U33B-01 INVITED 13:40h
Sublithospheric small-scale convection and its role in global mantle dynamics and evolution
Sublithospheric small-scale convection is an important part of global mantle dynamics. It was proposed as an explanation for the bathimetry, the heat flux and the gravity anomalies in the oceanic region. It was also proposed to occur in the continental region. Sublithospheric small-scale convection could be a major convective mode on other terrestrial planets where plate tectonics is absent. I will discuss fluid dynamical constraints on small-scale convection, including the heat flux and temperature variations, the lithospheric stresses, and the length scales associated with small-scale convection. I will also discuss the possible role of small-scale convection in the origin of plate tectonics and in the secular cooling of the Earth.
U33B-02 13:55h
The Fate of Subducted Basaltic Crust in the Earth's Lower Mantle : an Experimental Petrological Study
Several models have attempted to reconcile observed geochemical data with a whole mantle convection. Such a convection pattern would be in agreement with geochemical constraints, provided the crustal component of subducted lithospheric slabs is denser than the surrounding mantle and reaches the lowermost mantle, where it could accumulate and eventually generate ascending plumes, producing the particular geochemical signatures observed at the surface [Albarede, {\it Chem. Geol.} 145, 413-429, 1998; Coltice and Ricard, {\it Earth Planet. Sci. Lett.} 174, 125-137, 1999]. In such models, the key point is that the relatively minor crustal component of the subducting lithosphere is able to reach the base of the mantle. The density differential between the oceanic crust and the mantle is thus a crucial parameter we would like to address experimentally. In this work, we describe synchrotron X-ray diffraction experiments carried out {\it in situ} at high-pressure and high-temperature (to 90 GPa and 2800 K), and coupled with the study of recovered samples by Analytical Transmission Electronic Microscopy (ATEM). Quenched samples were prepared with the FIB (Focused Ion Beam) technique for ATEM investigations, so as to obtain chemical compositions of observed phases identified by selected area electron diffraction (SAED) patterns study. Under lower mantle pressure and temperature conditions, the high aluminium budget of MORB cannot be totally accommodated in an orthorhombic perovskite structure and two Al-rich phases are observed in our experiments, in addition to Mg-perovskite, Ca-perovskite, and stishovite. The chemical compositions of individual phases were taken into account in the Rietveld refinements made to the X-ray diffraction patterns, from which phase relative abundances, molar volumes, hence densities could be extracted. We provide an estimate for the density profile of a MORB oceanic crust at lower mantle pressure and temperature conditions as well as a detailed phase relationships picture for this chemical composition. We show that the density profile of a MORB oceanic crust may intersect that of a normal lower mantle at depths greater than 2000 km.
U33B-03 14:10h
Lithospheric thickness, preservation of recycled oceanic lithosphere and the importance of water
It has been hypothesized that the viscosity structure of the mantle might be influenced by spatial variations of water content in the mantle. Thus, to what extent does the presence (or lack) of water play a role in controlling the dynamic and chemical evolution of Earth and other terrestrial planets? Fundamental to this question is how chemical differentiation processes control the distribution of water in the mantle. One place where dehydration of the mantle may be occurring is during the melting process associated with passive upwelling beneath mid-ocean ridges, which leaves a mantle residue that is depleted in meltable components and likely dehydrated. This depleted and dehydrated residuum manifests itself as a chemical and rheologic boundary layer, whose thickness should correlate with mantle potential temperature. Here, we suggest that the viscosity of this dehydrated chemical boundary layer would be high enough that vertical heat transfer through the boundary layer might occur only by conduction and not by convection. We further show that its thickness beneath oceans may limit the thickness of the oceanic thermal boundary layer, which includes the strong chemical boundary layer and the underlying convective sublayer. The presence of a depleted and dehydrated mantle residuum layer may also explain why the thickness of continents is lower than that predicted by simple conductive cooling arguments. If these conclusions are correct, scaling arguments and model results suggest that even after thermal re-equilibration with the ambient mantle, large segments of subducted oceanic lithosphere could be preserved for much longer times than predicted if oceanic lithospheric mantle was assumed to be hydrated and hence isoviscous with the ambient mantle. This implies further that there would be segments of preserved oceanic lithosphere throughout the mantle and that such lithosphere (crust + depleted residuum) might be occasionally sampled in plumes. We show using the first series transition metals (Fe, Mn, Ni, Co, Sc) that the oceanic lithospheric mantle, e.g., the complement to the oceanic crust, may indeed be a component in the source regions of many hotspot magmas. This is based on the fact that most trace-element and isotopic tracers are sensitive to crustal components, but the transition metals are sensitive to the mantle residuum due to their moderately incompatible to compatible behaviors. Compared to MORBs, most hotspot magmas have low Sc/Mn ratios and high (Fe, Co,Ni)/Mn ratios, which cannot be explained by high degree partial melting of a fertile mantle source. Instead, Sc/Mn and (Fe,Co,Ni)/Mn ratios respectively decrease and increase in peridotite residues as melt is extracted. Re-melting these depleted residues produce melts having metal systematics remarkably similar to those of hotspot magmas. One component in hotspot source regions may thus be represented by highly melt-depleted mantle, which we interpret to be recycled oceanic lithospheric mantle. These interpretations are consistent with the preservation of coherent sections of oceanic lithosphere in the mantle for long times and that dehydration of oceanic lithosphere may be a requisite for preservation. If the above interpretations and arguments are correct, it seems that water must play an important role in the dynamics and chemical differentiation of the Earth. A combination of seismology, geodynamics and geochemistry/petrology is needed to expand our understanding of how water influences mantle dynamics and chemical evolution.
