2009 Portland GSA Annual Meeting (18-21 October 2009)

Paper No. 7
Presentation Time: 3:25 PM

UPLIFT ASSOCIATED WITH SLOW SLIP EVENTS ON THE CASCADIA SUBDUCTION ZONE RESOLVED IN TIDAL RECORDS


WELDON, Ray J., Department of Geological Sciences, University of Oregon, 1272 University of Oregon, Eugene, OR 97403, LIVELYBROOKS, Dean, Department of Physics, University of Oregon, 1371 E 13th Avenue, Eugene, OR 97403 and SCHMIDT, David, Department of Geological Sciences, University of Oregon, 1272 University of Oregon, Eugene, OR 97403-1272, ray@uoregon.edu

We have determined the average relative uplift between Port Angeles and Port Townsend, Washington, associated with the past 10 slow slip earthquakes on the Cascadia subduction zone by analyzing hourly water level records from the NOAA tide gauges at those sites. The average relative uplift of Port Angeles, 4.4 mm, and duration of the transient, 16 days, are identical to that inferred from modeling horizontal GPS data for the same events. Verified hourly water levels at the two sites were compared with a transfer function to remove correlated ocean and atmospheric “noise”. The transfer function is constructed by mapping Fourier coefficients determined from the de-tided time series at the two sites. Analysis includes coherence filtering for individual transfer function frequencies and complex multipliers that account for amplitude and phase shifts (necessary because weather and other noise sources vary in time and magnitude between sites). The de-noised hourly records are averaged into days, and 100 day windows centered on the 10 known slow slip events were stacked to produce the average signal. Approximately 20 usable tidal records exist along the Cascadia subduction zone, with the longest record, Seattle, exceeding 100 years. Relative uplift between these numerous sites should provide an image of the vertical component of deformation associated with slow slip events, complementing the more precise horizontal motion measured by GPS. While our stacking approach takes advantage of the known age of the events, 7 of the 10 individual records contain identifiable signals. This suggests that individual events may be identifiable in the tidal records with additional improvement in our processing technique, and that this approach could be extended back in time for Cascadia events and to other subduction zones with tidal records.