Lithosphere formation in the central Slave Craton (Canada): plume subcretion or lithosphere accretion?

Major-element compositions of minerals in peridotite xenoliths from the Lac de Gras kimberlites provide constraints on the mode of lithosphere formation beneath the central Slave Craton, Canada. Magnesia contents of reconstructed whole rocks correlate positively with NiO and negatively with CaO contents, consistent with variable partial melt extraction. Alumina and Cr₂O₃ contents are broadly positively correlated, suggestive of melt depletion in the absence of a Cr–Al phase. Garnet modes are high at a given Al₂O₃ content (a proxy for melt depletion), falling about a 7 GPa melt depletion model. These observations, combined with high olivine Mg# and major-element relationships of FeO-poor peridotites (<7.5 wt%) indicative of melt loss at pressures >3 GPa (residual FeO content being a sensitive indicator of melt extraction pressure), and similar high pressures of last equilibration (∼4.2 to 5.8 GPa), provide multiple lines of evidence that the mantle beneath the central Slave Craton has originated as a residue from high-pressure melting, possibly during plume subcretion. Apparent low melt depletion pressures for high-FeO peridotites (>7.5 wt%) could suggest formation in an oceanic setting, followed by subduction to their depth of entrainment. However, these rocks, which are characterised by low SiO₂ contents (<43 wt%), are more likely to be the result of post-melting FeO-addition, leading to spuriously low estimates of melt extraction pressures. They may have reacted with a silica-undersaturated melt that dissolved orthopyroxene, or experienced olivine injection by crystallising melts. A secular FeO-enrichment of parts of the deep mantle lithosphere is supported by lower average Mg# in xenolithic olivine (91.7) compared to olivine inclusions in diamond (92.6).     

Geochemical constraints on the origin of the late Jurassic proto-Caribbean oceanic crust in Hispaniola

The nature of the oceanic crust produced through rifting and oceanic spreading between North and South America during the Late Jurassic is a key element for the Caribbean plate tectonic model reconstruction. Located in the Cordillera Central of Hispaniola, the Loma La Monja volcano-plutonic assemblage (LMA) is composed of gabbros, dolerites, basalts, and oceanic sediments, as well as metamorphic equivalents, which represent a dismembered fragment of this proto-Caribbean oceanic crust. Petrologic and geochemical data show that the LMA have a relatively broad diversity in composition, which represent the crystallization products of a typical low-pressure tholeiitic fractionation of mid-ocean ridge basalts (MORB)-type parental magmas, ranging from N- to E-MORB. Three geochemical groups have been distinguished in the volcanic sequence: LREE-flat to slightly LREE-enriched basalts of groups II and III occur interlayered in the lower stratigraphic levels; and LREE-depleted basalts of group I in the upper levels. Mantle melt modeling suggests that group III magmas are consistent by mixing within a mantle melt column of low-degree (<1%) melts of a deep garnet lherzolite source and high-degree (>15%) melts of a shallow spinel source, and groups II and I magmas are explained with moderate to high (14–18%) and very high (>20%) fractional melting degrees of a shallower spinel mantle source, respectively. Thus, upward in the volcanic sequence of the LMA, the magmas represent progressively more extensive melting of shallower sources, in a plume-influenced spreading ridge of the proto-Caribbean oceanic crust. Nb/Y versus Zr/Y systematics combined with recent plate tectonic model reconstructions reveal that Caribbean Colombian oceanic plateau fragments in Hispaniola formed through melting of heterogeneous mantle source regions related with distinct plumes during at least from Aptian–Albian (>96 Ma) to Late Campanian.     

峨眉山大火成岩省：地幔柱活动的证据及其熔融条件

对苦橄岩中橄榄石斑晶及其中熔体包裹体的电子探针分析表明，峨眉山大火山岩省的原始岩浆具高镁（ MgO > 16％）特征。玄武岩的 REE反演计算揭示，参与峨眉山玄武岩岩浆作用的地幔具有异常高的潜能温度（ 1 550℃）。这些特征以及峨眉山玄武岩的大面积分布和一些熔岩所显示的类似于洋岛玄武岩 (OIB)的微量元素和 Sr- Nd同位素特征均为地幔热柱在能量和物质上参与峨眉山溢流玄武岩的形成提供了确凿证据。峨眉山两个主要岩类（高钛和低钛玄武岩）可能是不同地幔源区物质在不同条件下的熔融产物。低钛玄武岩形成于温度最高、岩石圈最薄的地幔柱轴部。地幔（ ISr≈ 0.705,ε Nd(t)≈＋ 2）熔融始于 140 km，并一直延续到较浅的深度（ 60 km,尖晶石稳定区 ),部分熔融程度为 16％，这类岩石可能代表了峨眉山玄武岩的主体。而高钛玄武岩的母岩浆的形成基本局限在石榴子石稳定区（ > 70 km），其源区特征为 : ISr≈ 0.704,ε Nd(t)≈＋ 5，可能代表了热柱边部或消亡期地幔小程度部分熔融（ 1.5％）的产物。     

