Search

Melt transport across the ductile mantle is essential for oceanic crust formation or intraplate volcanism. However, mechanisms of melt migration and associated chemical interaction between melt and solid mantle remain unclear. Here, we present a thermo-hydro-mechanical-chemical (THMC) model for melt migration coupled to chemical differentiation. We consider melt migration by porosity waves and a chemical system of forsterite-fayalite-silica. We solve the one-dimensional (1D) THMC model numerically using the finite difference method. Variables, such as solid and melt densities or MgO and SiO2 mass concentrations, are functions of pressure (P), temperature (T), and total silica mass fraction (urn:x-wiley:15252027:media:ggge22741:ggge22741-math-0001). These variables are pre-computed with Gibbs energy minimization and their variations with evolving P, T, and urn:x-wiley:15252027:media:ggge22741:ggge22741-math-0002 are implemented in the THMC model. We consider P and T conditions relevant around the lithosphere-asthenosphere boundary. Systematic 1D simulations quantify the impact of initial distributions of porosity and urn:x-wiley:15252027:media:ggge22741:ggge22741-math-0003 on the melt velocity. Larger perturbations of urn:x-wiley:15252027:media:ggge22741:ggge22741-math-0004 cause larger melt velocities. An adiabatic or conductive geotherm cause fundamentally different vertical variations of densities and concentrations, and an adiabatic geotherm generates higher melt velocities. We quantify differences between melt transport (considering incompatible tracers), major element transport and porosity evolution. Melt transport is significant in the models. We also quantify the relative importance of four porosity variation mechanisms: (a) mechanical compaction and decompaction, (b) density variation, (c) compositional variation, and (d) solid-melt mass exchange. In the models, (de)compaction dominates the porosity variation. We further discuss preliminary results of 2D THMC simulations showing blob-like and channel-like porosity waves.

Harzburgite–dunite associations in the Yushigou ophiolitic complex (North Qilian suture, NW China) offer insights into melt migration and melt–rock interaction above a subduction zone. Based on detailed petrographic and in situ analyses, we propose a three-stage model to describe the history of melt channel formation in oceanic arc mantle. In the first stage, high-degree melt extraction was followed by widespread infiltration of small-volume melts, which formed the harzburgite that makes up most of the Yushigou complex. The Yushigou harzburgites thus have highly refractory major-element compositions (olivine Fo = 91·1–93·2, spinel Cr# [atomic 100 × Cr/(Cr + Al)] = 55·2–62·9, whole-rock Mg# = 90·5–92·0, Al2O3 = 0·29–0·74 wt%, CaO = 0·35–0·54 wt%, and low Na2O = 0·02–0·05 wt%) but slight enrichment in the light rare earth elements (LREE). In the second stage, high-volume focused melts infiltrated into the mantle wedge and produced two types of dunite, with high-Cr# (mainly 63·1–73·6) versus low-Cr# (23·6–33·7) spinel. In the low-Cr# dunites, spinels have significantly fewer inclusions and lower Fo values (88·7–90·1) in olivine than in the high-Cr# dunite (Fo = 89·9–90·5), together with higher modal abundances of interstitial sulfides. The two groups of dunite show similar extremely low TiO2 (10 in high-Cr# dunite), indicating distinct differences in the infiltrating melts. The low-Cr# dunite is a cumulate from an anhydrous Al- and S-enriched basaltic melt, whereas the high-Cr# dunite was produced by reaction of harzburgite with a carbon-bearing, alkaline hydrous silicate magma. The confluence of these melt migrations finally formed a channelized dunite network, and later magmatic activity was restricted to these channels. Pulsated melt supply after chromite crystallization formed pervasive sieve-textured rims around spinel in both kinds of dunite, but these are rare in the harzburgite. The third stage was marked by fluid-dominated metasomatism recorded by millimeter-scale veined conduits, which contain carbonate (dolomite and magnesite), amphibole, phlogopite and compound crystal assemblages in both kinds of dunite. CH4–N2–graphite-dominated fluid inclusions are widespread in all lithologies, recording the volatiles transported during the last metasomatic event. The Yushigou complex thus provides a detailed example of interaction between multiple batches of melt or fluid and a lithospheric mantle wedge dunitic channel system. The results of this study further suggest that the formation of podiform chromite requires melt participation, and the nature and origins of the melts can be diverse.

