Seismic anisotropy provides important constraints on deformation in the lithosphere and mantle flow patterns in the asthenosphere. Radial anisotropy can be measured from Rayleigh and Love wave dispersions; and azimuthal anisotropy is often obtained from shear-wave splitting of SKS phases and from Rayleigh wave tomography. However, the results of azimuthal anisotropy from the two methods do not agree in many areas, such as in the Colorado Rockies, Iceland, and Australia. Part of the discrepancy is due to the different properties of body and surface waves. For example, SKS splitting reflects an integral result of anisotropy from the core mantle boundary to the surface while Rayleigh waves at periods less than 100 s primarily sample the earth above 200 km. The discrepancy can also be attributed to many other factors, including the tradeoff between lateral heterogeneity and anisotropy, 3-D variations of anisotropy, and the finite frequency. The goal of our proposed study is to investigate the discrepancy between shear-wave splitting anisotropy and Rayleigh wave anisotropy in general cases. The suspected reasons listed above will be studied individually. We propose to adopt the parallel pseudo-spectral method in computing synthetic waveforms of SKS phases and surface waves in various 3-D heterogeneous and anisotropic models. This method has been successfully applied in modeling shear-wave splitting. More experiments will be performed in modeling surface waves through the pseudo-spectral method. The shear-wave splitting technique and the two-plane-wave inversion method will be applied to the synthetic data to retrieve anisotropy and heterogeneity. The proposed study will improve our understanding about the discrepancy in azimuthal anisotropy from body and surface waves.
