Showing papers on "Mid-ocean ridge published in 1993"

Heat flow from the Earth's interior: Analysis of the global data set

Abstract: We present a new estimate of the Earth's heat loss based on a new global compilation of heat flow measurements comprising 24,774 observations at 20,201 sites. On a 5 o x 5 o grid, the observations cover 62% of the Earth's surface. Empirical estimators, ref- erenced to geological map units and derived from the observations, enable heat flow to be estimated in areas without measurements. Corrections for the effects of hydrothermal circulation in the oceanic crest compen- sate for the advected heat undetected in measurements of the conductive heat flux. The mean heat flows of continents and oceans are 65 and 101 mW m -2, re- spectively, which when areally weighted yield a global mean of 87 mW m -2 and a global heat loss of 44.2 x 10 2 W, an increase of some 4-8% over earlier esti- mates. More than half of the Earth's heat loss comes from Cenozoic oceanic lithosphere. A spherical hat- monic analysis of the global heat flow field reveals strong sectoral components and lesser zonal strength. The spectrum principally reflects the geographic dis- tribution of the ocean ridge system. The rate at which the heat flow spectrum loses strength with increasing harmonic degree is similar to the decline in spectral strength exhibited by the Earth's topography. The spectra of the gravitational and magnetic fields fall off much more steeply, consistent with field sources in the lower mantle and core, respectively. Families of con- tinental and oceanic conductive geotherms indicate the range of temperatures existing in the lithosphere under various surface heat flow conditions. The heat flow field is very well correlated with the seismic shear wave velocity distribution near the top of the upper mantle.

Palaeozoic and Cenozoic lithoprobes and the loss of >120 km of Archaean lithosphere, Sino-Korean craton, China

Abstract: In eastern China Palaeozoic kimberlites and Cenozoic basalts have been erupted through the same Archaean crust, thus providing deep probes of the cratonic lower lithosphere over a period of 400 Ma. While Palaeozoic diamondiferous kimberlites point to the existence of thick, refractory lower lithosphere in the east, Cenozoic basalt-borne xenoliths reveal the presence of hot, thin, less refractory lower lithosphere. Remnants of the Archaean lithosphere may have survived as harzburgites which are chemically similar to those from the Kaapvaal craton but very different from recently accreted lherzolites. In the absence of convincing evidence for supra-subduction or intraplate processes it is believed that the dramatic change of lithosphere architecture in the Phanerozoic was caused by indentor tectonics resulting from the collision of India and Eurasia. Passive reactivation and remobilization of the Archaean lower lithosphere, in particular metasome horizons, contributed to Cenozoic magmatism aligned along major lithospheric faults. Traditionally the oldest Archaean cratonic nuclei are thought of as the most stable, inert parts of the Earth's surface. In the case of South Africa, Canada and Western Australia, Archaean cratons (Liu et al. 1992) lie atop a thick mechanical boundary layer characterized by high velocity anomalies (Anderson et al. 1992). In addition, the occurrence of Archaean P-type diamonds in on-craton kimberlites confirms the presence of an ancient thick lithospheric keel that, in some cases, was stabilized to depths of 200 km in the first billion years of Earth's history (Boyd & Gurney 1986). However, not all cratons have retained their structural integrity. In the case of the GreenlandHebridean craton, elevated mantle temperatures associated with the Iceland plume and tectonic forces related to the opening of the North Atlantic, may have been responsible for erosion of the craton margin. This would account for the existence of thinned Archaean crust (< 30 km) on the eastern Atlantic margin (i.e. Hebridean craton) and the survival of a thick cratonic nucleus in Greenland (Scott-Smith 1987). Similarly a thick cratonic keel does not underlie the SinoKorean Archaean craton, eastern China. Detailed seismic tomography (Chen et al. 1991; Liu 1992) indicates that the 'present-day' lithosphere is < 80 km thick (see Fig. 4) with greatly thinned lithosphere around the Bohai Sea (Ma & Wu 1981). The presence of thin lithosphere with a low velocity structure similar to an ocean ridge is substantiated by heat flow studies in eastern China (Teng et al. 1983) which reveal a region of very high heat flow on the craton in the vicinity of the Bohai Sea and Beijing (Fig. 1). The measured heat flow (1.2-2.53 HFU) corresponds to geotherms observed in tectonically active continents or ocean basins (50-105 mW m-Z). The aim of this paper is to review the temporal evolution of the lower lithosphere beneath the Sino-Korean craton, a crustal province known to contain some of the oldest crustal rocks on Earth (Jahn et al. 1987). In this review we will: (a) present petrological and geochemical evidence for the character of the Palaeozoic and Cenozoic lithosphere; (b) review the available geological and geochemical data on eastern China pertinent to lithosphere evolution, (c) outline a model to explain the temporal changes in lithosphere architecture. Palaeozoic kimberlite-borne xenoliths Palaeozic (400 Ma) kimberlites entrain a variety of xenoliths and megacrysts including diamonds (Lu et al. 1991; Zhang et al. 1991; Chi et al. 1992) (Fig. 1). While the petrology, mineralogy and thermal history of peridotite xenoliths and heavy mineral concentrates have been determined across the Sino-Korean craton, very little geochemical data are available for these xenoFrom Prichard, H. M., Alabaster, T., Harris, N. B. W. & Neary, C. R. (eds), 1993, Magmatic Processes and Plate Tectonics, Geological Society Special Publication No. 76, 71-81. 71 at Royal Holloway University of London on July 10, 2013 http://sp.lyellcollection.org/ Downloaded from

The genesis of oceanic crust - Magma injection, hydrothermal circulation, and crustal flow