International audience; The exposure of gabbroic sequences at Oceanic Core Complexes (OCC) along ultraslow-to slow-spreading ridges permits the study of the processes forming the lower oceanic crust. On top of the Atlantis Bank OCC along the ultraslow-spreading Southwest Indian Ridge, IODP Expedition 360 drilled Hole U1473A, mainly composed of primitive olivine gabbros interspersed with more evolved Ti-Fe oxide-bearing gabbros and minor felsic veins. These rocks record a complex history of protracted magmatism during continuous uplift and deformation of the gabbroic sequence. Extensive crystal-plastic deformation is dominantly recorded in the shallower sections of the drillhole, whereas the deeper sections better preserve primary magmatic features. We focus on microstructures, including intra-crystalline deformation of rock-forming minerals, and plagioclase crystallographic preferred orientations of olivine gabbros lacking evidence for exhumation-related crystal plastic deformation, to gain insights on the relationship between compaction, melt migration and melt accumulation during the early magmatic history of this section of lower oceanic crust. Olivine gabbros are characterized by ubiquitous grain-size variations, from coarse-to fine-grained intervals. Minerals in coarse-grained intervals show intra-crystalline deformation, while fine-grained crystals lack internal strain. Bent coarse-grained plagioclase associated with weak magmatic foliation and lack of lineation suggest that the coarse-grained intervals were deformed under weak compaction. On the other hand, crystallographic preferred orientations of undeformed fine-grained plagioclase show weak lineations, likely indicative of non-coaxial strain. We thereby infer that the coarse-grained intervals underwent ongoing weak compaction from the stage of olivine + plagioclase ± clinopyroxene crystal mush to the melt-poor stage, and that this process likely aided melt extraction and accumulation in discrete melt-rich zones where crystals orientated in the direction of magmatic flow. Crystallization of melts in the melt-rich zones ultimately formed the fine-grained intervals at different depths in Hole U1473A. This indicates that processes of compaction can lead to local chemical and grain-size heterogeneities in a lower crustal section, while had a minor role in the melt movement at larger scales (e.g., the whole crystal mush) within the oceanic crust.

To investigate channelization during migration of a reactive melt, we performed a series of Darcy-type experiments in which an alkali basalt infiltrated partially molten harzburgites and lherzolites at a confining pressure of 300 MPa, temperatures of 1200°C and 1250°C, and pore pressure gradients of ~2 to 60 MPa/mm. We compare our results to those from previously published experiments performed on wehrlites. In all experiments, irrespective of the exact mineralogy, a planar reaction layer composed of olivine + melt developed in which all of the pyroxene was consumed. Under specific conditions controlled primarily by the melt flow velocity, finger-like channels composed of olivine + melt also developed. In wehrlites, these reaction infiltration instabilities formed at 1200°C and 1250°C at pressure gradients >25 and >5 MPa/mm, respectively. In harzburgites, channelization occurred only at 1250°C at a pressure gradient of 35 MPa/mm. In lherzolites, a planar melt-filled vein developed at 1250°C; no finger-like channels formed under a pressure gradient of ~25 MPa/mm. Both the finger-like channels and the planar vein led to very efficient extraction of melt from the reservoir. Channelization established large compositional variations over short distances in the crystalized phases as well as in the local melt and greatly enhanced the abundance of the reaction product, olivine, similar to dunite channels in the Earth. The range of chemical-mechanical responses displayed by this array of compositions provides a set of targets for reactive transport and mechanical modeling studies., National Science Foundation (Grant OCE-1459717)

Author Posting. © American Geophysical Union, 2020. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research-Solid Earth 125(2), (2020): e2019JB018203, doi:10.1029/2019JB018203.

