Surface wave dispersion, regionalized velocity models, and anisotropy of the Pacific crust and upper mantle

Summary The Rayleigh and Love wave phase and group velocities of Yu & Mitchell (1979) have been inverted after applying anelasticity corrections corresponding to laterally varying Q structures beneath the Pacific Ocean. The derived models indicate that the lithosphere thickens with age at the expense of the asthenosphere, that no resolvable variations of velocity occur in the asthenosphere, and that resolvable variations of velocity in the lithosphere occur only in the younger portions of the Pacific. Azimuthal anisotropy is small (less than 1 per cent) throughout the Pacific. Polarization anisotropy is larger, being resolvable in the lithosphere, but not in the asthenosphere, of all the regions of this study, using data at periods between 20 and 100 s. Limitations of our inversion procedure for anisotropic models were investigated by doing forward calculations assuming that the lithosphere is transversely isotropic with symmetry axes oriented vertically. The results for our models, in which anisotropy is restricted to the lithosphere, indicate that SH and SV velocities in the lithosphere differ only slightly from the velocities derived from our formal inversion procedure. A two‐stage model is proposed for the generation of anisotropy in the oceanic lithosphere. Azimuthal anisotropy is generated at the ridges by preferential glide in olivine crystals (Francis 1969) and is thus restricted to the uppermost mantle. As the plates spread and grow older, however, they thicken by the accretion of crystals at their bases which have a preferred vertical orientation. This model is capable of explaining all of the observations of anisotropy presently available from oceanic seismic refraction and surface wave studies.