Geoscience Reference
In-Depth Information
very short time. This can be shown by replacing in (5.17) the values for residence time
obtained from the mass of the two parts of the mantle and the flow of plates across the
transition zone, to give a chemical relaxation time of 200-400 million years. Under these
conditions, how could the geochemical identity and, above all, the isotopic identity of
MORB be conserved relative to OIB?
It was virtually impossible for a long time to find incontrovertible evidence of deep
seismic activity that could indicate plate penetration into the lower mantle.
With a mantle homogenized by convection, the radioactive element contents of the total
mantle are equal to those of the upper mantle. So we can calculate the ratio between the
heat produced by the Earth's radioactivity and that (43 TW) given off from the Earth's
surface. This ratio, the Urey ratio, is then much smaller than unity (
0.4). This would
imply that the Earth still contains a considerable amount of the initial heat derived from
the gravitational energy of accretion and core segregation locked at depth. If this condi-
tion is calculated back to the time our planet formed, it leads to an absolutely infernal
temperature of formation, much higher than the vaporization temperatures of silicates.
The argument from radiogenic argon has played a very significant role. The atmosphere
contains a large quantity of radiogenic argon 40 Ar. If the total quantity of 40 Ar produced
by decay of terrestrial 40 K is estimated, and if allowance is made for the very low levels
of this gas in the degassed upper mantle, an inventory of it for the planet can be drawn
up. We quickly reach the conclusion that nearly 50% of the mantle was never degassed
and so has never risen near the surface beneath ridges.
These powerful arguments were the basis for the canonical two-layer convection model
(lower panel of Fig. 11.14 ), which still pervades the literature. The upper and lower mantle
are separated by the transition zone at 660 km and convect separately. Plates are recycled
within the upper mantle and hot spots and their basalts (OIB) rise from the transition zone.
The lower mantle is of virtually primeval composition and has very little involvement,
supplying only the argon and helium found in OIB.
A number of observations cloud this ideal picture. Except for rare gases, there is little
evidence that large expanses of primitive mantle contribute to the genesis of MORB or
OIB. The second is that the terrestrial balance of some elements or isotopic systems is not
at all consistent with the idea of a primitive lower mantle: Nb and Ta are depleted in both the
continental crust and the MORB source with respect to similarly incompatible elements,
the U-Pb, Re-Os, and Lu-Hf isotopic systems are not complete when known reservoirs are
added up. The third observation was decisive. Modern high-resolution seismic tomography
frequently shows lithospheric plates crossing the boundary between the upper and lower
mantle.
All the arguments against whole-mantle convection have also been severely weakened
by recent observations and concepts. The canonical two-layer convection model of the
1980s with separation at the 660 km transition zone seems no longer viable. This does not
mean that the discontinuity does not hamper, and even locally stop transfer of material.
What is the current situation? As we just discussed, it is probable that convection has
not fully homogenized the initial mantle of the planet and that streaks of rocks inherited
from the earliest time, have survived convective stirring at very great depths (1500 km)
 
 
Search WWH ::




Custom Search