U33B-04 14:25h
Evolving Mantle Regimes Revealed by $^{143}$Nd Isotopic Compositions of 3850Ma to Modern Mafic rocks
The ground truth of mantle models, particularly for the early Earth, relies on accurate determinations of original isotopic compositions. The $^{147}$Sm-$^{143}$Nd isotopic system can provide the most complete record of mantle differentiation processes, when combined with intensive field and geochronologic studies leading to judicious sample selection. Our on-going field and laboratory investigations have identified the islands near Nuuk, southwest Greenland, which include Akilia and Innersuartuut, as a key repository of information on early Earth chemistry. The oldest orthogneisses, with SHRIMP U-Pb zircon dates of 3850-3840Ma now determined at four localities cross-cut and provide minimum ages for varied mafic/ultramafic rock sequences. The initial (at 3850Ma) $^{143}$Nd isotopic compositions determined from 22 samples from the best preserved rocks in low strain areas, all have positive $\epsilon$Nd with a narrow range of initial compositions from +2 to +4 ($\pm$0.5). No negative values have been identified. This isotopic homogeneity at 3850Ma constrasts with the the large range of positive and negative $\epsilon$Hf measured in $>$4.0Ga zircons from Western Australia (Harrison et al., 2004; Goldschmidt abs.). Comparison of the new Nd data with compilations of literature data from younger crust, including 3.0 Ga and 2.7 Ga rocks indicates rapid early mantle depletion, with 3850Ma compositions of ca. +3 followed by little change in $\epsilon$Nd for greater than 1 billion years, from 3.85 to at least 2.7 Ga. This view of the mantle contrasts with the well established near linearly increasing $\epsilon$Nd values throughout much of the Proterozoic and Phanerozoic to values of +10 characterising the MORB source mantle today. The apparent large scale changing pattern of isotopic evolution must record mantle regimes characterised by different mixing times and/or variable (size, number and origin) enriched and depleted reserviors. Early Archean rocks record rapid early depletion of some portions of the upper mantle. The extended period of near constant isotopic compositions records balancing of the effects of crust extraction by either rapid crustal recycling, or mixing with deep mantle enriched domains; post-Archean Nd isotopic compositions reflect the transition to a modern style tectonic regime.
U33B-05 14:40h
Numerical modeling of noble gas recycling into the mantle
Numerical simulations of mantle convection have provided a unique insight into the use of geochemical tracers such as 3He/4He and 40Ar [1]. Almost no attention has been paid to information about convective mixing in the mantle from integrated analyses of Ne, Ar, Kr and Xe: Solar Ne, Xe and now Ar [2] have been resolved in upper mantle volatiles. Holland and Ballentine [2] show how the convecting mantle Ar and Xe isotopic composition can be accounted for by mixing between a Solar noble gas component trapped during the accretionary process, with atmosphere-derived noble gases, most likely recycled into the mantle dissolved in seawater. From mass balance, He and Ne isotopic compositions are little affected by this process. We present here numerical models that investigate the subduction of seawater-derived noble gases into the mantle as a constant proportion of recycled oceanic crust. These models constrain the amount of unmodified seawater that can be subducted and are compared with observed concentrations in oceanic crust to assess the efficiency of volatile return in arc processes. We show how a whole-mantle convective regime efficiently mixes these recycled heavy noble gases with residual Solar noble gases to provide the He, Ne, Ar, Xe abundance and isotopic composition observed in the convecting mantle. Low 40Ar/36Ar and low 129Xe/130Xe isotopic ratios in OIB relative to MORB are observed and has been used to argue that the OIB source is rich in primitive Ar and Xe. We further use the numerical simulations to test the hypothesis that because of the higher proportion of recycled material in the OIB-source, low 40Ar/36Ar and low 129Xe/130Xe isotopic ratios would be produced as a result of seawater recycling into this portion of the mantle. [1] Van Keken, P.E., C.J. Ballentine and E.H. Hauri, Convective mixing in the Earth's mantle, in The Mantle and Core, edited by R.W. Carlson, pp. 471-492, Elsevier, New York, 2003. [2] Holland. G and C.J. Ballentine, AGU abstract, This meeting.
