Journal of Petrology Volume 41 Number 7 Pages 1071-1098 2000
© Oxford University Press 2000
Processes in High-Mg, High-T Magmas: Evidence from Olivine, Chromite and Glass in Palaeogene Picrites from West Greenland
1GEOLOGICAL SURVEY OF DENMARK AND GREENLAND, THORAVEJ 8, DK-2400 COPENHAGEN NV, DENMARK
2GEOLOGICAL MUSEUM, ØSTER VOLDGADE 5–7, DK-1350 COPENHAGEN K, DENMARK
Received September 2, 1999; Revised typescript accepted February 10, 2000
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
Uncontaminated volcanic rocks from the 60 Ma Vaigat Formation, West Greenland, contain 6·5–30 wt % MgO, averaging 15·5 wt % MgO. Olivine (mg-number 77·4–93·3) forms diverse assemblages of zoned phenocrysts and xenocrysts showing evidence for equilibrium and fractional crystallization, oxidation, partial to complete re-equilibration, as well as magma mixing. The olivine crystals contain glass inclusions and have high contents of Ca and Cr, indicating that all olivines with up to mg-number 93·0 crystallized from melts. Associated chromites (mg-number 45·4–77·2) are essentially unzoned and in equilibrium with the olivines. Matrix glasses from pillow breccias have 6·7–8·8 wt % MgO and quenched close to 1200°C with oxidation states one log-unit above the NNO (nickel–nickel oxide) buffer. Compositional differences between the glasses from different volcanic members are inherited from the primary melts. The magmas erupted as crystal-charged melts, and liquids with more than
14 wt % MgO were not erupted. The compositions of the unerupted parental melts were calculated by stepwise addition of equilibrium olivine to the matrix glasses, and these melts had 20–21 wt % MgO and liquidus temperatures of 1515–1560°C. They had lower FeO* than the erupted rocks with 20–21 wt % MgO because the rocks contain more iron-rich olivine than the melts would have had. The accumulated primary melts ascended through a lithospheric lid of 100 km thickness, and we envisage that major crystallization took place from 45 km depth and to the surface. Magma batches ascended in narrow dyke-like conduits and fractionated high-Mg olivine and chromite at deep levels and less magnesian crystals at shallower levels. At high ascent velocities numerous olivine crystals were carried in suspension to the surface, even some from deep levels, whereas during slower ascent most olivines settled out before eruption. Pulsating ascent rates led to mixing of magma batches in various stages of fractionation. On average, 13 wt % olivine was left within the crust, presumably as olivine-plated conduit walls. The conduit systems are similar to the crystal-rich narrow magma chambers suggested for mid-ocean ridges but are of much greater vertical extent. KEY WORDS: chromite; glass; Greenland; olivine; picrite
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
Highly magnesian volcanic rocks have been used to estimate mantle melting conditions and mantle compositions because their composition may be close to that of the primary mantle-derived melts. There is, however, a long-standing discussion as to how well bulk-rock compositions actually reflect melt compositions (Drever & Johnston, 1957
; Clarke, 1970; Hart & Davis, 1978, 1979; Clarke & O’Hara, 1979; Elthon & Ridley, 1979; Maaløe, 1979; Francis, 1985, 1995; Chen, 1993). Highly magnesian volcanic rocks are generally very olivine rich and may have accumulated olivine; this also applies to spinifex-textured komatiite lavas (Arndt, 1986).
Highly magnesian volcanic rocks are found in the early parts of flood basalt successions such as Karoo (Bristow, 1984; Cox et al., 1984), Deccan (Krishnamurthy & Cox, 1977; Melluso et al., 1995), West Greenland (Drever & Johnston, 1957; Clarke, 1970; Clarke & Pedersen, 1976) and East Greenland (Nielsen et al., 1981; Fram & Lesher, 1997; Hansen & Nielsen, 1999). Assuming that these rocks represent high-Mg melts, they have been viewed as the products of partial melting in the hottest parts of mantle plumes, at temperatures considerably above normal mantle temperatures (McKenzie & Bickle, 1988; Campbell & Griffiths, 1990). The Archaean komatiites, with estimated liquid MgO contents up to 29 wt % (Nisbet et al., 1993), have been taken as evidence that mantle plumes existed already in Archaean times (Campbell & Griffiths, 1992). The youngest known komatiites, the 88-my-old Gorgona komatiites, have estimated liquid MgO contents of 18–20 wt % (Aitken & Echeverria, 1984).
During continental break-up in the North Atlantic area in the Palaeogene, rifting in the Davis Strait–Baffin Bay region led to eruption of mafic and ultramafic magmas in central West Greenland and on Baffin Island [reviewed by, for example, Upton (1988) and Saunders et al. (1997); see also Chalmers et al. (1999)]. These volcanic successions contain an unusually high proportion of Mg-rich, presumably primitive high-temperature rocks (Clarke, 1970; Clarke & Pedersen, 1976). They are situated peripherally within the circular area of 2000 km diameter proposed to be influenced by the ancestral Iceland mantle plume (White & McKenzie, 1989), and this has caused debate because it is contrary to predictions from mantle plume models such as that of Campbell & Griffiths (1990), according to which Mg-rich melts should form only above the hot-spot at the centre of the plume-influenced area (Hill, 1991; Gill et al., 1992).
The highly magnesian Palaeogene volcanic rocks of West Greenland contain up to 30 wt % MgO. Whereas the most magnesian rocks are probably cumulates, Clarke (1970) and Clarke & O’Hara (1979) argued that some rocks with 19–20 wt % MgO represent primary liquids. This was disputed by Hart & Davis (1978, 1979) who, on the basis of Ni partitioning between olivine and liquid, argued that rocks with more than 11–13 wt % MgO are olivine accumulative. Pedersen (1985) estimated liquid MgO contents of 18–19 wt % based on the occurrence of very magnesian olivines (up to Fo92·5). Francis (1985) produced similar estimates for the similar and coeval volcanic rocks on Baffin Island and noted that the use of bulk-rock compositions as proxies for liquid compositions requires that the phenocrysts are continuously equilibrated with the liquid. Bulk-rock compositions may indicate whether the Mg-rich members in a suite of rocks are cumulates, because if they are, the suite will in many geochemical diagrams, for example, FeO* vs MgO, form linear trends that project back to the composition of the cumulus olivine (Albarède, 1992). Individual units in the volcanic succession in West Grenland do define a number of such linear trends; these have variable slopes that project back to olivine compositions varying from Fo86 to Fo89, suggesting the presence of accumulated olivine in some of the rocks. However, as pointed out by Hart & Davis (1979), elucidation of the true relation between rocks and liquids requires knowledge of the compositions of the phenocrysts involved. In this work, we use the compositions of olivines, chromites and glasses in the picrites of the Vaigat Formation in West Greenland to study this relation and to show how the above conundrum can be resolved. Primary high-Mg liquids did exist at depth but were not erupted; the erupted liquids had lower MgO and were charged with crystals acquired during open-system processes in the crystal-rich conduit systems during ascent to the surface.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
The volcanic rocks in the West Greenland Basin cover onshore areas of
22 000 km2 mainly on Disko island, Nuussuaq peninsula and Svartenhuk Halvø (Clarke & Pedersen, 1976), and an even larger submerged area on the shelf (e.g. Escher & Pulvertaft, 1995; Chalmers et al., 1999). The overall vertical thickness of the exposed volcanic succession is 2–3 km on Disko and Nuussuaq but because the depocentres shifted laterally with time the stratigraphic thicknesses are considerably higher. In all the onshore areas the volcanic succession can be divided into a lower part with Mg-rich olivine-phyric rocks (the Vaigat Formation) and an upper part with plagioclase-phyric rocks that have been assigned to a number of local formations. The Vaigat Formation constitutes roughly one-third of the erupted volume.