Abstract: In this study we construct a thermal and mechanical model for the genesis of oceanic crust. Magma is halted in its ascent within the oceanic crust when it reaches a freezing horizon, where the dilational volume change associated with magma freezing leads to viscous stresses that favor magma ponding near the freezing horizon. To model the steady state thermal impact of crustal accretion via dike injection and pillow flows, we treat all crustal accretion in rocks cooler than a magma "solidus" to occur in a narrow 250-m-wide dike-like region centered about the ridge axis. The rest of the oceanic crust is modeled to be emplaced as a steady state magma lens directly beneath the "solidus" freezing horizon where the steady state emplacement rate is determined by the constraint that this lens supply all crust that is not emplaced through diking/extrusion above the magma lens. If hydrothermal heat transport within crustal rocks cooler than 600oC removes heat 8 times as efficiently as heat conduction, then we find that a steady state magma lens will only exist within the crust for ridges spreading faster than a 25 mm/yr half rate. The depth dependence of the magma lens with spreading rate is in good agreement with seismic observations. These results suggest that a fairly delicate balance between magmatic heat injection during crustal accretion and hydrothermal heat removal leads to a strongly different crustal thermal structure at fast and slow spreading ridge axes. Our results support the hypothesis that median valley topography is due to extension of strong ridge axis lithosphere; it is the difference in thermal regime that is directly responsible for the striking difference between the typical median valley seen at slow spreading ridges (e.g., Mid-Atlantic Ridge) and the axial high seen at fast spreading ridges (e.g., East Pacific Rise). This paradigm for the origin of a median valley at a slow spreading ridge predicts that along-axis variations in median valley topography of a slow spreading center reflect variations in recent magmatic heat input along a segment, that is, that the axial topography is a good time-averaged indicator of the relative importance of hydrothermal cooling and magmatic injection along a given section of a ridge segment. We determine the accumulated crustal strain associated with lower crustal flow which supports the hypothesis that the Oman Ophiolite crust was created at a paleo-analogue to a fast spreading ridge and also suggests that crustal strain, and not cumulate layering, may be the dominant physical process that generates "layered gabbros" within the Oman Ophiolite.

Emplacement of mantle rocks in the seafloor at mid-ocean ridges

Abstract: This paper discusses the geological and geophysical data available on mid-ocean ridges with outcrops of serpentinized mantle peridotites, with the objective of better constraining the modes of emplacement of these rocks in the seafloor. Ridges with serpentinized peridotites outcrops are in most cases characterized by slow-spreading rates, and in every case by deep axial valleys. Such deep axial valleys are thought, based on geophysical constraints and on mechanical modelling results, to characterize ridges with a thick axial lithosphere. A predictable effect of a thick axial lithosphere is that it should prevent magmas from pooling at crustal depths in a long-lasting magma chamber: gabbroic magmas should instead form shortlived dike or sill-like intrusions. Samples from axial outcrops of serpentinized peridotites are often cut by dikelets of evolved gabbros which are interpreted as apophyses of such dike and sill-like intrusions. This observation leads to a discontinuous magmatic crust model, in which mantle-derived peridotites form screens for numerous gabbroic intrusions. This discontinuous magmatic crust is expected to form in magma-poor ridge regions, where there is not enough magma to produce a 4-to 7-km-thick magmatic crust, and where the uppermost kilometers of oceanic lithosphere therefore have to be at least partially made of tectonically uplifted mantle material. Because the dimensions of individual mantle-derived ultramafic screens may be smaller than seismic experiments detection limits, the discontinuous magmatic crust model discussed in this paper may produce a layer 3-type seismic signature, even without extensive serpentinization of its ultramafic component. It therefore provides an alternative to Hess's [1962] serpentinite layer 3 model, for the geological interpretation of seismic data from oceanic areas with frequent outcrops of deep crustal and mantle-derived rocks.

Ultra-depleted primary melt included in an olivine from the Mid-Atlantic Ridge

Abstract: MODELS of magma genesis at mid-ocean ridges1, together with recent experimental data2 and observations of trace element abundances in clinopyroxenes from abyssal peridotites3, suggest that small-volume melt fractions can be efficiently extracted from the melting mantle. As shown in ref. 3, residues of this type of melting (fractional melting) display extremely low abundances of incompatible trace elements and extreme fractionation amongst them, especially at advanced stages of the process because of the compounded effects of differences in the partition coefficients. If this process operates beneath mid-ocean ridges, one would expect to sample melts that are correspondingly depleted and fractionated in trace elements. Indeed, the existence of very depleted melts in mid-ocean ridges was predicted previously4–6 in order to explain the presence of magnesian pyroxene and calcic plagioclase in mid-ocean-ridge basalts. We report here the discovery of melt that has major and trace element characteristics consistent with these predictions4–6, occurring as an inclusion in an olivine phenocryst in a typical mid-ocean-ridge basalt from the Mid-Atlantic Ridge. Although our preferred model for the origin of this 'ultra-depleted' melt is critical (continuous) melting, we cannot at this stage rule out other models. Our results underscore the importance of trapped melt inclusions as recorders of the processes involved in melting and melt extraction, and also as pointers to primary melt compositions.

Anomalous CH 4 and NH 4 + concentrations at an unsedimented mid-ocean-ridge hydrothermal system

Abstract: SINCE the discovery in 1977 of sea-floor hydrothermal systems, the study of the chemistry of the venting fluids has transformed our understanding of the geochemical cycles that influence the composition of sea water and the ocean crust. With few exceptions (Guaymas basin being the most notable), the vent systems studied so far are free of sedimentary influence and the chemistry of the fluids can be explained on the basis of interactions between sea water and basalt. Such fluids typically contain low methane concentrations, ranging from 50 to 120 μM (refs 1–7), and ammonium concentrations less than 10 μM (ref. 8). Here we report CH4 and NH4+ concentrations from the Endeavour segment of the Juan de Fuca Ridge which are many times greater than those measured previously at any unsedimented mid-ocean ridge. The 13C/12C ratio of this CH4 is the lowest yet found in any hydrothermal environment, implying an unusual source. We attribute these high CH4 and NH4+ concentrations to the decomposition of sub-sea-floor organic matter associated with sediments buried at an earlier stage of the ridge's evolution. These data illustrate that the organic geochemistry of unsedimented ridges may be more complex than suspected hitherto.