The effect of the large-scale mantle flow field on the Iceland hotspot track

Fluid dynamical simulations were carried out in order to investigate the effect of the large-scale mantle flow field and the depth of the plume source on the structure of the Iceland plume through time. The time-dependent location and shape of the plume in the Earth's mantle was calculated in a global model and it was refined in the upper mantle using a 3D Cartesian model box. Global flow was computed based on density heterogeneities derived from seismic tomography. Plate motion history served as a velocity boundary condition in both models. Hotspot tracks of the plume conduits and the plume head were calculated and compared to actual bathymetry of the North Atlantic. If a plume source in the lowermost mantle is assumed, the calculated surface position of the plume conduit has a southward component of motion due to southward flow in the lower mantle. Depending on tomography model, assumed plume age and buoyancy the southward component is more or less dominating. Plume models having a source at the 660 km discontinuity are only influenced by flow in the upper mantle and transition zone and hence rather yield westward hotspot motion. Many whole-mantle plume models result in a V-shaped track, which does not match the straight Greenland–Iceland–Faroe ridge. Models without strong southward motion, such as for a plume source at 660 km depth, match actual bathymetry better. Plume tracks were calculated from both plume conduits and plume heads. A plume head of 120 K anomalous temperature gives the best match between plume head track and bathymetry.     

Petrological geodynamics of mantle melting III. AlphaMELTS + multiphase flow: The effect of water

The influence of water is evaluated in this last contribution of a series aiming to study the petrological and dynamic evolution of mantle melting. Water is considered to be either a chemical component in the melt or solid assemblage but it can also be present as a pure water phase in a oversaturated environment. A three-phase-flow model was developed for this purpose.Only a limited set of conditions has been applied to the 1-D upwelling mantle column. A range of fixed temperatures (1150–1450 °C) and water contents in the solid mantle (0, 0.02 wt.%, 0.2 ​wt.%) have been imposed at the entry point (120 ​km deep) for the two melting models introduced in the previous installments, dynamic equilibrium melting (DEM) and dynamic fractional melting (DFM) model.As expected, for a given temperature at the base of the mantle column, the depth of the first melt formation increases with higher water content in the mantle. After the first melt is created, very negligible amount of melt is formed over a certain depth interval which approximately ends at the depth where the first melting of the dry mantle would take place. However melt is present as a dynamic phase thorough the entire region regardless whether the DEM or DFM model has been applied.Under a quasi-steady state regime, the melt and residual mantle compositions vary significantly over depth, depending on the conditions imposed to the model (DEM, DFM, bottom temperature and water content). Several distinctions can be made at the extraction point (top of the mantle column ​= ​15 ​km deep). For DEM and DFM models at this lowest depth, the most influential factor affecting the melt composition after the quasi-steady state condition has been reached is the temperature at the base of the column. In general, for a high temperature model, the input water in the mantle does not seem to play a significant role on the bulk composition of the melt (except for the water content in melt). But at low temperature water does have some noticeable influence on the variation of some chemical components in melt ($Si{O}_2$, $F{e}_2{O}_3$, $CaO$, $N{a}_{2}O$ at T ​= ​1250 °C or lower). A similar conclusion can be made also for the residual mantle composition. The presence of a dynamic free water phase is detected only in absence of melt or in coexistence with a melt phase when the mantle is relatively cold (bottom temperature $\le$1250 °C) and the input water content at the base of the model is relatively high (0.2 ​wt.%).Complete output data for several numerical simulations and six animations illustrating various melting models are available following the instructions in the supplementary material.     