The Paleocene volcanic succession on Disko and Nuussuaq has been dated at 60·7 ± 0·5 to 59·4 ± 0·5 Ma (Storey et al., 1998). There are no discernible age differences within the Vaigat Formation, which was probably erupted within around half a million years with very high eruption rates.
Vaigat Formation
The Mg-rich rocks of the Vaigat Formation form a widespread series of light grey, thin pahoehoe lava flows and hyaloclastite breccias. Vertical thicknesses reach 1800 m. Subordinate intercalated horizons of brownish, aphyric, less-magnesian rocks, enriched in silica, are interpreted to have resulted from reaction between mafic magmas and crustal rocks, often in high-level magma chambers within the underlying sedimentary succession (Pedersen, 1979a, 1985; Pedersen & Pedersen, 1987; Goodrich & Patchett, 1991; Lightfoot et al., 1997). The contaminated rocks are not considered further in this paper, which is solely concerned with the uncontaminated parts of the Vaigat Formation.
The volcanism of the Vaigat Formation occurred in three main cycles, which are represented by the three main members. These are, from older to younger, the Anaanaa Member, Naujánguit Member and Ordlingassoq Member. Each cycle started with olivine-rich magmas and ended with more evolved, often contaminated magmas. At the initiation of the volcanism the environment was a sea-covered shelf, and after build-up to above sea level the volcanic pile prograded eastwards into the sea and gradually filled the basin (Pedersen et al., 1993, 1996; Pedersen et al., 1998). Because of the progradational deposition, hyaloclastite breccias formed in abundance during all three cycles, often by entry of subaerial lava flows into water.
The uncontaminated rocks of the Vaigat Formation contain 6·5–30 wt % MgO with an average of 15·5 wt % MgO (Fig. 1). Following the IUGS nomenclature (Le Bas et al., 2000) the majority of the rocks are picrites (MgO > 12 wt %), and the remaining rocks are basalts. All the rocks are tholeiitic.
|
|
PETROGRAPHY AND MINERALOGY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
The majority of the uncontaminated rocks of the Vaigat Formation have olivine and chromite as the only phenocryst phases. The most magnesian rocks have abundant equant to prismatic phenocrysts of olivine up to 3–4 mm in size, and chromite up to 0·1 mm in size (Figs 2 and 3). With declining content of MgO in the rock the phenocrysts become smaller. In rocks with less than
15 wt % MgO most olivines are <1 mm in size and the rocks are olivine microphyric or aphyric in hand sample. Rocks with <9 wt % MgO are aphyric or olivine–plagioclase phyric. Rocks with plagioclase phenocrysts are rare in the Vaigat Formation. Glassy rocks have phenocrysts of equant to elongate–skeletal olivine with included or attached chromites set in clear pale yellow glass (Fig. 2a and b; Fig. 3), sometimes with plagioclase microlites. The rare evolved rocks also have plagioclase glomerocrysts in the glass. The glassy rocks typically consist of 70% glass and 30% crystals; no glassy rocks are truly aphyric.
|
|
|
|
, 17 analyses); two xenocrysts and a late groundmass crystal are different.
|
|
Groundmass minerals are plagioclase, olivine, clinopyroxene and Fe–Ti oxides. Pillows from pillow breccias have chilled glass rims of 0·5–1 cm width and interiors with aphanitic to intersertal textures (Fig. 2a and b). Groundmasses in lava flows are variably fine-grained intersertal, coarse-grained subophitic, or intergranular. Although the crystalline groundmasses frequently contain abundant elongate–skeletal olivine crystals (Fig. 2c) and sometimes show feathery quench textures, spinifex textures have not been observed. Vesicle fillings and alteration minerals include various zeolites and Ca-hydrosilicates, clays, carbonate and silica phases.
Microprobe analyses of phenocrysts, glass inclusions and matrix glasses were made on a JEOL Superprobe at the University of Copenhagen, using either wavelength-dispersive (WDS) or energy-dispersive (EDS) analytical systems, and a combination of natural and synthetic standards. Normal operating conditions were 15 kV accelerating voltage, 15 nA beam current, 20 s total counting time for WDS and 60 s lifetime for EDS analyses. High-precision analyses of the minor elements Ca, Cr, Ni and Ti in olivine and Ni, Ti and V in chromite were made using 100 nA beam current and 40 s total counting time (WDS); the lower limits of detection for these conditions are 15–30 ppm (Pedersen, 1985, p. 35). Tiny glass inclusions in olivine were analysed with 2 nA beam current (EDS) to minimize volatilization during analysis. Data on the tabulated samples are given in the Appendix. Large datasets for olivines, chromites and glasses are given as supplementary data on the Journal of Petrology web site, at http://www.petrology.oupjournals.org.
Olivine
The olivines are texturally complex (Fig. 3), and several types of olivine are found within a single sample. Crystals of all types may contain scattered inclusions of glass and chromite.
The most frequent olivines are equant to somewhat elongated, clear euhedral to subhedral crystals of all sizes (Fig. 3a, b and d). The larger crystals often form clusters of several individuals. Strained, kink-banded crystals are occasionally observed but are rare. Some crystals have rounded or embayed outlines and may enclose patches of groundmass, making them more or less skeletal. Following Drever & Johnston (1957) and Donaldson (1976) this is interpreted as a result of rapid primary growth and generally not as a result of resorption. The morphological types described by Donaldson (1976) as hopper olivine are frequently observed.
Some olivine crystals have inclusion-filled zones (‘necklaces’) separating the crystals into cores with rounded outlines and rims (Fig. 3e). The compositional zoning continues with little or no interruption across the inclusion zone.
Other olivine crystals have dark, dusty cores crowded with tiny inclusions (often below microprobe resolution) of oxide and glass (Fig. 3b, f and g). Such inclusion-filled cores may be either euhedral or rounded, and they are surrounded by clear rims. Cores of this type show some within-grain compositional variability, partly but not completely because of variable success in avoiding the inclusions during microprobe analysis. The compositional range is mostly within the range of the clear olivines, and the clear rims are of normal groundmass composition. A few such inclusion-filled crystals or clusters are found in more than half of all thin sections.
Rare anhedral olivines have broken or ragged outlines and may be overgrown with a fringe of small olivine crystals (Fig. 3c and d). Reverse compositional zoning confirms the xenocrystic nature of these crystals.
Quenched olivine crystals may be tiny and near-equant hoppers (Fig. 2b) but are often elongated up to 8 mm long and 0·5 mm wide, and skeletal to highly skeletal, consisting of parallel-set rods and blebs in optical continuity (chain and parallel-growth olivines, Donaldson, 1976) (Fig. 2a).
Complex olivine populations similar to those in the Vaigat Formation have been described from Hawaii by, for example, Maaløe & Hansen (1982), Helz (1987), Baker et al. (1996) and Garcia (1996), one difference being the scarcity of kink-banded olivine in the Vaigat Formation. Drever & Johnston (1957) discussed and illustrated many of the same features in olivines from the Vaigat Formation on Ubekendt Ejland, West Greenland, north of Nuussuaq.
Olivine compositions
Representative olivine compositions are given in Table 1. The olivines are denoted by their mg-number, i.e. atomic 100Mg/(Mg + Fe). The Vaigat Formation olivines span the compositional range mg-number 76·5–93·3, with the major part of the analyses in the range mg-number 85–89 (Fig. 4). Around half of the samples with microprobed phenocrysts contain olivines with very Mg-rich cores with mg-number >90 (see Appendix). These crystals are always near-equant and usually (but not always) among the larger crystals in a section (Fig. 2); they are mostly undeformed and contain inclusions of chromite and glass (Fig. 3h). They are visually indistinguishable from similar olivine crystals with mg-number <90. The Mg-rich olivines normally constitute <1% of a thin section, and in the histogram (Fig. 4) they are overrepresented because they were searched for.
|
Olivines with mg-number >90 are a common feature of komatiites (Arndt et al., 1977; Nisbet et al., 1977; Echeverria, 1980; Lesher, 1989) and are also reported from magnesian lavas in arc settings (e.g. Eggins, 1993; Kamenetsky et al., 1995a, 1995b). They are only rarely reported from mid-ocean ridge basalt (MORB) (Donaldson & Brown, 1977; Kamenetsky et al., 1998), ocean island basalt (OIB) or continental basalts. Olivines with up to mg-number 93 have been described from the picrites on Baffin Island (Francis, 1985; Kent et al., 1998). On Hawaii, the well-analysed olivine populations do not exceed mg-number 90·7 (Maaløe & Hansen, 1982; Yang et al., 1994; Baker et al., 1996; Garcia, 1996), and crystals with mg-number >90 appear to be very rare.