High-resolution global upper mantle structure and plate tectonics

Abstract: A global high-resolution S wave velocity model RG5.5 is obtained for the upper 500 km of Earth's mantle using a 5° × 5° equal-area block parameterization. The data set consists of some 18,000 seismograms associated with 971 events with magnitudes larger than 5.5. Fundamental modes (Love and Rayleigh waves) are used with periods from 75 to 250 s. The horizontal resolution length is around 1000 km, and the vertical resolution varies with depth from 60 to 250 km. Model RG5.5 has many features consistent with previous three-dimensional global and local seismic studies, but many new features are found. The S waves under mid-ocean ridges have broad slow velocity and have very slow velocity in the upper 100 km below the surface. The minimum velocity is at depths near 50 km or shallower. The lateral extent of the slow velocity region across ridges increases with spreading rate. The S wave velocities under ridges are strongly correlated with spreading rates at shallow depth, but the correlation decreases with depth and almost disappears at 100 km. The slow velocities shift off the current spreading positions below 100 km depth under the Mid-Atlantic Ridge and may record past positions of the ridge and/or be related to hotspots near the ridge. Some major hotspots are associated with slow-velocity anomalies with magnitudes of about 1–2% slower than the global average and with lateral dimension larger than 1000 km at depths between 100 and 200 km. Differences in the upwelling structure between ridges and hotspots are indicated. The S wave velocity structures may suggest an active mechanism for the East African Rift Valley and a plate extension mechanism for the Baikal Rift Valley.

Metallogenesis in back-arc environments; the Lau Basin example

Abstract: Geologic investigations with the submersible Nautile in the Lau basin represent one of the first detailed studies of hydrothermal activity on a modern back-arc volcanic ridge, the Valu Fa Ridge. Three major hydrothermal fields (Hine Hina, Vai Lili, and White Church) were discovered in the areas of the greatest differention of volcanic rocks. The type of hydrothermal deposit is controlled by the type of volcanism and by tectonic activity which increases from south to north. Three stages of sulfide formation are proposed and described. Vent fluids at Valu Fa have much higher metal contents than those at midocean ridges. Cl enrichment is best explained by mixing with deep brine rather than by subcritical phase separation and there is no evidence for a magmatic fluid contribution. Some characteristics of the fluids, such as low pH (2) and low concentration of H[sub 2]S, can be explained by subsea-floor sulfide formation. Vertical mineralogical zonation within the mound differs from midocean ridge deposits but is similar to that of a kuroko deposit. At the surface of the deposit, the virtual absence of pyrite and the high amount of sphalerite, barite, tennantite, galena, and locally, native gold are remarkable. Compared to midocean ridges, Lau basinmore » deposits are enriched in Ba, Zn, As, Pb, Ag, Au, and Hg and depleted in Mo, Se, and Co. Their mineralogy, chemical composition, and geologic setting show that the southern Lau basin deposits are intermediate between oceanic and continental back-arc deposits.« less

An episodic hypothesis for Venusian tectonics

Abstract: It is suggested that episodic plate tectonics occurs on Venus; episodes of rapid plate tectonics are separated by periods of surface quiescence. For the last 500 ± 200 m.y. it is postulated that the surface of Venus has been a single rigid plate that has been thickening due to conductive cooling. A near-uniform surface age is consistent with observed crater densities and the relatively small number of craters modified by surface tectonics or embayed by lava flows. A lithosphere that has conductively thickened for some 500 m.y. has a thickness of about 300 km, nearly an order of magnitude greater than the thickness associated with steady state conductive heat loss. Such a thick lithosphere can support the high topography and associated gravity anomalies on Venus as well as the unrelaxed craters; studies of lithospheric flexure at coronae are also consistent with a thick elastic lithosphere. Incipient subduction associated with large coronae may represent the onset of a new episode of rapid plate tectonics. On the Earth, 75–90% of mantle heat transport is attributed to the creation of new oceanic lithosphere at ocean ridges. This process is not operative on Venus. This paper suggests that episodic plate tectonics on Venus constitutes the primary mechanism for mantle heat transport on that planet.

Melting study of a peridotite KLB-1 to 6.5 GPa, and the origin of basaltic magmas

Abstract: With a newly established multi-anvil press in the Tokyo Institute of Technology, we have carried out a series of melting experiments on peridotite KLB-1 up to 6.5 GPa. Melt fractions of the peridotite were determined in a wide P-T range using extensive X -ray mapping analysis of run products by EPMA and a digitalized back-scattered electron image technique. Compositions of partial melts and solid residues were determined in the whole melting range up to 6.5 GPa. Given quantitative information on mantle melting, we discuss conditions of melting of various basalt magmas and the nature of their source materials. Our conclusions are consistent with the hypothesis that typical mid oceanic ridge basalts represent low pressure ( ca . 1 GPa), low temperature ( T p ≈ 1300 °C) partial melting products of mantle peridotite. Island arc picritic tholeiites may also be regarded as partial melts of a peridotitic source, at 1-2 GPa pressures and T p ranging from 1400 to 1500 °C. However, proposed primary magmas for Hawaiian tholeiite are difficult to produce by partial melting of typical mantle peridotite at any depth under anhydrous conditions. Source materials for magmas in large hotspots (e.g. Hawaii, Iceland and some continental flood basalts (CFBS)) may be anomalously enriched in FeO and TiO 2 relative to typical upper mantle peridotites such as KLB-1.

