U41A-0701 0800h
Volatile Origin and Recycling Into the Mantle
Noble gas isotopes in magmatic CO2 well gases provide a unique insight into mantle volatile origin and dynamics. Previous work has resolved mantle 20Ne/22Ne ratios and suggests a Solar wind irradiated meteoritic source for mantle He and Ne. This constrains the mechanism of volatile addition to the mantle during accretion and the relationship between volatiles in MORB and OIB (Ballentine et al., 2003). We present here complimentary high precision analyses of Ar and Xe isotopes from three CO2 gas fields from the SW USA; Bravo Dome, Sheep Mountain and McElmo Dome. All fields contain 129Xe excesses from 129I decay. In addition, Sheep Mountain and Bravo Dome contain correlated non-radiogenic 124Xe, 126Xe and 128Xe excesses indicating up to a 10% non-air contribution to these gases. Whilst mass dependent fractionation can theoretically produce correlated excesses in 124Xe-128Xe, it cannot generate the correlated excess of 124Xe-128Xe and radiogenic 129Xe therefore this excess must be due to a resolvable primordial component (Caffee et al., 1999). Preliminary results for the first time show a 38Ar/36Ar ratio in Bravo Dome lower than air implying a 5% contribution from a primordial reservoir with Solar 38/36Ar. The resolved primordial component in Ne, Xe and now Ar produces 132Xe/36Ar and 22Ne/36Ar in the well gases that are similar to those of irradiated protoplanetary material. This is distinct from Solar values and provides crucial corroborative evidence for the interpretation of the mantle 20Ne/22Ne. Furthermore, preservation of the relative amounts of primordial Ar and Xe is most simply explained by recycling and homogenisation of seawater-derived noble gases into the mantle. Mass balance calculations show that seawater recycling would have a negligible impact upon the Ne isotopic ratios. These results suggest that the noble gas subduction barrier (Staudacher et al., 1988) is not effective; that the dominance of air-like non-radiogenic Ar (Kr) and Xe isotopes in the mantle is due to seawater recycling; that this process results in a relatively homogeneous air contribution to the mantle system; and is further consistent with low 129Xe excess observed in OIB being explained by a significant recycled volatile component. Ballentine, C.J. et al., AGU, V52D-06, 2003 Caffee, M.W. et al., Science, 285, 2115-2118, 1999 Staudacher, T. and C.J. Allegre, Earth Planet. Sci. Lett., 89, 173-183, 1988
U41A-0702 0800h
Whole-Mantle 3D Seismic Attenuation: Evidence for Global Processes
We present a three-dimensional tomographic model of the seismic shear wave attenuation in the mantle. The model was obtained from all available digital broadband records during 1990-2001. The data consist of over 90,000 differential attenuation measurements of ScS-S, SS-S, ScS-SS, S-S, sScS-sS, sSS-sS phase pairs. The use of differential measurements removes contaminating source effects, which is the same in both phases, as well as near-source and near-receiver structural anomalies. The result is that our data have the best resolution in the lower mantle, which has previously been unexamined with 3D modeling. The differential attenuation measurements (taken as t* values) are inverted using the LSQR routine over a long-wavelength 5 x 5 degree grid for a 3D model of shear wave quality factor (Q). These t* star measurements are also used to generate a whole-mantle 3D shear wave velocity model. While the velocity model does not have resolution equal to other models that incorporate surface wave and normal mode data, because it uses the same paths as went into the Q model, it is of interest in interpreting the attenuation anomalies. There is a very strong suggestion of whole-mantle flow in the attenuation tomography. High-Q sheet-like anomalies extend from the surface to the core-mantle boundary region (CMBR) at subduction zones, dominating the model. The African and Pacific superplumes involve low-Q anomalies that extend from the CMBR to the surface. For the Pacific superplume, the low-Q anomalies extend vertically from core to crust, but with the African superplume, the low-Q anomalies do not extend directly up into the sub-African upper mantle, but rather branch east and west up towards the Atlantic and Indian ocean spreading centers. While the interpretation of seismic Q in terms of anelasticity is challenging (as there are other factors such as seismic scattering, water content, grain size, and deviatoric strain that can influence it), if we assume that the Q anomalies are influenced to the first order by temperature variations, then this model strongly supports whole-mantle mass flux between the surface and the base of the mantle.
U41A-0703 0800h
Numerical thermo-mechanical model predicts new observable magmatic and tectonic signatures of lithospheric delamination at active continental margins
Major modification of the continental crust occur at active continental margins. An outstanding but poorly understood modification process is delamination of lower crust and mantle lithosphere, which can be expected if the overriding continental plate is subjected to compression and tectonic shortening. Delamination was previously suggested beneath the Southern Puna plateau in the Central Andes during the last 3-5 Myr, based on the occurrence of distinctive back-arc magmas, rapid recent uplift and high seismic attenuation in the upper mantle. These features are absent in the main plateau region of the Central Andes, the Altiplano and northern Puna; however, unusually low seismic velocities in the crust and the lack of a thick mantle lithosphere to match the double-thick crust beneath the Altiplano suggests that delamination may have been active in this region in the geologic past. We examine the physical conditions for lithospheric delamination and search for possible signatures of the process in the geologic record based on coupled thermo-mechanical modelling of the dynamic interaction between subducting slab and overriding continental plate. The models predict that, when delamination takes place, corner flow in the mantle wedge can move the relatively cool delaminating material toward the trench. Eventually, the delaminating material blocks the corner flow, reduces the temperatures in the mantle wedge and impedes slab roll-back. The consequence of these processes is a simultaneous reduction of arc magmatism and an increase in the rate of tectonic shortening in the overriding plate. The time scale for these effects is a few to several Myr, depending on the volume of the delaminating material. Some 5-10 Myr later, the models predict melting in the middle crust in the delamination region, with a beginning and then intensification of crustally-derived magmatic activity in the back arc. From a compilation of tectonic shortening rates and a statistical analysis of Cenozoic arc productivity for the Altiplano region at 19-22 œS, we show that the predicted sequence of delamination signatures may indeed show up in the geologic record, suggesting an episode of major delamination in the Central Andes at ca. 20-25 Ma.
U41A-0704 0800h
Is a pyrolitic adiabatic mantle compatible with seismic data?
We test the simplest average physical model of a mantle convecting as a whole (i.e., following an adiabatic temperature gradient) with a single composition with phase transitions (pyrolite) directly against global seismic data, instead of against spherically symmetric seismic models. Although the seismic models have significant uncertainties, looking at the data reveals some very strong constraints which are hard to reconcile with an adiabatic pyrolitic mantle, given the current state of knowledge of elastic and anelastic mineral parameters and their extrapolation to high pressure and temperature. This physical model generally gives (a) a stronger baseline offset between upper and lower mantle average travel-time residuals than allowed by the data and (b) an insufficient decrease in velocity gradient with depth in the deeper lower mantle (above 2500 km). 10$^5$ upper and 10$^5$ lower mantle models selected randomly within the mineral physics parameter uncertainties were tested. Only 2 lower mantle models and 24 upper mantle models yield whole mantle seismic structures that are compatible with global ISC P and S travel times and central frequencies of toroidal and spheroidal fundamental modes with angular order higher than 18. To improve the fit to the seismic data, the physical model would require (a) a lower velocity transition zone composition than dry pyrolite (at least around continents and subduction zones) as well as (b) a gradual change in physical state of the lower mantle that decreases the velocity-depth gradient, e.g., a superadiabatic temperature increase.