The oceanic crust, covering two-thirds of the Earth's surface, is formed along mid-oceanic ridges by crystallization at shallow levels of melts formed at depth by partial melting of mantle peridotite. Yet, the process of melt transport to the ridge axis remains poorly understood. Ophiolites, which provide a window into the uppermost mantle, contain dunite bodies often interpreted as relics of melt flow conduits, formed by pyroxene dissolution during melt-peridotite interaction. Here, we present structural and geochemical data on peridotites from the southeastern Oman ophiolite where three types of dunite, corresponding to the Moho transition zone (MTZ), the main and basal mantle sections, are identified. We focus on osmium isotopes, which are particularly well-adapted to tracing melt flow through peridotites. Osmium isotope signatures from host harzburgites accord with abyssal peridotite values and do not vary systematically with setting. In contrast, dunite Os compositions depend on structural context. Basal dunites display compositions similar to harzburgite values, while MTZ dunites have highly radiogenic compositions similar to those of the overlying crust, requiring extensive interaction with melts more radiogenic than MORB. Modeling shows that melts percolating through and equilibrating with dunite channels would acquire unradiogenic compositions, inconsistent with the observed Os signatures of MTZ dunites and lower crust. Thus, our findings require melt transport without equilibration with dunite or harzburgite, arguing for rapid or at least chemically isolated melt migration from the mantle source to the Moho.

The interior structures of planetary bodies beyond the Earth are broadly unknown. Our observational capacity is largely limited to surface imagery from spacecraft. The work presented in this dissertation uses novel modeling methods of melt migration and melt focusing processes to illuminate the thermal and structural characteristics of the lithospheres of Mars and Jupiter’s moon Io. Models are constrained by, and inform observations of, surface volcanism. Coupled petrological-geodynamical models of one-dimensional melt migration are performed to determine the depth of impermeable boundaries, known as permeability barriers, in the lithosphere of Mars. Relatively deep permeability barriers are found to be prevalent throughout Martian history unless in regions of high strain rate (10^-13 s^-1), or a wet mantle (25-1000 ppm H2O). Permeability barrier depth is suggested to be linked to the style of volcanic edifice seen at the surface, with deep barriers creating larger edifices like shield volcanoes, and shallower barriers creating widespread flows. Similar petrological-geodynamical models performed for the lithosphere of Io reveal that permeability barriers always form at the base of the lithosphere due to the cold temperatures caused by geologically rapid resurfacing (~1 cm/yr) and subsidence. Melt may ascend closer to the surface in areas with a low subsidence rate (0.02 cm/yr) Two-dimensional numerical models of melt migration in the Martian lithosphere suggest that convection in a highly porous layer beneath a permeability layer (a decompaction channel) focuses melt over the convective wavelength. Melt ascends in the lithosphere at this wavelength which is reflective of volcano spacing at the surface for Hesperian aged terrains. Numerical and analytical models of melt flow through the asthenosphere and lithosphere of Io constrain the lifespan of its volcanic plumbing systems. A 1 km conduit will fully close within ~10,000 years while a 25 km conduit of melt will close within 6-7 million years. Solid convection in the asthenosphere is found to be necessary for melt focusing to heat pipe centers at the base of the lithosphere, however it is counterintuitively found that an arrangement with downwelling undeath the eruptive center is the most efficient for melt extraction.

Continuous tow investigations have shown that the present vent field inventory along fast to intermediate spreading ridges may be underestimated by at least 3–6 times, while the limited towed line investigations of venting sites along slow to ultra-slow spreading ridges make it impossible to determine their distribution. The Chinese Dayang cruise has conducted detailed towed line surveys of hydrothermal activity on segment 27 of the ultra-slow spreading southwest Indian ridge in 2015. The results have identified as many as 9 hydrothermal fields along 85-km-long segment, including one confirmed hydrothermal field, three inferred hydrothermal fields and five suspected fields. Hydrothermal activities are not only limited along-axis but also found approximately 10 km away from the axis. These vent fields are likely powered by a seismically identified axial magma chamber, including melt migration along normal faults to flank areas. The calculated hydrothermal activity frequency on segment 27 is approximately 3.6–8 times higher than that calculated from the Interridge database, suggesting that careful system exploration can reveal more hydrothermal activities even on ultra-slow spreading ridges effected by hotspot.

Porosity waves arise naturally from the equations describing fluid migration in ductile rocks. Here, we show that higher-dimensional porosity waves can transport mass and therefore preserve geochemical signatures, at least partially. Fluid focusing into these high porosity waves leads to recirculation in their center. This recirculating fluid is separated from the background flow field by a circular dividing streamline and transported with the phase velocity of the porosity wave. Unlike models for one-dimensional chromatography in geological porous media, tracer transport in higher-dimensional porosity waves does not produce chromatographic separations between relatively incompatible elements due to the circular flow pattern. This may allow melt that originated from the partial melting of fertile heterogeneities or fluid produced during metamorphism to retain distinct geochemical signatures as they rise buoyantly towards the surface.

