Tsunamis and sea levels of the past millennium in Puget Sound, Washington
MetadataShow full item record
Tidal marsh deposits in the Puget Sound area contain evidence for multiple earthquakes and tsunamis over the past 1,000 years. This dissertation focuses on evidence beneath a salt marsh at Lynch Cove, at the head of the Hood Canal about 40 kilometers southwest of Seattle. Previous work at this marsh described stratigraphic evidence for coseismic uplift and liquefaction from a crustal earthquake or earthquakes about 1,000 years ago. New findings from Lynch Cove include two anomalous silt layers interpreted as tsunami deposits that postdate the earthquake uplift and liquefaction. These layers are better explained by tsunamis than by storms or river floods, based on the layer morphology, extent, sedimentology, and microfossils. Radiocarbon ages of the two silt layers at Lynch Cove are 1690–1830 A.D. (120–260 cal yr BP, layer A), and 1170–1230 A.D. (720–780 cal yr BP, layer B). These ages more closely align with the ages of two Cascadia earthquakes than with any other known earthquake in the Puget Sound region within the last 1,000 years. These the silt layers may have been deposited by tsunamis generated by Cascadia subduction thrust earthquakes, as were likely correlative deposits at another tidal marsh at Discovery Bay, along the tsunami path between the Pacific Ocean and Hood Canal. This study improves the age ranges of the youngest six tsunami deposits at Discovery Bay, and compares them to layers A and B at Lynch Cove, and to the ages of known earthquakes and their secondary effects, including tsunamis and slope failures, of the last 1,200 years in the Pacific Northwest. Beds 1 and 3 at Discovery Bay are attributed to Cascadia subduction thrust tsunamis, and have radiocarbon ages that overlap with the ages of layers A and B at Lynch Cove. Discovery Bay Bed 2 has now been dated to 560-630 cal yr BP (1320-1390 A.D.). It is unclear why no corresponding deposit is present between layers A and B at Lynch Cove, and why no known 14th-century coseismic subsidence or tsunami is preserved at any of the Pacific coast estuaries of southern Washington. The source of the tsunami that deposited Discovery Bay Bed 2 remains to be determined. If the source was a rupture along the Cascadia subduction thrust, it may have been limited to an area offshore southern British Columbia and northern Washington, on the northern end of the subduction zone. To test whether Cascadia tsunamis could have deposited the silt layers at Lynch Cove and Discovery Bay, numerical tsunami simulations were run for three different rupture styles of great Cascadia earthquakes, a local Seattle fault tsunami, and a transoceanic tsunami from Alaska. The Cascadia earthquake tsunami simulations produced flow depths and current speeds sufficient to deposit the silt layers at both Lynch Cove and Discovery Bay, while the Seattle fault simulation did not. The Alaska tsunami simulation also produced flooding at Lynch Cove and Discovery Bay, in agreement with historical observations from 1964. Using the inferred tsunami deposits at Lynch Cove as time markers for great Cascadia earthquakes, the paleoecology of the last ~1,000 years is reconstructed using fossil diatoms to test whether Lynch Cove, 240 km inland of the deformation front, records any Cascadia earthquake cycle deformation. A diatom transfer function was developed by statistically comparing the fossil diatoms at Lynch Cove to a training set of modern intertidal diatoms from Puget Sound. Using this method, 31 paleomarsh surface elevations were reconstructed, and with radiocarbon ages, a relative sea level curve was constructed. An overall rise in relative sea level of about 2.5 m is estimated at Lynch Cove over the last 1,000 years, a rate that is faster than rates estimated by other Puget Sound studies. Superimposed on this overall relative sea level rise, paleomarsh surface elevations are observed to rise by about 25 cm prior to the deposition of both layers A and B. While these may record Cascadia preseismic deformation, these rises are within the error range of adjacent data points, so are inconclusive. Because of this, Lynch Cove marsh is interpreted as a location that probably does not record Cascadia earthquake cycle vertical deformation. Lynch Cove is the only forearc data point of vertical interseismic deformation for the Cascadia subduction zone, and these negative results provide an inland limit of earthquake cycle deformation. The findings of this research help to better understand hazards from Cascadia earthquakes and tsunamis in the Puget Sound region. The identification of paleotsunami deposits in Hood Canal identifies a tsunami hazard that was previously unknown. The tsunami simulations corroborate the geological evidence, and identify some areas in Puget Sound with greater tsunami hazard. This study also places constraints on the inland limit of Cascadia earthquake deformation. This is important for accurate estimates of areas of strong shaking, which influence earthquake hazard maps, and for geophysical models. This research also influences estimates of earthquake recurrence. If Bed 2 at Discovery Bay is from a northern Cascadia earthquake, recurrence rates at the northern end of the subduction zone may be shorter than current estimates.