U33B-06 14:55h
Noble Gases in the Earth's Core?
Chemical inertness, surface volatility and low abundance have made the noble gases a unique trace elemental and isotopic system for constraining the formation and evolution of the solid Earth and its atmosphere. This geochemical role parallels extensive physical-property measurements on the condensed rare gases alone at the pressures equivalent to those of the Earth's deep mantle and core from diamond-anvil cell (DAC) experiments. Traditional geochemical approaches to the processes of planetary evolution have involved crystal-melt partitioning at low pressures relevant more to near-surface degassing. The degree of compatibility has fluctuated among different studies and largely rests with the conclusion that, for common upper mantle phases, the noble gases are highly incompatible. But the long-known high $^3$He/$^4$He ratios for some ocean-island basalts and more recent observations for some of the rare gases (Ne, Ar and possibly Xe) that there is a solar component emanating from the Earth, continue to raise questions on the source reservoir as well as on accretionary and incorporation processes. Changes in models of mantle convection style have made it harder to rely on the deep mantle as a reservoir, and the core has remained a particularly unfavourable location either because of difficulty in constructing a retention mechanism during planetary accretion or simply because of lack of data: Partitioning studies at pressure are rare and complicated by the difficulty in reproducing not only absolute concentrations, but confinement of gas in high-pressure apparatus and post-run analysis. We have investigated noble gas solubility in silicate liquids at high pressures in a DAC (relevant to a magma-ocean model of the early Earth) that suggests that the detailed composition and structure of silicate liquids may act as an important control on the level of incompatibility. The long-held idea of partial melting as a single-stage, efficient process for extracting noble gases from the Earth's mantle at all depths, may well be oversimplified. For molten metal compositions interacting with silicate melt, Matsuda et al. (1993) defined the near-zero limits of noble gas solubility expected in metal with increasing pressure. We re-visit the phenomenological aspect of (saturated) noble gas solubility in metals with new experiments in noble gas pressure-transmitting media in the laser-heated DAC. First results with argon analysed with SEM methods suggest up to an order of magnitude higher partition coefficient (D(Ar)$_{Fe/sil}$ $\sim$ 0.1) for liquids in the DAC at 5 GPa. We have also recovered samples for analysis with more sensitive UV laser-ablation mass spectroscopic techniques that provide additional, depth-resolved constraints on noble gas solubility at moderate pressures.
U33B-07 15:10h
Liquid Sodium Laboratory Models of the Earth's Outer Core
Experiments in liquid sodium have become important to understanding several aspects of the Earth's outer core. I'll present a discussion of liquid metal experiments and the theory they have stimulated to better understand dynamo magnetic field generation and liquid metal turbulent convection. Three aspects will be discussed: (1) international efforts on liquid metal experiments and resulting theoretical models (2) a discussion of the mismatch between parameters in numerics, experiments and the Earth, and (3) a presentation of observations of liquid metal rotating convection and the resulting predictions regarding the Earth's core. The most recent rotating convection experiments yield an estimate for the power dissipation, relevant small scales in core processes, and a causal mechanism for the Earth's magnetic field strength. These different approaches to using laboratory experiments are opening up a new direction to understanding the dynamics of the Earth's outer core and other Planetary interiors.
<a href='http://complex.umd.edu' >http://complex.umd.edu
U33B-08 15:25h
Deep mantle heat flow and thermal evolution of the Earth's core based on thermo-chemical mantle convection
A coupled core-mantle evolution model that combines the global heat balance in the core with a fully-dynamical thermo-chemical mantle convection [Nakagawa and Tackley, 2004 published in EPSL] is used to investigate the deep mantle heat flow that is required to sustain the magnetic field generated by the geodynamo process. Effects of a radioactive heat source due to potassium in the core are also included in the global heat balance in the Earth??s core. Two important parameters are checked in this study; (1) density variation between depleted hartzbergite and basaltic material (0 to 3 percent) and (2) concentration of radioactive potassium in the core alloy (0ppm to 400ppm). The parameter set that most closely satisfies the criteria of size of the inner core (1220km at present time) is around 2 percent of density difference in a convecting mantle and 200ppm of radioactive heat source in the core. The concentration of potassium in the core is consistent with the geochemical approach [Murthy et al., 2003] but smaller than other successful thermal evolution models [Labrosse, 2003; Nimmo et al., 2004]. Heat flow through the core-mantle boundary and the contribution of radioactive heat sources in the core are consistent with theoretical estimates [e.g. Buffett, 2002] and geochemical constraints [Gessmann and Wood, 2002]. The power available to the geodynamo, based on the predicted heat flow through the core-mantle boundary, is approximately four times greater than the value predicted by numerical models of the geodynamo [Christensen and Kutzner, 2004] but closer to theoretical estimates [e.g. Buffett, 2002].