Geology Reference
In-Depth Information
the mantle. The uppermost mantle would therefore be less fertile, produce less
melt, and result in a thinner oceanic crust.
(There are minima in the curves in Figure 9.8(a) that correspond with the melting
zone, but the gradient below that is clear. The results in Figure 9.8(b) are a more
direct and reliable measure of the depletion.)
Figure 9.8(b) shows the crustal thickness calculated from the model, based on
the number of tracers removed from the melt zone into the oceanic crust ('Actual').
Remarkably, it averages only 5-8 km and is never greater than about 10 km, being
smaller in the past. For comparison, the crustal thickness that would be gener-
ated by mantle of normal basaltic content is also shown ('Fertile'). The fertile
thickness declines steadily from almost 40 km. The difference between these two
curves is a measure of the degree of depletion relative to the present mantle fer-
tility, and it shows a depletion by a factor of 3-6 early in Earth history. Other
models of this type have shown comparable degrees of depletion [123, 155, 156].
This has important geochemical implications as well, which will be discussed in
Chapter 10.
The implication of this result is that the buoyancy of the oceanic crust may not
have been an important factor inhibiting subduction. Thus the possibility that plate
tectonics has operated through most of Earth history is revived. Of course, there
may be other reasons why plate tectonics did not operate, but this difficulty does
not seem to be as serious an obstacle as it first appeared.
9.4.3 A mid-mantle basalt barrier and early mantle overturns
The other depth range in which oceanic crust is believed to be less dense than
average mantle is below the transition zone, 660-750 km depth. In this depth
range, it will therefore tend to rise, whereas in the upper mantle it tends to sink.
Thus there will be a tendency for old subducted oceanic crust to accumulate around
660 km depth. This turns out to have strong effects early in Earth's history.
An example is shown in Figure 9.10. In the first two panels of this model, tracers
of subducted oceanic crust have accumulated around 660 km depth to such an
extent that they have blocked vertical flow, and the convection is separated into two
layers. There is a large temperature difference between the layers. These features
are not present in the later two panels.
A fuller picture of the behaviour of this model is conveyed in Figure 9.11, which
shows summary plots of the evolution of the model. Immediately evident are very
large swings in the upper-mantle temperature (panel (a)). The first billion years of
these swings are shown on an expanded timescale in panel (c), and it is apparent that
the upper-mantle temperature makes sudden jumps upwards, followed by smooth
declines leading into another jump.
