ANISOTROPY IN THE OCEANIC LITHOSPHERE - THE NGENDEI SEISMIC REFRACTION EXPERIMENT (PACIFIC)

P and S-wave travel times and synthetic seismogram modeling of seismic refraction data from the 1983 Ngendei expedition to the south Pacific indicate seismic anisotropy both within the top kilometer of oceanic crust and in the uppermost mantle. The anisotropy at both depths is characterized by a 20 azimuthal dependence of velocity, with the fast crustal direction at N120(DEGREES)E and the fast mantle direction at N30(DEGREES)E. We calculate average P and S-wave velocity versus depth functions for these two orthogonal azimuths by using WKBJ synthetic seismogram modeling of phase and amplitude information. Probable lateral heterogeneities at the Ngendei site prevent perfect matching of data and synthetic waveforms. Crustal anisotropy is indicated by a 2(theta) pattern of both P and S-wave travel times as a function of azimuth, and is consistent with an approximate 0.2 km/s difference in P-wave velocities and 0.1 km/s difference in S-wave velocities. Upper mantle anisotropy is characterized by P(,n) velocity differences of 7.95 to 8.4 km/s and a nearly uniform S(,n) velocity of 4.65 km/s. We use these velocity profiles and travel time data to calculate bounds on the properties of the elastic constants of the crust and upper mantle. The crustal anisotropy can be explained by a model involving aligned cracks parallel to the original spreading ridge; the upper mantle anisotropy is consistent with aligned olivine models, in which the fast direction is parallel to the fossil spreading direction. For both of these models, we use bounds from the Ngendei data to place constraints on the physical properties of the lithosphere. We also examine polarizations of arriving phases to look for P-wave polarization anomalies and S-wave splitting, but find that the Ngendei polarization data are too noisy to resolve anisotropy.