The olivines of the Vaigat Formation contain significant amounts of the minor elements Ca, Cr and Ni. There is a very good correlation between the contents of these elements and the mg-number of the olivine, with Ca increasing and Cr and Ni decreasing with decreasing mg-number (Fig. 5). The Mg-rich olivines with mg-number >90 contain 0·23–0·35 wt % CaO and 0·09–0·19 wt % Cr2O3. Together with the glass inclusions this clearly shows that the Mg-rich olivines crystallized from melts. They are not xenocrysts of mantle olivines, which have very low contents of Ca and Cr (Fig. 5; Gurenko et al., 1996). Further, the smooth and continuous trends through the whole olivine compositional range shown in Fig. 5 strongly suggest that the Mg-rich olivine crystals are cognate phenocrysts and not accidental xenocrysts.
|
Olivine zoning
Most olivine phenocrysts are normally zoned with steepening or linear compositional gradients towards the margins (Fig. 6a–d); most of the larger cores have compositions of mg-number 87–89. Even the skeletal quench olivines are normally zoned (Fig. 6e). In the glassy rocks the most iron-rich olivine rims bordering the matrix glass have compositions of mg-number 82–85, whereas more iron-rich rims occur in the crystalline lava samples. The Mg-rich olivines (mg-number >90) differ from the other olivines in having cores with constant composition and normally zoned rims, giving flat-topped compositional profiles (Fig. 6a and b). The significance of the different zoning profiles in Fig. 6 is discussed later in the section on phenocryst–melt relations. In a few samples the olivines are unzoned with all compositions within a narrow interval around mg-number 86–87, even though these rocks show exactly the same petrographic diversity of the olivines as other samples.
|
Olivine crystals even within the same sample have individual zoning trends with respect to the minor elements: for the same mg-numbers the crystals have different levels of Ca, Cr and Ni (Fig. 7). A few normally zoned crystals have thin reversely zoned rims. The petrographically identified xenocrysts have outlying compositions and are often reversely zoned. The different levels of minor elements suggest that olivine crystals within the same sample did not crystallize from completely identical melts, as discussed later.
Chromite
Chromite forms euhedral to rounded crystals most frequently 10–50 µm in size, enclosed in or attached to olivine, and sometimes also ‘free’ in the matrix glass (Fig. 3). The crystals enclosed in olivine are unzoned or slightly zoned, but individual crystals differ somewhat in composition from one host olivine to the other. The chromite crystals enclosed in matrix glass may show more pronounced zoning and in some cases signs of oxidation and re-equilibration along the rims.
Chromite compositions
Representative chromite analyses are given in Table 2. The compositional range encountered in unaltered chromites is mg-number 45·4–77·2, cr-number 44·9–69·5. TiO2 is <1 wt % in the primary chromites but may be up to 3 wt % in re-equilibrated rims. There is a positive correlation between mg-number in coexisting olivine and chromite (Fig. 8). The chromites enclosed in olivine with mg-number >90 form a distinct high-Mg group with more Mg and less Fe2+, and more Al + Cr and less Fe3+, than the other chromites (Fig. 9). The major substitutions from the high-Mg group to the main group are Mg 
Fe2+, (Al + Cr) Fe3+ and Al 
Cr. The high-Mg chromites have unusually low contents of total iron (FeO* = 12–20 wt %). They have somewhat lower Fe3+/Fe2+ ratios than the main group of chromites although the intervals overlap. In a plot of cr-number in chromite vs mg-number in coexisting olivine (Fig. 8) the high-Mg chromites fall within the array for mantle spinels defined by Arai (1994a, 1994b), whereas the main group of chromites fall outside the mantle array at near-constant cr-number, the fractional crystallization trend of Dick & Bullen (1984) and Arai (1994b).
|
|
|
The main group of chromites are broadly similar to some Cr-spinels from MORBs (Dick & Bullen, 1984
; Allan, 1992, 1994). They have higher cr-number than spinels from most abyssal peridotites and fall on an extension of Dick & Bullen’s (1984) abyssal peridotite trend. The high-Mg chromites plot on the high-temperature side of the abyssal peridotite trend, suggesting very high crystallization temperatures.
The high-Mg chromites do not have many counterparts in the literature. Chromites from the Gorgona komatiites (Echeverria, 1980) are broadly similar. In Archaean komatiites, most chromites have higher Fe and Cr and lower Mg and Al (Arndt et al., 1977; Nisbet et al., 1977; Lesher, 1989). The chromite populations from Belingwe and Mt Keith extend to Mg-rich and Fe-poor compositions similar to the high-Mg chromites in West Greenland (Zhou & Kerrich, 1992; Barnes, 1998). However, the high-Mg chromites from Belingwe have much higher cr-numbers whereas those from Mt Keith have lower fe-numbers. Chromites included in olivine mg-number >90 from Vulsini have similar Fe and Mg but higher Cr and lower Al (Kamenetsky et al., 1995a). Some of the best analogues from natural rocks are chromites from harzburgites (Haggerty, 1976, 1991; Arai, 1994b; Allan & Dick, 1996; Bernstein et al., 1998), which might suggest that the primary melts left a harzburgitic residue. It is, however, improbable that the Vaigat Formation chromites are un-equilibrated restites from mantle melting. Some of them are euhedral crystals sitting in melt inclusions within the Mg-rich olivines (Fig. 3h).
The very wide range of mg-number in the chromites from West Greenland is matched only by chromites in komatiites (Barnes, 1998). Likewise, the only matching experimental chromites are those formed by melting experiments on a komatiite from Kambalda (Murck & Campbell, 1986). In Fig. 10 the West Greenland and the experimental chromites are compared. The Mg and Fe2+ variations, which are temperature dependent, are similar, especially considering that the lowest crystallization temperatures for the West Greenland chromites are around 1200°C (see later) compared with 1300°C as the lowest for the Kambalda experiments. High-Mg chromites and olivines with mg-number >90 formed in the experiments at temperatures of 1350–1500°C in equilibrium with melts with 18·2–25·6 wt % MgO, and similar conditions are expected to apply for the West Greenland magmas. However, Cr and Al behave differently in the high-temperature ends in Fig. 10. In the 1 atm experiments, rising temperature leads to uniformly increasing Cr and decreasing Al. The two Greenland examples behave like the experiments with rising temperature (increasing olivine mg-number) only up to a certain level (olivine mg-number 91 and 88) above which Cr decreases and Al increases. We suspect that this is an effect of pressure, the high-Mg olivines and enclosed chromites having crystallized at considerable depths.
Glass
Matrix glasses
The fresh glass in the samples is pale yellow and transparent, and often shows curved fractures (Fig. 3a–d). Incipient devitrification is seen as dark brown feathery patches, and alteration to green clay is often concentrated in and around vesicles.