南大西洋慢速洋中脊与地幔柱相互作用研究进展与展望

揭示洋中脊与地幔柱(脊-柱)之间的可能联系为认识地球深部物质组成与深部地幔动力学过程提供了重要窗口,也是过去40多年以来固体地球科学研究领域的前沿与热点。在绵延八万多千米的全球洋中脊系统中,部分洋脊片段会受到地幔柱作用不同程度的影响。研究显示,大西洋的形成演化与地幔柱作用之间具有密切联系,尤其在南大西洋的裂解、打开演化过程中,南大西洋中脊系统始终与其周围地幔柱(如圣赫勒拿、阿森松、特里斯坦、高夫、发现等地幔柱)之间具有不同程度的相互作用关系,导致沿脊出露玄武岩在地球化学组成上呈现出明显的不均一性特征。本文在系统性总结脊-柱相互作用研究现状与南大西洋地区地质构造演化特征的基础上,详细阐述了南大西洋中脊13.2°S～24.2°S地区玄武岩的岩石地球化学特征;揭示了南大西洋中脊研究区的岩浆演化、地幔源区性质;指示出圣赫勒拿地幔柱物质向南大西洋中脊系统传播的主要方向;圈定了圣赫勒拿地幔柱对南大西洋中脊系统地幔源区性质在沿脊方向的影响范围(14.2°S～20.4°S);同时推测了南大西洋中脊系统与圣赫勒拿地幔柱之间受地幔柱影响的软流圈地幔物质在大洋岩石圈底部的空间展布。最后本文提出了关于南大西洋...     

沙茨基海隆(ShatskyRise)白垩纪玄武岩的地球化学特征及其地幔柱-洋中脊相互作用过程

沙茨基海隆(Shatsky Rise)是白垩纪早期形成的西北太平洋大火成岩省, 其成因和演化过程目前仍存在较大争议。本次研究对沙茨基海隆白垩纪玄武岩进行了全岩主量、微量元素、Sr-Nd-Pb同位素的分析。沙茨基海隆玄武岩主要属于拉斑玄武岩, 具有较亏损的大离子亲石元素和轻稀土元素以及较富集的重稀土元素的特征, 没有明显的Eu异常(δEu=0.99~1.29), 与正常洋中脊玄武岩(N-MORB)的微量元素配分模式较为相似。然而该系列玄武岩却具有相对较富集的初始⁸⁷Sr/⁸⁶Sr(0.702986~0.703991)和¹⁴³Nd/¹⁴⁴Nd(0.513034~0.513194)同位素比值、较富集的²⁰⁷Pb/²⁰⁴Pb(15.439~15.508)和²⁰⁸Pb/²⁰⁴Pb(37.853~38.488)同位素比值, 与富集的洋岛玄武岩(OIB)和岛弧火山岩的同位素成分较为相似, 且源区混入高U/Pb比值(HIMU型)的富集地幔成分。稀土元素部分熔融模拟反演表明沙茨基海隆火山岩的原始岩浆可能起源于尖晶石相二辉橄榄岩源区, 且具有较高程度的部分熔融作用(>10%)。在以上研究基础上, 本文提出地幔柱-洋中脊相互作用模型来解释沙茨基海隆拉斑玄武岩较亏损的不相容元素成分和较富集的同位素成分这一特殊地球化学特征。由于来自扩张洋中脊的强大拉张应力的影响, 地幔柱岩浆物质将流向洋中脊并发生减压部分熔融, 导致不相容元素的高度亏损, 但由于放射成因元素(Sm、Rb和U)的半衰期相对较长, 同位素成分则难以在较短时间内被改变, 因此本文推测沙茨基海隆同位素富集的N-MORB拉斑玄武岩可能是地幔柱-洋中脊相互作用的产物。

     

Chalcophile element partitioning between sulfide phases and hydrous mantle melt: Applications to mantle melting and the formation of ore deposits