Seismic Structure of the Southern East Pacific Rise

Abstract: Seismic data from the ultrafast-spreading (150 to 162 millimeters per year) southern East Pacific Rise show that the rise axis is underlain by a thin (less than 200 meters thick) extrusive volcanic layer (seismic layer 2A) that thickens rapidly off axis. Also beneath the rise axis is a narrow (less than 1 kilometer wide) melt sill that is in some places less than 1000 meters below the sea floor. The small dimensions of this molten body indicate that magma chamber size does not depend strongly on spreading rate as predicted by many ridge-crest thermal models. However, the shallow depth of this body is consistent with an inverse correlation between magma chamber depth and spreading rate. These observations indicate that the paradigm of ridge crest magma chambers as small, sill-like, midcrustal bodies is applicable to a wide range of intermediate- and fast-spreading ridges.

Orogenic uplift and collapse, crustal thickness, fabrics and metamorphic phase changes: the role of eclogites

Abstract: Coesite-bearing eclogites in several deep crustal metamorphic assemblages now exposed in extensionally-collapsed orogens indicate the tectonic denudation of more than 90km of crustal rocks and pre-collapsed crustal thicknesses of at least 120km. For mountain ranges and orogenic plateaux up to 5 km in elevation and average crustal densities of about 2.8, crustal thickness cannot exceed about 80 km unless pre-shortening crustal/ lithosphere thickness ratios were less than 0.135 or some way can be found to preferentially thicken the lithospheric mantle. This problem can be avoided and very thick orogenic crusts built up if granulite facies rocks transform to denser eclogite facies during shortening, where the petrographic Moho is continuously depressed below a density/seismic velocity Moho buffered at about 70 km and mountains at about 3 km. Advective thinning of the lithosphere combined with the resultant heating and eclogite to sillimanite-granulite/amphibolite transformation causes surface uplift of about 2 km, a rapid change in isostatic compensation level, and a switch from a shortening to an extensional/collapse regime. We have developed a simple numerical model based upon field observations in southwestern Norway in which coherent regional-scale transformation of lower crustal rocks to eclogite facies during lithospheric shortening is followed by heating, transformation of eclogite to amphibolite and granulite, extension, and crustal thinning by coaxial then non-coaxial mechanisms. The model also explains strong lower crustal layering (eclogite and other lenses in horizontallyextended amphibolites), regionally horizontal gneissic fabrics, rapid return from orogenic to 'normal' crustal thickness with minor erosion, the lateral and vertical juxtaposition of low-grade and high-grade rocks and rapid marine transgression shortly after orogeny. Young orogens have maximum regional average elevations of about 5 km; their crustal thicknesses, determined directly from seismic reflection/refraction studies or indirectly from gravity anomalies, do not exceed about 70 km (Meissner 1986). The pre-plate tectonic view was that the surface elevation (e) of a mountain belt in perfect isostatic balance is related to its compensating root (r) and crustal thickness (Cz) by the relative densities of crust (Pc) and mantle (Pro), the floating iceberg principle in which the mantle was considered to behave as a fluid (Stokes 1849; Airy 1855; Heiskanen & Vening Meinesz 1958). The plate tectonic view is that the crust forms part of the lithospheric boundary conduction layer (thickness lz) with a mantle density of Pm" The whole l z column of crust and mantle is compensated to the asthenospheric mantle of density Pa" The crust 'floats', whereas the mantle portion 'sinks', the lithosphere; therefore, the level at which the surface sits relative to the oceanic ridges depends upon l~ and CJl~. Also, the cooler upper portion of the boundary layer acts as a strong flexural beam to support loads, which allows a departure from pure zerostrength Airy isostasy and the support of loads and elevations higher than those for a pure Airy model and the development and maintenance of substantial positive gravity anomalies. The elevation of a mountain range is, therefore, a function of the vertical density distribution within and the flexural strength of the lithosphere, although, for most mountain belts, there is insufficient knowledge of these parameters yet to draw conclusions that clearly relate crustal thickness to elevation. In this paper, we wish to draw attention to some general principles of vertical density distribution in relation to elevation and crustal/lithospheric thickness that may be important in understanding orogenic evolution, although we emphasize that crust/lithospheric thickness beneath mountain ranges is poorly constrained. Assuming homogenous bulk strain and preshortening crustal (Cz) and lithospheric (lz) values of 32 km and 120 km respectively, crustal thickening, caused by vertical stretching, is buffered at about 70 km beneath a surface elevation of about 3 km by vertical compression caused by isostatic compensation in a thickened crust that From Prichard, H. M., Alabaster, T., Harris, N. B. W. & Neary, C. R. (eds), 1993, Magmatic Processes and Plate Tectonics, Geological Society Special Publication No. 76, 325-343. 325 326 J .F . DEWEY E T A L . balances horizontal compression caused by plate convergence (Dewey 1988; England & Houseman 1988, 1989). Elevations above 3 km may be achieved by crustal underplating or by mechanisms that thin the lithosphere without thinning the crust such as hot-spot jetting, delamination, or rapid advective thinning. The general picture that has emerged for the wide Tibetan Himalayan zone of crustal thickening is one of Palaeogene vertical plane strain crustal thickening to about 65 km, caused by the India/Eurasia convergence, succeeded by a phase of horizontal plane strain followed, in turn, by a phase of post-Miocene uplift to the present 5 km average elevation accompanied and followed by lithospheric extension and magmatism (Dewey et al. 1988), a phase of so-called orogenic extensional collapse (Dewey 1988). Many Phanerozoic orogens appear to follow this general pattern of shortening followed by extensional collapse (Dewey 1988), which is related probably to lithospheric shortening followed by the rapid advective removal of the lithospheric mantle root. (Houseman et al. 1981; England & Houseman 1988, 1989). During shortening by roughly vertical plane strain and vertical bulk stretching, buoyant crustal and negatively buoyant lithospheric mantle roots develop, rocks are progressively buried and geothermal gradient and heat flow are reduced. During shortening, the principal axis of compression is horizontal and the axis of least compression vertical. Crustal thickness is buffered at about 70 km, and the intermediate axis of compression becomes vertical, a wrench regime is developed and shortening must spread laterally if convergence continues, one possible explanation of the lateral progradation of thrust sheet complexes. Advective removal of the mantle root leads to uplift to about 5 km, a rapid increase in geothermal gradient and vertical shortening/extensional collapse in a stress regime of vertical compression. There is sufficient seismic data to indicate that the Moho, as defined by that data, does not exceed depths of about 70 km beneath Cenozoic mountain ranges (Meissner 1986). If 70 km is the limit to which normal density (2.8) continental crust may be thickened by bulk vertical stretching, there are substantial implications for the structural and metamorphic history of orogenic belts. First, gross orogenic shortening values achieved by bulk vertical plane strain expressed by orogenic structure and fabrics cannot exceed about 50% if we start with a 'normal' thickness continental crust of about 35km. Many orogenic belts have shortening values, at least locally, greatly in excess of 50%. Of course, shortening may be increased in several ways such as by substantial erosion during shortening, and/or by starting with a thin crust, which would be expected at a colliding rifted margin. Both of these have operated in the Himalayas; some 20 km has been lost over large areas by erosion and some of the highest thrust sheets probably had a thin starting crust, although these factors alone are insufficient to account for Himalayan shortening. Horizontal plane strain by lateral escape tectonics will also allow increased shortening values without increasing C z although it is likely to generate steep structures and cannot generate thrust regime structures and fabrics. A C z limit of 70 km places even greater constraints on crustal metamorphism. Rocks now at the surface in older orogens have up to 35 km of continental crust below them so that, if orogenic thickness was 70 km, erosional denudation can have removed a maximum of only 35 km and extensional collapse reduces this figure still further. Tectonic denudation mechanisms such as motion towards the surface in the footwalls of major normal faults above zones of 'replacive' lower crustal flow may allow deeper crustal levels to be exposed but we are still limited by a C z maximum of 70 km. Therefore, we would expect the regional exposure of metamorphic rocks of greenschist/amphibolite granulite facies recording pressures of around 10-12kb with localized zones of blueschist/eclogite facies indicating maximum pressures of about 20 kb. The rocks of the Tauern window in Austria experienced peak pressures of 20 kb and yet appear to be part of a regionally-coherent metamorphic terrain now at the surface of a 50km crust suggesting a peak Cz of 100 km. In the western Alps, Chopin (1984, 1987) and Chopin et al. (1991) have described coesite/pyrope-bearing rocks and, in the Dabie Shan, Enami & Zhang (1990), Hirajima et al. (1990), Okay & Seng6r (1992), Wang & Liou (1991), Wang et al. (1989, 1992) and Xu et al. (1992) have recorded and described diamondand coesite-bearing eclogites indicating pressures of 39-40 kb. In southwestern Norway, coesite-bearing eclogites form part of the regionally-coherent Western Gneiss Region (Smith 1984; Austrheim 1987, 1991; Smith & Lappin, 1989; Andersen & Jamtveit 1990; Andersen et al. 1991), and experienced minimum peak pressures of 28 kb (Fig. 1) and have 35 km of subjacent crust indicating a peak C~ of at least 120 kin. We are, therefore, forced into the ~pparent paradox that observation and theory indicate that C~ cannot exceed about 70 km and yet metamorphic rocks indicate C~ values of at least 120 km. The problem is how to get rocks down OROGENIC UPLIFT & COLLAPSE: THE ROLE OF ECLOGITES 327