Rocky planetesimals in the early solar system melted internally and evolved chemically due to radiogenic heating from Al-26. Here we quantify the parametric controls on magma genesis and transport using a coupled petrological and fluid mechanical model of reactive two-phase flow. We find the mean grain size of silicate minerals to be a key control on magma ascent. For grain sizes larger than $\approx$ 1 mm, melt segregation produces distinct radial structure and chemical stratification. This stratification is most pronounced for bodies formed at around 1 Myr after formation of Ca,Al-rich inclusions. These findings suggest a link between the time and orbital location of planetesimal formation and their subsequent structural and chemical evolution. According to our models, the evolution of partially molten planetesimal interiors falls into two categories. In the magma ocean scenario, the whole interior of a planetesimal experiences nearly complete melting, which would result in turbulent convection and core-mantle differentiation by the rainfall mechanism. In the magma sill scenario, segregating melts gradually deplete the deep interior of the radiogenic heat source. In this case, magma may form melt-rich layers beneath a cool and stable lid, while core formation would proceed by percolation. Our findings suggest that grain sizes prevalent during the internal heating stage governed magma ascent in planetesimals. Regardless of whether evolution progresses toward a magma ocean or magma sill structure, our models predict that temperature inversions due to rapid Al-26 redistribution are limited to bodies formed earlier than $\approx$ 1 Myr after CAIs. We find that if grain size was smaller than $\approx$ 1 mm during peak internal melting, only elevated solid-melt density contrasts (such as found for enstatite chondrite compositions) would allow substantial melt segregation to occur., 15 pages, 6 figures, 1 table; associated video file at http://goo.gl/V1UK1i; associated blog post at http://bit.ly/2GZZ0e9

Recently detected converted phases in upwelling mantle have been attributed to seismic velocity gradients at the onset of melting. In this study, we investigate conditions required to generate a melting onset phase by combining melt migration models with synthetic receiver functions. We find that increasing upwelling velocity, increasing mantle viscosity, and decreasing water content in the mantle source increases the predicted strength of a melting onset phase. Differences in these parameters between study regions may explain observed variations in converted phase magnitude. For a wet mantle source, the amplitude of synthetic receiver functions, calculated using a standard relationship for dependence of seismic velocity on melt fraction, is lower than observed amplitudes. One possibility is that the observed receiver functions indicate a heterogeneous mantle source in which wet and dry components melt independently. Alternatively, the dependence of seismic velocity on melt fraction at very low melt fractions may be much stronger than that used here, as suggested in recent studies.

[1] We conducted geophysical surveys, including bathymetry, gravity, and magnetism, within a first-order segment of the Southwest Indian Ridge (SWIR) between the Prince Edward and Eric Simpson fracture zones (FZs) (latitude 35°–40°E, segment PE), in the vicinity of the Marion hotspot. Segment PE includes four orthogonally spreading second-order segments (PE-1, PE-2, PE-3, and PE-4) and a long, oblique axial valley (NTD-1). Segments PE-1, PE-2, and PE-4 are magmatic, whereas segment PE-3 and NTD-1 are characterized by low magmatic activity. Segment PE-3 is a nascent segment and NTD-1 contains three tiny magmatic sections. Each low-magmatic interval along the axis of segment PE lies between two magmatic segments. This segmentation pattern is similar to the SWIR between the Gallieni and Melville FZs; therefore, a strong melt-focusing process can be expected. Different characteristics of second-order magmatic segments suggest that the magmatic activity in each segment varies among each other as well as that of the other segments of SWIR. Continuous seafloor morphology and isochrons over off-axis areas of segment PE-1 and NTD-1 suggest that PE-1 shortened after the C2An chron. The V-shaped bathymetric structure between segment PE-1 and NTD-1 suggests that the melt supply center has migrated westward. This westward melt migration would have reduced magmatic activity at NTD-1 after C2An. Ridge obliquity may also have reduced magmatic activity. Geophysical characteristics of second-order segments suggest that magmatic activity of segment PE is mainly controlled by a strong melt-focusing process and a comparatively low contribution of melt supply from Marion hotspot.

