HR: 08:00h
AN: T41F-01    [Abstracts]
TI: Seismic Velocity Structure of the Southern Cascadia Subduction Zone
AU: * Roth, J B
EM: jeffrey.roth@asu.edu
AF: Arizona State University, Department of Geological Sciences Box 871404, Tempe, AZ 85281 United States
AU: Fouch, M J
EM: fouch@asu.edu
AF: Arizona State University, Department of Geological Sciences Box 871404, Tempe, AZ 85281 United States
AB: The goal of this study is to examine seismic velocity structure within the southern Cascadia subduction zone in order to place bounds on thermal and compositional variations across the region. These fundamental constraints are necessary to enable a better understanding of the thermal, mechanical, and lithologic structure of the Cascadia subduction system. Data used in this study were recorded by the Cascadia Array (CASC), a dense linear array of broadband seismometers with 4-8 km spacing that spanned from the western coast of central Oregon 300 km inland. We used a multi-channel cross correlation technique to accurately determine relative delay times for 157 regional and teleseismic P wave events, resulting in 2906 individual measurements. We assumed the IASP91 velocity model to correct raw relative delay times for moveout and also applied an elevation correction. The distribution of events provides good azimuthal coverage and excellent incidence angle coverage for many backazimuthal ranges. Clear variations in relative delay times exist across the region. Maximum peak-to-peak variations in delay times are ~2 seconds. The most conspicuous feature of the data is a relative delay time anomaly recorded at stations around 122.5 W longitude in which first arrivals are 0.5-1.0 s early relative to the mean relative delay time. This feature is consistently observed for events from all backazimuths; however, the magnitude and exact location of the peak varies as a function of backazimuth. Events arriving from the east show the greatest magnitude of variation (~1.5-2.0 s) while events arriving from the west show the least variation (~0.5-1.0 s). Preliminary analysis of S wave delay times show the same general pattern, but with greater peak-to-peak variations of up to twice that of the P data. Using these relative delay time data, we have generated relative velocity perturbation models using a linear inversion applying the VanDecar [1990] method. Results of these preliminary models are consistent with the raw data which exhibit low velocities flanking a central region of higher than average velocities near 122.5 W longitude. We are currently evaluating the possible causes for the observed rapid increase in velocity with slower flanking velocities. Strong seismic anisotropy has also been observed in data from this array and likely accounts for some, but not all, of the variation observed in the delay times. Preferential propagation of seismic energy along the subducting Juan de Fuca slab could also account for the early relative arrival times. Other possibilities for the observed variations include structure due to a serpentinized mantle wedge or thermal variations within the subduction zone.
DE: 7203 Body waves
DE: 7240 Subduction zones (1207, 1219, 1240)
DE: 7270 Tomography (6982, 8180)
DE: 8120 Dynamics of lithosphere and mantle: general (1213)
DE: 8124 Earth's interior: composition and state (1212, 7207, 7208, 8105)
SC: Tectonophysics [T]
MN: Fall Meeting 2005