The compositions of representative matrix glasses and their corresponding bulk rocks are presented in Table 3 and plotted in Fig. 11. The analytical totals of the fresh glasses are in the range 97·5–100·2% (average 98·8%), indicating that the glasses are not significantly hydrated. Separated glasses have contents of H2O + CO2 in the range 0·15–0·37 wt %. Within any sample the matrix glass is fairly homogeneous. The glasses span a relatively narrow range of MgO contents, 7·5–8·8 wt % for the olivine-phyric rocks and 7·1–7·6 wt % for the plagioclase-phyric rocks. There is a wider relative range in FeO*, 9·4–11·6 wt %, and an even larger relative range in TiO2, 1·3–1·9 wt %. The compositional ranges of TiO2 and FeO* in the glasses reflect corresponding ranges in the bulk rocks, as shown by the near-parallel glass–bulk-rock tie-lines in Fig. 11: the most Ti- and Fe-rich bulk rocks also have the most Ti- and Fe-rich glasses, and the tie-lines simply reflect the presence of olivine in the bulk rocks. The compositional ranges in TiO2 and FeO* in the glasses and bulk rocks must reflect similar ranges in the primary mantle melts.
|
|
The CaO contents in the glasses are high, 12·5–14·0 wt % CaO, as expected after prolonged olivine crystallization. The reason for the quenching of the melts within a narrow interval of MgO is probably the approach of plagioclase saturation. The steep trends defined by the glasses in Fig. 11 are interpreted as olivine–plagioclase cotectics (Francis, 1985), which the olivine-fractionating liquids reached along the individual liquid lines of descent. Only the plagioclase-phyric melts have started to move along the cotectic trend with steep TiO2 and FeO* enrichment. The plagioclase-phyric melts have precipitated plagioclase of composition An88–74.
Glass inclusions
The olivine phenocrysts contain glass inclusions, which have been analysed in four samples (Table 4). The inclusions are rounded or elongate, up to 70 µm in size, and sometimes contain a shrinkage bubble. Most inclusions contain a euhedral to rounded chromite crystal (Fig. 3h), which, because of its size, must have been present before the entrapment. It appears that many glass inclusions were formed in connection with inclusion of a chromite crystal into the growing olivine.
In contrast to the homogeneous matrix glasses the inclusion glasses show considerable compositional variation, with MgO = 3·0–8·7 wt %, FeO* = 4·2–9·3 wt %, and SiO2 = 49·1–55·0 wt %. They have high CaO (12·6–15·5 wt %) and Al2O3 (13·9–17·8 wt %), with CaO/Al2O3 ratios fairly constant around 0·80–0·90. Compositional variability of glass inclusions in contrast to matrix glasses seems to be a common phenomenon (Sobolev & Shimizu, 1994; Kamenetsky et al., 1995a, 1998; Gurenko et al., 1996; Kent et al., 1998). It has been ascribed partly to initial melt differences and partly to post-entrapment modification of the inclusions. Post-entrapment modification arises through two processes: (1) precipitation of olivine from the melt onto the inclusion walls; (2) diffusional homogenization of the olivine, with diffusion of Fe from the more Fe-rich olivine of the precipitated inclusion rim into the mass of the host olivine, and ensuing re-equilibration of the trapped melt (Gurenko et al., 1996). By these processes the glass inclusion loses Fe to the large reservoir of the host olivine, and the lost Fe is not available for dissolution back into the inclusion during high-T homogenization experiments. Therefore both homogenized and unhomogenized glass inclusions often have lower Fe contents than expected from bulk-rock compositions (Gurenko et al., 1996). To arrive at original melt compositions the lost iron has to be calculated back in by a stepwise procedure (Gurenko et al., 1996).
Figure 12 shows the results of the combined effects of olivine crystallization and diffusional re-equilibration on the composition of the glass inclusions. As a result of re-equilibration, the Fe contents in the glass inclusions and the host olivines are strongly positively correlated, whereas the increased SiO2 in the inclusions relative to the matrix glasses is a result of crystallization of olivine on the inclusion walls. The glass inclusions in the most Mg-rich olivines tend to have the highest SiO2 contents and thus to have crystallized most olivine, which is explicable because these inclusions have been entrapped through the largest temperature interval before quenching.
|
, from the same samples as the glass inclusions). (Note the general trend of decreasing FeO* and increasing SiO2 for inclusions in increasingly magnesian olivines, whereas the MgO contents vary more irregularly.) In both diagrams the Mg-rich olivines and their inclusions are end members of smooth series, indicating that all are comagmatic. (See text for discussion.)The contents of MgO in the glass inclusions are not dependent on the composition of the host olivines (Fig. 12). Rather, the smaller inclusions have lost most MgO and the larger inclusions (>50 µm) the least. The small inclusions also tend to have the highest SiO2 but some large ones also have high SiO2.
Figure 13 shows clearly that there is no magmatic equilibrium between most of the inclusion glasses and their host olivines. Equilibrium requires an olivine–liquid FeO–MgO distribution coefficient (KD) close to 0·3 (Roeder & Emslie, 1970; Ulmer, 1989), and oxidation of 15% of the Fe in the glasses (see later) brings only the matrix glasses and a few inclusion glasses down to the equilibrium KD line. It is necessary to oxidize up to 60% of the iron in some of the glass inclusions to bring them all down to the equilibrium line, and this is not realistic for magmatic conditions.
|
|
CRYSTALLIZATION CONDITIONS FOR THE MATRIX GLASSES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
Whereas the composition and P–T relations for the parental melts may be subject to discussion, the end points of crystallization, represented by the glassy rocks (
70 vol. % glass), are fairly certain. The melts were quenched close to 1 atm pressure to glasses in equilibrium with olivine and chromite, and the quench temperatures and oxygen fugacities can be calculated.
Glass quench temperatures
Quench temperatures can be calculated from equilibrium pairs of matrix glass and olivine rims bordering the glass, if the oxidation state of the melt is known or inferred. For reasons given below we used atomic Fe3+/
Fe = 0·15 for the glasses. Ford et al. (1983) provided an olivine–liquid thermometer together with a useful method for calculating a theoretical equilibrium olivine and the temperature (Tsum) for any given melt composition. We used this method for calculating the temperatures of all the analysed matrix glasses because in many samples the necessary olivine rims were not analysed. As shown in Fig. 14, quench temperatures are in the range 1180–1210°C for the olivine-phyric glasses and 1160–1180°C for four plagioclase–olivine-phyric glasses. The calculated equilibrium olivines have mg-number 81·5–86·5 for the olivine-phyric glasses and mg-number 78·3–81·9 for the plagioclase–olivine-phyric glasses. In samples where calculated and measured olivines can be compared, they are concordant within less than one mg-number.
|
Glass oxidation states
Matrix glass–spinel equilibrium pairs provide independent estimates of the oxidation state of the quenched melts. Maurel & Maurel (1982) found for 1 atm experiments at T = 1180–1300°C and fO2 = 10–9–10–3 a simple relation between Fe2+/Fe3+ in spinel and coexisting melt:
 |
Fe for the glasses of 0·10–0·23. Eight glasses have Fe3+/Fe = 0·16–0·18, and the coexisting chromites have Fe3+/Fe = 0·38–0·41. The relatively high degree of oxidation is confirmed by Mössbauer analysis of one of the 12 glasses (264137, Table 3), which gave Fe3+/Fe(glass) = 0·18, where the Maurel & Maurel (1982) algorithm for the same sample gave Fe3+/Fe(glass) = 0·16. The lower and higher degrees of oxidation calculated for some samples are possibly real but can also be due to lack of equilibrium between the analysed phases.
Given the temperatures and the oxidation states of the quenched glasses, the oxygen fugacities can be calculated following Kilinc et al. (1983). The resulting fO2 values are in the range 10–6·3–10–7·1, which around 1180–1200°C corresponds to around one log10 unit above the nickel–nickel oxide (NNO) buffer and 1·5–2 log10 units above the fayalite–magnetite–quartz (FMQ) buffer. Comparison of the chromites with the experimental spinels of Murck & Campbell (1986) and Balhaus et al. (1991) similarly suggests that spinels with Fe3+/Fe around 0·4 at 1200°C formed above the NNO buffer. The experiments of Balhaus et al. were performed at high pressures, but the Fe3+/Fe ratios of the spinels are primarily dependent on the oxidation state of the melts and are only very slightly pressure dependent.