We present rheological data on samples of partially molten mantle rocks deformed in torsion at 1200 °C and 300 MPa in a gas-medium deformation apparatus. Samples in which stress-driven melt segregation occurs are significantly weaker than those with a homogeneous melt distribution, with strain rate increased by up to a factor of 6. The degree of melt segregation, which is influenced by the permeability of the sample material (and hence the compaction length), is controlled by changing chromite fraction as a minor solid second phase. In general, a higher degree of melt segregation causes a greater strain rate enhancement. Although we focus here on the steady-state rheological properties, a simple model of the evolution of melt segregation and rheological properties is presented. This model fits well the latter part of the stress–strain evolution but fails to fit the early, transient stages of deformation. One implication is that weakening occurs very quickly due to the emergence of smaller-scale melt bands. The model (constitutive and evolution equations) is designed to be incorporated into geodynamic-scale continuum models to explore the effects of segregation without resolving the length scales of the process directly.

To better understand the process of geofluid (melt, water) migration in the earth’s mantle, we have investigated the mathematical formulation of the two-phase theory for compaction, which implicitly accounts for a porosity dependent effective (bulk) viscosity of the two-phase mixture. Following previous studies, we have searched for solitary waves solution of permanent shape. Our results confirm analytically the existence of such solitary waves. To highlight their characteristics, a comparison of the properties of these waves with the waves produced in a two-phase mixture with constant effective (bulk) viscosity is presented. In particular, the solitary waves that are observed in porosity dependent effective viscosity simulations are steeper and their speed decreases as a function of one over the background porosity. These characteristics may have important implications for the interpretation of the geochemistry and morphology of melt-related processes. Additionally, our analytical solutions can be used for numerical benchmarking of two-phase codes.

Knowledge of melt transfer within the mantle is primarily from the study of mid-ocean ridge settings, leaving elusive the process of sub-arc melt transfer. The sub-arc mantle section of the Sapat Complex (Kohistan-Pakistan) exposes coarse-grained dunites that contain clinopyroxene-enriched zones, which, in turn, contain gabbroic lenses. Structural, petrological, and chemical relationships indicate that the clinopyroxene-enriched zones reflect a continuum of melt transport mechanisms between pervasive percolation and fully segregated melt flow. This spectrum of mechanisms and the petrographic features at Sapat are explained by mechanical flow instabilities (porosity waves) that cause channeling of pervasively distributed melt.

Our ability to monitor volcanoes (using seismic signals, ground deformation, gas fluxes, or other ground and satellite based observations) as well as our understanding of melt reservoirs that feed eruptions have evolved tremendously in recent years. The complex plumbing systems that are thought to feed eruptions are, however, difficult to relate to the monitoring signals. Here we show that the record preserved in compositional zoning of erupted minerals may be used to reconstruct sections of the plumbing system. Kinetic modeling of such zoning can yield information on the residence time of magma in different segments of the plumbing systems. This allows a more nuanced evaluation of the link between observed monitoring signals or eruption styles and the magmatic processes and movement of batches of melts at depth. The approach is illustrated through a study of the compositional zoning recorded in olivine crystals from the 1991–1993 SE-flank eruption products of Mt. Etna (Sicily). The zoning patterns in crystals reveal that the plumbing system of the volcano consisted of at least three different magmatic environments between which magma was transported and mixed in the year or two preceding the start of eruption. Quantification of this history indicates that two main pathways of melt migration and three timescales dominated the dynamics of the system. Combination of this information with the timing of observation of various monitoring signals allows a reconstruction of the dynamic evolution of this section of the plumbing system during the early stages of the 1991–1993 eruption. It is seen, for example, how the migration of melt through the same sections of the plumbing system can cause pre-eruptive triggering, enhance Strombolian activity, and through the ensuing eruption cleanse and flush the plumbing system. Different kinds of mixing occur simultaneously at different sections of the plumbing system on different timescales (a few days up to two years).

Coupled thermal–orbital histories of early lunar evolution are considered in a simple model. We consider a plagioclase lid, overlying a magma ocean, overlying a solid mantle. Tidal dissipation occurs in the plagioclase lid and heat transport is by conduction and melt migration. We find that large orbital eccentricities can be obtained in this model. We discuss possible consequences of this phase of large eccentricities for the shape of the Moon and geochronology of lunar samples. We find that the orbit can pass through the shape solution of Garrick-Bethell et al. (Garrick-Bethell, I., Wisdom, J., Zuber, M. [2006]. Science 313, 652), but we argue that the shape cannot be maintained against elastic deformation as the orbit continues to evolve.

