Geology Reference
In-Depth Information
these zones are quite variable, as they are pushed around by sinking subducted
lithosphere. This picture corresponds quite well in general terms with the picture
from seismology of a thin D
zone, perhaps 200 km thick, and larger anomalies
extending hundreds of kilometres upwards under Africa and the western Pacific,
as described earlier (Figure 8.8).
This kind of model also yields important geochemical results that will be taken up
in Chapter 10. Indeed, that was a primary motivation of Christensen and Hofmann's
original models. This model, and others shown here, assume the operation of plate
tectonics throughout the Earth's history. It is not known if plate tectonics really
did operate in the first half of Earth history, but any behaviour of the top thermal
boundary layer that formed and recycled a mafic crust would probably yield mafic
accumulations like those found here. One can treat the assumption of plate tectonics
as a hypothesis. If it is not true, then discrepancies with evidence ought to emerge. In
the meantime it yields well-defined models that provide a benchmark for alternative
hypotheses.
9.4.2 Buoyancy of oceanic crust
The oceanic crust is significantly less dense than the mantle within about 60 km
of the surface. As lithosphere begins to subduct, the compositional buoyancy of
the oceanic crust will counteract some of the thermal negative buoyancy of the
lithosphere. Except for lithosphere younger than about 20 Ma, the negative thermal
buoyancy dominates and the lithosphere can subduct. The implications are fairly
minor for the present Earth. Presumably any lithosphere younger than about 20 Ma
that subducts must be pulled down by attached older lithosphere, or by other forces
in the system.
However, in the past the implication might be very important. If the mantle is
hotter, then there ought to be more melting under mid-ocean ridges and that should
produce thicker oceanic crust. At the same time, convection will be faster in a
hotter mantle and plates will be younger, on average, when they reach subduction
zones. Thus the negative thermal buoyancy is lower and the positive compositional
buoyancy is greater. This means that plates will have to be older than 20 Ma before
their net buoyancy becomes negative. Positively buoyant plates cannot subduct.
They would have to age further until they were able to subduct. Thus plate tectonics
might still operate, but more slowly than required to cool the mantle. This could
lead to a thermal runaway in which the mantle gets hotter in time, and this leads
to the paradox that its present cool state cannot be explained. Thus, presumably,
something else would have to have happened.
This argument was made in 1992 [154] and for some time seemed to preclude
plate tectonics before about 2 Gyr ago. However, models like that in Figure 9.5