Christie et al. (1986) found that MORB pillow glasses have oxidation states around two log10 units below the NNO buffer, corresponding to Fe3+/Fe around 0·07, and that the crystalline pillow interiors are more oxidized with Fe3+/Fe around 0·19, 1–1·5 log10 units higher than the glasses but still below the NNO buffer. They suggested that the crystalline parts are oxidized as a result of hydrogen loss during crystallization.
The relatively high oxidation states of the Vaigat Formation glasses could perhaps be explained by the fact that most of the pillow breccias were formed after the subaerial lavas flowed into water so that they may have been degassed and oxidized before quenching. However, chromites enclosed in pre-eruption olivine phenocrysts with mg-number >89 have only slightly lower Fe3+/Fe than those in equilibrium with the glasses (Table 2). They show some variation in oxidation state (Fe3+/Fe = 0·23–0·48) but still correspond to redox conditions close to the NNO buffer. Thus, in accordance with Pedersen (1985), we suggest that the Vaigat Formation magmas were inherently more oxidized than MORB magmas and crystallized close to and above the NNO buffer. In the olivine–liquid calculations we have consequently used a value of Fe3+/Fe of 0·15 for the matrix glasses, which is a minimum estimate but facilitates the comparison with other datasets where this value is commonly used, for example, that of Albarède (1992). Similarly, Francis (1985) noted that in the Baffin Island picrites 15% iron in oxidized form is required for equilibrium between glass and olivine rims. Indeed, we can in several cases improve the match between measured and calculated olivine by oxidizing up to 20% of the iron in the glass.
|
Mg-RICH HIGH-TEMPERATURE MELTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
Phenocryst–melt relations
The occurrence of olivine-rich picritic rocks does not necessarily prove that Mg-rich melts existed because the rocks may consist of relatively Mg-poor melt with accumulated olivine crystals. However, the highly magnesian olivines with high contents of Ca and Cr and with glass inclusions, precluding the possibility that they are mantle xenocrysts, indicate that highly magnesian parental melts did exist. For an olivine–liquid Fe–Mg KD of 0·33 (see below), olivines with mg-number 92–93 are in equilibrium with melts with mg-number 79–81·5. The bulk rocks of the Vaigat Formation with mg-number 79–81 have 19–22 wt % MgO and could thus be representative of the parental melts, as concluded by Clarke (1970)
. However, the observed modal proportions of Mg-rich olivine in the rocks are far from the proportions expected for such melts. The expected modal proportions of olivine in a melt can be calculated by simple stepwise subtraction of equilibrium olivine. Such calculations show that a melt with a composition like a typical bulk rock with 20 wt % MgO will by fractional crystallization precipitate >10 vol. % olivine with mg-number >90. In strong contrast, such olivines constitute <1 vol. % of the actual rocks. The majority of the olivine phenocrysts in the rocks have compositions of mg-number 86–89 (Fig. 4), which would be in equilibrium with melts with 10–14 wt % MgO. There are two possible explanations for this (which are not mutually exclusive). Either, the ultramafic parental melts fractionated olivine with mg-number 93–86 at depth until there was 14–10 wt % MgO in the melts; most of the early-formed high-Mg olivines were left at depth, and the melts erupted charged with accumulated later-formed olivine with mg-number 89–86, bringing many of the bulk compositions back to as high MgO contents as those of the parental melts and sometimes higher (Fig. 1). Alternatively, the parental melts underwent closer to equilibrium crystallization at depth, with the earlier formed olivines being re-equilibrated by diffusion to successively more iron-rich compositions. The second explanation was favoured by Francis (1985), who concluded that the bulk rocks therefore can be considered representative of the original melts, and the Mg-rich olivines are xenocrysts. Our data show, however, that the Mg-rich olivines are integral parts of the cognate mineral assemblage of the West Greenland picrites (Figs 5, 7 and 12). Further, the olivine and chromite phenocrysts are to a large extent not re-equilibrated. The olivines are strongly zoned, and crystals with individually different compositions are preserved (Fig. 7).
The zoning profiles of the olivine phenocrysts shown in Fig. 6 provide clues to crystallization mechanisms. The smoothly curved profile in Fig. 6c, with steepening gradients towards the rims, closely matches the theoretical profile calculated for simple fractional crystallization. The very magnesian cores with mg-number >92 in Fig. 6a and b are virtually unzoned, suggesting crystallization closer to equilibrium conditions. The sharp slope breaks between cores and rims in these two crystals indicate change of mechanism. Sharp slope breaks and flat profile centres may also be seen in acentral crystal sections (Pearce, 1984), but this would mean that the true centres must have even higher mg-numbers, which is improbable. The two olivine crystals shown in Fig. 6a and d, from the same sample, have near-linear marginal zoning profiles. This can be produced either by an accelerating crystallization rate that fortuitously leads to linear profiles, or, more probably, by re-equilibration by diffusional exchange between olivine and liquid, as shown by Maaløe & Hansen (1982). The centre of the smaller crystal (as cut by the section) is modified throughout. The thin rims on the mg-number 92·5 core in Fig. 6b are probably produced by growth of new olivine during and after entrainment and eruption.
Maaløe & Hansen (1982) and Maaløe et al. (1989) described olivine zoning profiles in Hawaiian basalts very similar to those in Fig. 6b and c, and they concluded that the olivines formed under supercooled conditions and variable growth rates during successive stages of evolution of the melts. They did not find rims with linear zoning profiles like those in Fig. 6a and d, and they concluded that diffusional re-equilibration was minimal. The West Greenland data indicate that the olivines in some samples did partly re-equilibrate (e.g. 362148): the crystal shown in Fig. 6a consists in plane section, as revealed by scanning electron micrographs, of nearly equal areas of homogeneous core and re-equilibrated rims, and in three dimensions thus more than three-quarters of the original crystal is re-equilibrated. In rare samples (four found among 56 microprobed) the olivines equilibrated completely with the liquid (e.g. 264217, right panel in Fig. 7). In most cases, however, re-equilibration was minimal, and this process cannot account for the lack of olivine with mg-number >90. We must conclude that a large amount of high-Mg olivine fractionated out and was left behind in the conduit systems at depth during the magma’s ascent towards the surface.
Composition and temperature of the parental melts
The loss of most of the high-Mg olivines (mg-number >90) and the accumulation of more iron-rich olivines (mg-number <90) in the conduit systems during magma ascent show that the bulk rocks do not directly represent the parental melt compositions. By parental melts we understand the most primitive melts recorded in the mineral compositions; however, because the olivines go up to mg-number >92 these may be in equilibrium with mantle olivine and the parental melts may be close to also being primary mantle melts.
Because olivine and chromite are the only fractionating phases the compositions of the parental magnesian melts may be calculated by starting from the matrix glasses and adding equilibrium olivine back in successive steps until the equilibrium olivine has the same composition as the most magnesian olivine in a sample. This is a standard method used by many workers, e.g. Albarède (1992) and Sobolev & Nikogosian (1994). Results of such calculations for the Vaigat Formation are shown in Tables 5 and 6. The results are dependent on assumptions with regard to the most magnesian olivine, the oxidation state of the melts, and the pressure increase with depth. The most magnesian olivine was defined as mg-number 92·5 for all samples from the Anaanaa and Naujánguit Members, and mg-number 92·0 for all samples from the Ordlingassoq Member, because these values are the normal maximum values for the olivines in the respective formations. The minor elements Cr and Ni in the added olivine were adjusted in steps in accordance with the measured concentrations. In the matrix glasses, 15% of the iron was oxidized in accordance with the measured oxidation states; this corresponds to fO2 values one log-unit above the NNO buffer (NNO + 1). No further adjustments of the oxidation state were made, which means that fO2 decreased during olivine addition to close to and slightly above the NNO level at the end of calculations. Another possibility is to adjust the oxidation stage to a buffer curve, and this has pronounced effects on the results (Table 6), but there is no geological reason why the melt’s fO2 path should be parallel to the synthetic curves. The pressure increase with depth, required for the calculation of olivine KD values (Ulmer, 1989), was assumed to be linear at 0·2 kbar per 0·5 wt % olivine added because this ends the procedure at depths in the uppermost lithospheric mantle (13–16 kbar), but other choices are possible (Table 6). Equilibrium with olivine mg-number 92·0 or 92·5 was obtained when the calculated melts consisted of 28–34 wt % olivine and 66–72 wt % glass. Figure 15 illustrates that the composition of a calculated melt is rather sensitive to the composition of the olivine at which the procedure is stopped, and Table 6 illustrates the effects of varying fO2 and pressure in the calculations for one sample. However, the differences between the samples calculated with one set of assumptions are maintained when the assumptions are changed.