Melting experiments on a gneissic protolith using the diamond trap technique produced an accumulation of crystals of biotite in the boundary of the diamond trap. Several hypotheses can be inferred to explain this feature, but textural evidence and calculations of the settling and drag velocities of the crystals suggest a model of extraction of melt due to a pressure gradient inside the capsule as the most likely origin of this biotite layer. This biotite layer and the trapped melt resemble the melanosome and leucosome, respectively, of stromatic metatexite migmatites. The melt extraction model developed is similar to the “filter pressing” model for migmatite formation. The diamond trap technique is a suitable method to develop migmatites experimentally.

This paper presents a model for the genesis of migmatites and granites during orogenic evolution based on the analysis of several Phanerozoic crustal segments and a review of the physical properties of partially-molten rocks and magmas. This model inventories the modes of bulk flow of partially-molten rocks and magmas and the mechanisms of melt/solid segregation for each mode. Partial melting of rocks is associated with a strength decrease of two to three orders of magnitude leading to strain partitioning expressed by channel flow driven either by forces related to plate tectonics (vertical channel flow) or by the gravity force associated with lateral variations of crustal thicknesses (horizontal channel flow). Although the structural characteristics of migmatites indicate that deformation plays a role in melt migration, the emplacement of laccoliths of leucogranites above migmatites attests to the efficiency of the buoyancy force. Another decrease in apparent strength of about ten orders of magnitude is associated with the loss of continuity of the solid framework marking the transition from partially-molten rocks to magmas. The breakdown of the solid framework also allows for settling of the solid in suspension increasing the buoyancy of the remaining magma. Accordingly, domes cored by diatexites (former heterogeneous magmas) and mantled by metatexites (former partially-molten rocks) are interpreted as gravitational instabilities driven by the relative buoyancy of the magma and permitted by the weakness of the partially-molten rocks. This model provides a first order framework to elucidate the development of a crustal-scale horizontal layering during the evolution of orogenic belts that are affected by partial melting. In this case, the middle crust is dominated by migmatites with domes cored by diatexites and mantled by metatexites that correspond to a partially-molten and magmatic zone, respectively. The granitic dikes and sills and the associated laccoliths of leucogranites correspond to an intrusive zone overlying the partially-molten zone. The refractory lower crust is potentially in part formed by accumulation of solids segregated from the overlying heterogeneous magmas.

The state of the art of the problem of mantle melt transport through the upper mantle beneath spreading zones is reviewed. The geochemical data and physical properties of the mantle constraining the melt migration processes are considered. The review concentrates on the ways of mantle magma transport and mechanisms of the localization of diffuse intergranular porous melt percolation into the channel flow. The hypothesis of the reaction of migrating mantle magmas with wall peridotites and the formation of replacive dunites as a result of this process are discussed. We examine the publications on the field, theoretical, and geochemical studies of the reactive melt migration and its role in the dunite formation in the mantle sections of the ophiolites. Alternative models of the dunite origin in the mantle are also tested. The results of experimental studies of the melt/peridotite interaction are presented.

The amount of magmatism that accompanies the extension and rupture of the continental lithosphere varies dramatically at rifts and margins around the world. Based on widely spaced geophysical transects, some margins are known to preserve a transition from magmatically robust to magmatically starved rifting along strike, but the nature of the transition is unknown. Wide-angle seismic data from the Black Sea provide the first direct observations of such a transition and show that it is abrupt, occurring over only ~20–30 km, and coincides with a transform fault. This abrupt transition cannot be explained solely by gradual along-margin variations in mantle properties, since these would be expected to result in a smooth transition from magma-poor to magma-rich rifting over hundreds of kilometers. We suggest that the abruptness of the transition results from the development of three-dimensional (3-D) melt migration due to along-strike variations in extension and thus the thickness of the lithosphere at the time of rifting. Localized magmatic addition attributed to melt focusing has been observed in modern mid-ocean ridges and active rift environments, but here we show that such processes can also produce abrupt along-strike changes from magma-poor to magma-rich rifting.