U41A-0705 0800h
A Lower Mantle Origin for Megacryst Suite Pyroxene-Ilmenite Xenoliths in Kimberlites: High-Pressure Experimental Constraints and Geodynamic Significance
Megacryst suite xenoliths (MSX's) in kimberlites, alnoites and alkali basalts are an important and poorly understood association. MSX's comprise medium- to coarse-grained monomineralic, or rare, multi-grain aggregates of low Cr, high Ti-Na pyrope, Mg ilmenite, sub-calcic pyroxene, Fo85$\pm$3 olivine, orthopyroxene and zircon. Some MSX's exceed 30 cm in diameter. They are interpreted to form by fractional crystallization from their host magma, near the base of the lithosphere [1-2]. However, majorite, and other high-pressure phases in some garnetite MSX's, indicates a mantle transition zone (TZ) origin [3]. A sub-lithospheric, deeply subducted slab source is also supported by Hf isotopic data [4]. A common member of the MSX suite, are graphic intergrowths of pyroxene and Mg-ilmenite interpreted to reflect cotectic, or non-equilibrium crystallization [5-6] from the kimberlite magma. However, Pb isotopic data for Monastery [7], and Namibian [8] megacrysts shows that MSX's and their host magmas are unrelated. Thus the mineralogy of the Ti-rich px-ilm MSX's needs to be determined at TZ and higher P. We have conducted multi-anvil (MA) and diamond anvil (DA) experiments on natural px-ilm xenoliths from Monastery and Malaita with different TiO$_{2}$ contents (17% and 12%), in an attempt to synthesize the pre-exsolution phase. MA experiments were carried out on both starting compositions at 18 and 21 GPa, at 1800$\deg$C and 2100$\deg$C. None of the experiments yielded a single phase. Phases identified (EPMA, Raman & XRD) include: majorite, Si-rich ilmenite and Ca-Si-Ti Pv. At 21 GPa and 2100$\deg$C wadsleyite formed part of the assemblage, and melt was locally developed. Majorite is the most abundant phase in all experiments. Maximum majorite TiO$_{2}$ occurs at 18 GPa (i.e., 5.4% - Malaita and 6.2% - Monastery). In the lower Ti Malaita composition, at 25 GPa and 1800$\deg$C, the assemblage is dominated by almost equal amounts of majorite (TiO$_{2}$ 1.3% to 2.1%) and Ca-Si-Ti Pv, with a small amount of titaniferous Na$_{2}$O-bearing magnesiowustite. Conclusions from MA experiments are: (1) a single homogeneous Ti-rich silicate phase is not stable in the TZ; (2) closest approach to a homogeneous phase occurs in the lower TiO$_{2}$ Malaita composition at 25 GPa; (3) maximum solubility of TiO$_{2}$ in majorite is 5-6% at 18 GPa; (4) at 2100$\deg$C and 21 GPa, melting occurs, and majorite plus Si-ilmenite reacts to form Ti-Ca-Si Pv and wadsleyite. Furthermore, in a DA experiment at 30 GPa and 1800$\deg$C a multi-phase, Ti-Ca-Si Pv dominated assemblage still prevails. Pyroxene-Mg ilmenite MSX's are therefore interpreted to form from Ti-rich protoliths, possibly ilmenite-rich cumulates, in subducted slabs of oceanic lithosphere. The homogeneous phase from which the pyroxene-Mg-ilmenite intergrowths were exsolved, is thus interpreted to be Ti-Ca-Si Pv, not Ti-rich majorite [cf., 9-10]. Results confirm large-scale transport of oceanic lithosphere into the upper lower mantle during subduction. [1] Nixon, et al., (1963) {\it Am. Mineral.} 48, 1090. [2] Jones et al., (1987) In: {\it Mantle Xenoliths} 711. [3] Collerson et al., (2000) {\it Science}, 288, 1215. [4] Nowell et al., (2004) {\it J Petrol.} 45, 1583. [5] Wyatt (1977) {\it CMP} 61, 1-9. [6] Mitchell (2004) {\it Lithos} (in press) [7] Collerson et al., (2001) {\it EOS} 82. [8] Davies et al., (2001) {\it J Petrol.} 42, 159. [9] Ringwood & Lovering (1970) {\it EPSL,} 7, 371. [10] Zhang et al., (2003) {\it EPSL,} 216, 591.
U41A-0706 0800h
Testing Models of Thermo-chemical Convection Against Models from Probabilistic Tomography
The main difficulty of a thermo-chemical interpretation of seismic tomography is the existence of a strong trade-off between temperature and composition. This trade-off can never be fully resolved, but we can identify all possible models compatible with data. In previous work, we have developed the technique of probabilistic tomography, which gives independent probability density functions for long wavelength models (spherical harmonic degree 2, 4 and 6) of bulk-sound and shear wave speed, density and boundary topography in the mantle. Using appropriate sensitivities (which take into account our ignorance on the thermodynamic reference state and the published range of mineral physics data), we have converted the results from probabilistic tomography into likelihoods of variations in temperature, perovskite and iron content throughout the lower mantle. Several robust features emerge which shed a new light on the nature of the lower mantle. Throughout the mantle temperature variations are much weaker than classically inferred from shear wave speed alone. Compositional variations are essential to explain the seismic data. In most places, the inferences are robust, i.e. the amplitudes of chemical and temperature variations are much larger than the uncertainties inferred from the width of the likelihoods. Below 2000 km, the correlation between relative shear wave speed variations and temperature is quite low, and we find that the much debated superplumes beneath the Pacific and Africa are due to an enrichment in perovskite and iron, rather than to high temperatures. These features are therefore denser than the surrounding mantle, and hence not buoyant. Our results clearly show that chemical variations are a key ingredient to model mantle dynamics, but they cannot yet discriminate between different possible models of thermo-chemical convection. We report first statistical comparisons between some chosen models of thermo-chemical convection, and likelihoods of seismic parameters observed by probabilistic tomography.
U41A-0707 0800h
Convective Patterns in the Indo-Atlantic Mantle over the last 260 Myr
Using recent fluid mechanics results as a framework, we reinterpret the images of the Indo-Atlantic mantle obtained from global tomography studies together with geochemical, geological and paleomagnetic observations to unravel the pattern of convection in the Indo-Atlantic "box" and its temporal evolution over the last 260 Myr. Seismic tomography sections at different depths show that the Earth's mantle seems to be divided in two "boxes" by the subducted plates, the Pacific and the Indo-Atlantic boxes. The latter presently contains a) a broad slow seismic anomaly at the CMB, which divides into several branches higher in the lower mantle, b) one "superswell" centered on the western edge of South Africa, c) at least 5 "primary hotspots" with long tracks related to trapps, and d) numerous smaller hotspots. Moreover, in the last 260 Myr, this mantle box has undergone 10 traps events, 7 of them related to continental break up. Several of these past events are spatially correlated with present-day seismic anomalies and/or upwellings, suggesting episodicity. Laboratory experiments show that superswells, long-lived hot spot tracks and traps probably represent three stages of a same phenomenon: the episodic destabilization of a hot thermal boundary layer lying close to the bottom of the mantle and containing chemical heterogeneities. This could produce thermochemical plumes, either with large heads (trapps) and thin tails (tracks) if less viscous than the ambient mantle, or fingerlike domes if more viscous than the ambient mantle. The recurrence of such phenomenon in the mantle would be 100-200 Myr with a wavelength of 2000-4000 km at the CMB, which is in agreement with the observations.
U41A-0708 0800h
Effects of Aluminium on the compressibility of silicate perovskite
Among the elements present in all mantle compositional models, aluminium oxide Al$_{2}$O$_{3}$ is estimated to amount 4 to 5 mole %. While aluminium is incorporated in specific minerals such as garnet and majorite in the upper mantle, it is believed to be incorporated into (Mg,Fe)SiO$_{3}$-perovskite under the pressure and temperature conditions of the lower mantle. Recently, the effect of Al on the elastic parameters of perovskite has received extreme attention, since Zhang and Weidner [1] presented for an Al-bearing silicate perovskite a bulk modulus 10% smaller than that of the end-member MgSiO$_{3}$-perovskite. However, the bulk moduli obtained afterwards at various alumina content were highly scattered. We present here a series of volume measurements up to 40 GPa, for Mg-perovskites with respectively 5, 7.7 and 20 mol % aluminium. Samples were either synthesized in a multi-anvil apparatus or in a diamond anvil cell by laser heating a 20 $\m$um thin glass slide of the relevant composition. Samples were studied in a diamond anvil cell with neon as a hydrostatic pressure transmitting medium and were annealed with a YAG laser within their pressure stability filed. Angle dispersive diffraction patterns were collected upon compression to the peak pressure and during decompression, at the ID9 dedicated high-pressure beamline of the ESRF. From the various data sets fitted to a Birch-Murnaghan equation of state, we deduce that the the relationship between the bulk modulus of perovskite and the Al content is not a linear decrease. Whereas 5 mol % aluminium has no noticeable effect on the compressibility of silicate perovskite, increasing slightly the aluminium content of perovskite strongly decreases its bulk modulus. Moreover, as the aluminium is added into perovskite, the compression/decompression behaviour of the samples displays different regimes. These results can be related to the substitution mechanism of Al into perovskite, suggesting that the coupled substitution mechanism is probably efficient at low aluminium content, but might then replaced by the oxygen vacancy one at higher aluminium content. This indicates that the investigations of the chemical and petrological compositions of the uppermost lower mantle and of heterogeneities should definitely take into account the effect of Al on the thermoelastic properties of perovskite. This also suggests that the partitioning behaviour of some trace elements between perovskite and the other lower mantle phases might be different in deep subducted lithospheric plates than in the `mean' lower mantle. [1] Zhang, J., and D.J. Weidner,{\it Science},{\it 284}, 782, 1999.
U41A-0709 0800h
The role of convective mixing in degassing the Earth's mantle
Because the chemical evolution of the mantle depends on the way rocks are processed at the surface of the Earth and on the efficiency of convective mixing, the goal of the present paper is to quantify how mixing and processing depend on convection parameters (Rayleigh number, heating mode and viscosity stratification). An extensive set of simple 2D convection simulations with passive tracers is built to compute bulk (a) Lagrangian strain rates to evaluate mixing efficiency and (b) processing histories. At high Rayleigh number, the strain rate is only a function of the velocity of the flow, whatever the heating mode or viscosity stratification. Though this is not the case in low Rayleigh number experiments with basal heating where a transition regime is observed with inefficient mixing though chaotic. The simulations show that the processing efficiency depends a lot on the heating mode : the probability of sampling primordial rocks is larger with internal heating than with basal heating. Scaling laws are proposed to parameterize mantle processing histories and it is predicted that in a 300~K hotter Earth, free convection can account for the mantle early degassing (more than 90% in less than 100~Myrs).