Active and relict sea-floor hydrothermal mineralization at the TAG hydrothermal field, Mid-Atlantic Ridge

Abstract: The TAG hydrothermal field is a site of major active and inactive volcanic-hosted hydrothermal mineralization in the rift valley of the slow-spreading Mid-Atlantic Ridge at 26[degree]N. The axial high is the principal locus of present magmatic intrusions. The TAG field contains three main areas of present and past hydrothermal activity: (1) an actively venting high-temperature sulfide mound; (2) two former high-temperature vent areas; (3) a zone of low-temperature venting and precipitation of Fe and Mn oxide deposits. The volcanic centers occur at the intersections between ridge axis-parallel normal faults and projected axis-transverse transfer faults. The intersections of these active fault systems may act as conduits both for magmatic intrusions from sources beneath the axial high that build the volcanic centers and for hydrothermal upwelling that taps the heat sources. Radiometric dating of sulfide samples and manganese crusts in the hydrothermal zones and dating of sediments intercalated with pillow lava flows in the volcanic center adjacent to the active sulfide mound indicate multiple episodes of hydrothermal activity throughout the field driven by heat supplied by episodic intrusions over a period of at least 140 [times] 10[sup 3] yr. The sulfide deposits are built by juxtaposition and superposition during relatively long residence times near episodic axial heat sources counterbalanced by mass wasting in the tectonically active rift valley of the slow-spreading oceanic ridge. Hydrothermal reworking of a relict hydrothermal zone by high-temperature hydrothermal episodes has recrystallized sulfides and concentrated the first visible primary gold reported in a deposit at an oceanic ridge.