|
|
|
The calculated parental melts (Table 5) contain 20–21 wt % MgO, similar to or slightly higher than the bulk rocks. The differences are also small for elements that are incompatible during olivine fractionation, and all ratios of such elements (e.g. CaO/Al2O3) are preserved. The major difference between the calculated melts and the bulk rocks lies in FeO*, this being 0·5–1 wt % lower in the calculated liquids than in the bulk rocks. As shown by three examples in Fig. 16, this is because the equilibrium olivines added in the calculations become successively more Fe poor, and in liquids with more than 13 wt % MgO the added olivines (mg-number >88·5) contain less iron than the liquid. These Mg-rich, Fe-poor olivines constitute a major part of the olivine assemblage in the calculated liquids, whereas the olivines in the bulk rocks are more Fe rich, with a frequency peak at mg-number 86–88·5 (Fig. 4).
|
The content of Cr2O3 in the parental melt was estimated from Crol–Crliq partitioning because the modal amounts of chromite involved, which completely dominate the Cr inventory, are unconstrained. The most Mg-rich olivines contain 0·15 wt % Cr2O3, and a partition coefficient Crol–Crliq of 0·5 (Hanson & Jones, 1998) thus gives 0·30 wt % Cr2O3 in the parental melt. Therefore, 0·6 wt % chromite was added to the calculated parental melt to increase the Cr2O3 content to 0·30 wt %. This value is distinctly higher than that of the bulk rocks with 20–21 wt % MgO, which have on average 0·22 ± 0·03 wt % Cr2O3 (Table 3 and unpublished data, 2000). The result suggests that more chromite has been lost by fractionation than has been acquired by accumulation—which is geologically reasonable.
The content of NiO in the parental melt should be higher than in the bulk rocks because the bulk rocks contain less magnesian olivines with less NiO than the melts would have done. The difference is, however, only 0·02 wt % as estimated from the measured contents of NiO in the olivines. Accordingly, the calculated parental melts have 0·12 wt % NiO (Table 5), whereas the bulk rocks with 20–21 wt % MgO have on average 0·11 ± 0·01 wt % NiO (Table 3 and unpublished data, 2000). The NiO contents in the bulk rocks thus approximate those of the parental melts fairly well. With 0·46 wt % NiO in olivine mg-number 92 and 0·12 wt % NiO in the parental melt, olivine–liquid partition coefficients for Ni close to four are indicated. This is less than found in the experiments by Hart & Davis (1978), but similar to other high-temperature results summarized by Elthon & Ridley (1979).
The calculated temperatures for the parental melts are very high, 1500–1560°C. These results are supported by the composition of the Mg-rich olivines and chromites. Olivines as Cr rich as those in the Vaigat Formation are found only in komatiites and other ultramafic igneous rocks equilibrated at high temperatures, in accordance with the experimental data of Li et al. (1995) that suggest 1 atm temperatures of 1450°C for olivines with 0·1 wt % Cr2O3 at FMQ. Similarly, the 1 atm experiments on the Kambalda komatiite by Murck & Campbell (1986) produced olivine–chromite pairs similar to the high-Mg mineral pairs in the Vaigat Formation only at 1400°C or higher temperatures. Attempts to simulate crystallization of the measured olivine–chromite pairs from the calculated parental melts with the program SPINMELT (Ariskin & Nikolaev, 1996) resulted in temperatures of 1450–1570°C, but the calculated spinel compositions were variable and extremely dependent on small variations in the estimated Cr contents in the melts and the assumed pressures.
Mantle melting conditions
The West Greenland lavas were produced beneath a rifting continent. Holm et al. (1993) showed that the melts were produced in the asthenospheric mantle and found Sr and Nd isotopic differences between the Naujánguit and Ordlingassoq Members indicating more of a depleted MORB-type component in the Naujánguit Member and more of an Iceland-plume-type component in the Ordlingassoq Member. The calculated parental melts in Table 5 show compositional differences between the Anaanaa and Naujánguit Members on one hand and the Ordlingassoq Member on the other hand, with Ordlingassoq Member having higher FeO* and TiO2. These differences are present in the matrix glasses as well as in the bulk-rock analyses (Fig. 11) and must have arisen during mantle melting. Because the isotopes indicate variable amounts of different mantle components, whereas the calculated temperatures for all members are similar (Table 5), we suggest that the main cause for the differences in parental melt composition lies in different mantle source compositions whereas the melting processes were similar.
The parental melts calculated here have FeO* contents that are lower than those in the bulk rocks by 0·5–1·0 wt %. This is significant for the estimation of the depth of melting (e.g. Elliott et al., 1991). In melting models such as that of Langmuir et al. (1992) it would correspond to a pressure decrease of 5–10 kbar. Holm et al. (1993) suggested that a bulk-rock data trend extending to high levels of FeO* at constant Al2O3/TiO2 is a consequence of deep melting of the hot mantle plume source. However, we consider that it is solely a consequence of iron-rich olivine accumulation.
Segregation pressures for the primary melts have been estimated by Clarke (1970), Herzberg (1995) and Herzberg & Zhang (1996) at 30–40 kbar, based on bulk-rock CaO/Al2O3, and this result is still valid because CaO/Al2O3 is insensitive to compositional and modal variations in the olivines. Herzberg & O’Hara (1998) estimated average segregation pressures for the West Greenland picrites of 38 kbar and initial melting pressures of 45 kbar, and these estimates are also valid because the calculated parental liquids in Table 5 plot similarly to the bulk rocks with 20–21 wt % MgO in the CMAS projections used. The segregation pressures calculated with the simple algorithms of Albarède (1992) are 28–36 kbar, that is, somewhat lower in comparison.
Segregation temperatures calculated for the parental melts after Albarède (1992) are 1563–1606°C (Table 5). For the relevant pressures, this corresponds to mantle potential temperatures of 1520–1550°C, in excellent agreement with the estimates by Gill et al. (1992) and Herzberg & O’Hara (1998). The 50°C difference between the temperatures calculated at segregation depths and in the upper lithospheric mantle is an effect, partly of the pressure differences and partly of the latent heat of fusion.
|
PROCESSES IN THE CONDUIT SYSTEMS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
If the final pressures of segregation were
30 kbar the melts segregated at 140–100 km depth, indicating the existence of a lithospheric lid of 100 km thickness that curtailed the melting column. The melts thus ascended through roughly 100 km of conduit systems in the lithospheric mantle and crust. The depth interval of the conduit systems for which evidence of crystallization is preserved in the erupted lavas is unconstrained. We have put the deepest part at pressures of 13–16 kbar (Table 5), i.e. in the uppermost lithospheric mantle. Considering the shallower parts, the fractionated melts always crystallized plagioclase before clinopyroxene, as evidenced by the plagioclase-phyric basalts and rare small cognate xenoliths of troctolitic gabbros. Modelling with the computer program COMAGMAT (Ariskin et al., 1993) indicates that the two-phase cotectic was reached at pressures of 3 kbar or lower because at higher pressures clinopyroxene would crystallize before plagioclase. Thus, the vertical extent of the zone of major crystallization is estimated to be situated at depths from 40–50 km to <9 km, and in principle it reached to the surface.