In the upper part of the Earth there exists a region where the rock is partially molten. The magmatic melt rises through the partially molten region as a form of permeable flow. The timescale and efficiency of the melt migration are controlled by the permeability, which is a function of the melt fraction and the distribution of the melt at the grain scale. Using gel as an analog of the solid phase of the partially molten system, the permeability and the electrical conductivity were measured. With deformation the gel deforms into a polyhedron due to its high deformability, and the fluid phase exists at edge and corner regions. The morphology of the mixture system is similar to that of the partially molten system controlled by the interfacial energy. Both of the permeability and the electrical conductivity obey power-law functions of the liquid fraction. The permeability is close to Kozeny–Carman type permeability and the electrical conductivity obeys Archie's law at higher liquid fraction (about 10% to 30%). They reduce drastically with reduction of the liquid fraction at lower liquid fraction (about 2% to 10%). These behaviors are considered to result from dispersion of pseudo-dihedral angle. The permeability in the upper mantle would behave like that of the gel–fluid system because the dispersion of dihedral angle exists due to poly-mineralic solid phase of the upper mantle and anisotropic interfacial energy. Also, the permeability is found to be proportional to the square of the electrical conductivity. This relation is useful to estimate the permeability at depth from the electrical conductivity measurement.

(MREE) depletion in clinopyroxenes in the harzburgite suggests that the harzburgites are residues of two-stage fractional melting, which operated initially in the garnet field, and subsequently continued in the spinel lherzolite field. The early stage melting produced the depleted harzburgite. The later stage melting was responsible for the gabbroic rocks and dunite. Strongly LREE^ MREE-depleted clinopyroxene in the lower harzburgite and HREE-enriched clinopyroxene in the upper harzburgite and plagioclase dunite were formed by later reactive melt migration occurring in the harzburgite.

Summary Using a two-phase flow model, we investigate the formation of a high-percentage melt layer beneath the oceanic lithosphere and focusing of the melt towards the ridge axis, taking into account freezing and melt extraction. Melt migration is modelled dynamically within a viscous permeable media that includes the asthenosphere and viscous part of the lithosphere. Due to a much faster melt migration in the brittle part of the lithosphere, the melt migration is simulated by instantaneous melt extraction from an assigned melt extraction region beneath the ridge axis. It is shown that a high-percentage melt layer forms and successfully focuses melt to a narrow zone beneath the mid-ocean ridge. Performance of the melt focusing mechanism is not significantly sensitive to the size of the melt extraction region, the melt extraction threshold and the spreading rate. In all of our models, about half of the total melt production freezes beneath the base of the lithosphere and the rest is focused towards the ridge and forms the crust.

[1] Throughout the global mid-ocean ridge system, transform faults offset spreading centers. Conductive cooling may be more efficient beneath transform faults, producing a thickened lithosphere that directs melt away from the transform. However, recent observations of thickened crust along transform faults at fast ridges suggest melt redistribution toward transforms, intra-crustal melt production, or efficient extraction of melt. We apply a 3-D model of melt migration and extraction along an oceanic transform domain bounded by ridge segments. Melt is assumed to travel vertically before collecting and migrating beneath a low-permeability boundary inclined towards the ridge axis. A melt extraction zone, which may be geologically interpreted as the presence of faults and/or dikes leading to rapid lateral and vertical melt migration toward plate boundaries, affects the pattern of crustal accretion at segmented ridges. First, we examine a generic ridge-transform-ridge geometry and then a model that represents the Siqueiros transform on the East Pacific Rise. On the basis of crustal thickness variations within the intra-transform spreading centers along the fast-slipping Siqueiros fault, we constrain the presence of a melt extraction zone within 10 km of the transform zone.

Lower crustal earthquakes (12–25 km depth) have been detected since August 2005 in the Askja volcanic system along the north Iceland rift, in the normally ductile part of the crust. The earthquakes occur in three clusters, which have stable dimensions and locations through time and are interpreted as positions of repeated melt supply from the mantle to the lower crust. Seismic velocity Vp/Vs ratios are consistent with the presence of partial melt in the lower crust at Askja. The spatial separation of the clusters shows that there are multiple positions of melt injection within this one magmatic segment and all three positions are currently active. This pattern of melt supply is more like that observed on fast spreading ridges than slow spreading ridges and is probably a consequence of the increased melt production beneath Iceland compared to the rest of the Mid-Atlantic Ridge. However, the relative number of earthquakes in each cluster shows that two thirds of the melt is supplied to the central volcano Askja (i.e. segment center). During the last major rifting episode shallow lateral melt migration occurred from the magma chamber beneath the volcano. Therefore on long time scales melt supply is probably greater at the segment center, with melt redistribution in the upper crust, even though there are multiple points of lower crustal injection along the segment.