Hydrothermal methane venting between 12°N and 26°N along the Mid‐Atlantic Ridge

Abstract: Hydrographic surveys along the Mid-Atlantic Ridge (MAR) between 12°N and 26°N, carried out from 1984 to 1990, show a variable pattern of CH4-rich water column plumes. The vertical distribution of CH4 at stations located every 20–40 km is presented along this 1200-km-long section of the MAR. CH4 venting is clearly demonstrated. CH4-enriched fluids rise from vents as plumes; spreading is confined to the axial valley due to the topography of the MAR. CH4 contents from 45 nmol to 675 nmol/kg are measured in the buoyant plumes above the two active hydrothermal sites (MARK 23°N; TAG 26°N) known at present, whereas CH4 anomalies up to 3.6 nmol/kg are typically observed in plumes emitted either on the inner floor, on the walls, and/or at the top of the rift mountains along the studied ridge section. CH4 concentrations (45 μmol to 144 μmol/kg) in MARK and TAG vent fluids are of the same order of magnitude as those found in the East Pacific Rise fluids. Even though CH4 is known to be unstable with respect to oxidation by dissolved oxygen, and in spite of its microbial oxidation in plumes, these results confirm CH4 to be a good indicator to track hydrothermal plumes and to map the variation of hydrothermal activity along mid oceanic ridges. Moreover, between 12° and 26°N along the MAR, CH4 results show that while hydrothermal activity is present everywhere along the ridge, it is predominant near fracture zones (FZ) (Kane FZ, 15°20′N FZ). Comparison of CH4 tracer with total dissolvable manganese (TDM) tracer in plumes allows us to differentiate subseafloor hydrothermal processes. The high TDM/CH4 found above TAG and MARK areas is indicative of basalt-seawater interaction, while at 15°N the low TDM/CH4 provides evidence of fluid circulation in ultrabasic rocks. CH4 data confirm the association between mantle degassing, hydrothermal activity, and serpentinization along this 12°–26°N section of the Mid-Atlantic Ridge.

Variation in cross-sectional area of the axial ridge along the East Pacific Rise: Evidence for the magmatic budget of a fast spreading center

Abstract: Along the fast and ultrafast spreading East Pacific Rise, the cross-sectional area of the axial ridge varies significantly over length scales similar to its morphologic segmentation. Using an automated method, we measure the ridge area (volume per kilometer along axis) where there is complete bathymetric coverage. Along 3500 km of the northern and southern East Pacific Rise, our two study areas encompass numerous transform faults, large overlapping spreading centers, small overlapping spreading centers, and smaller discontinuities (first-, second-, third-, and fourth-order discontinuities). The cross-sectional area variation mimics the undulation of the ridge crest depth; local area maxima occur toward the middle of segments, and the axial area decreases by 40% or more at first- and second-order discontinuities. Third-order discontinuities are generally marked by smaller disruptions in the ridge area, and fourth-order offsets do not systematically correspond with features of the cross-sectional area profile. The correlation between shallower ridges and larger ridge areas breaks down at some locations because axial cross-sectional area represents a longer term average of the ridge's magmatic state than axial depth. A correlation between large ridge areas and negative residual gravity anomalies indicates that inflated ridges are underlain by low-density crust and mantle. Also, a correlation between larger area and higher MgO content of axial basalts suggests that inflated areas generally erupt hotter magmas which are presumably supplied more rapidly to the neovolcanic zone. The cross-sectional area of the axial ridge appears to correlate with the width of the axis-centered low-velocity zone in the crust. These observations, as well as the absence of large, relict axial ridges off-axis, indicate that the axial ridge originates from buoyancy due to thermal expansion and the presence of melt in the crust and mantle within about 10 km of the rise axis. Portions of the northern East Pacific Rise underlain by a magma chamber reflector generally occur where the ridge cross-sectional area is greatest; this supports the connection between processes which inflate the axial ridge and those which heat the crust and upper mantle and produce melt. Thus, while the axial high on fast spreading centers resembles a constructional volcano in cross section, it is more like a long, narrow balloon whose cross-sectional area is a sensitive indicator of magma supply. Using this relationship, we predict with 75% confidence that at least 45% of the unsurveyed northern East Pacific Rise (18°N to 13°N and 9°N to 5°N) and at least 60% of the southern East Pacific Rise (4°S to 23°S) is underlain by a magma chamber reflector.

Tomographic image of the Mid‐Atlantic Plate Boundary in southwestern Iceland

Abstract: The 170 km South Iceland Seismic Tomography (SIST) profile extends from the west and across the Mid-Atlantic Ridge spreading center in the Western Volcanic Zone and continues obliquely through the transform zone (the South Iceland Seismic Zone) to the western edge of the Eastern Volcanic Zone. A total of 11 shot points and 210 receiver points were used, allowing precise travel times to be determined for 1050 crustal P wave rays and 180 wide-angle reflections. The large amplitudes of the wide-angle reflections and an apparent refractor velocity of 7.7 km/s are interpreted to be from a relatively sharp Moho at a depth of 20-24 kin. This interpretation differs from the earlier models (based on data gathered in the 1960s and 1970s), of a 10-15 km thick crust underlain by a upper mantle with very slow velocity of 7.0-7.4 km/s. Nevertheless, these older data do not contradict our new interpretation. Implication of the new interpretation is that the lower crust and the crust-mantle boundary are colder than previously assumed. A two-dimensional totoographic inversion of the compressional travel times reveals the following structures in the crust: (1) a sharp increase in thickness of the upper crust ("layer 2A") from northwest to southeast and (2) broad updoming of high velocity in the lower crust in the Western Volcanic Zone, (3) depth to the lower crust ("layer 3") increases gradually from 3 km at the northwestern end of the profile to 7 km at the southeastern end of the profile, (4) a low-velocity perturbation extends throughout the upper crust and midcrust into the lower crust in the area of the transform in south Iceland (South Iceland Seismic Zone), and (5) an upper crustal high-velocity anomaly is associated with extinct central volcanos northwest of the Western Volcanic Zone. The travel time data do not support the existence of a large (> 0.5 km thick) crustal magma chamber in this part of the Western Volcanic Zone but do not exclude the possibility of a smaller one.