U41A-0710 0800h
The Dynamics of Layer Formation in the Earth's Mantle
The internal structure of the Earh is made up by a series of layers, though it is unclear how many layers exist and if there are layers invisible to remote sensing techniques. Layering can not be explained by simple gravitational settling. Double-diffusive convection (d.d.c) is considered as a vital mechanism behind the generation of layered structures. We demonstrate that d.d.c can lead to layer formation on a planetary scale in the diffusive regime where composition stabilizes the system while heat provides the destabilizing force. Choosing initial conditions in which a stable compositional gradient overlies a hot reservoir we mimic the situation of a planet in a phase after core formation. Differently from earlier studies we fixed the temperature rather than the heat flux at the lower boundary, resembling a more realistic condition for the core-mantle boundary. We have carried out extended series of numerical experiments, ranging from 2D calculations in constant viscosity fluids to fully 3D experiments in spherical geometry with strongly temperature dependent viscosity.The buoyancy ratio R and the Lewis number Le are the important dynamical parameters. In all scenarios we could identify a parameter regime where the non-layered initial structure developed into a state consisting of several, mostly two layers.Initially plumes from the bottom boundary homogenize a first layer which subsequently thickens. The bottom layer heats up and then is initiated in the top layer. This creates dynamically (i.e. without jump in the material behavior) a stack of separately convecting layers.The bottom layer is significantly thicker than the top layer. Strongly temperature dependent viscosity,. leads to a more complex evolution The formation of the bottom layer is followed by the generation of several layers on top. However this layers generally collapse into one layer, again resulting in a two layer system. We employed a numerical technique, allowing for a diffusion free treatment of the compositional field. In each case a similar evolution has been observed. This indicates that a temporary formation of layered structures in planetary interiors is a typical phenomenon. The presence of phase boundaries may further help to stabilize the boundaries between the layers against overturning
U41A-0711 0800h
The Role of Plate Boundaries in Degassing: Dynamic Mantle Models
In recent years, the geodynamical community has put forth a number of models which examine the chemical evolution of the mantle with respect to noble gases. These models are able to satisfy general constraints such as heat flow and overall degassing rate, but have difficulty in reproducing the noble gas isotopic heterogeneity observed in oceanic basalts. Such models generally employ a free slip surface boundary condition and thus lack the formation of sharp and long-lived zones of convergence and divergence that are associated with subduction zones and mid-oceanic ridges. This makes it difficult in the models to distinguish mid-oceanic ridge volcanism from ocean island volcanism. Model approximations of plate tectonics include the prescription of a kinematic surface boundary condition, or by using an advanced rheological description with imposed weak zones or a finite yield strength criterion. Here, we use another approach, which is based on the force balance method by Gable (JGR, 1991). This is essentially an advanced kinematic boundary condition in which segments of constant boundary velocity are prescribed in a way that is dynamically consistent with the overall convective flow. By doing this, we explicitly link degassing to zones of plate divergence in our model, thereby consistently following the degassing associated with mid ocean ridge volcanism. Our model simulates mantle convection by the numerical solution of the time dependent Boussinesq equations on a two dimensional cylindrical finite element mesh, with mantle noble gas inventories discritized to a large number of passive tracers. We present the results from a suite of model runs with temperature- and pressure-dependent viscosity, and with a variable number of plates.
U41A-0712 0800h
Characterizing Deep ($>$ 500 km) Earthquake Regions to Investigate the Fate of Subducting Slabs
Subducting oceanic lithosphere is the main driving force of plate tectonics and the main source of upper mantle chemical heterogeneity. Here we present a comprehensive characterization of all regions of deep seismicity ($>$ 500 km): Japan, Izu-Bonin, Marianas, Philippines, Java, Solomon, New Hebrides, Tonga, and South America. Regional tomographic studies in subduction zones find fast velocity anomalies associated with the Wadati-Benioff zone of earthquakes, indicating continuity of some slabs to the bottom of the transition zone. Global tomographic studies find evidence of fast velocity anomalies in the mid-mantle, but often these anomalies are not obviously connected to descending slabs. The data we have assembled include the regional patterns of seismicity, CMT focal mechanisms, and discontinuity topography. Tomographic and dynamic models are also available for some regions. Most dynamic models show that, in the absence of complicated plate geometries and reorganizations, slabs should remain continuous into the lower mantle retaining a seismically detectable shear velocity anomaly of at least 1%. However, this expected anomaly is not observed below 700 km in many of these regions of deep seismicity. Rather, in most of the Western Pacific we find large-scale fast anomalies above 700 km along with focal mechanisms that indicate slab ponding rather than slab penetration. Most high velocity anomalies that do exist below the transition zone are no longer planar features, also indicating past slab ponding. In South America and New Hebrides there is deep seismicity seemingly isolated from the shallower slab seismicity. The New Hebrides seismicity is difficult to explain as it occurs outside of the current slab geometry with no consistent pattern in the focal mechanisms. Whereas, the South American seismicity is likely happening within the slab and exhibits down-dip compression. In Java and the Marianas, the steep angle of the slabs and the nearly vertical compression axes of the focal mechanisms indicate current slab penetration. Further dynamic modeling is needed to understand the stability of slabs with regard to their thermal and chemical structure as they pass through the transition zone. The unique structure of seismicity and mantle heterogeneity in each region suggests that the subduction process is highly time dependent and difficult to fully interpret over short time scales.
U41A-0713 0800h
Migration of SS precursor Data to Image Fine-scale Structure on the Upper Mantle Discontinuities Beneath Hawaii
Using a wavefield migration technique, we use the precursors to the seismic phase SS to study the reflectance and topography of the 410 and 660- km discontinuities in the region of the Hawaiian hotspot. Our dataset consists of broadband waveforms from 9 shallow focus events (0-75 km depth) in the southwestern Pacific recorded by seismic arrays in the Canadian National Seismic Network and by POLARIS, with central bouncepoints of the SS phase located within and outside of the proposed location of Hawaiian hotspot mantle plume. Migration allows us to constrain the topography, depth, and sharpness of these discontinuities in a region predicted to have a high degree of heterogeneity, and is not restricted to the assumption of horizontal reflectors used in slowness stacking approaches. Measurement of these parameters provides useful information for understanding the dynamics, thermal structure, and composition of a plume rising through the Transition Zone. Seismograms are instrument deconvolved, low-pass filtered at $\sim$ 6 seconds, and aligned on the SS phase to study the coherency of the precursors at depth. The wavefield is then migrated to each node in a 3-D grid of potential reflectors over a region spanning 170 to -140 deg E and 0-30 deg N, in increments of 1 deg laterally, and 10 km vertically. Initial migration results indicate a strong reflector at a depth of 630 km near the projected location of the hotspot, and at a depth of 660 (i.e. close to the global average) away from the hotspot. A 30 km elevation of the 660 km phase boundary is consistent with an excess plume temperature of $\sim$ 350 K for the negative Clapeyron slope of the ringwoodite to perovskite phase transition. The migrations show strong lateral amplitude anomalies in SS precursor Fresnel zones, indicating significant off great circle path contributions to observed waveforms. The migrated results are compared with slowness stacking of a much larger SS dataset consisting of 4500 broadband and filtered records, with extended coverage over most of the central Pacific. Stacks for these regions show a correlation with the migration results, though the effects of Fresnel zone sampling and azimuthal dependence bias the slowness stacks. Additional upper and lower mantle reflectors are imaged using our migration technique and will be discussed.
