Geology Reference
In-Depth Information
is no hint of a bulge, due to putative buoyant upwelling under the spreading centre,
extending across the transform fault. This topography is difficult to reconcile with
a buoyant upwelling from depth, but is readily explicable if the mid-ocean ridge
topography is due to the near-surface and local process of conductive cooling,
thickening and thermal contraction of the thermal boundary layer (that is, of the
lithosphere).
If upwelling under normal mid-ocean rises is passive, then there is no problem
with spreading centres that move relative to other parts of the system: they merely
pull up whatever mantle is beneath them as they move around the Earth. This solves
the puzzle that led Heezen (Chapter 3) to postulate an expanding Earth in order to
try to explain how spreading centres could exist simultaneously on both sides of
Africa.
8.2.3 There is no significant 'decoupling' layer
There is a zone of low seismic velocity in the upper mantle due to the close approach
of the geotherm to the melting temperature [94]. It is plausible that mantle viscosity
is also a minimum in this zone. This led to the idea that the zone would 'lubricate'
the plates, decoupling them from the deeper mantle and allowing them to move
independently. However, a viscosity 4-5 orders of magnitude lower than ambient
mantle would be necessary for such decoupling, and there is no expectation of or
evidence for such a drastic drop in viscosity. The idea of lubrication also gained
support from efforts to model forces acting on plates [95], which concluded that
there was little 'drag' acting on the base of plates. However, this can be explained
by the mantle under plates moving essentially with the plates, both driven by
the sinking lithosphere. We were able, in Chapter 5, to explain the velocities of
plates without having to appeal to any decoupling from the underlying mantle
(Figures 5.2 and 6.3).
8.2.4 Return flow is not shallow
There was an early assumption that mantle flow implied by plate tectonics was
confined to the upper mantle, because the lower mantle had been believed to be so
stiff that it was essentially static, as we noted in Chapter 7. Some versions even
assumed that the 'return flow' is confined to the presumed low-viscosity zone, only
about 100 km thick, under the plates. This idea is closely related to the decoupling
layer idea, and the arguments against it are the same. Numerical models show that
the return flow of plates penetrates into the lower mantle, even though the viscosity
of the upper mantle is a factor of about 30 less than that of the lower mantle, as
will be seen in Chapter 9.