A lithium isotope study of hot springs and metabasalts from Mid-Ocean Ridge Hydrothermal Systems

Abstract: The Li isotopic compositions of seawater and of fresh and altered basalts are distinct and therefore applicable to the study of the hydrothermal processes in the oceanic crust and the Li balance in the ocean. High-temperature fluids from seven vents on the East Pacific Rise (EPR) at 21°N and 11°–13°N have δ6Li values ranging between −6 and −11‰, i.e., 3–7‰ heavier than fresh basalt values. The intervent variations in Li concentration and isotopic composition correlate well with water/rock ratios. No temporal change in the isotopic value was observed between 1981 and 1985. Metabasalts show both Li depletion and enrichment relative to fresh basalt. They display light Li isotopic values which reflect incorporation of Li mobilized from fresh basalt. From these observations we conclude that the Li isotopic composition of submarine hot springs is controlled by a path-dependent process involving dissolution of primary minerals and precipitation of Li with alteration phases. δ6Li values of fluids from the Mid-Atlantic Ridge (−6 to −8‰) fall in the range of the EPR, indicating similar reaction controls at the two ridge systems. The lack of 7Li enrichment in the fluids from slower spreading ridges indicates that Li is not recycled from older weathered crust. Thus the difference between the 3He and 87Sr/ 86Sr based hydrothermal flux and the crustal transfer rate of Li cannot be reconciled by the inclusion of secondary Li from older crust.

Fault strain and seismic coupling on mid‐ocean ridges

Abstract: The contribution of extensional faulting to seafloor spreading along the East Pacific Rise (EPR) axis near 3°S and between 13°N and 15°N is calculated using data on the displacement and length distributions of faults obtained from side scan sonar and bathymetric data. It is found that faulting may account for of the order of 5–10% of the total spreading rate, which is comparable to a previous estimate from the EPR near 19°S. Given the paucity of normal faulting earthquakes on the EPR axis, a maximum estimate of the seismic moment release shows that seismicity can account for only 1% of the strain due to faulting. This result leads us to conclude that most of the slip on active faults must be occurring by stable sliding. Laboratory observations of the stability of frictional sliding show that increasing normal stress promotes unstable sliding, while increasing temperature promotes stable sliding. By applying a simple frictional model to mid-ocean ridge faults it is shown that at fast spreading ridges (≥90 mm/yr) the seismic portion of a fault (Ws) is a small proportion of the total downdip width of the fault (Wƒ). The ratio Ws/ Wƒ interpreted as the seismic coupling coefficient X, and in this case X≈ 0. In contrast, at slow spreading rates (≤40 mm/yr), Ws≈Wƒ, and therefore X≈ 1, which is consistent with the occurrence of large-magnitude earthquakes (mb= 5.0 to 6.0) occurring, for example, along the Mid-Atlantic Ridge axis.

Equilibration during mantle melting: a fractal tree model

Abstract: Many basalts from oceanic islands, ridges, and arcs show strong trace element evidence for melting at great depths, where garnet is a stable phase in mantle peridotites. If partial melts ascend to the surface by porous (intergranular) flow processes, the high-pressure garnet signature will be obliterated by diffusive reequilibration at shallower depths in the mantle. Spiegelman and Kenyon [Spiegelman, M. & Kenyon, P. (1992) Earth Planet Sci. Lett. 109, 611-620] argued that partial melts must therefore be focused into a coarser transport network, for high-speed delivery to the surface. Numerous natural network systems, such as rivers and the human vascular and bronchial systems, have fractal structures that are optimal for minimizing energy expenditure during material transport. I show here that a fractal magma "tree" with these optimal properties provides a network in which magma rapidly loses diffusive chemical "contact" with its host matrix. In this fractal network, magma conduits combine by twos, with the radius and flow velocities scaling as (2)n/3, where n is the generation number. For reasonable values of volume diffusivities, viscosities, and aspect ratios, melts will experience only limited diffusive reequilibration once they have traveled some hundreds of meters from their source. Melts thus represent rather local mantle domains, and there is little problem in delivering melts with deep (<100 km) geochemical signatures to the surface.

Building the crust at the Mid-Atlantic Ridge

Abstract: The morphology of the sea floor at mid-ocean-ridge spreading centres provides a key to understanding how ocean crust is constructed. Images of the axial zone of the slow-spreading Mid-Atlantic Ridge, obtained at a range of spatial scales, show that crustal construction is complex and highly variable, reflecting an underlying three-dimensional variability in magmatic, tectonic and hydrothermal processes.