U41A-0714 0800h
Seismological Constraints on Geodynamics
Earth is an open thermodynamic system radiating heat energy into space. A transition from geostatic earth models such as PREM to geodynamical models is needed. We discuss possible thermodynamic constraints on the variables that govern the distribution of forces and flows in the deep Earth. In this paper we assume that the temperature distribution is time-invariant, so that all flows vanish at steady state except for the heat flow $J_{q}$ per unit area (Kuiken, 1994). Superscript 0 will refer to the steady state while x denotes the excited state of the system. We may write $\sigma$$^{0}$=(J${_q}^{0}$$\cdot$X${_q}^{0}$)/T where $X_{q}$ is the conjugate force corresponding to $J_{q}$, and $\sigma$ is the rate of entropy production per unit volume. Consider now what happens after the occurrence of an earthquake at time t=0 and location (0,0,0). The earthquake introduces a stress drop $\Delta$P(x,y,z) at all points of the system. Response flows are directed along the gradients toward the epicentral area, and the entropy production will increase with time as (Prigogine, 1947) $\sigma$$^{x}$(t)=$\sigma$$^{0}$+$\alpha$${_1}$/(t+$\beta$)+$\alpha$${_2}$/(t+$\beta$)$^{2}+etc A seismological constraint on the parameters may be obtained from Omori's empirical relation N(t)=p/(t+q)$\space$ where N(t) is the number of aftershocks at time t following the main shock. It may be assumed that p/q$\sim\alpha_{1}/\beta$ times a constant. Another useful constraint is the Mexican-hat geometry of the seismic transient as obtained e.g. from InSAR radar interferometry. For strike-slip events such as Landers the distribution of $\Delta$P is quadrantal, and an oval-shaped seismicity gap develops about the epicenter. A weak outer triggering maximum is found at a distance of about 17 fault lengths. Such patterns may be extracted from earthquake catalogs by statistical analysis (Lomnitz, 1996). Finally, the energy of the perturbation must be at least equal to the recovery energy. The total energy expended in an aftershock sequence can be found approximately by integrating the local contribution over volume V: $\int$${_V}$$\sum$${_k}$$\nu$${_k}^{2}$(grad$\Delta$P)$^{2}$dV$\space$ where the $L_{ik}$ are Onsager coefficients and $\nu_k}$ is the stoichiometric coefficient of phase k. This enables us to calibrate the process. Note that there are contributions to the excited state that include material flows $J_{k}^{x}$ (flows of water and electrons into the epicentral area), in addition to the excited heat flow values $J_{q}^{x}$. A geodynamic model of the earthquake process would be essential for earthquake prediction. Kuiken, G.D.C., Thermodynamics of Irreversible Processes (John Wiley & Sons, New York, 1994). Lomnitz, C., On thermodynamics of planets, Geophys. J. Roy. Astr. Soc., 5, 157-161, 1961. Lomnitz, C., Search of a worldwide catalog for earthquakes triggered at intermediate distances, Bull. Seismol. Soc. Am., 86, 293-298, 1996.
U41A-0715 0800h
Scales of mantle heterogeneity
A long-standing question in mantle dynamics concerns the scale of heterogeneity in the mantle. Mantle convection tends to both destroy (through stirring) and create (through melt extraction and subduction) heterogeneity in bulk and trace element composition. Over time, these competing processes create variations in geochemical composition along mid-oceanic ridges and among oceanic islands, spanning a range of scales from extremely long wavelength (for example, the DUPAL anomaly) to very small scale (for example, variations amongst melt inclusions). While geochemical data and seismic observations can be used to constrain the length scales of mantle heterogeneity, dynamical mixing calculations can illustrate the processes and timescales involved in stirring and mixing. At the Summer 2004 CIDER workshop on Relating Geochemical and Seismological Heterogeneity in the Earth's Mantle, an interdisciplinary group evaluated scales of heterogeneity in the Earth's mantle using a combined analysis of geochemical data, seismological data and results of numerical models of mixing. We mined the PetDB database for isotopic data from glass and whole rock analyses for the Mid-Atlantic Ridge (MAR) and the East Pacific Rise (EPR), projecting them along the ridge length. We examined Sr isotope variability along the East Pacific rise by looking at the difference in Sr ratio between adjacent samples as a function of distance between the samples. The East Pacific Rise exhibits an overall bowl shape of normal MORB characteristics, with higher values in the higher latitudes (there is, however, an unfortunate gap in sampling, roughly 2000 km long). These background characteristics are punctuated with spikes in values at various locations, some, but not all of which are associated with off-axis volcanism. A Lomb-Scargle periodogram for unevenly spaced data was utilized to construct a power spectrum of the scale lengths of heterogeneity along both ridges. Using the same isotopic systems (Sr, Nd and Pb), we correlated isotopic data from oceanic islands and seamounts with geophysical measurements and observations in order to determine whether the mantle source component end-members are related to geophysical structure. For this, data obtained from the GeoROC database and other sources were first carefully examined for consistency and then correlated with recent global upper mantle tomographic models of seismic attenuation and velocity structure. Complementing this approach is a fractal analysis of dynamical models of mantle mixing. A variety of published numerical mixing models show a fractal distribution of heterogeneities, consistent with a marble-cake mantle.
U41A-0716 0800h
Water, Mid-ocean Ridges, and Dynamic Geochemical Layering of the Earth's Mantle
A model has been developed that could help reconcile apparently conflicting observations about the extent of geochemical layering of the Earth's mantle. Seismological observations and numerical simulations both suggest that there is significant material exchange across the upper mantle (UM; defined as above the 660 km seismic discontinuity) and the lower mantle (LM). Geochemical observations indicate that the mantle that melts at mid-ocean ridges (MOR) is different from that melted at intraplate hot spots (ocean island basalts, OIB), suggesting that the LM is chemically distinct from the UM. An important additional observation is that typical MOR mantle is "depleted," but only in incompatible trace elements, indicating that it has been affected by removal of only small degrees of partial melt (ca. 1%). We propose that a key component of the mantle differentiation cycle is the effect of water on melting under MORs, and possibly also in island arcs. The presence of ca. 200-500 ppm of H$_{2}$O causes melting to begin at greater depth (ca. 100-150 km). The resulting oceanic lithologic column is composed of not only the basaltic crust (4-9km) and depleted harzburgite/lherzolite (DHL; ca. 40-70 km thick) sections, but also a third major section of incipiently melted lherzolite (IML; 30-70 km thick). If during subduction the IML is preferentially retained in the UM since it remains at the ambient mantle temperature, while the BC-DHL package is preferentially carried to the LM (as a consequence of its negative buoyancy due to lower temperature and phase assemblage, as well as high viscosity), the result will be a dynamical geochemical layering in terms of trace elements, isotopes, and heterogeneity. Over 2 billion years a quasi-steady-state UM will develop that is highly "depleted" and relatively homogeneous, and the complementary LM will be slightly depleted to enriched, and more heterogeneous. Application of a simple box model to the Sm-Nd system gives a steady-state UM-LM difference in $\epsilon$_{Nd}$ of 4 units, successfully reproducing the observed difference between mean MORB and mean OIB. This value occurs for mass fluxes between UM and LM corresponding to about 50% of the current rate of generation of oceanic lithosphere. Therefore, this model accounts for observed $\epsilon$_{Nd}$ values if slab penetration into the LM occurs but is ca. 50% efficient. The LM would also tend to have more variable $\epsilon$$_{Nd}$, by about a factor of 2. The results are not sensitive to the relative size of UM and LM, but the overall Nd mass balance for the Earth requires the UM be small.
U41A-0717 0800h
3-D seismic imaging of the D" region beneath the Cocos Plate
The 3-D velocity structure of the deep mantle has been recently inferred from imaging procedures such as migration, tomography, stacking, and waveform modeling, all which utilize localized 1-D reference structures. As these methods often have limiting assumptions it is essential to assess to what extent 3-D solution models are self-consistent with the imaging procedures from which they were produced; this is possible through synthesizing waveforms in laterally varying media. We use a 3-D axi-symmetric finite difference algorithm to model SH-wave propagation through cross-sections of recent 2- and 3-D lower mantle models along a north-south corridor roughly 700 km in length beneath the Cocos Plate. Synthetic seismograms with dominant periods up to 3 sec are computed to assess fit of 3-D model predictions to data. Models considered include: (1) a D" reflector 264 km above the CMB with varying S-wave velocity increases of 0.9% to 2.6% by {\it Lay et al., }[2004] obtained using a double-array stacking method; (2) an undulating D" reflector with large topography obtained using a migration method by {\it Thomas et al., }[2004]; (3) cross-sections through the mantle shear velocity tomography model TXBW of {\it Grand }[2002]; and (4) cross-sections through a regional lower mantle shear wave tomography model based on finite frequency kernels from {\it Hung et al., }[2004]. Model predictions show strong waveform variability. We apply double-array stacking to synthetics to assess recovery of starting model D" discontinuity depths and also demonstrate the tradeoff between D" topography and velocity heterogeneity. We compare synthetics of these models to each other as well as to high quality broadband observations. The dependencies of waveform features on the various scale lengths of structures in the different models tested are also explored.
U41A-0718 0800h
Origin of Tungsten Excess in Komatiites
The limited database available for W abundances in komatiites (n=7, Newsom et al., 1996) suggests that when melting and fractional crystallization effects are filtered out, these komatiites have about 10 times higher W, compared to other mantle-derived mafic-ultramafic magmas (MORB, OIB). The excess of W in the komatiites relative to lithophile highly incompatible elements becomes obvious when compared with the low concentrations of the light REE Ce and Nd (about 1-2 ug/g in many komatiites, compared to $>$ 10 ug/g in most MORB and OIB). In order to increase the komatiite W database, komatiite samples from Phanerozoic (Gorgona Island) and Archean terraines (Boston Creek/Canada, Belingwe/South Africa, 2.7 Ga) were dissolved and W was separated in order to obtain W concentrations by isotope dilution. Except for one sample from Gorgona Island with low W (23 ng/g), samples from all three locales show high W (516 to 2643 ng/g), with most samples containing near 700 ng/g W. Three Hawaiian picrites (H23, LO-02-04, MK-1-6) were also analyzed for comparative purposes and contain 75, 163 and 418 ng/g W, respectively. The W concentrations in the Hawaiian picrites are comparable or lower than W concentrations in Hawaiian tholeiites (Newsom et al., 1996). Mass balance considerations suggest that it is unlikely that the W excess in komatiites reflects W contributions to the mantle sources of komatiites from the outer core. The W enrichment could result from shallow-level alteration processes if primary W abundances of komatiites were low and W was added via fluids, containing W and other fluid-mobile elements derived from crustal rocks. Because most W in such samples would be of crustal origin, small contributions from the outer core may be difficult to detect using 182W systematics (Schersten et al., 2003).
