Variations in the density and rigidity of the mantle induce changes in the speed at which seismic waves move through it. Mapping mantle regions with slowed and faster waves in three-dimensions is the basis for assessing temperature anomalies within the deep Earth. It has been such tomography that has begun to test ideas about the depth from which mantle plumes rise and the fate of subducted slabs of oceanic lithosphere, and an increasingly certain model for mantle motions has evolved with improvements in the resolution of seismic analyses. However, the P and S waves used in tomography have other properties than simply speed. These include direction, polarization, signs of conversion of P to S waves, and even interference properties for which the birefringence observed in petrography is an analogue.
Analysing these properties reveals that there are deviations in the structure of the minerals that make up mantle rocks from random arrangement; there are anisotropies (Park, J and Levin, V. Seismic Anisotropy: tracing plate dynamics in the mantle. Science, v. 296, p. 485-489). Deformation lines up minerals in such a way that the bulk rock structure affects the propagation of seismic waves in different directions – again, the way in which crystallographic anisotropy of minerals affects light passing through them is a means of visualizing what happens on vastly larger scales. In their review, Park and Lewin describe how this novel approach is revealing aspects of convection in the upper mantle, how lithospheric plates have formed and features spatially related to accretionary boundaries in continents.
Field studies of ophiolites have shown that the dominant olivines of mantle peridotite are commonly aligned, probably as convection dragged it at right angles to the axes of lithospheric spreading. Indeed, seismic anisotropy confirms that view with trends normal to the mid-Atlantic, Pacific and Indian Ocean spreading centres. Destructive margins show two trends, those parallel to trenches and those in the direction of subduction, but there are complex variations depending on depth. Once resolved into indicators of past motions, that complexity may tell volcanologists a lot about large-scale variations in magmatism. The Hawaiian hot spot has associated vertical anisotropy, that is consistent with a disturbance of the overall flow of shallow mantle. Several ancient orogens in continents, dating back to the Precambrian, show anisotropy in the mantle beneath them, often parallel to the orogenic trends, but occasionally more complex. Clearly, this use of natural earthquake signals has a lot to contribute, but depends on much more complex computations than “conventional” tomography and awaits the wider distribution of software and powerful hardware.
The latest significant development from tomography based on detection of wave-speed anomalies relates to the Earth’s two major mantle plumes, beneath Africa and the Pacific Ocean (Romanowicz, B. and Gung, Y. 2002. Superplumes from the core-mantle boundary to the lithosphere: implications for heat flux. Science, v. 296, p. 513-516). Both apparently persist through the transition zone of mantle wave speeds at 670 km below the surface, to become deflected laterally beneath the lithosphere. They may well be supplying heat to the asthenosphere that could find its way to spreading ridge systems. The lowering of viscosity in the asthenosphere as a result of this heat originally from the core-mantle boundary (some of it may be heat lost by the core) would act as a lubricant for plate motions. In particular, it could enhance the influence of slab-pull force at subduction zones, such as those around the Pacific, thereby speeding up tectonics. The mantle beneath the African lithosphere has probably been heated. The huge topographic and gravitational anomaly generated by massive flood basalt eruptions in Kenya and Ethiopia may more easily have been able to convert the resulting extensional stresses into extensional deformation, thereby driving the East African Rift system above a zone of thermal lubrication. Far more gravitationally unstable lithosphere beneath young orogens does undergo lateral collapse, but the lack of associated plumes makes it impossible for the entire lithosphere to fail through lack of such lubrication. And when superplumes eventually wane, as perhaps have those beneath Iceland and western North America, that too would influence both plate tectonics and that on more local scales by increasing viscous drag in the asthenosphere.