(2001) and Ivanov
(2003) suggest that recycled slabs may convert
to low-velocity anomalies as they heat up. The latter
authors propose that radioactivity in the slab is
the source of heating and calculated that subducted
slabs can be heated on a time-scale of 1-2 billion
years and can be converted from high-velocity to low-velocity
structures. A much faster conversion from low temperature
slabs to low-velocity anomalies can be accomplished
simply by conducting heat into the slab from the ambient
mantle into which it settles. Slabs are composed in
part of oceanic crust and in part of serpentinized
peridotite. Both CO2 and water occur in
the upper part of the slab. All these effects serve
to lower the melting point and seismic velocities
of slabs compared to dry refractory peridotite. After
the conversion of basalt to eclogite the melting point
is still low. Even small amounts of fluid or melt
can drastically lower the seismic velocity, even if
the density remains high.
The time scale for heating a slab
to above the dehydration and melting points is a small
fraction of the age of the plate upon subduction.
The low seismic velocities at the top of and above
descending slabs are the clearest manifestation of
this effect. Recycled slabs can indeed be converted
to features that are not only low-velocity but also
fertile, thus having the characteristics normally
associated with plumes (Meibom
2003). However, the time scale of this conversion
can be much less than the time scale of radioactive
heating, and probably less than 20 million years.
As far as the slab is concerned the surrounding mantle
is an infinite heat source.
The idea that low-velocity material
in the mantle can be low-temperature also is typically
overlooked in most interpretations of seismic tomography
images. Seismic velocity is controlled by composition,
mineralogy, and volatile content as well as by temperature.
If the melts and volatiles completely leave a reheated
slab then it can become a high-seismic-velocity anomaly
again, at least between depths of ~ 60 to 600 km (Anderson,
1989). Deeper than that the eclogite in a slab becomes
low-density and low-velocity compared to normal mantle.
The abundances of K, U and Th decrease
with time because of radioactive decay. This decay
is the major heat source in the mantle. Higher mantle
temperatures are expected for Archean mantle (Korenaga,
2003; Ernst et al., 2001). Since the shallow
mantle is close to or above the solidus today this
higher temperature means that more extensive and deeper
melting should have occurred in the past. In fact,
in the Archean mantle, radioactive heating was twice
as high as today and the mantle had experienced less
secular cooling. Although the need for localized high
temperatures today may be debated, the evidence for
higher temperatures and widespread volcanism in the
Archean should not come as a surprise. If the surface
of the Archaean Earth was blanketed with buoyant basalt
then the recycling mechanism was depression of the
buoyant lid into the upper mantle melting zone as
new basalt overplated the old. The Archean Earth may
have resembled Mars, Venus and some of the Jovian
satellites in having non-plate-tectonic styles of
heat removal. Emplacement of giant dikes and planetary
resurfacing are styles of heat removal in such bodies.
Even in a planet with plate tectonics, the plates
may be permeable. Dikes can deliver heat to the near
surface even if they do not erupt.
Simple heat conduction, rather than
radioactivity, may thus explain the observations described
(2003). Lord Kelvin neglected radioactivity in
his calculations of the age of the Earth, and most
subsequent workers have neglected terrestrial secular
cooling (the Kelvin effect). Demonstrations of the
Second Law of thermodynamics often invoke an infinite
heat bath, as used above. Neglect of secular cooling,
and the assumption of steady-state balance between
heat production and heat loss, also violate the Second
Law. The outer shells of Earth, at least, have certainly
cooled and radioactive heating has certainly declined.
Geological evidence against substantial cooling (survival
of cratonic roots, komatiitic temperatures, and the
Archaean catastrophe) have been addressed already
(Korenaga, 2003; Lenardic, 1998).
In summary, both high-velocity and
low-velocity tomographic anomalies can be generated
by recycled slabs. Volatile and basalt-containing
slab fragments are likely to be low-velocity anomalies,
at least in the upper part of the mantle.
Anderson, D.L., Where on Earth
is the crust?, Physics Today, 42,
- Ernst, R.E., Grosfils, E.B., Mege, D. (2001) Giant
Dike Swarms: Earth, Venus and Mars, Ann. Rev.
Earth Planet. Sci., 29, 489.
T. (1997) A model for the layered upper mantle, Phys.
Earth Planet. Int., 100, 197.
Korenaga, J. (2003) Energetics
of mantle convection and the fate of fossil heat,
Geophys. Res. Lett., 30,
Lenardic, A. (1998) On the partitioning
of mantle heat loss below oceans and continents
over time and its relationship to the Archean paradox,
Geophys. J. Int., 134,