No abstract

[1] Seismic reflection profile data are used to determine the stratigraphic ''architecture'' of the flexural moat that flanks the Cape Verde Islands. The moat region is characterized by upward of 1-2 km of poorly to well-stratified material. The two lowermost units thicken from west to east and are attributed to sediment loading and flexure at the nearby West Africa continental margin during Early Cretaceous to early Miocene. The two uppermost units, in contrast, thicken concentrically around the islands and are attributed to the infilling of a flexural moat that formed by volcanic loading since the early Miocene. Flexure modeling shows that the thickness of the moat infill cannot be explained only by surface loading and requires that the downward flexure due to surface loads is opposed by an upward acting subsurface load. The best fit combined surface and buried loading model is for an elastic thickness, T e , of 29 km, a load and infill density of 2700 kg m À3 , and a ratio of surface to subsurface loads, f, of 0.2. These results are in accord with spectral studies of free-air gravity anomaly and topography data based on the admittance technique. The subsurface loads are spatially limited to the islands and their submarine flanks. Nevertheless, they are associated with a broad regional uplift of up to $400 m. The uplift is large enough to explain why the moat infill is generally tilted away from, rather than toward, the islands. The uplift is too small, however, to account for the height of the swell, upon which the Cape Verde Islands are superimposed. The origin of the Cape Verde swell is not known. However, a normal T e and a modest heat flow anomaly suggest that the swell cannot be fully explained by uplift due to thermal reheating of the lithosphere by an underlying ''hot spot'' and that other, deep-seated, mantle processes must be involved.

Extensive marine geophysical surveys within the central Scotia Sea have revealed two areas of h e a t e d magnetic anomalies. The limited length of anomaly sequences makes unique correlation with the reversal timescale difficult, but a model is derived for the longer sequence requiring slow spreading during the period 21-6 Myr about a ridge trending 085". The north-south h e a t e d anomalies in the other sequence are well formed, but their identification is unavoidably ambiguous.The longer spreading sequence is interpreted as back-arc spreading behind the 'Discovery Arc', which was active during at least the period 20-12 Myr ago. The spreading was asymmetric, as is spreading at the present South Sandwich back-arc ridge. More than two-thirds of the oceanic crust of the Scotia Sea has now been shown to be less than 30 Myr old, and it is probable that it formed almost entirely withm this period, as a complication of the South American-Antarctic plate boundary.

No abstract

The nature of back-arc extension in the East Scotia Sea is re-examined with the use of an enlarged geophysical data set. Well developed oceanic magnetic lineations confirm that the present spreading episode started about 8 Ma ago, that spreading is asymmetric, and that the total rate increased from 50 to 70 m m /a about 1.5 Ma ago. Most of the currently active South Sandwich volcanic island arc lies upon ocean floor only 6-8 M a old and generated at the current spreading ridge. Subsequent extension has not modified the curvature of the arc. East-west magnetic lineations of Miocene age in the Central Scotia Sea and contemporaneous low-K arc tholeiites dredged from the eastern South Scotia Ridge (Discovery Bank) indicate a regime of coupled subduction and back-arc extension preceding that occurring now. A speculative model involving a series of collisions of parts of this earlier Discovery trench with ridge crest sections of the South American-Antarctic plate boundary explains the transformation of this earlier regime into the present, self-contained Sandwich plate regime. The considerable small-scale variability observed in the back-arc region may be seen as an inevitable consequence of the action of the ridge-trench collision mechanism. The entire Scotia Sea could have formed by a similar kind of back-arc extensional modification of the South American-Antarctic plate boundary.