Seismic tomography – the processing of records of  seismic waves from many earthquakes that arrive at the world-wide network of receiving stations – continues to add detail to structures in the mantle. It is based on 3-dimensional mapping of variations in wave speeds that gives clues to variations in temperature and rheological properties at depth. One of its most fascinating outcomes has been the detection of thick, steeply dipping sheets of anomalous material well below the 660 km mantle discontinuity where earthquakes cease to occur, i.e. where the whole mantle behaves in a ductile manner. These show signs of linkage to near-surface destructive plate margins, and have been ascribed to lithospheric slabs that continue to be subducted as discrete entities to as deep as the core-mantle boundary (CMB). If that were the case, it follows that their accumulation in this D” region might displace other deep material laterally, perhaps to set mantle-wide convective plumes in motion.

One such sheet occurs deep beneath the Caribbean, and is attributed to the remnants of  a lithospheric plate, once forming the foundation of the eastern Pacific, which ceased to form once North America had overridden the East Pacific Rise. By analogy with the 160 Ma width of the West Pacific plate, this one would have been sufficiently extensive to reach the CMB once subducted. New tomography beneath the region no only suggests that it did, but that in doing so it accumulated as a heap of buckled material (Hutko, A.R. et al. 2006. Seismic detection of folded, subducted lithosphere at the core-mantle boundary. Nature, 441, p. 333-336). The reconstruction from tomographic results is highly reminiscent of the folding that occurs when honey or treacle is tipped into a tumbler of hot tea and falls to the bottom.. If the interpretation is correct, part of .the D” zone is made up of gigantic recumbent folds of former oceanic lithosphere.

Afar and the African superplume

Seismic tomography has also played a role in mapping zones in which hot, low-density mantle is likely to be rising – a contribution to understanding how plumes give rise to near-surface hot spots and major intra-plate volcanism. One of the largest active and long-lived zones of such thermal and magmatic activity is that of Ethiopia and Yemen, connected somehow with the opening of the Red Sea, Gulf of Aden and the East African Rift system; the Afar plume. This began about 45 Ma ago in Kenya and southern Ethiopia, reached its climax with the rapid extrusion of vast continental flood basalts of the Ethiopia-Yemen province around 30-26 Ma, and continues today in the Afar Depression. Thought by some to be a classic example of how a single upwelling of hot, low-density mantle generated a magmatic and tectonic hotspot, an alternative view is that the Afar plume is a mere near-surface part of a vast and complex system of anomalous mantle beneath the whole of southern and eastern Africa. Tomography based on the world-wide network of seismic observatories is unable to resolve the matter one way or the other. Geophysicists of the Pennsylvanian University and Carnegie Institution in the USA have analysed data from a more closely spaced network of temporary seismic stations around the famous RRR triple junction of Afar (Benoit, M.H. et al. 2006. Upper mantle P-wave speed variations beneath Ethiopia and the origin of the Afar hotspot. Geology, v. 34, p. 329-332).

The results outline a wide (>500 km), elongated region of low P-wave speeds below 400 km that trends south-west from Djibouti, roughly parallel to the Ethiopian Rift. This is far too large to represent a classic plume, whose tails are thought to be no more than 100-200 km diameter, and whose heads on reaching the base of the lithosphere are no more than 100-200 km thick, despite spreading laterally to a radius of up to 2000 km. The huge structure is more consistent with a broad mantle upwelling that penetrates down to the lower mantle. Lower-resolution tomography does show anomalous low-speed mantle in a broad zone, which is deep in the mantle below southern Africa then rises obliquely towards the vicinity of Afar. The more detailed results support the influence of this African ‘superplume’.

Crustal spreading from the Tibetan Plateau

In the mid 1970s Peter Molnar and Paul Tapponnier proposed that the active tectonics of eastern Asia were driven by gravitational collapse and lateral spreading of the huge mass of thickened crust that had accumulated beneath Tibet after India collided with Eurasia. The driving forces for such lateral spreading are variations in gravitational potential energy (GPE) due to regional differences in surface elevation. In the oceans, such GPE adds to plate driving forces as sliding from oceanic ridge systems that are elevated relative to abyssal plains because ridges are underlain by warmer, lower density oceanic lithosphere. Partly because the continental surface is not covered by water up to 4 km deep, the stresses resulting from GPE associated with Tibet’s high elevation are about twice as large as those connected with ridge slide. Computing the variations in GPE in eastern Asia and the adjoining oceans allows the magnitudes and directions of stresses due to gravitational spreading to be mapped (Ghosh, A. et al. 2006. Gravitational potential energy of the Tibetan Plateau and the forces driving the Indian plate. Geology, 34, p. 321-324).

One of the oddities discovered by Ghosh et al. is that the dominant stresses resulting from GPE differences in Tibet are oriented N-S and would tend to cause crustal spreading in those directions. Yet the surface of the Tibetan Plateau is riven with numerous N-S normal faults that indicate current spreading in E-W directions, as Molnar and Tapponier surmised. Somehow the N-S gravitational extension forces must be cancelled out, probably by traction between the lithosphere and motion of the underlying mantle driven by sea-floor spreading from the ridges in the Indian Ocean. One possibility is that the known buckling and thrusting within the oceanic part of the Indian Plate is a reflection of this balance. However, the stresses that emerge from the GPE calculations are simply not large enough to account for this intraplate deformation.