Melting of upwelling mantle beneath spreading centers generates the basaltic crust that covers a large fraction of the Earth's surface. Crustal emplacement of mantle melt and faulting associated with lithosphere extension in a narrow, few kilometer wide region along the axis of spreading centers largely defines the structural and morphological fabric of the entire seafloor. Given this important role in shaping the earth, it is not surprising that spreading centers have been a major focus of earth science research over the past four years. The papers that have appeared even during this short time are too numerous to review completely in this report. Fortunately, following the previous report of Phipps Morgan (1991), several papers (Forsyth, 1992; Solomon and Toomey, 1992) provide a good review of many important aspects of the geophysical characteristics of spreading centers. Consequently this report will focus on recent developments in understanding the mechanisms and processes that shape spreading centers. An extensive (but not exhaustive) bibliography of papers that have appeared in the last four years is included, but not all of these can be discussed in the text.

A partial melt undergoing large scale deformation is shown to be unstable with respect to small scale redistribution of melt, provided the shear viscosity of the matrix depends on melt fraction. In the physically realistic case where melt “softens” the matrix, melt migrates along the direction parallel to the axis of minimum compressive stress and accumulates in “veins” (melt-rich lenses). The maximum growth rate of the instability is of order —⋵_od ln η/df (⋵_o ≡ largest component of the principal strain tensor, η = shear viscosity, f = melt fraction), and plausibly fast enough to invalidate the conventional applications of Darcy's law (e.g., mid-ocean ridges, subduction zones) since melt migrates into veins in preference to being pervasively flushed vertically. The preferred lengthscale of the instability is poorly determined but probably ∼ meters. Veins may eventually form an interconnected drainage network, allowing rapid vertical flushing of melt.

Selective enrichment or depletion in either Zr and Hf (HFSE) INTRODUCTION or Nb and Ta (HFSE) is a feature commonly observed in many Because of their similar crystal-chemical properties (i.e. mantle-derived melts and amphiboles occurring as either disseminated high charge and ionic radius), high field strength elements minerals in mantle xenoliths and peridotite massifs or in vein (HFSE: Ti, Zr, Hf, Nb and Ta) are considered to behave assemblages cutting these rocks. The fractionation of Nb from Zr as a coherent group of substituents in rock-forming silicate seen in natural mantle amphiboles suggests that their incorporation minerals, and thus are not expected to fractionate greatly is governed by different crystal-chemical mechanisms. An extensive during partial melting and crystallization processes (Sun set of new partitioning experiments between pargasite–kaersutite and & McDonough, 1989). Nevertheless, fractionation of melt under upper-mantle conditions shows that HFSE incorporation HFSE from HFSE is a common feature in many and fractionation depends on amphibole major-element composition primitive mantle melts such as island-arc volcanics and and the presence or absence of dehydrogenation. Multiple regression intra-plate alkaline melts, suggesting either the presence analysis shows that DNb/Zr is strongly dependent on the mgof a solid phase with DNb,Ta significantly different from number of the amphibole as a result of a combination of amphibole DZr,Hf, or of two distinct phases controlling these element and melt structure effects, so that the following generalizations pairs. Experimental investigations of HFSE decoupling apply: (1) high-mg-number amphiboles crystallized from unmodified have been so far focused on Ti minerals (rutile and mantle melts more easily incorporate Zr relative to Nb leading to ilmenite), because of their high compatibility for HFSE an increase of the Nb/Zr ratio in the residual melt; (2) low-mgand their consequent capability of altering HFSE signumber amphiboles, such as those found in veins cutting peridotites, natures of equilibrium melts. Rutile has been shown to may strongly deplete the residual melt in Nb and cause very low have partition coefficients for Ta and Nb exceeding that Nb/Zr in residual melts. Implications and applications to mantle for Zr by one to two orders of magnitude, with Zr a environments are discussed. further one to two orders higher than other trace elements ( Jenner et al., 1993; Foley et al., 2000). Rutile could thus cause significant decoupling between Nb and Zr, and has the potential to produce the pattern typical of islandarc basalts (Foley & Wheller, 1990; Foley et al., 2000), in