Influence of Upper Mantle Anisotropy on Isotropic P-Wave Tomography Images Obtained in the Eastern Mediterranean Region

Seismic body-wave tomography studies typically assume an isotropic upper mantle, possibly mapping anisotropy into artificial isotropic velocity anomalies in the resulting images. The Eastern Mediterranean with its oceanic, continental, and extinct subduction systems, as well as dense station coverage, provides an ideal setting to explore this issue. To examine the influence of seismic anisotropy, our study deals with both synthetic and real data inversions in which realistic seismic anisotropy models derived from 3D mantle convection simulations and shear wave splitting measurements are taken as a priori constraints. Spatial large-scale velocity perturbations are mostly consistent between models derived with and without considering anisotropy. Small differences in the magnitude (up to 2%) and shape of velocity perturbations occur, and some structures are less diffuse when including anisotropy. Additionally, good backazimuthal coverage of teleseismic events and a larger data set improve the resolution of our model with respect to previous tomography studies and allow us to better interpret first-order isotropic velocity anomalies. Key features, such as the half-arc subducting oceanic plate in the southern Aegean and a wide and deep tear in the slab beneath southwestern Turkey, are clearly visible in all models. Our final tomography images also provide evidence for a shallow horizontal tear in the northern Hellenides and a vertical tear between two parts of the Cyprian slab. In eastern Anatolia, slab-related high-velocity anomalies are absent due to the continental collision and break-off.