Understanding the geochemical behavior of chalcophile elements in magmatic processes is hindered by the limited partition coefficients between sulfide phases and silicate melt, in particular at conditions relevant to partial melting of the hydrated, metasomatized upper mantle. In this study, the partitioning of elements Co, Ni, Cu, Zn, As, Mo, Ag, and Pb between sulfide liquid, monosulfide solid solution (MSS), and hydrous mantle melt has been investigated at 1200 °C/1.5 GPa and oxygen fugacity ranging from FMQ−2 to FMQ+1 in a piston-cylinder apparatus. The determined partition coefficients between sulfide liquid and hydrous mantle melt are: 750–1500 for Cu; 600–1200 for Ni; 35–42 for Co; 35–53 for Pb; and 1–2 for Zn, As, and Mo. The partition coefficients between MSS and hydrous mantle melt are: 380–500 for Cu; 520–750 for Ni; ∼50 for Co; <0.5 for Zn; 0.3–6 for Pb; 0.1–2 for As; 1–2 for Mo; and >34 for Ag. The variation of the data is primarily due to differences in oxygen fugacity. These partitioning data in conjunction with previous data are applied to partial melting of the upper mantle and the formation of magmatic-hydrothermal Cu–Au deposits and magmatic sulfide deposits.I show that the metasomatized arc mantle may no longer contain sulfide after >10–14% melt extraction but is still capable of producing the Cu concentrations in the primitive arc basalts, and that the comparable Cu concentrations in primitive arc basalts and in MORB do not necessarily imply similar oxidation states in their source regions.Previous models proposed for producing Cu- and/or Au-rich magmas have been reassessed, with the conclusions summarized as follows. (1) Partial melting of the oxidized (fO₂ > FMQ), metasomatized arc mantle with sulfide exhaustion at degrees >10–14% may not generate Cu-rich, primitive arc basalts. (2) Partial melting of sulfide-bearing cumulates in the root of thickened lower continental crust or lithospheric mantle does not typically generate Cu- and/or Au-rich magmas, but they do have equivalent potential as normal arc magmas in forming magmatic-hydrothermal Cu–Au deposits in terms of their Cu–Au contents. (3) It is not clear whether partial melting of subducting metabasalts generates Cu-rich adakitic magmas, however adakitic magmas may extract Cu and Au via interaction with mantle peridotite. Furthermore, partial melting of sulfide-bearing cumulates in the deep oceanic crust may be able to generate Cu- and Au-rich magmas. (4) The stabilization of MSS during partial melting may explain the genetic link between Au-Cu mineralization and the metasomatized lithospheric mantle.The chalcophile element tonnage, ratio, and distribution in magmatic sulfide deposits depend on a series of factors. This study reveals that oxygen fugacity also plays an important role in controlling Cu and Ni tonnage and Cu/Ni ratio in magmatic sulfide deposits. Cobalt, Zn, As, Sn, Sb, Mo, Ag, Pb, and Bi concentrations and their ratios in sulfide, due to their different partitioning behavior between sulfide liquid and MSS, can be useful indices for the distribution of platinum-group elements and Au in magmatic sulfide deposits.     

Geochemical systematics of the Mauranipur-Babina greenstone belt,Bundelkhand Craton,Central India:Insights on Neoarchean mantle plume-arc accretion and crustal evolution

The Neoarchean Bundelkhand greenstone sequences at Mauranipur and Babina areas within the Bundelkhand Gneissic Complex preserve a variety of magmatic rocks such as komatiitic basalts, basalts,felsic volcanic rocks and high-Mg andesites belonging to the Baragaon, Raspahari and Koti Formations.The intrusive and extrusive komatiitic basalts are characterized by low SiO_2(39-53 wt.%), high MgO(18-25 wt.%).moderately high Fe_2O_3(7.1-11.6 wt.%), Al_2O_3(4.5-12.0 wt.%), and TiO_2(0.4-1.23 wt.%)with super to subchondritic(Gd/Yb)N ratios indicating garnet control on the melts. The intrusive komatiitic suite of Ti-enriched and Al-depleted type possesses predominant negative Eu and positive Nb, Ti and Y anomalies. The chemical composition of basalts classifies them into three types with varying SiO_2, TiO_2, MgO, Fe_2O_3, Al_2O_3 and CaO. At similar SiO_2 content of type Ⅰ and Ⅲ basalts, the type II basalts show slightly high Al_2O_3 and Fe_2O_3 contents. Significant negative anomalies of Nb, Zr, Hf and Ti, slightly enriched LREE with relatively flat HREE and low ∑REE contents are observed in type Ⅰ and Ⅱ basalts. TypeⅢ basalts show high Zr/Nb ratios(9.8-10.4), TiO_2(1.97-2.04 wt.%), but possess strikingly flat Zr, Hf, Y and Yb and are uncontaminated. Andesites from Agar and Koti have high SiO_2(55-64 wt.%), moderate TiO_2(0.4-0.7 wt.%), slightly low Al_2O_3(7-11.9 wt.%), medium to high MgO(3-8 wt.%) and CaO contents(10-17 wt.%). Anomalously high Cr, Co and Ni contents are observed in the Koti rhyolites. Tholeiitic to calc alkaline affinity of mafic-felsic volcanic rocks and basalt-andesite dacite-rhyolite differentiation indicate a mature arc and thickened crust during the advanced stage of the evolution of Neoarchean Bundelkhand greenstone belt in a convergent tectonic setting where the melts were derived from partial melting of thick basaltic crust metamorphosed to amphibolite-eclogite facies. The trace element systematics suggest the presence of arc-back arc association with varying magnitudes of crust-mantle interaction. La/Sm, La/Ta,Nb/Th, high MgO contents(20 wt.%), CaO/Al_2O_3 and(Gd/Yb)_N 1 along with the positive Nb anomalies of the komatiite basalts reflect a mantle plume source for their origin contaminated by subductionmetasomatized mantle lithosphere. The overall geochemical signatures of the ultramafic-mafic and felsic volcanic rocks endorse the Neoarchean plume-arc accretion tectonics in the Bundelkhand greenstone belt.     

