Intrinsic versus extrinsic seismic anisotropy

Evidence for seismic anisotropy (radial or azimuthal anisotropy) in the Earth’s mantle has been steadily growing over several decades. Different mechanisms including LPO and CPO (related to intrinsic anisotropy), aligned cracks, partial melting and fine layering (related to extrinsic anisotropy) can give rise to seismic anisotropy. They make the interpretation of observed anisotropy at the same time important but non-unique. In this thesis, we concentrate on the separation and interpretation of seismic anisotropy in terms of intrinsic and extrinsic anisotropy. For radial anisotropy in the popular one-dimensional reference Earth model PREM, we find that that as well as intrinsic anisotropy, fine layering can be considered to explain its lithospheric anisotropy. By investigating anisotropy produced by the tilted finely layered isotropic model whose effective model can be described by 13 elastic parameters, we find that it is hard to jointly explain the apparent radial and azimuthal anisotropy in seismic data by purely intrinsic anisotropy (e.g. LPO) or purely extrinsic anisotropy (e.g. fine layering), and a combination of intrinsic and extrinsic anisotropy can be a solution of the explanation of the discrepancy between radial (about 10%) and azimuthal anisotropy (within 4%). We image a three dimensional anisotropic velocity structure in the Africa and Arabian plates from regionalized surface wave data by using our iterative quasi-Newton method together with GMRES method. We find several two-layered or three-layered stratified structures especially in the East African Rift System (EARS), the Afar regions and the Persian Gulf. Finally, we analyse the mechanism of anisotropy in Africa and Arabia, by combining the information of the mantle flow movement, the stress distribution, and the absolute plate motion (APM).