Geoscience Reference
In-Depth Information
elements toward the surface. The drawn-out cool-
ing also implies that mantle viscosity is not just
a function of temperature or that surface bound-
ary conditions may also regulate the vigor and
style of mantle convection. A chemically strati-
fied mantle, even with the radioactivity toward
the top, will cool slowly. On the other hand, if
the deep layers are undepleted or enriched, they
will warm up with time, and the system may
overturn.
continents. The Americas and Australia are cur-
rently over-riding mantle cooled by recent sub-
duction. The African crust and lithosphere have
little knowledge of temperatures and convective
vigor of the Pacific mantle. Realistic thermal
history calculations include the coexistence of
regional forms of convection and the multiple
branches available to mantle evolution and cool-
ing. Continents also introduce a long-wavelength
lateral temperature gradient at the surface of the
mantle. This alone can drive mantle flow. These
lessons show us that secular cooling -- the Kelvin
effect -- may not be lightly dismissed as an impor-
tant contributor to surface heat flow, that the
heat loss in oceans is affected by the presence of
continents, and that mantle convection can be
organized and driven by conditions at the sur-
face. The plates may control the cooling of the
mantle.
Regional Earth models
The mantle loses two to three times as much heat
per unit area in oceanic areas as in continen-
tal areas. One cannot discuss heat flow through
continents, at present or in the past, without
taking into account this partitioning. A corol-
lary is that geological evidence from continents
provides little constraint on average mantle tem-
peratures or secular cooling rates. Continents
affect the style of mantle convection, the rate of
heat loss and the locations of downwellings. The
growth rate of continents
and the partition-
ing of mantle heat loss below oceans
and continents
may serve to maximize man-
tle heat loss. The motions of continents, plate
reorganizations and the creation of new plate
boundaries permit pent-up heat to escape and
these introduce temporal variability into global
heat flow.
The
petrogenesis and inferred tem-
peratures of Archean komatiites
and the
survival of cratonic roots may argue against
an extremely hot mantle in the Archean and,
thus, against substantial secular cooling. It has
been thought that such cooling could be ruled
out since the very high temperatures and con-
vective vigor predicted for the mantle do not
show up in continental geology; this is the
Archean paradox
. These arguments are based
on one-dimensional and static continent consid-
erations. Continents both affect the distribution
of heat flux and move around. A large fraction
of the heat flow, and the evidence for possi-
bly elevated mantle temperatures in the past,
is therefore in the missing ocean basins and
does
Thermal history
There can be a long delay between the generation
of heat in Earth's interior and its arrival at the
surface. A tabulation of current heat sources and
heat losses is therefore only part of the story. To
understand the Earth's thermal budget requires
information about thermal history. Some esti-
mates of global heat flow appear to greatly
exceed that available from current radioactive
heating in the mantle. The ratio of heat pro-
duction to total heat flux is called the
Urey
ratio
.
The viscosity of the mantle is an important
issue. Viscosity is temperature dependent and
high temperatures result in low viscosity and
increased convective vigor. This negative feed-
back serves to buffer mantle temperature. How-
ever, water content also affects viscosity, and the
melting point. Magmatism extracts water from
the mantle causing the viscosity to increase, even
if the temperature is high. Therefore, high con-
vective vigor and extensive magmatism in the
past could actually lead to decreased convective
vigor and lower heat flow. Magmatism plays a
minor role in current estimates of heat flow
but it may have been a more important mech-
anism for removing heat in the past. The current
not
show
up
in
the
thermal
history
of