Crustal thermal state and origin of silicic magma in Iceland: the case of Torfajökull,Ljósufjöll and Snæfellsjökull volcanoes

Pleistocene and Holocene peralkaline rhyolites from Torfajökull (South Iceland Volcanic Zone) and Ljósufjöll central volcanoes and trachytes from Snæfellsjökull (Snæfellsnes Volcanic Zone) allow the assessment of the mechanism for silicic magma genesis as a function of geographical location and crustal geothermal gradient. The low δ¹⁸O (2.4‰) and low Sr concentration (12.2 ppm) measured in Torfajökull rhyolites are best explained by partial melting of hydrated metabasaltic crust followed by major fractionation of feldspar. In contrast, very high ⁸⁷Sr/⁸⁶Sr (0.70473) and low Ba (8.7 ppm) and Sr (1.2 ppm) concentrations measured in Ljósufjöll silicic lavas are best explained by fractional crystallisation and subsequent ⁸⁷Rb decay. Snæfellsjökull trachytes are also generated by fractional crystallisation, with less than 10% crustal assimilation, as inferred from their δ¹⁸O. The fact that silicic magmas within, or close to, the rift zone are principally generated by crustal melting whereas those from off-rift zones are better explained by fractional crystallisation clearly illustrates the controlling influence of the thermal state of the crust on silicic magma genesis in Iceland.     

Rayleigh wave tomography in the North Atlantic: high resolution images of the Iceland, Azores and Eifel mantle plumes

Presented in this paper is a high resolution S_v-wave velocity and azimuthal anisotropy model for the upper mantle beneath the North Atlantic and surrounding region derived from the analysis of about 9000 fundamental and higher-mode Rayleigh waveforms. Much of the dataset comes from global and national digital seismic networks, but to improve the path coverage a number of instruments at coastal sites in northwest Europe, Iceland and eastern Greenland was deployed by us and a number of collaborators. The dense path coverage, the wide azimuthal distribution and the substantial higher-mode content of the dataset, as well as the relatively short path-lengths in the dataset have enabled us to build an upper mantle model with a horizontal resolution of a few hundred kilometers extending to 400 km depth. Low upper mantle velocities exist beneath three major hotspots: Iceland, the Azores and Eifel. The best depth resolution in the model occurs in NW Europe and in this area low S_v-velocities in the vicinity of the Eifel hotspot extend to about 400 km depth. Major negative velocity anomalies exist in the North Atlantic upper mantle beneath both Iceland and the Azores hotspots. Both anomalies are, above 200 km depth, 4–7% slow with respect to PREM and elongated along the mid-Atlantic Ridge. Low velocities extend to the south of Iceland beneath the Reykjanes Ridge where other geophysical and geochemical observations indicate the presence of hot plume material. The low velocities also extend beneath the Kolbeinsey Ridge north of Iceland, where there is also supporting geochemical evidence for the presence of hot plume material. The low-velocity upper mantle beneath the Kolbeinsey Ridge may also be associated with a plume beneath Jan Mayen. The anomaly associated with the Azores extends from about 25°N to 45°N along the ridge axis, which is in agreement with the area influenced by the Azores Plume, predicted from geophysical and geochemical observations. Compared to the anomaly associated with Iceland, the Azores anomaly is elongated further along the ridge, is shallower and decays more rapidly with depth. The fast propagation direction of horizontally propagating S_v-waves in the Atlantic south of Iceland correlates well with the east–west ridge-spreading direction at all depths and changes to a direction close to NS in the vicinity of Iceland.     