The scarcity of picrites in most flood basalt successions has been ascribed to density trapping of the primary Mg-rich melts in sill-like complexes at the crust–mantle boundary where the olivines are deposited in ultramafic cumulates, and only the evolved basaltic melts are erupted (Cox, 1980). The eruption of picrites requires that a density trap does not exist or is circumvented, and the conduits are most likely to be steep and dyke like. In West Greenland, the conduits were also very crystal rich, and the preserved olivines and chromites provide evidence for complex open-system processes of equilibrium and fractional crystallization, magma mixing, oxidation, and partial to complete re-equilibration. We envisage that crystal-charged magma batches ascended through irregular, dyke-like conduits, precipitating high-Mg olivines (mg-number 90–93) and chromites (mg-number 69–77) at deeper levels and less magnesian olivines (mg-number 86–88) and chromites (mg-number <69) at shallower levels. At high magma ascent rates many olivine crystals were kept in suspension all the way to the surface, even some of the deep high-Mg olivines, whereas at low ascent rates only few and small olivines were suspended and the magmas erupted as olivine-microphyric to aphyric basalts. Pulsating magma ascent would lead to mixing of magma batches in various stages of fractionation and equilibration. In stagnating magma pockets crystallisation proceeded to low-Mg olivines (mg-number 85–78) and rare troctolites later picked up in passing magmas. However, the majority of the magmas passed so fast that equilibration was far from complete, and their crystal assemblage consists of mechanical mixtures of crystals both from several magma batches and picked up from the sidewall at all depths. The significant variability of the contents of the minor elements Ca, Cr and Ni in olivine crystals in some samples (Fig. 7) can be explained as an effect of fractionation and mixing of magma batches. This process introduces variabilities that are directly proportional to the bulk distribution coefficients (Albarède et al., 1997), and the variabilities should thus decrease in the order Cr (bulk D 6) – Ni (bulk D = 4) – Ca (bulk D = 0·028). This is in accordance with the observed variabilities in non-equilibrated samples (Fig. 7). Excess enrichment of incompatible elements in the melt need not be detectable because the magmas did not go through many cycles and a steady state was not attained, as indicated by the large spread in MgO content of the erupted magmas down to 6·5 wt % MgO (Fig. 1). It is clear, however, that the back-calculation with stepwise addition of equilibrium olivine to the matrix glass (Table 5) is a simplification that works best for elements with partition coefficients around one.
The magma ascent velocity can be estimated in one case. The olivine crystal shown in Fig. 6a has rims of 0·4 mm width with linear zoning profiles, which we ascribe to diffusional re-equilibration with the melt (Maaløe & Hansen, 1982). The diffusion data of Jurewich & Watson (1988) indicate that rims of this width at 1300°C would have formed in a period of 2 months. If the crystal formed in the deep part of the conduit system around, say, 30 km depth, this corresponds to a magma ascent velocity of 500 m/day, i.e. 0·6 cm/s. In comparison, Maaløe (1973) calculated ascent velocities of 1–10 cm/s for Hawaiian lavas carrying 2–20 cm olivine nodules. Considering that the olivines in the West Greenland picrites are normally not larger than 0·5 cm, the lower ascent velocity of 0·6 cm/s is in accordance with the Hawaii data.
Even the rare magma batches that erupted with completely re-equilibrated olivines apparently passed fast and without complete equilibration through the greater part of the conduits because their olivine assemblage is petrographically just as diverse as in the other rocks. The complete equilibration to olivine mg-number 86·5 must have taken place in high-level ephemeral magma chambers where the melts evolved to MgO contents around 10 wt % at 1250°C. The few plagioclase-phyric basalts would have developed in similar magma chambers.
The crystal-rich conduit systems envisaged here are sketched in Fig. 17. They have many features in common with the crystal-dominated ‘slim-line’ magma chambers proposed by Sinton & Detrick (1992) for mid-ocean ridges. These magma chambers are vertically extensive and consist of a central crystal-rich zone within which the magma can move, surrounded by a rigid, largely crystalline (>60% crystals) transition zone to the wall rocks. Despite the very different tectonic setting and lithosphere thickness of West Greenland, the magma conduits were apparently similar. The rifting continent may be viewed as a slow-spreading ridge environment with high magma supply rates, where the large magma volumes contributed less to lateral crustal accretion than they would do in an oceanic setting, and mainly formed new crust by aggradation at the surface. Features seen in West Greenland and noted by Sinton & Detrick (1992) as characteristic of slow-spreading ridges are the scarcity of evolved magmas, the frequency of mixed populations of crystals, and the association of magma conduits with syn-volcanic faults. However, there is no rift valley because of the high magma supply rate and the aggradation of the lava plateau. The vertical extent of the conduit system is much greater than at mid-ocean ridges because the lithosphere is much thicker. Sinton & Detrick (1992) noted that the MgO content of the cumulate crystals should increase with depth, similar to our interpretation for West Greenland. The applicability of similar magma chamber models to both oceanic spreading centres and a continental rift zone is supported by the work of Donaldson & Brown (1977), who described complex olivine populations very similar to those described here, but from MORBs. Despite the differences in lithosphere thickness, the processes in the conduit systems appear to be the same.
|
The conduit systems envisaged here for the Vaigat Formation would not have led to simple underplating but rather to an irregular vertical distribution of olivine cumulates across the mantle–crust boundary and throughout the crust, concentrated in zones of crustal extension (Fig. 17). Judged from the average MgO content of 15·5 wt % in the erupted lavas, on average 13 wt % of the parental melts was left behind as cumulate crystals in the conduit systems.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
The uncontaminated picritic volcanic rocks of the Vaigat Formation contain 6·5–30 wt % MgO (average 15·5 wt %) and contain olivines with up to mg-number 93. The olivine crystals show zoning patterns indicating equilibrium and fractional crystallization, oxidation, partial and complete re-equilibration, and magma mixing. All the olivine crystals contain glass inclusions and have high contents of Ca and Cr, indicating that they precipitated from melts; even the most Mg-rich olivines are cognate phenocrysts. Mantle xenocrysts have not been found. The most Mg-rich rocks were erupted as melt–olivine mixtures. The erupted melts contained no more than up to 14 wt % MgO and were charged with olivine crystals with mg-number 78–93, most frequently mg-number 86–88·5, and associated chromites. The melts quenched in pillow breccias to glasses with
8 wt % MgO at around 1200°C and log fO2 at NNO + 1, close to the olivine–plagioclase cotectic. The melts evolved along a number of individual liquid lines of descent; the differences were inherited from the primary melts and probably reflect compositional differences in the mantle sources.
The parental melts were not erupted. Their compositions were calculated by stepwise addition of equilibrium olivine to the matrix glasses; they had 20–21 wt % MgO at 1500–1560°C and fO2 around NNO, and they are close to being primary magmas. Erupted rocks with around 20 wt % MgO as a result of accumulated olivine are similar to the parental melts in many respects, especially in incompatible and moderately compatible elements, but the rocks have higher FeO* contents than the melts by 0·5–1 wt % because most of the high-Mg olivines with mg-number 90–93 were lost in the conduit systems. Therefore, petrogenetic conclusions based on bulk-rock FeO* contents (Holm et al., 1993) should be viewed with caution. Conclusions on melting depths based on CaO/Al2O3 and CMAS projections (Clarke, 1970; Herzberg, 1995; Herzberg & Zhang, 1996; Herzberg & O’Hara, 1998) are still valid.
The primary melts ascended in conduit systems of around 100 km vertical extent in the lithospheric mantle and crust, and no density traps such as sill-like magma pools at the crust–mantle boundary were in effect. Crystallization of olivine and chromite, preserved in the eruption products, took place from assumed depths of around 45 km and to the surface. The crystal-charged magmas precipitated high-Mg olivines and chromites at deep levels and less magnesian crystals at shallower levels. Rapidly ascending magma batches carried large amounts of suspended crystals to the surface whereas slowly ascending magma batches erupted free of accumulated crystals. Pulsating magma ascent rates led to mixing of magma batches in various stages of fractionation and equilibration. On average, 13 wt % of the parental melts were left behind in the conduit systems as cumulates, irregularly distributed across the mantle–crust boundary to high levels in the crust. At this early stage of the volcanism, crustal underplating would be insignificant.