U41A-0719 0800h
Finite-Frequency Tomography of $D^{\prime\prime}$ Shear Velocity Heterogeneity beneath the Caribbean
The shear velocity structure in the lowermost 500 km of the mantle beneath the Caribbean and surrounding areas is determined by seismic tomography applied to a suite of Sdiff-SKS, ScS-S, (Scd+Sbc)-S, and ScS-(Scd+Sbc) differential times, where (Scd+Sbc) is a pair of overlapping triplication arrivals produced by shear wave interaction with an abrupt velocity increase at the top of the $D^{\prime\prime}$ region. The inclusion of the triplication arrivals in the inversion, a first for a deep mantle tomographic model, is possible because of the widespread presence of a $D^{\prime\prime}$ velocity discontinuity in the region. The additional raypath sampling provided by the triplication arrivals yields improved vertical resolution of velocity heterogeneity within and above the $D^{\prime\prime}$ region. The reference velocity model, taken from a prior study of waveforms in the region, has a 2.9% shear velocity discontinuity 250 km above the CMB. Effects of aspherical structure in the mantle at shallower depths than the inversion volume are suppressed by applying corrections for several different long-wavelength shear velocity tomography models. Born-Fr\'{e}chet kernels are used to characterize how the finite-frequency data sample the structure for all of the differential arrival time combinations; inversions are performed with and without the kernels. The use of 3-D kernels stabilizes the tomographic inversion relative to a ray theory parameterization, and a final model with 60 and 50 km correlation lengths in the the lateral and radial dimensions, respectively, is retrieved. The resolution of the model is higher than that of prior inversions, with 3 to 4% velocity fluctuations being resolved within what is commonly described as a circum-Pacific ring of high velocities. A broad zone of relatively high shear velocity material extends throughout the lower mantle volume beneath the Gulf of Mexico, with several percent lower shear velocities being found beneath northern South America. Concentrated low velocity regions extend through the $D^{\prime\prime}$ layer under the Caribbean, Colombia, and Ecuador, suggestive of small-scale plumes in the boundary layer. One scenario consistent with the imaged features involves subducted Farallon plate ponding at the base of the mantle and laterally displacing hot boundary layer material that piles up and destabilizes on its margins.
U41A-0720 0800h
Ridge-like upwelling in the uppermost lower mantle beneath eastern Africa from finite-frequency seismic tomography
Global seismic tomographic imaging has shown a large and coherent low-velocity anomaly in the lower part of the mantle beneath Africa. It has been suggested that this anomalous feature may be responsible for the unique geological history of Africa. But the link between the African superplume in the deep mantle and surface tectonics remains unclear. In this study we carry out tomographic inversions for the seismic velocity structure beneath southern Africa, utilizing "banana-doughnut" traveltime sensitivity kernels of body waves recorded by temporary and permanent broadband seismic stations in the region. Preliminary results show a ridge-like low-velocity anomaly in the upper part of the lower mantle (900 - 1200 km depth) beneath eastern Africa. It is 300-400 km wide beneath the Kaapvaal craton and becomes broader beneath Tanzania. At shallower depth and in the mantle transition zone, the low-velocity anomaly becomes localized and cylindrical beneath Tanzania. This is consistent with an anomalously thin mantle transition in the area. A weak, localized low-velocity anomaly is also found in the transition zone beneath the western end of the Limpopo Mobile Belt (between the Zimbabwe and Kaapvaal cratons). Our observations may have important implications for the deep earth engine. Models of mantle convection driven by a combination of basal and internal heating predict a transition in the shape of a buoyant upwelling from tabular to cylindrical near the basal boundary layer. The ridge-like low-velocity anomaly in the upper part of the lower mantle beneath southern Africa may thus represent upwelling from a boundary layer in the mid-mantle, possibly above the top of the African superplume. The fact that the ridge-like low-velocity anomaly is present beneath southern African cratons suggests that volcanism in eastern Africa is not the result of passive upwelling driven by plate separation but mainly the consequence of deep, buoyant mantle convection.
U41A-0721 0800h
Temporal Variations in the Convective Style of Planetary Mantles
Investigations of mantle convection with temperature- and stress-dependent viscosity have revealed the existence of fundamentally different convective styles: By varying e.g. the Rayleigh number, the viscosity contrast or the stress dependency of viscosity, the planform of convection in the asymptotic stationary state changes from the so-called stagnant lid regime to an episodic behaviour and further to a state characterized by a mobile surface. Our studies suggest that this transition may not only be induced by a change of parameters but also occurs temporally, i.e. a system that initially shows an episodic behaviour falls into the stagnant lid mode of convection as time proceeds. Such a temporal variation can probably describe the evolution of terrestrial planets like Mars, which is assumed to have undergone a change of convective style in its early history. In this contribution we present a 3D fluiddynamical model that exhibits a change of the convective style for temporally constant parameters. We also describe the methods developed to characterize different regimes and to identify the transitions observed.
U41A-0722 0800h
An Early Formed D'' Reservoir Reconciles Geochemical Mass Balance With Whole Mantle Convection Models
One of the most intriguing present-day problems in Earth sciences is reconciling whole mantle convection models (that follow from seismic tomography and dynamic modeling) and the chemical and isotopic mass balance of continents and depleted mantle, which favor partial-mantle convection. Specifically, geochemical observations point to an apparently isolated, early-formed reservoir deep in the Earth. The most important of these observations are: (1) The occurrence of solar noble gases in the mantle, which is in contrast with the extreme degassing of this reservoir indicated by mantle xenology; (2) specific isotopic compositions of mantle He, Ne and Xe point to a reservoir with low U/3He and 136Xe(Pu)/129Xe(I) ratios, implying both early formation and low degassing of this reservoir. We suggest that the core-mantle transition zone (termed D'') is the reservoir indicated by these observations. The material of D'' could comprise an early gabbroic-basaltic crust loaded with chondrite-like, late-accreting matter including a solar-wind irradiated regolith. If subducted, this material should accumulate above the metal core due to an intrinsic density contrast. Provided that it was not hydrated at the surface, so that subduction did not entail volatile loss, it could have retained its geochemical characteristics. We examined the consequences of this scenario by transport models envisaging: (1) Earth accretion accompanied by mantle melting and fractionation, core segregation, formation and recycling of mafic crust, degassing, and gas loss from the atmosphere, followed by (2) crust-mantle evolution involving continent growth and recycling. Comparison of calculated and observed parameters allows a solution of the model. The D'' is formed within a time interval from 40 to 80 Ma after formation of the solar system and comprises about 20% of the BSE inventory of incompatible (including heat-producing) elements. Because the bulk of the D'' material (basalt) is fractionated, its apparent isolation allows a mass balance for 147Sm-144Nd and U-Th-Pb systematics to be achieved with whole-mantle convection (apart from D''. D'' is an important 40Ar*-, 129Xe*- and 3He- bearing reservoir in the Earth. After accretion a small amount of D'' material was entrained by convective flow thus contributing noble gases to the mantle: Rare gas modeling yields a low flux from D'' into the overlying mantle, about 20% of the D'' mass per 4.5 Ga, which is about 100 times lower than the rate of ridge magmatism. These and other results of the modeling characterize D'' as a geochemically important reservoir. formed within a time interval from 40 to 80 Ma after formation of the solar system and comprises about 20% of the BSE inventory of incompatible (including heat-producing) elements. Because the bulk of the D'' material (basalt) is fractionated, its apparent isolation allows a mass balance for 147Sm-144Nd and U-Th-Pb systematics to be achieved with whole-mantle convection (apart from D''). D'' is an important 40Ar*-, 129Xe*- and 3He- bearing reservoir in the Earth. After accretion, a small amount of D'' material was entrained by convective flow thus contributing noble gases to the mantle: Rare gas modeling yields a low flux from D'' into the overlying mantle, about 20% of the D'' mass per 4.5 Ga, which is about 100 times lower than the rate of ridge magmatism. These and other results of the modeling characterize D'' as a geochemically important reservoir.
