Lithosphere-asthenosphere boundary and radial anisotropy of the Australian continent from multi-mode surface waves

The lithosphere-asthenosphere boundary (LAB) is a key to the understanding of plate tectonics, but its seismological detection is not very simple due to its nature as a mechanical or thermal boundary rather than as a material boundary. Some recent works on the LAB using receiver functions of body waves have revealed evidence for clear converted signals at the bottom of lithosphere in oceanic region, whose depth is consistent with the thickness of fast velocity lid estimated from surface wave studies. But this is not the case in continental regions (in particular, cratons). While the S wave speed profile from surface waves shows fast anomalies in the top 200 km, receiver functions normally do not show clear converted signals from such depth. In this study we estimate the depth and thickness of the LAB from a radially anisotropic S-wave speed model derived from multi-mode surface waves. We have employed a fully non-linear inversion scheme to estimate path-specific multi-mode phase speeds of surface waves to map the high-resolution 3-D anisotropic shear wave model of Australia, using permanent and transportable seismic stations deployed across the continent. The LAB beneath the Australian continent is then estimated from the final SV wave speed model. Although surface waves are inherently not very sensitive to the sharpness of boundaries due to their long-wavelength features, the depth of LAB can be roughly estimated from the depth of either the negative peak of velocity gradient or the slowest shear velocity beneath the lithosphere. The former represents, in this study, an upper bound of LAB and the latter a lower bound. The thickness (or sharpness) of LAB can be deduced from the differences between the upper and lower bounds. Our new anisotropic Australian model has provided us with an insight into the relationship between the lateral variations of the LAB and radial anisotropy. In particular, anomalous radial anisotropy (SH>SV) are found within the lithosphere as well as beneath the LAB in central Australia, where we can find thinner transition to the asthenosphere, indicating the effects of past deformation of the lithosphere as well as horizontal flow in the asthenosphere. The estimated thickness of LAB is considerably thicker (over 60-80 km) beneath the West and North Australian Cratons, where receiver function studies did not show any clear signals from the base of the continental lithosphere (e.g., Ford et al., 2010, EPSL). In contrast, the eastern half of the continent tends to show a relatively sharp LAB and its depth is consistent with the results of receiver functions, which show apparent converted signals from LAB. From these results, we can deduce that the transition from lithosphere to asthenosphere in the western half of the continent (especially beneath Archean and Proterozoic cratons) is fairly gradual so that the converted signals cannot be produced efficiently in western and northern Australia.