Maps of shear-wave velocity in the upper mantle show fast propagation velocities to depths of order 200-300 km beneath parts of the continents. Lithospheric thickness estimates based on such measurements show that large variations exist in the thickness of the continental mantle lithosphere. Most of the older cratons are associated with lithospheric thickness of order 250 to 300 km, whereas younger regions may have thickness more comparable to that of oceanic lithosphere ( ∼100 km). Major contrasts in thickness exist across recognised geological boundaries such as the Tornquist-Teisseyre line in eastern Europe and the Tasman line of eastern Australia.
The apparent stability of these domains of thickened continental lithosphere tells us something about the properties of the lithospheric mantle in these regions. If the thickened lithosphere is of the same composition as normal mantle, and it is relatively colder, one might expect that the steep gradients of density implied by the horizontal gradients of temperature would cause a gravitational overturn to occur in the upper mantle, replacing the cold lithospheric root with hot asthenosphere. On the other hand, even if a positive buoyancy causes this thick lithosphere to be gravitationally stable, then the mantle lithospheric roots must also be stabilised against displacement caused by the shear on horizontal planes associated with horizontal movement of lithosphere relative to asthenosphere.
One might argue then that, relative to the surrounding mantle, these mantle lithospheric roots have either anomalously higher viscosity, or intrinsically lower density, or both. There are good reasons why both may be true. More surprising, however, is that a thick lithospheric root can persist beneath a region where active convergence has thickened the upper layers by a factor of order two. It has been argued previously that convergent deformation of the lithosphere can effectively trigger gravitational instability and replacement of thickened lithosphere by asthenosphere. The Tibetan lithosphere is cited as an example of lithospheric convective thinning or delamination triggered by the continental collision between India and Asia (England and Houseman, J. Geophys. Res., 1989).
We argued that the simplest explanation of the onset of normal faulting in the late Miocene Plateau required an increase in gravitational potential energy that is most simply explained by replacing thickened mantle lithosphere with hot asthenosphere. McKenzie and Priestley (Lithos, 2007), however, have determined, on the basis of surface-wave analysis, that Tibet still possesses an apparently cold lithospheric root. In attempting to resolve these apparently conflicting results, I will review the various arguments that have been put forward for and against a lithospheric gravitational instability having occurred beneath Tibet.