U41A-0723 0800h
New Search for 142Nd Anomaly in Kimberlites and Carbonatites
The short-lived chronometer 146Sm-142Nd (T1/2 = 103 Ma) has been successfully used to improve our understanding of early planetary differentiation on asteroids, the Moon, Mars and Earth. On the Earth, all reported 142Nd excesses (8 to 30 ppm) are in early-Archean samples from SW Greenland (Harper and Jacobsen, Nature 1992; Sharma et al., GCA 1996; Boyet et al., EPSL 2003; Caro et al., Nature 2003). This evidence for silicate differentiation during the first hundred Myr of Earth's history is now robust, but traces of this major fractionation event are still very scarce and only the light-REE depleted reservoir has been identified so far. To determine whether the chemically distinct reservoirs created in this early differentiation still exist, we examined the 146Sm-142Nd system in kimberlites and carbonatites, the most incompatible element rich melts on Earth, some of which are believed to be of deep mantle origin. Nd was measured with the Finnigan TRITON installed at DTM in April 2004. The two standards analyzed, La Jolla (n=41) and JNdi-1 (n=20) give an external reproducibility for the 142Nd/144Nd ratio of 5 ppm, 2-sigma. Our data show no correlation between 150Nd/144Nd and 142Nd/144Nd, hence a second-order fractionation correction (Caro et al., 2003) has not been applied. JNdi-1 appears to have a 142Nd/144Nd ratio 6 ppm higher than the La Jolla standard. Though requiring more investigation, this result raises the question of which, if either, of these standards best represents the chondritic, and hence bulk-earth, 142Nd/144Nd. Repeat analyses of terrestrial samples show 142Nd/144Nd reproducibility of better than 8 ppm 2-sigma. We confirm previously reported (Boyet et al., 2003) positive anomalies (up to 20 ppm) in some metabasalts from Isua. However, no 142Nd anomalies were detected in 20 samples of Group I kimberlite from Siberia, Brazil, SW Greenland, Montana, Colorado and S. Africa, nor in two mid-ocean ridge basalts (all of which lie within the range -8 to 7 ppm of the average 142Nd/144Nd in La Jolla Nd). Among the three carbonatites from SW Greenland analyzed, only one is characterized by a very small, and still inadequately resolved, positive 142Nd excess of 11 ppm. This new investigation resolves no trace of the early silicate differentiation in the source of these rocks. Either the effects of this differentiation were mixed away by early Earth convection, or the distinct reservoirs created in this event do not produce magmas that reach the surface.
U41A-0724 0800h
Comparing lower mantle compositions and temperatures inverted from different seismic observations
Most constraints on lower mantle composition and temperature come from comparison of seismic profiles with their equivalents deduced from mineral physics. For seismologists, the parameters directly observable are the velocities $V_P$ and $V_S$, whereas the most readily determined properties experimentally are density $\rho$ and bulk modulus $K_S$. In this study, we apply a generalized inverse method and high-quality experimental datasets to infer the lower mantle composition and temperature profile from seismic observations. Inversion from density and bulk sound velocity highlights the following points: (1) equally satisfactory fits to seismic profiles can be obtained either for pyrolite-type composition with a cool geotherm, or for perovskite-rich composition with a hot geotherm, (2) consistent features in all inversions are a total iron content of 0.1 and a subadiabatic temperature gradient, with a peculiar correlated behavior of these two parameters below the 660 km discontinuity (3) the results of inversions are unaffected by the partitioning of iron between perovskite and magnesiow\"ustite (4) the inversion does not constrain the Al$_2$O$_3$ and CaO contents of the lower mantle. Using the compressional and shear velocities as additional constraints decreases the a posteriori uncertainties on the inverted parameter and removes the dependence of the final results on the a priori model. However, the inverted composition and temperature profiles drastically depend on the shear properties of lower mantle minerals, particularly of silicate perovskite. The experimental uncertainties yield a wide range of compositional and thermal profiles ranging from a uniform pyrolite-type lower mantle with closely adiabatic geotherm to a lower mantle whose composition gradually increases from pyrolite at 660 km depth to pure perovskite at the CMB with a strongly superadiabatic profile. A better understanding of lower mantle composition and temperature requires better constraints on shear properties, especially their pressure and temperature dependence.
U41A-0725 0800h
Solubility of Helium in Olivine at 1 Atmosphere
We have measured the solubility of He in olivine at 1 atm. Previous measurements of noble gas solubility in mantle minerals have found mineral/melt partition coefficients (D) that are higher [1;2] or close to [3] the D values of U and Th in the mantle. In contrast, geochemical systematics suggest that D$^{He}$ is lower than D$^{U}$ and D$^{Th}$. Our experiments were specifically designed to avoid gas trapped in melt pockets or between sintered grains, which may have affected previous studies [1;2]. The starting materials are gem quality San Carlos olivine and synthetic pure forsterite. These materials were examined before and after the experiments for inclusions and bubbles using optical, scanning electron and scanning transmission electron microscopic techniques. No bubbles were found. The primary experiments were performed on cm size grains to avoid trapping of gas in sintered powders. The experiments were run for 17-21 days at 1350$\deg$C, in crucibles made from large San Carlos olivine crystals, in an atmosphere of 50% He and 50% a mix of CO$_{2}$ and H$_{2}$ (to control fO$_{2}$ at NNO and QFM). At no time was the olivine in contact with a melt phase. To examine the effects of powder sintering, experiments that span a range of smaller grain sizes (100-1000 microns) were also performed. He concentrations in the olivines were measured by noble gas mass spectrometry using sequential in vacuo crushing followed by melting of the powders. The experimental results are consistent: 8.3(+/-2.6) e$^{-7}$ cc STP He/g are released by crushing and 6.2(+/-1.3) e$^{-7}$ cc He STP/g are released by melting. Over 50% of the total gas is released by crushing. Powdered samples release unusually high amounts of gas on the first crush step (interpreted to be trapped between grains), but subsequently follow the same release pattern as the unpowdered samples and yield the same solubility values (excluding the first step). The consistency of the results over a range of compositions, grain sizes and fO$_{2}$ conditions, and our careful examination of the experimental materials indicate that the gas released during crushing is not trapped in bubbles or inclusions, but is loosely bound within the crystal lattice. Where it resides in the olivine crystal structure is not clear. Using only the gas released by melting, along with published solubilities of He in basaltic melts [4], the ol/melt D$^{He}$ is 0.003(+/-0.001) at 1 atm. This is most likely an upper limit because the crushing experiments may not have completely eliminated the loosely bound helium. If the gas from both the crushing and melting steps are combined, the apparent D$^{He}$ is 0.006(+/-0.001). We caution against extrapolating these preliminary values to mantle melting conditions until the location of He in olivine is identified and the effects of pressure and temperature are quantified. [1] Broadhurst et al. (1990) GCA 54: 299-309 [2] Hiyagon and Ozima (1986) GCA 50: 2045-2057 [3] Brooker et al. (2003) Nature 423: 738-741 [4] Jambon et al. (1986) GCA 50: 401-408
U41A-0726 0800h
Diamond Formation From Organic Carbon in a Subducting Slab at Depths Between 250 and 450 km
Diamonds from the Jagersfontein kimberlite in South Africa contain inclusions of majorite garnet with eclogitic major element composition, providing evidence for diamond formation in ``basaltic'' environments at depths of up to 450 km. All 12 majoritic garnets recovered show distinct negative europium anomalies linking their ``basaltic'' source rocks to subducted oceanic crust. The carbon isotope composition ($\delta$$^{13}$C) of the host diamonds is restricted to values between -17 and -24 $\permil$. This tight distribution is completely different to the carbon isotope composition of ``normal'' (non-majoritic) eclogitic diamonds from Jagersfontein, which exhibit a broad range from -1 to -24 $\permil$ with a pronounced mode at -4 $\permil$. The very distinct isotopic distributions show that diamonds of majoritic and normal eclogitic paragenesis form from different carbon sources. The small range in carbon isotopic composition of diamonds with majorite inclusions suggests that they directly reflect the isotopic signature of their carbon source. The isotopically light carbon isotopic composition of these diamonds is, therefore, consistent with a derivation from organic matter within a subducting slab. The main mechanism for the formation of the majorite bearing diamonds probably is direct conversion from graphite, which is expected to occur deep within the diamond stability field due to very sluggish kinetics.
U41A-0727 0800h
Mineral Physics in Thermo-Chemical Mantle Models
The mantle structures observed by seismic tomography can only be linked with convection models by assuming some relationships between temperature, density and velocity. These relationships are complex and non linear even if the whole mantle has a uniform composition. For example, the density variations are not only related to the depth dependent thermal expansivity and incompressibility, but also to the distribution of the mineralogical phases that are themselves evolving with temperature and pressure. The geochemical observations indicate that the mantle cannot be homogeneous but is composed with various reservoirs of different compositions, although their sizes, origins and topologies are still questionable. Here, we present a stoichiometric iterative method to compute the equilibrium mineralogy of mantle assemblages by Gibbs energy minimization. The numerical code can handle arbitrary elemental composition in the system MgO, FeO, CaO, Al$_2$O$_3$ and SiO$_2$ and reaches the thermodynamic equilibrium by choosing the abundances of 31 minerals belonging to 14 possible phases. The code can deal with complex chemical activities for minerals belonging to solid state solutions. We illustrate our approach by computing the phase diagrams of various compositions with geodynamic interest (pyrolite, harzburgite and oceanic basalt). Our simulations are in reasonable agreement with high pressure and high temperature experiments. We predict that subducted oceanic crust remains significantly denser than normal mantle even near the core mantle boundary. We then provide synthetic tomographic models of slabs. We show that properties computed at thermodynamic equilibrium are significantly different from those computed at fixed mineralogy. Although the accuracy of our results is limited by the uncertainties on the thermodynamic parameters and equations of states of each individual mineral, future geodynamic models will need to include these mineralogical aspects to interpret the tomographic results as well as to explain the geochemical observations.
