Perspective - Mantle Convection for Geologists - page 118

Pribac [91] later offered a mechanism for such a burst of plumes, as discussed in
Section 8.2.5 and illustrated in Figure 8.1.
The term superplume quickly gained currency, but not a physical or observational
basis. It is of course not difficult to adjust the parameters of a numerical model
to produce a very strong upwelling, but that may or may not be consistent with
observations, a question that was not given much attention.
As the deep seismic anomalies under Africa and the Pacific began to be resolved
[103], the term superplume was applied to them. However, they are very broad,
narrowing towards the top, and bear no obvious resemblance to either a plume
tail or a plume head. Because such large volumes would rise very rapidly if they
were purely thermal, it seemed that they must be compositionally heavier, an
inference later quantified by Simmons et al. [104]. In that case they can plausibly
be identified with the piles of denser material that develop in numerical models that
include subducted oceanic crust [122, 123], which is denser than normal mantle
and tends to settle to the base of the mantle. We will encounter these models in
Chapter 10. They suggest that the anomalies are connected with the D seismic layer
in the lowest 200-300 km of the mantle, and that both are due to the accumulation of
old subducted oceanic crust. Where deep-mantle flow converges and turns upwards,
the accumulation is swept into piles, which can quantitatively be identified with
the deep seismic anomalies [119]. Such piles might be called thermochemical piles
or superpiles , but not plumes.
There is one case in which the much-abused term superplume might apply.
Simmons et al. [104] have imaged the sub-African structure and also separated
thermal and compositional contributions to the seismic anomaly. Their results
(Figure 8.8) show a structure that is rounded at the top and thinner part way down.
This they plausibly interpret as having a positive net buoyancy and an upward
velocity, though it might not continue to ascend all the way to the top of the mantle.
Because this structure actually has some resemblance to the head-and-tail structure
of a thermal plume, it is not unreasonable to call it a superplume.
8.4.3 Small-scale convection
The idea that the lithosphere approaches a constant thickness contributed to the
development of the idea of a pervasive mode of 'small-scale' convection confined
to the upper mantle. The constant-thickness lithosphere model implied a heat input
to the base of the lithosphere, in order to maintain an asymptotic steady-state heat
flux at the surface, and it was proposed that this was due to some form of sub-
lithospheric small-scale convection. Initially this was supposed to be convection
cells of the scale of the upper mantle, to which this mode of convection was assumed
to be confined [124-126].