2015 GSA Annual Meeting in Baltimore, Maryland, USA (1-4 November 2015)

Paper No. 197-12
Presentation Time: 11:00 AM

REPEATED TSUNAMIS DURING THE LAST 4000 YEARS AT SANAK ISLAND, ALASKA


ENGELHART, Simon E., Department of Geosciences, University of Rhode Island, Woodward Hall, 9 East Alumni Avenue, Kingston, RI 02881, WITTER, Robert C., U S Geological Survey, Alaska Science Center, 4210 University Drive, Anchorage, AK 99508, BRIGGS, Richard W., Geologic Hazards Science Center, U.S. Geological Survey, 1711 Illinois St., Golden, CO 80401, VANE, Christopher H., British Geological Survey, Environmental Science Centre, Keyworth, United Kingdom, DURA, Tina, Marine and Coastal Science, Rutgers University, New Brunswick, NJ 08901, CORBETT, D. Reide, East Carolina University & UNC Coastal Studies Institute, Greenville, NC 27858, NELSON, Alan R., Geologic Hazards Science Center, U.S. Geological Survey, 1711 Illinois St, Golden, CO 80401, HAEUSSLER, Peter J., U.S. Geological Survey, 4210 University Dr, Anchorage, AK 99508, KOEHLER, Rich D., State of Alaska, Division of Geological and Geophysical Surveys, 3354 College Road, Fairbanks, AK 99709 and GELFENBAUM, Guy, Coastal and Marine Geology Program, U.S. Geological Survey, Menlo Park, CA, engelhart@uri.edu

Sanak Island lies above the Alaska-Aleutian megathrust, arcward of the rupture zone of the 1946 M8.6 earthquake and above a hypothesized larger rupture patch of 1788. Accompanying the 1946 earthquake, the “Scotch Cap” tsunami was the deadliest in US history, with more than 150 fatalities in Hawaii. Despite the historical hazard, recurrence intervals and magnitudes for prehistoric great earthquakes along this portion of the Alaska-Aleutian megathrust remain unknown. Because geodetic data suggest that the megathrust beneath Sanak Island is mostly uncoupled, current seismic hazard models assume that great (>M8) earthquakes do not rupture beneath it. Here we report new evidence for repeated inundation of Sanak Island by tsunamis, which we infer to be products of the underlying megathrust.

Sediment cores from three stream valleys and one freshwater wetland reveal at least three anomalous sandy deposits and two distinctive tephras that we correlate among sites spanning >8km. The sandy deposits are compositionally similar to beach sand, and the presence of marine diatoms and geochemical indicators, including n-alkane distributions, suggests a marine source. Cs-137 data suggest the youngest sand deposit was most likely deposited by the 1946 tsunami, consistent with drift logs flanking the valley. Although we find no geological evidence of the hypothesized 1788 earthquake, C-14 ages suggest two older sand sheets were deposited in the last 4000 years. The stratigraphy, micropaleontology, geochemistry, and ages demonstrate that at Sanak Island, where the megathrust is currently creeping, repeated tsunamis have inundated >500m inland and >4m above modern mean sea level during the last 4000 years. The evidence for large tsunamigenic earthquakes near Sanak Island calls into question the working assumption that uncoupled sections of subduction zones are incapable of producing large ruptures.

This work was coauthored by Benjamin P. Horton.