U41A-0728 0800h
A Comparison of Methods for Modeling Geochemical Variability in the Earth's Mantle
Numerial models of isotopic and chemical heterogeneity of the Earth's mantle fall into three categories, in decreasing order of computational demand. First, several authors have used chemical tracers within a full thermo-chemical convection calculation (e.g., Christensen and Hofmann, 1994, van Keken and Ballentine, 1999; Xie and Tackley, 2004). Second, Kellogg et al.\ (2002) proposed an extension of the traditional geochemical box model calculations in which numerous subreservoirs were tracked within the bulk depleted mantle reservoir. Third, All\`egre and Lewin (1995) described a framework in which the variance in chemical and isotopic ratios were treated as quantities intrinsic to the bulk reservoirs, complete with sources and sinks. Results from these three methods vary, particularly with respect to conclusions drawn about the meaning of the Pb-Pb pseudo-isochron. We revisit these methods in an attempt to arrive at a common understanding. By considering all three we better identify the strengths and weaknesses of each approach and allow each to inform the other. Finally, we present results from a new hybrid model that combines the complexity and regional-scale variability of the thermochemical convection models with the short length-scale sensitivity of the Kellogg et al.\ approach.
U41A-0729 0800h
Terrestrial Magma Ocean and Hadean Crust Formation: Insights From $^{146}$Sm-$^{142}$Nd and Trace Element Partitioning in High-Pressure Mantle Minerals
The effect of the formation of a magma ocean during terrestrial accretion on the composition and stratification of the Earth mantle remains a major unanswered question. Although partial or complete melting of the mantle appears unavoidable as a result of heating by giant impacts, core segregation, and possibly a blanketing atmosphere, ratios of refractory lithophile elements in modern upper mantle peridotites do not show the characteristic fingerprint of magma ocean crystallization. A likely interpretation is that convective mixing would have hampered efficient segregation of liquidus mineral phases during the early stages of cooling. However, differentiation may have proceeded by segregation of a residual melt during the final stage of solidification, when convection would have slowed down in the crystallizing mush. Here, we address this issue using short-lived $^{146}$Sm-$^{142}$Nd chronometry. $^{146}$Sm decays to $^{142}$Nd with a half-life of 103 Myr, such that $^{142}$Nd is a selective tracer of early mantle differentiation. Using ultrahigh-precision Nd isotope measurements, we show that 3.6-3.8 Gyr old rocks from West Greenland carry positive $^{142}$Nd anomalies ranging between 8 and 15 ppm. This indicates that the Earth mantle underwent large-scale differentiation near the final stage of terrestrial accretion, at 4.35-4.5 Gyr. Using trace element partition coefficients for high-pressure mantle minerals, we show that the observed Nd isotopic signature of early Archaean rocks and combined Hf signature of Hadean zircons are consistent with segregation of small amounts of melt from a crystallizing mush at upper mantle pressure. To match our observations, the protocrust resulting from crystallization of this residual melt must have had a lifetime of ca. 500 Ma. This suggests that crustal recycling was less efficient during the Hadean era, despite vigorous mantle convection. Alternatively, the earliest terrestrial crust may have undergone efficient recycling followed by long-term storage at the core-mantle boundary.
U41A-0730 0800h
Cryptic Striations in the Indian Ocean Mantle Revealed by Hf Isotopes in Southeast Indian Ridge MORB Glasses
The Southeast Indian Ridge (SEIR) provides a unique opportunity to study the consequences of along-axis variation in upper mantle temperature and flow. Over a distance of 2500 km, between $86\deg$E and $120\deg$E, axial depth increases from 2300m to 5000m. The axial depth gradient occurs at intermediate and constant spreading rate (70-75 mm/y full rate) and in the absence of large transform offsets and nearby hotspots. There is also a morphologic transition from axial high to axial valley going eastward, due to the decrease in melt production rate and crustal thickness. The range in axial depth and ridge morphology is similar to the global range for spreading ridges away from hotspots, making the SEIR a regional-scale analogue of the 50,000 km-long global ocean ridge system. Previous work has established that He, Pb, Sr and Nd isotope variations along the SEIR are primarily controlled by changes in the depth of melting of isotopically heterogeneous mantle, and that all SEIR lavas west of the Australian-Antarctic Discordance are true Indian-type as indicated by their high $^{208}$Pb/$^{206}$Pb ratios. New Hf isotope results have been obtained for 48 MORB glasses recovered from the SEIR between the Amsterdam-St. Paul Plateau and the Australian-Antarctic Discordance. $\epsilon$$_{Hf}$ ranges from +5.5 to +17.8. The extreme values occur on the ASP Plateau (5.5) and in the westernmost AAD (17.8), and are similar to values measured previously in those regions. Overall, Hf and Nd isotopes along the SEIR show the typical positive correlation seen in oceanic basalts. However, over a stretch of 2000 km, between $88\deg$E and $110\deg$E, the Hf isotope compositions are strikingly bimodal, with a "gap" of more than 1 epsilon unit between the two groupings ($\epsilon$$_{Hf}$ = 9.5 to 11.5 and $\epsilon$$_{Hf}$ = 12.5 to 14.6, respectively). Although this Hf isotope bimodality occurs within the band of globally correlated $\epsilon$$_{Hf}$ - $\epsilon$$_{Nd}$ data, our sample suite has a strong spatial resolution, suggesting the presence of mantle striations (streaks) beneath the SEIR. The bimodality in Hf isotopes is not observed in other geochemical parameters, indicating that the streaks carry a cryptic memory of augmented Lu/Hf fractionation at some time in the past. The streaks defined by the Hf isotope bimodality display a Poisson spatial distribution, (i.e., number of streaks observed is proportional to the length of ridge sampled) with an average computed streak width of 50-100 km. This is the expected spatial distribution for a well-stirred upper mantle in which geochemical heterogeneity is continually created by tectonic recycling and diminished by convective refolding and stretching.
U41A-0731 0800h
Experimental Insights on the Thermochemical State of the Lower Mantle: Convection, Chemical Heterogeneities, and the D" layer
We report recent findings in the field of high-pressure mineral physics with important implications for Earth's lower mantle. We find that the two main constituents of the lower mantle, namely (Mg,Fe)SiO$_3$ - magnesium silicate perovskite - and (Mg,Fe)O - magnesiow\"ustite -, undergo electronic transitions at lower mantle pressures, in which iron is transforms from the high-spin state to the low-spin state. The transformations should profoundly alter the thermochemical state of Earth's lower mantle. Minerals bearing high-spin iron have characteristic absorption lines in the near-infrared, hindering radiative conductivity at lower-mantle temperatures. These absorption lines shift in low-spin iron-bearing minerals to the visible range (green to violet), and their intrinsic intensities decrease; the minerals thus become transparent in the near-infrared and their radiative conductivity (and therefore total thermal conductivity) increases. Other issues at stake are that of melting temperature or iron partitioning between mineral phases or even between various crystallographic sites of a mineral phase. The two transition pressures correspond to the bottom third of the lower mantle (70 GPa, 1700 km depth), and to the last 300 km above the core-mantle boundary (120 GPa, 2600 km depth); these regions have very special geophysical signatures, since chemical heterogeneities have been reported in the bottom third of the lower mantle, and that the bottom 300 km of Earth's mantle is constituted by the D" layer. Our observations could provide a mineral physics basis for these two regions of Earth's lower mantle. The implications of these transitions on the dynamics of the lower mantle will be discussed.
U41A-0732 0800h
The Effects of Increased Thermal Conductivity and Viscosity on Mixing Rates and Convection Patterns in the Deep Lower Mantle.