Historic magmatism on the Reykjanes Peninsula,Iceland: a snap-shot of melt generation at a ridge segment

We present new compositional data on a suite of historic lava flows from the Reykjanes Peninsula, Iceland. They were erupted over a short time period between c. 940 and c. 1340 ad and provide a snap-shot view of melt generation and evolution processes beneath this onshore, 65 km long, ridge segment. The lavas are tholeiitic basalts (MgO 6.5–9.2 wt%) and sparsely (≪5%) olivine and/or plagioclase phyric (±trace clinopyroxene). Individual eruptive events show remarkable compositional homogeneity. Despite a limited variation in Sr–Nd isotope compositions, high-precision double-spike Pb isotope data show tight coherent arrays that, together with correlations with incompatible trace element ratios, indicate control by binary mixing processes. Poor correlations with elemental abundances require that this mixing took place prior to extensive fractional crystallisation. Olivines in the historic lavas have light δ¹⁸O values (+4.2 to +4.3‰), which is likely to be a feature of the enriched mantle source to Reykjanes Peninsula lavas. High precision Pb isotope analyses of other post-glacial Reykjanes Peninsula lavas show significant variability in ²⁰⁷Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb at lower ²⁰⁶Pb/²⁰⁴Pb values than in the historic lavas. This variation demonstrates that at least three compositionally distinct components within the mantle are required to explain the Pb isotope variations within the Reykjanes Peninsula as a whole. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

The Darfur Dome,western Sudan: the product of a subcontinental mantle plume

Field investigations, K-Ar age determinations and chemical data were used to describe the development of an intraplate volcanic province, the Darfur Dome, Sudan. Magmatism started 36 Ma ago at a small subvolcanic complex (Jebel Kussa) in the center of the dome and was active in the same area between 26 and 23 Ma. Two major volcanic fields (Marra Mountains and Tagabo Hills) developed between 16 and 10 Ma. Volcanism started again at 6.8 Ma with a third volcanic field (Meidob Hills) and at 4.3 Ma in the Marra Mountains and with the reactivation of the center. Activity then continued until the late Quaternary. Having started in the center of the Darfur Dome, volcanism moved in 36 Ma 200 km towards the NNE and 100 km SSW No essential difference in the alkaline magma types (basanitic to phonolitic-trachytic, with different amounts of assimilation of crustal material) in the different fields, was observed. Magmatism is thought to have been produced by a rising mantle plume and volcanism was triggered by stress resolution along the Central African Fault Zone.     

Evidence of diverse depletion and metasomatic events in harzburgite–lherzolite mantle xenoliths from the Iberian plate (Olot, NE Spain): Implications for lithosphere accretionary processes

Mantle xenoliths from the Olot volcanic district (NE Spain) comprise a bi-modal suite consisting of protogranular spinel lherzolites (cpx 12–14%) sometimes with pargasitic amphibole, and highly refractory spinel harzburgites (cpx ≤ 1%) with coarse-grained granular textures. The lherzolites range from slightly depleted to moderately LREE-enriched with flat HREE patterns between 1.5 and 2.7 × chondrite (Ch). In contrast, the harzburgites are extremely depleted in HREE (down to 0.2 × Ch) and strongly LREE-enriched (La_N/Yb_N = 12.3–17.2). LA-ICP-MS analyses of clinopyroxene and amphibole of the lherzolites highlight variable degrees of LREE depletion (HREE up to 13 × Ch, La_N/Yb_N down to 0.01), with the exception of a single sample in which both clinopyroxene and amphibole are LREE-enriched (La_N/Yb_N up to 19). In the harzburgites, clinopyroxenes display totally different REE distributions, characterized by extreme HREE depletion (down to 0.4 × Ch) and upward convex positively fractionated middle-light REE patterns (Nd_N/Yb_N up to 20.7 × Ch; La_N/Yb_N up to 12 × Ch). Sr–Nd–Hf isotopic data for both whole-rocks and cpx separates, coherently indicate depleted mantle (DM) compositions for the lherzolites (εSr = − 15 to − 26, εNd = + 9 to + 17, εHf = + 18 to + 68) and enriched mantle (EM) compositions for the harzburgites (εSr = − 10 to + 36, εNd = − 1 to − 6, εHf = + 3 to + 8). Modelling of the clinopyroxene REE data and isotopic systematics suggest that some lherzolites were affected by pre-Paleozoic (0.6–1 By) low-degree partial melting processes, while others probably reflect some extent of refertilization of the mantle protolith by metasomatizing melts similar to the Triassic rift-related tholeiites reported from several Pyrenean localities. The harzburgites represent extreme refractory residua, resulting from a complex depletion history due to multistage melt extraction as often observed in the cratonic mantle. The distinctive REE patterns and isotopic systematics of their clinopyroxenes suggest that the harzburgites were formed by the interaction of an ultra-depleted peridotite matrix with highly alkaline basic melts similar in composition to the Permo-Triassic alkaline lamprophyres which are widespread within the Iberian plate. Lherzolites possibly represent younger lithosphere (accreted asthenosphere?) up-lifted and juxtaposed to the older subcontinental lithospheric mantle (harzburgites) during the post-Variscan rifting of the Iberian margin. These two genetically different, but adjoining, mantle domains intimately mingled along the northern Iberian margin during the subsequent plate convergence processes, leading to the close association of harzburgites and lherzolites observed in the Olot mantle xenoliths and in some Pyrenean peridotite massifs.     

