Geoscience Reference
In-Depth Information
and the planform of mantle convection. Such
perturbations include stress changes in a plate,
and ridge--trench and continent--continent colli-
sion and separation. Changes in boundary condi-
tions are transmitted essentially instantaneously
throughout the system and new plates and new
boundaries form. These topside changes cause
mantle convection to reorganize or reverse. The
lithosphere is not
rigid
, as generally assumed, but
is fractured and is weak in extension. It reassem-
bles itself into new plates, held together by lat-
eral compression. Regions of extension become
volcanic chains and plate boundaries. The sign
of the horizontal stress field determines what
becomes a plate and what becomes a plate bound-
ary. Plate tectonics is, therefore, an example
of
self-organization
. If the convecting mantle was
bounded by isothermal free-slip surfaces it would
be free to self-organize and to organize the sur-
face motions, but it is not and cannot.
Near-equilibrium systems respond to a fluc-
tuation by returning to equilibrium. Systems
far from equilibrium may respond by evolv-
ing to new states. The driving forces of plate
tectonics are thermal and gravitational and
therefore change slowly; mass distributions,
mantle temperature and viscosity are slow to
change. Dissipation forces, such as friction and
continent--continent collision, however, change
rapidly as do normal stresses across subplate
boundaries and tensile stresses in plate interiors.
These stress-related fluctuations can be large and
rapid and can convert the slow steady thermal
and gravitational stresses into episodic 'catastro-
phes'. This is the essence of
self-organization
. This
is the reverse of the view that it is mantle con-
vection that causes breakup and reorganization
of plates and giant igneous events, i.e. a rapidly
changing and localized event in the convecting
mantle or at the core--mantle boundary.
In plate tectonics and mantle convection, as
in other slow viscous flow problems, there is
a balance between buoyancy forces and dissipa-
tion. Basic questions are: where is most of the
buoyancy -- or negative buoyancy -- that drives
the motion, and where is most of this motion
resisted? In normal Rayleigh--Bénard convection
the buoyancy is generated in thermal boundary
layers and is due to thermal expansion. The dissi-
pation -- resistance to vertical flow -- is the viscos-
ity of the mantle. In a homogenous layer of fluid
heated from below and cooled from above the
upper and lower TBLs contribute equally to
the buoyancy and there is a basic symmetry to
the problem. In a spherical container with mostly
internal heating, the upper TBL removes more
heat than the lower TBL and the heated regions
move around so all parts of the interior can
deliver their heat to the surface. If viscosity
and strength are temperature dependent then
the cold surface layers will resist motions and
deformation, more so than the hotter interior.
The main resistance to plate motions and man-
tle convection may therefore be at the surface.
In one limit, the surface does not participate
in the motions and acts simply as a rigid shell. In
another limit, the plate acts as a low-viscosity
fluid and has little effect on interior motions.
Plates and slabs are driven by gravity and
resisted by mantle viscosity, plate bending and
friction between plate subunits. The driving
forces change slowly but resisting and dissipat-
ing forces in and between plates can change
rapidly, and are communicated essentially instan-
taneously throughout the system. Episodic global
plate reorganizations are inevitable. Even slow
steady changes can reverse the normal stresses
across fracture zones and can start or shut off
volcanic chains. Construction and erosion of vol-
canoes change the local stress field and can gen-
erate self-perpetuating volcanic chains; the load
of one triggers the next. All of these phenomena
are controlled by the lithosphere itself, not by
a hot convective template from the underlying
mantle or introduction of core--mantle plumes
into the shallow mantle.
Self-organized systems evolve via dissipation
when a large external source of energy is avail-
able, and the systems are far from equilibrium.
This self-ordering is in apparent violation of
the second law of thermodynamics. This process
is called
dissipation controlled self-
organization
. A fluid heated uniformly from
below is an example of far-from-equilibrium self-
organization. The static fluid spontaneously orga-
nizes itself into convection cells but one cannot
predict when, or where the cells will be, or their
sense of motion. In contrast, if the heating or