Changes in the spin state of iron in both magnetowustite and perovskite at lower mantle conditions may result in increases in radiative thermal transport and viscosity that could suppress convection in the lowermost mantle (Badro et al. 2003, 2004). It has been suggested that such a stagnant layer in the lower mantle could serve as a reservoir for a significant portion of the mantle's incompatible elements, accounting for the isotopic characteristics of hot spots linked to proposed deep-rooted mantle plumes. We investigate the possible effects on mantle dynamics of increases in thermal conductivity and viscosity, using finite-element models of mantle convection in 2-D. Our previous results (Naliboff et al. 2003) showed that increases in thermal conductivity in the lower mantle up to 250 times that in the upper mantle, with otherwise uniform physical properties, fail to isolate a stagnant layer beneath a mid-mantle phase change. When both the viscosity and thermal conductivity increase in the lower mantle, flow velocities through the lower layer and across the boundary decrease. To investigate the rate of mass exchange and mixing in the presence of a partially stagnant layer, we injected tracer particles into the models. We examine mixing in three different classes of models: two models have a viscosity and thermal conductivity change at the mantle mid-point; the third has a viscosity increase at 660 km and a viscosity and thermal conductivity change near 2000 km depth. In models in which the viscosity and thermal conductivity increases by a factor of 10 at the mid-mantle, multi-cell whole-mantle convection rapidly produces a marble cake mantle, leaving no isolated reservoir of material in the lower mantle. Increasing the viscosity and thermal conductivity in the lower mantle by a factor of 50 or 100 produces a relatively stable pattern of convection with a few strong upwellings and downwellings. Although mixing rates decrease and the residence time of material in the lower mantle increases with increasing viscosity and thermal conductivity in these models, no long-term stagnation of material occurs beneath the phase change. When the lower mantle viscosity and thermal conductivity increase by a factor of 150 in the lower mantle, the size of upwellings decreases, with multiple small plume-like structures forming at the interface. This transition marks the change to a more stable mid-mantle boundary between material above and below the interface. Increasing the viscosity and thermal conductivity in the lower mantle by a factor of 250 further decreases the size and material transport rate, leading to longer residence times of material in the bottom half of the mantle. In summary, a phase change with large magnitude changes in viscosity and thermal conductivity at the mid-mantle would fail to maintain isolated lower mantle geochemical reservoirs over significant geologic time. Mixing rates, however, decrease with increasing viscosity and thermal conductivity.
U41A-0733 0800h
Core Formation Timescale, Silicate-Metal Equilibration, and W Diffusivity
The extent to which material accreted to the proto-Earth and segregated to form the core was chemically and isotopically equilibrated with the silicate mantle is an outstanding problem in planetary science. This is particularly important when attempting to assign a meaningful age for planetary accretion and core formation based on Hf-W isotope systematics. The Earth and other terrestrial planets likely formed by accretion of previously differentiated planetesimals. For the planetesimals themselves the most important energy source for metal-silicate differentiation is the combined radioactive heating due to decay of 26Al (half-life 0.7 Ma) and 60Fe (half-life 1.5 Ma). It is expected that the fractionation of Hf and W during planetesimal core formation will lead to a divergence in the W isotopic compositions of the core and silicate portions of these bodies. This expectation is supported by the enormously radiogenic 182W signatures reported for basaltic eucrites. The observation that the W isotopic compositions of the silicate portions of Earth, Moon and Mars are similar and markedly less radiogenic than eucrites suggests that during planet accretion the pre-differentiated metallic core material containing low 182W must have equilibrated extensively with the more radiogenic (high 182W) silicate material to subdue the ingrowth of 182W in the silicate mantle of the planets. The standard theory of planet formation predicts that after runaway and oligarchic growth, the late stage of planet formation is characterized by impact and merging of Mars-sized objects. This is a tremendously energetic process estimated to raise the temperature of the proto-Earth to about 7000K (a temperature equivalent to a mass spectrometer's plasma source, which indiscriminately ionizes all incoming elements). After the giant impacts, the proto-Earth had a luminosity and surface temperature close to a low mass star for a brief period of time. Stevenson (1990) argued that emulsification caused by large-scale Rayleigh-Taylor instabilities following giant impacts breaks up the metallic core of the impactor into centimeter-sized droplets within minutes. We have constructed a simple model to study the kinetics of isotope equilibration between metal-silicate during the "rain-fall" of metal droplets descending through the terrestrial magma ocean. This model highlights the importance of the kinetics of ion mobility of W for assessing quantitatively the degree of metal-silicate equilibration during core formation. We have determined for the first time that the W self-diffusion coefficient in basaltic liquid is 4.98E-7 cm2/s at 3GPa, 1500 C. We assume this is a minimum value in the magma ocean scenario, and the equilibration is rate-limited by diffusion in the silicate liquid. Applying this value and taking a reasonable estimate of viscosity for silicate liquids from the literature, we show that the degree of equilibration asymptotically approaches 100% within the timescale of metal-silicate segregation when the metallic droplets are $<$20 cm in diameter.
U41A-0734 0800h
Testing Models of Core-Mantle Exchange
Compositional estimates of Earth reservoirs (i.e., crust, mantle core) allow us to critically evaluate models of core-mantle exchange from an elemental and isotopic perspective. For a significant number of elements the composition of the Earth's core and mantle is well constrained by the data of chondritic meteorites (undifferentiated, primitive materials) and mantle samples. These constraints are most robust for elements with high condensation temperatures $>$1100 K (e.g., Fe, Mg, Si, etc), assuming a chondritic bulk planet. Isotopic models are the most sensitive to assessing small core contributions to the mantle (e.g., $<$1%), whereas chemical and isotopic data can be brought to bear on assessing larger amounts of core additions to the mantle (e.g., $<$0.5%). Chemically, the core and mantle have been isolated from each other since core separation, which occurred during and immediately following accretion (i.e., $<$30 Ma of solar system formation). Support for this isolation is demonstrated by W isotopic composition of the mantle and the constancy of mantle ratios of lithophile/siderophile elements (mantle/core stored elements). However, neither of these approaches have sufficient resolution to decisively test the alternative hypothesis that argues for limited core-mantle exchange in the sources of hotspot basalts based on Os isotopic evidence. A series of models are tested that encompass both elemental and isotopic consequences of core-mantle exchange for a range of plausible outer core compositions. Compositional models for the outer core are dependent on inner/outer core fractionation parameters and growth curves for the formation of the inner core. Results from these models permit limited transfer of core material (up to ~0.5%, by mass) into the mantle source regions of hotspot magmas. Future measurements and developments with improvements in precision will provide greater resolution and evaluation of these models.
U41A-0735 0800h
Helium Isotopes in hot Springs and (hot) Wells of the Basin and Range
The Basin and Range (BR), a major tectonic province of the western margin of the North American plate, is characterized by an anomalous thermal gradient, large heat flux, high regional elevation, and volcanism and extensional tectonics that have varied in time and space. He isotopes provide a definitive distinction between crustal and magmatic fluid sources and in the case of crustal fluids, allow an evaluation of the presence, size, and involvement of mantle melts; important factors for understanding tectono-magmatic processes and their influence on BR evolution. Since elevated $^{3}$He fluxes, as measured by He isotope compositions, reflect mantle melting, He isotopes provide constraints on models for regional high heat flux anomalies, such as lithospheric stretching accompanied by accretion of crystalline mantle, thinning accompanied by accretion of mantle melts, or thinning associated with intrusions of mantle melts into the upper crust. To evaluate these models, we are in the process of constructing a He isotope 'map' of the greater BR. Fluids from the western margin of the BR have He isotope ratios as high as 6-7 Ra, indicating upper crustal intrusion of mantle melts consistent with recent and current volcanism. Moving away from these areas, He isotope ratios decrease rapidly to 'background' values of around 0.6 Ra, and then gradually decrease eastward to low values of 0.1 Ra at the eastern margin of the BR. Assuming a steady-state relationship for the transport of heat and He through the crust, such low ratios are inconsistent with regional underplating or intrusion by mantle melts into the upper crust. Superimposed on this regional trend are isolated features with elevated He isotope ratios (0.8-2.1 Ra) compared to the local background. Spring geochemistry and local geology indicate that these 'He-spikes' are not related to current or recent magmatic activity, suggesting that the spikes may reflect either small localized zones of deep mantle melting or deep permeable pathways (faults) with high vertical fluid flow rates. A detailed study of one of the He-spikes (Dixie Valley and the Stillwater Range Front Fault system), indicates that features with high He isotope ratios are confined to the range front normal faults characteristic of the extensional regime in the BR, suggesting that such faults can be permeable pathways for deep fluid circulation.
U41A-0736 0800h
Mineral Physics Constraints on the Composition and Lateral Heterogeneity in the Lower Mantle
New experimental results from a simultaneous determination of the pressure and temperature derivatives of the bulk and shear moduli and thermal expansivity of magnesium silicate perovskite at high pressure and high temperature enabled us to carry out an up-to-date modeling of both P and S wave velocities and density to the lower mantle depths to constrain the composition and temperature of the Earth's lower mantle. We found that a pyrolite composition along 1600K adiabatic geotherm can reproduce P and S wave velocities, Poisson' ratio, and density of PREM and AK135 reasonably well within experimental uncertainties. Discrepancies in composition and geotherm reported in previous studies could be attributed to the choice of thermoelastic properties of perovskite obtained from different experimental techniques. According to current modeling, the values of the dlnVs/dlnVp due to thermal effect are close to those observed by majority of seismic tomographic observation, ranging from 1.7 to 2.1 from the top to the lower part of the lower mantle. Other effects, including variations of perovskite and magnesiowustite proportions, Fe enrichment, as well as iron and magnesium partitioning between lower mantle phases on the lateral variation of S and P wave velocity change and density to shear wave scaling factor are also explored while searching for causes for anomalously high lateral heterogeneities observed in tomographic studies at the base of the lower mantle.