2002 Denver Annual Meeting (October 27-30, 2002)

Paper No. 4
Presentation Time: 1:30 PM-5:30 PM

OXCAL ANALYSES AND VARVE-BASED SEDIMENTATION RATES CONSTRAIN THE TIMES OF 14C-DATED TSUNAMI IN SOUTHERN OREGON


NELSON, Alan R., U.S. Geol Survey, Geologic Hazards Team, MS 966, PO Box 25046, Denver, CO 80303, KELSEY, Harvey M., III, Department of Geology, Humboldt State Univ, 1 Harpst St, Arcata, CA 95521, HEMPHILL-HALEY, Eileen, U.S. Geological Survey at Department of Geological Sciences, Univ of Oregon, Eugene, OR 97403-1272 and WITTER, Robert C., William Lettis and Associates, Inc, 1777 Botelho Dr., Suite 262, Walnut Creek, CA 94596, anelson@usgs.gov

Low-lying freshwater lakes beyond the reach of storm surges contain records of local tsunamis that can be used to reconstruct the history of great earthquakes along the Cascadia subduction zone. A 7300-yr record from Bradley Lake, formed when a sand dune blocked a stream 28 km north of Cape Blanco, contains 17 disturbance events (DEs), 13 of which record tsunami inundation higher than 5 masl. We dated the 17 DEs in Bradley Lake using 63 AMS 14C ages measured on detrital plant fragments from disturbance event beds in 11 of the 21 piston and vibracores that we described from the lake.

To evaluate the stratigraphic consistency of the 14C ages we used the sequence-analysis feature of the program OxCal using the INTCAL98 radiocarbon calibration data set. Ages were grouped in stratigraphic order using 19 phases; the ages within each phase (DE) were unordered. Ages similar enough to be considered from the same population at the 95% level were averaged. Calibrated-age distributions were calculated for means and single ages within each phase. In the sequence analysis, ages that had probabilities of <10% of being in the correct stratigraphic order were eliminated from further analyses.

In later sequences analyses, we added estimates of the minimum and maximum number of years between DEs to further constrain the 14C age distributions associated with DEs 2-16. The estimates were calculated by dividing mean sediment thicknesses between DEs by sedimentation rates determined by counting varves in finely laminated, 1- to-7-cm-thick beds of clay-gyttja that overlie each DE. Varve preservation is favored by anoxic, brackish conditions at the bottom of the lake that lasted for decades following many DEs. Times for DEs 2-9 derived by summing the number of years (estimated from thicknesses and sedimentation rates) for each interval between DEs compare favorably (r2=0.99) with the 14C ages for the same DEs. The consistency in the two types of independently derived ages supports our inference that light-dark couplets in finely laminated beds above DEs are varves. The age analyses and varve stratigraphy suggest intervals between local tsunami inundating the lake of 22-1100 years; most inundation (DE) times correlate with times inferred for great earthquakes from evidence at the nearby Sixes River estuary and at estuaries in Washington State.