Interaction of a thermochemical plume with free convection mantle flows and its influence on mantle melting and recrystallization

We present a thermophysical model for interaction between the conduit of a thermochemical plume and horizontal free convection flows in the mantle: The mantle flow incident on the plume conduit melts at the conduit boundary (front part) and crystallizes at its back. Geological data on the intensity of plume magmatism over the last 150 Myr are used to estimate the total thermal power of mantle plumes. A possible scenario for plume-related mantle recrystallization is proposed. Over the lifespan of a thermochemical plume, mantle melts and recrystallizes owing to the motion of the plume source and interaction between the plume conduit and horizontal free convection flows. The plume conduits can melt and recrystallize the entire mantle over a certain period of time. The model for the interaction of drifting plume conduits with mantle flows and the estimated total thermal power of mantle plumes are used to estimate the duration of plume-related melting and recrystallization of the entire mantle. The influence of mantle plumes on the convective structure of the mantle through melting is judged from the model for plume interaction with horizontal mantle flows.     

Immature and mature transform zones near a hot spot: The South Iceland Seismic Zone and the Tjörnes Fracture Zone (Iceland)

We describe and compare the two transform zones that connect the Icelandic rift segments and the mid-Atlantic Ridge close to the Icelandic hot spot, in terms of geometry of faulting and stress fields. The E–W trending South Iceland Seismic Zone is a diffuse shear zone with a Riedel fault pattern including N0°–N20°E trending right-lateral and N60°–N70°E trending left-lateral faults. The dominant stress field in this zone is characterised by NW–SE extension, in general agreement with left-lateral transform motion. The Tjörnes Fracture Zone includes three major lineaments at different stages of development. The most developed, the Húsavík–Flatey Fault, presents a relatively simple geometry with a major fault that trends ESE–WNW. The stress pattern is however complex, with two dominant directions of extension, E–W and NE–SW on average. Both these extensions are compatible with the right-lateral transform motion and reveal different behaviours in terms of coupling. Transform motion has unambiguous fault expression along a mature zone, a situation close to that of the Tjörnes Fracture Zone. In contrast, transform motion along the immature South Iceland Seismic Zone is expressed through a more complicate structural pattern. At the early stage of the transform process, relatively simple stress patterns prevail, with a single dominant stress field, whereas, when the transform zone is mature, moderate and low coupling situations may alternate, as a function of volcanic–tectonic crises and induce changes in stress orientation.     

