HR: 12:05h
AN: T21F-08 [PDF]
TI: Some Geochemical Implications of the Mantle Transition-Zone Water-Filter
Model
AU: * Karato, S
EM: shun-ichiro.karato@yale.edu
AF: Yale University, Department of Geology & Geophysics, New Haven, CT 06520 United States
AU: Bercovici, D
EM: david.bercovici@yale.edu
AF: Yale University, Department of Geology & Geophysics, New Haven, CT 06520 United States
AB:
Geochemical observations are usually interpreted in terms of the Earth's
initial composition and subsequent differentiation, including (1)
core-mantle separation, (2) continental crust formation and (3) partial
melting at mid-ocean ridges and/or subduction-zone upper mantle. In this
traditional picture, significant chemical differentiation occurs only in
the shallow upper mantle [process (3)], in which case it is difficult
to reconcile geochemical evidence for isolated mantle reservoirs with
geophysical evidence for mantle-wide convective circulation. However,
we recently proposed [Bercovici and Karato, Nature, Sept 4, 2003] that
dehydration-induced partial melting at 410km will act as a filter for
incompatible elements which are extracted into a dense melt that is
then returned to the deep mantle; the trace-element circulation is thus
partially decoupled from the convective circulation of the bulk of the
mantle which occurs at the whole-mantle scale. In this model, the OIB
source materials are drawn directly from the undepleted and heterogeneous
deep mantle (deeper than 410km), while the MORB source materials are
processed through the putative 410km dehydration-melting "filter".
The separation of MORB and OIB source materials is thus assumed to occur
at 410km and therefore it is important to examine if our hypothesis
is consistent with observations of isotopes with very slow radioactive
decays (i.e., $^{87}Sr$/$^{86}Sr$, $^{143}Nd$/${144}Nd$). The $^{87}Sr$/$^{86}Sr$ and $^{143}Nd$/$^{144}Nd$
ratios are distinct between OIBs and MORBs ( i.e., OIBs have a wider
distribution than MORBs, and MORB isotopic ratios correspond to "depleted"
chemistry). These isotopic ratios are controlled by (i) the chemical composition (Rb/Sr and Sm/Nd, respectively) and (ii) the
long-term radioactive decay
(half-life of 48 and 106 Gyrs, respectively). A simple analysis was
made to calculate the contrast in the $^{87}Sr$/$^{86}Sr$ and $^{143}Nd$/$^{144}Nd$ ratios
between OIBs and MORBs corresponding to our model. Two factors control
these isotopic ratios in our model: (i) chemical differentiation at 410km
between Rb/Sr and Sm/Nd and (ii) the volume fraction of OIB sources with
different isotopic ratios (and the degree of mixing of these materials
in the 410km dense melt layer).
We find that the observed isotopic ratios are consistent with
our model if (i) the source region of OIBs is volumetrically dominated by
materials with isotopic ratios similar to Hawaii or FOZO source regions
(but not EM1 or HIMU), (ii) significant chemical fractionation occurs by
partial melting at 410km for Rb/Sr but not for Sm/Nd and (iii) the mixing
at 410km is extensive. Implications for other geochemical signatures
including isotopic ratios of rare gas elements and their volumetric
amounts and the trace element distribution pattern will also be discussed.
DE: 1010 Chemical evolution
DE: 1025 Composition of the mantle
DE: 1040 Isotopic composition/chemistry
DE: 1065 Trace elements (3670)
DE: 3919 Equations of state
SC: Tectonophysics [T]
MN: 2003 Fall Meeting