Petrology of the East Pacific Rise crust and upper mantle exposed in Hess deep (Eastern Equatorial Pacific)

Abstract: The Hess Deep, a rifted oval-shaped depression located east of the Galapagos Triple Junction at the tip of the Cocos-Nazca ridge (about 101°W, 2°N), was explored in 1988 during 21 submersible dives. A total of 11 dives were devoted to the exploration of the E-W trending Intrarift ridge (15 km in length, 3000–5400 m in depth) north of the Hess Deep depression. The Intrarift ridge represents an outcrop of recent (1 m.y.) crustal and subcrustal material created at the axis of the East Pacific Rise (EPR), and emplaced during the lithospheric extention responsible for the westward propagation of the Cocos-Nazca rift (Francheteau et al., 1990). The lithospheric block has undergone cataclastic deformation and was dislocated by tectonic activity as attested to by the mixed and erratic distribution of rock types and by the occurrence of polygenic breccias and gabbroic mylonites. The samples are metamorphosed to varying degrees, but their protolith textures are generally well preserved. Their relic mineralogy indicates that they consist of harzburgites, dunites, gabbroic cumulates (gabbronorites and olivine gabbros), isotropic gabbros, dolerites, and basalts. Some samples of refractory harzburgites and most dunitic cumulates (with local accumulation of chromite) have been impregnated by wehrlitic and gabbroic primitive melts similar to those described from the mantle-crust transition zone of the Samail ophiolite complex (Oman). The mineral chemistry indicate that the ultramafics partly reequilibrated with the magmatic impregnations in the liquidus-solidus temperature range of 980–1100°C. The dolerites and basalts have been derived from mid-ocean ridge basalt primary melts presenting a broad range of incompatible element composition which suggests intermittent cycles of magmatic processes involving a progressive melting of a composite source with discontinous extraction of liquids as proposed for the EPR volcanics near 13°N (Hekinian et al., 1989). Most of the rocks underwent partial retrograde metamorphism and recorded several episodes of recrystallization from the upper greenschist facies (ultramafics and gabbros) to diagenetic alteration (volcanics). The cumulate gabbronorites, the isotropic gabbros, and some dolerites were partially albitized and amphibolitized during the penetration of seawater in the ocean crust prior to serpentinization. Several samples of unfoliated amphibolites are believed to be completely metamorphosed gabbroic rocks. The gabbroic cumulates and the plagioclase-rich melt impregnations were variably rodingitized (presence of various Ca-silicates such as epidote, prehnite, hydrogarnet, and zeolite) in relation to the serpentinization of the peridotites. One dive located on the scarps forming the northern wall of the Hess Deep to the east of the explored area, revealed the presence of in situ outcrops of isotropic gabbros, doleritic dikes, and extrusives and permitted to observe the contact between the sheeted dike complex and the high level isotropic gabbros.

Crustal structure of North Atlantic Fracture Zones

Abstract: Seismic studies have established that large-offset transforms along the slow spreading Mid-Atlantic Ridge exhibit anomalous crustal structures that fall well outside the range typically associated with oceanic crust. Seismically, fracture zone crust in the North Atlantic is extremely heterogeneous in both thickness and internal structure. It is frequently quite thin (<1–2 km thick) and is characterized by low compressional wave velocities and the absence of a normal seismic layer 3. A more gradual crustal thinning can extend up to several tens of kilometers from these fracture zones. Anomalously thin crust has also been inferred from both seismic and gravity studies at smaller ridge axis discontinuities along the Mid-Atlantic Ridge. The geological nature of the seismically anomalous crust found within Atlantic fracture zones, and how this crust forms, are still controversial. One interpretation consistent with available seismic observations is that the crust within North Atlantic fracture zones consists of a thin, intensely fractured, and hydrothermally altered basaltic section overlying ultramafics that are extensively serpentinized in places. Variations in apparent seismic crustal thickness along fracture zones may reflect different degrees of serpentinization of the upper mantle section or changes in the thickness of the igneous crust. The existence of a thinner crustal section in fracture zones can be explained by a reduced magma supply within a broad region near ridge offsets due to the three-dimensional nature of upwelling beneath a segmented spreading center and by tectonic dismemberment of the crust by large-scale detachment faults that form preferentially in the cold, brittle lithosphere near the ends of segments.

The paradox of the axial profile: Isostatic compensation along the axis of the Mid‐Atlantic Ridge?

Abstract: Mantle Bouguer gravity anomalies (MBA) and bathymetry on three profiles covering more than 1000 km along the axis of the Mid-Atlantic Ridge (MAR) are highly correlated, suggesting that along-axis topographic relief is locally compensated by variations in crustal thickness and/or mantle density structure. The quantitative relationship between topography and gravity on these profiles could be explained by the flexure of a thin, narrow elastic strip, representing the response to isostatic loads of an inner rift valley isolated from the rest of the plate by weak, bounding normal faults. The paradox is that across-axis profiles show that the median valley is an uncompensated feature, apparently created by a dynamic mechanism. New, extensive off-axis coverage of the MAR from 31° to 36°S shows that the high correlation does not persist outside the axial zone. We suggest that the on-axis correlation exists because the mechanism creating the median valley is controlled by the mantle thermal structure and along-axis variations in crustal thickness, both of which contribute to the MBA. If the mechanism is extension of a brittle-ductile lithosphere, the critical parameter controlling topographic relief is the thickness of the relatively stiff mantle layer immediately beneath the crust at the ridge axis. A three-dimensional (3-D) thermal model incorporating passive mantle flow, hydrothermal circulation, the plate boundary geometry, and variable magmatic heating associated with observed variations in crustal thickness predicts variations in the thickness of the stiff mantle layer that correlate with the observed axial topography. We model the expected topography using cross-axis sections of the 3-D thermal model and a 2-D finite element model of an extending lithosphere that incorporates temperature- and strain-dependent rheology, as well as the flexural response of a thickening plate. The predicted topographic signal produced by the combined dynamic and isostatic effects matches the amplitude of the observed axial bathymetry. At the 18 mm yr−1 spreading half-rate in our South Atlantic survey, the 2-D models also reproduce the observed rapid transition along the axis between rift valley/no rift valley morphology. We conclude that extensional forces acting on segmented oceanic lithosphere with varying rates of crustal production produce the highly variable morphology of the Southern MAR and the along-axis correlation between MBA and bathymetry.