The Afar Depression of Ethiopia and Eritrea is a feature of tectonic serendipity. It is unique in showing on land the extensional processes and related volcanism that presage sea-floor spreading. Indeed it hosts three rift systems and a triple junction between the existing Red Sea and Gulf of Aden spreading centres and the East African Rift System that shows signs of future spalling of Somalia from Africa. Afar has been a focus of geoscientific attention since the earliest days of plate theory but practical interest has grown rapidly over the last decade or so when the area has become significantly more secure and safe to visit. Two recent studies seem to have overturned one of the most enduring assumptions about what drives this epitome of continental break-up.

From the obvious thermal activity deep below Afar, linked with volcanism and high heat flow, a mantle host spot and rising plume of deep mantle has been central to ideas on the tectonics of the area. A means of testing this hypothesis is the use of seismic data to assess the ductility and temperature structure of deep mantle through a form of tomography. The closer the spacing of seismic recording stations and the more sensitive the seismometers are the better the resolution of mantle structure. Afar now boasts one of the densest seismometer networks, rivalling the Earthscope USArray. http://earth-pages.co.uk/2009/11/01/the-march-of-the-seismometers/ and it is paying dividends (Hammond, J.O.S. and 10 others 2013. Mantle upwelling and initiation of rift segmentation beneath the Afar Depression. Geology, v. 41, p. 635-638). The study  brought together geoscientists from Britain, the US, Ethiopia, Eritrea and Botswana, who used data from 244 seismic stations in the Horn of Africa to probe depths down to 400 km with a resolution of about 50 km.

The tomographic images show no clear sign of the kind of narrow plume generally aasociated with the notion of a ‘hot spot’. Instead they pick out shallow (~75 km depth) P- and S-wave  low-velocity features that follow the axes of the three active rift systems. The features coalesce at depth; in some respects the opposite of a classic plume that has a narrow ‘stem’ that swells upwards to form a broad ‘head’. If there ever was an Afar Plume it no longer functions. Instead, the rifts and associated lithospheric thinning are associated with a mantle upwelling that is being emplaced passively in the space made available by extensional tectonics. This is closely similar to what goes on beneath active and well-established mid-ocean spreading centres where de-pressuring of the rising mantle results in partial melting and basaltic magmatism along the rift system. Perhaps this is a sign that full sea-floor spreading in Afar is imminent, at least on geological timescales.

For once, mantle geochemists and geophysicists have data that support a common hypothesis (Ferguson, D.J. and 8 others 2013. Melting during late-stage rifting in Afar is hot and deep. Nature, v. 499, p. 70-73). This US-British-Ethiopian team compares the trace element geochemistry of Recent basaltic lavas erupted along the axis of the Afar rift that links with the Red Sea spreading centre with equally young lavas from volcanoes some 20 km from the axis. Both sets of lavas are a great deal more enriched in incompatible trace elements that are generally enriched in melt compare with source than are ocean-floor basalts sampled from the mid-Red Sea rift.  Modelling rare-earth element patterns in particular suggests that partial melting is going on at depths where garnet is stable in the mantle instead of spinel. This suggests that a strong layer, about 85 km down in the upper mantle is beginning to melt – magmas formed by small degrees of partial melting generally contain higher amounts of incompatible trace elements than do the products of more extensive melting. Estimates of the temperature of melting from lavas extruded at the rift axis than off-axis are significantly higher than expected at this depth suggesting that deeper mantle is rising faster than it can lose heat.

The depth of melting tallies with the thermal feature picked out by seismic tomography. The two teams converge on passively induced upwelling of hot asthenosphere while the Afar lithosphere is slowly being extended. The degree of melting beneath Afar is low at present, so that to become like mid-ocean ridge basalts a surge in the fraction of melting is needed. That would happen if the strong mantle layer fails plastically so that more asthenosphere can rise higher by passive means. The geochemists persist in an appeal to an Afar Plume for the 30 Ma old flood basalts that plaster much of the continental crust outside Afar. Those plateau-forming lavas, however, are little different in their trace element geochemistry from off-axis Afar basalts. Yet they are not obviously associated with an earlier episode of lithospheric extension and passive mantle upwelling.  Most geologists who have studied the flood basalts would agree that they preceded the onset of rifting but have little idea of the actual processes that went on during that mid-Oligocene volcanic cataclysm.