深部过程对埃达克质岩石成分的制约

埃达克岩、太古宙TTG和中国东部广泛出露的燕山期埃达克质中酸性火山-侵入岩在岩石地球化学特征方面有许多相似之处,也有一些显著的差异。与典型的埃达克岩相比,太古宙TTG具有相对高Si和低Mg^#的特点：中国东部埃达克质岩石多表现为低Mg^#贫A120,和高K特征。埃达克岩相对高Mg^#是由于俯冲洋壳部分熔融产生的原生埃达克岩熔体受到了地幔橄榄岩的混染,太古宙TTG多无明显的地幔混染印记,反映其可能主要形成于下地壳底侵玄武岩的部分熔融,而与洋壳俯冲没有直接联系。中国东部埃达克质岩石相对低Mg^#畜K,暗示其可能是下地壳底侵玄武岩部分熔融或拆沉-熔融的产物,而幔源富钾熔体的混合、壳内分异和混染过程都有可能影响其成分特征中国东部部分地区的高镁埃达克质岩石可能揭示了下地壳拆沉一熔融和地幔混染过程。钾质埃达克岩的源区可能是被小比例软流圈熔体交代富集的底侵玄武岩层(增厚的下地壳)。结合燕山期岩浆作用和构造转换的特点来看,埃达克岩的形成是中国东部晚中生代岩石圈强烈减薄和大规模岩浆作用产物的一部分,这一重大构造体制的转换可能与地幔柱上涌对岩石圈的侵蚀和导致的伸展作用有关。     

Constraints on crustal and mantle structure of the oceanic plate south of Iceland from ocean bottom recorded Rayleigh waves

From April to July 2002 we carried out a deployment of 6 ocean bottom seismometers and 4 ocean bottom hydrophones in the North Atlantic south of Iceland. During the deployment period we recorded clear Rayleigh waves from 2 regional and 14 teleseismic earthquakes. This corresponds to a Rayleigh wave detection rate of nearly 92% for events with M_W ≥ 6.06.0 and epicentral distance less than 110°, close to detection rate estimates based on noise level variability. We measured Rayleigh wave event-station group dispersion and inter-station phase dispersion for one Mid-Atlantic Ridge event. The group dispersion curve is sensitive to the structure of the North-East Atlantic with an average age of  39 Myr. The phase dispersion curve is sensitive to the structure just south of Iceland (average plate age 33 Myr). Both dispersion curves indicate faster velocities than previously postulated for oceanic plate generated at the Reykjanes Ridge. A grid search approach was used to constrain the range of models fitting the data. The high velocity seismic lid just south of Iceland in the model for the phase dispersion path is slower or thinner than in the group dispersion model, which averages over a larger area and a somewhat older plate age, but the velocities in the low velocity half space are similar. We further consider the residual bathymetry in the experimental area. The residual anomaly decreases by 300–400 m from the Reykjanes Ridge to the  30 Myr old plate south of Iceland. This decrease can be explained by the disappearance of a mantle thermal anomaly associated with the Iceland plume. Both the residual bathymetry and the surface wave data are thus consistent with the notion that the southward spreading of the Icelandic plume is channelised underneath the Reykjanes Ridge and does not spread far outside this channel.Based on the experience from the pilot experiment, we estimate that a minimum recording time of 13–15 months in favourable weather conditions (April–September) is required to record enough data to map the spreading plume with surface waves, and to produce a tomographic image to a depth of 1000 km using body waves. This can be achieved by a continuous deployment of at least  20 months, or by two or three deployments during the spring and summer of consecutive years.     

Review of melting experiments on carbonated eclogite and peridotite: insights into mantle metasomatism

Experimental studies on the partial melting of eclogite and peridotite provide important clues on mantle metasomatism. Here, we review results from some of the recent experiments and show that melting of carbonated eclogite and peridotite can produce carbonatitic to carbonated silicate melt, in which carbonates melt preferentially before Ti oxides and silicates. Low-degree melting results in carbonatitic melt coexisting with Ti oxides and silicates. This process also leads to the fractionation between some high-field strength elements (Nb, Ta, Zr, Hf, and HREE) and highly incompatible elements (U and Th) in the melt. When Ti oxides are nearly exhausted in eclogite, extremely high TiO₂ contents (e.g. 19 wt.%) are present in the melt with marked concentration of Nb and Ta. These results help to explain the features of carbonatitic metasomatism and the Nb–Ta spike in oceanic island basalts as identified in experimental studies. These studies also explain the reducing conditions that stabilize diamond in the deep mantle (>150 km) as well as the occurrence of diamond at different depths reported in various studies. Melting in such a reduced mantle can happen through redox reaction between diamond, pyroxene, and olivine, in which the initial liquid is a carbonated silicate melt. However, the theoretical oxygen fugacity (fO₂) in the asthenosphere is much lower than that predicted by the reaction and requires elevated fO₂, which can be caused by the addition of relatively oxidized materials from the lower mantle, deep asthenospheric material, and various recycled components. A combination of these processes generates locally oxidized domains in the deep mantle.