Geology Reference
In-Depth Information
magmas reaching the surface. Seismology has also directly resolved the oceanic
lithosphere. The bottom thermal boundary layer is inferred from less direct but still
robust arguments, principally that the core is likely to be hotter than the mantle.
The two thermal boundary layers behave very differently, and the reason is to
be found in their mechanical properties. The rheology of mantle material depends
strongly on temperature. At higher temperatures, the main effect is that the vis-
cosity decreases as temperature increases further. This means that upwellings from
the bottom thermal boundary layer take the form of columns (rather than sheets),
and that the columns start with a large head, with a thinner tail following. On the
other hand, at lower temperatures, mantle rheology changes its character, from a
deforming (approximately viscous) fluid through an intermediate range to effec-
tively a brittle solid at the surface. This means that the colder parts of the top
thermal boundary layer are strong, and move as 'rigid' units. The moving parts, the
plates, are separated by narrow faults or shear zones. If a plate descends into
the mantle, it usually retains its sheet structure at least within the upper mantle,
according to images from seismic tomography [90].
The plates, being pieces of the top thermal boundary layer, are the active com-
ponents in the plate mode of convection. Because they are strong and move as a
unit, the flow under them will be coherent. They also control where upwellings and
downwellings occur. The result is a mode of convection in which each plate drives
a 'cell' or roll-like flow.
The plumes are also active components, but deriving from the bottom thermal
boundary layer and driving a quite different form of flow. The plumes are narrow,
columnar upwellings rather than sheets. By conservation of mass there will be a
corresponding broad, slow downwelling driven by each plume. Plumes seem to be
capable of penetrating the rolls of the plate mode, judging by the low correlation
between volcanic hotspots and plate boundaries.
Plates and plumes must be regarded as separate agents playing different roles in
the mantle, though the flows they generate do interact, as we will discuss shortly.
A fundamental difference is that plates cool the mantle whereas plumes cool the
core. We have seen that the cycle of upwelling under a mid-ocean ridge, cooling
to form a plate, subduction and reheating in the mantle acts to remove heat from
the deep interior of the mantle. The upper thermal boundary layer is where heat is
lost from the mantle by conduction. On the other hand, the lower thermal boundary
layer is where heat enters the mantle from the core. Plumes rising from this thermal
boundary layer transport that heat through the mantle, and most of it stays under
the lithosphere and is stirred back into the mantle interior. Only a minor amount
conducts through the lithosphere to the surface.
A couple of implications follow immediately from recognising plates and plumes
as independent agents playing different roles. One is that 'plume tectonics' is not
