. The transplanted bed was quickly buried by sediment and sampled for microfossils after five years. Our reconstruction of this simulated “coseismic” subsidence using our foraminiferal transfer function was 0.61m, which yields the same magnitude in deformation models. Our analyses also showed that sediment mixing could explain some occurrences of the “pre-seismic” microfossil signal inferred from prehistoric records in Cascadia and presented further evidence for sediment mixing from a record of the AD 1700 earthquake at the Coquille River, Oregon.