Geoscience Reference
In-Depth Information
In the fluid outer core, a dynamo process con-
verts thermal and gravitational energy into mag-
netic energy. The power needed to sustain the
dynamo is set by ohmic losses.
Estimates of
ohmic losses in the core
coverawiderange,
from 0.1--3.5 TW with more recent estimates in
the range of 0.2--0.5 TW. The lower estimates
remove the need for radioactive heating in the
coreandallowsthe
age of the solid inner
core to exceed 2.5 billion years
.Inorder
to sustain the dynamo, the heat flow from the
core must be 5--10 times larger than the ohmic
dissipation of 1--5 TW.
The fate of core heat that enters the man-
tle is controversial. Some workers assume that
heat from the core is removed efficiently from
the mantle and taken directly to the surface via
narrow plumes or heat pipes. Heat from the man-
tle is removed by large-scale convection and plate
tectonics. If this is so, the thermal evolution of
the core and the mantle can be treated sepa-
rately. If mantle convection is strong, instabili-
ties at the core--mantle boundary (CMB) will be
swept away and entrained in mantle convection.
Core heat then just gets added to mantle heat.
Since more heat is generated near the surface of
the mantle, and more heat passes through the
surface boundary layer than passes through the
CMB, and since thermal expansion and viscosity
favor rapid turnover at the top, mantle convec-
tion must be primarily driven from the top and
not the bottom. Narrow instabilities are unlikely
to pass unperturbed through the whole mantle,
even if they can form in the first place, in the
presence of convection driven from above and
internal heating.
The most important considerations in the fate
of core heat are the thermal properties at lower
mantle conditions. Thermal conductivity is much
higher than at the top of the mantle because of
the combined effects of pressure on the lattice
conductivity and the effect of temperature, grain
size and composition -- including spin-transitions
-- on radiative conductivity. This promotes the
establishment of a thick, and sluggish, conduc-
tive TBL. The coefficient of thermal expansion is
verylowattheCMB,perhapsasmuchasanorder
of magnitude lower than at the top of the mantle.
This means that an increase in temperature -- and
heat from the core -- does not yield much ther-
mal buoyancy. Only very large features develop
enough buoyancy to rise. Pressure increases the
viscosity, making it difficult for even very large
features to rise. If the deep mantle is chem-
ically denser than the mesosphere it may be
trapped since temperature is ineffective in bring-
ing the density down low enough so it can escape.
These considerations are neglected in the plume
hypothesis for core cooling, which is based on
laboratory injection experiments where pressure
effects are minimal, and Boussinesq calculations
where pressure effects on material properties are
ignored. The net effect is that core heat is proba-
bly added to mantle heat, and is carried away by
conduction or radiation, or by large-scale slug-
gish upwellings.
Although small in magnitude, core heat may
dominate the heat budget at the base of the man-
tle if most of the radioactive elements are at the
top of the mantle, and most of the secular cool-
ing is in the top layers, the case for a chemically
stratified mantle. The core mainly cools by rapid
conduction into the base of the mantle rather
than by short-circuits to the top of the Earth. The
temperature gradients at the base of the man-
tle vary from place to place. The core will lose
heat most efficiently through the colder parts of
the lower mantle. These are not the places one
expects to find hot plumes.
Heat fluxes
Heat is removed from the interior of the man-
tle by conduction through the surface thermal
boundary layer (TBL), intrusion into the TBL,
hydrothermal circulation near the surface, and
volcanism. In addition, the interior is cooled
by subduction of cold slabs and warm delam-
inated lower crust. The relative importance of
these mechanisms changes with time. The con-
duction layer is, in part, chemical and perma-
nent -- although mobile -- and, in part, tran-
sient. This means that the heat-flux problem is
not one-dimensional (1D) or steady-state. Heat
can be diverted to regions having thin TBL, and
can, to some extent, be temporarily stored in the
mantle.