The conduit systems of the Vaigat Formation have many features in common with the crystal-rich magma chambers proposed by Sinton & Detrick (1992) for mid-ocean ridges. Despite the differences in lithosphere thickness, the rifting continent in West Greenland may be viewed as a slow-spreading ridge with high magma supply rates where the large magma volumes mainly formed new crust by aggradation of the lava plateau.
|
APPENDIX: SAMPLE DATA, VAIGAT FORMATION, WEST GREENLAND |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
|
|
ACKNOWLEDGEMENTS |
|---|
We are grateful to Jørn Rønsbo for assistance with the set-up of the microprobe, particularly for the trace-element analyses, and to Christian Bender-Koch for the Mössbauer analyses of glass separates. Stefan Bernstein, Henriette Hansen and Sven Maaløe provided constructive comments and discussions, as did the reviewers Nick Arndt, Miranda Fram and Don Francis. The Arctic Station in Godhavn, and the Geological Survey of Greenland’s expedition leaders Flemming G. Christiansen and Feiko Kalsbeek provided extensive support during field work. The Danish Natural Science Research Council provided the microprobe facility. Publication of this paper is authorized by the Geological Survey of Denmark and Greenland.
|
FOOTNOTES |
|---|
*Corresponding author. Telephone: +45-38142252. Fax: +45-38142050. e-mail: lml{at}geus.dk
Extended data set can be found at: http://www.petrology.oupjournals.org
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERALOGYCRYSTALLIZATION CONDITIONS FOR...Mg-RICH HIGH-TEMPERATURE MELTSPROCESSES IN THE CONDUIT...CONCLUSIONSAPPENDIX: SAMPLE DATA, VAIGAT...REFERENCES |
|---|
Aitken, B. G. & Echeverria, L. M. (1984). Petrology and geochemistry of komatiites and tholeiites from Gorgona Island, Colombia. Contributions to Mineralogy and Petrology 66, 415–427.
Albarède, F. (1992). How deep do common basaltic magmas form and differentiate? Journal of Geophysical Research 97(B), 10997–11009.
Albarède, F., Luais, B., Fitton, J. G., Semet, M., Kaminski, E., Upton, B. G. J., Bachelèry, P. & Cheminée, J.-L. (1997). The geochemical regimes of Piton de la Fournaise volcano (Réunion) during the last 530 000 years. Journal of Petrology 38, 171–201.
Allan, J. F. (1992). Cr-spinel as a petrogenetic indicator: deducing magma composition from spinels in highly altered basalts from the Japan Sea, Sites 794 and 797. In: Tamaki, K., Suyehiro, K., Allan, J. F. & McWilliams, M. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 127/128. College Station, TX: Ocean Drilling Program, pp. 837–847.
Allan, J. F. (1994). Cr-rich spinel in depleted basalts from the Lau Basin backarc: petrogenetic history from Mg–Fe crystal–liquid exchange. In: Hawkins, J., Parson, L. & Allan, J. F. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 135. College Station, TX: Ocean Drilling Program, pp. 565–583.
Allan, J. F. & Dick, H. J. B. (1996). Cr-rich spinel as a tracer for melt migration and melt–wall rock interaction in the mantle: Hess Deep, Leg 147. In: Mével, C., Gillis, K. M., Allan, J. F. & Meyer, P. S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 147. College Station, TX: Ocean Drilling Program, pp. 157–172.
Arai, S. (1994a). Compositional variation of olivine–chromian spinel in Mg-rich magmas as a guide to their residual spinel peridotites. Journal of Volcanology and Geothermal Research 59, 279–293.[Web of Science]
Arai, S. (1994b). Characterization of spinel peridotites by olivine–spinel compositional relationships: review and interpretation. Chemical Geology 113, 191–204.[Web of Science]
Ariskin, A. & Nikolaev, G. S. (1996). An empirical model for the calculation of spinel–melt equilibria in mafic igneous systems at atmospheric pressure: 1. Chromian spinels. Contributions to Mineralogy and Petrology 123, 282–292.
Ariskin, A. A., Frenkel, M. Y., Barmina, G. S. & Nielsen, R. L. (1993). COMAGMAT: a FORTRAN program to model magma differentiation processes. Computers and Geosciences 19, 1155–1170.
Arndt, N. T. (1986). Differentiation of komatiite flows. Journal of Petrology 27, 279–301.
Arndt, N. T., Naldrett, A. J. & Pyke, D. R. (1977). Komatiitic and iron-rich tholeiitic lavas of Munro Township, northeast Ontario. Journal of Petrology 18, 319–369.
Baker, M. B., Alves, S. & Stolper, E. (1996). Petrography and petrology of the Hawaii Scientific Drilling Project lavas: inferences from olivine phenocryst abundances and compositions. Journal of Geophysical Research 101(B), 11715–11727.
Balhaus, C., Berry, R. F. & Green, D. H. (1991). High pressure experimental calibration of the olivine–orthopyroxene–spinel oxygen geobarometer: implications for the oxidation state of the upper mantle. Contributions to Mineralogy and Petrology 107, 27–40.
Barnes, S. J. (1998). Chromite in komatiites, 1. Magmatic controls on crystallization and composition. Journal of Petrology 39, 1689–1720.
Bernstein, S., Kelemen, P. B. & Brooks, C. K. (1998). Depleted spinel harzburgite xenoliths in Tertiary dykes from East Greenland: restites from high-degree melting. Earth and Planetary Science Letters 154, 221–235.[Web of Science]
Bristow, J. W. (1984). Picritic rocks of the North Lebombo and South-East Zimbabwe. In: Erlank, A. K. (ed.) Petrogenesis of the Volcanic Rocks of the Karoo Province. Geological Society of South Africa, Special Publication 13, 105–123.
Campbell, I. H. & Griffiths, R. W. (1990). Implications of mantle plume structure for the evolution of flood basalts. Earth and Planetary Science Letters 99, 79–93.
Campbell, I. H. & Griffiths, R. W. (1992). The changing nature of mantle hotspots through time: implications for the chemical evolution of the mantle. Journal of Geology 92, 497–523.
Chalmers, J. A., Pulvertaft, T. C. R., Marcussen, C. & Pedersen, A. K. (1999). New insight into the structure of the Nuussuaq Basin, central West Greenland. Marine and Petroleum Geology 16, 197–224.
Chen, C.-Y. (1993). High-magnesium primary magmas from Haleakala Volcano, east Maui, Hawaii: petrography, nickel, and major-element constraints. Journal of Volcanology and Geothermal Research 55, 143–153.
Christie, D. M., Carmichael, I. S. E. & Langmuir, C. H. (1986). Oxidation states of mid-ocean ridge basalt glasses. Earth and Planetary Science Letters 79, 397–411.
Clarke, D. B. (1970). Tertiary basalts of the Baffin Bay: possible primary magma from the mantle. Contributions to Mineralogy and Petrology 25, 203–224.
Clarke, D. B. & O’Hara, M. J. (1979). Nickel and the existence of high-MgO liquids in nature. Earth and Planetary Science Letters 44, 153–158.
Clarke, D. B. & Pedersen, A. K. (1976). Tertiary volcanic province of West Greenland. In: Escher, A. & Watt, W.S. (eds) Geology of Greenland. Copenhagen: Geological Survey of Greenland, pp. 364–385.
Cox, K. G. (1980). A model for flood basalt volcanism. Journal of Petrology 21, 629–650.
Cox, K. G., Duncan, A. R., Bristow, J. W., Taylor, S. R. & Erlank, A. J. (1984). Petrogenesis of the basic rocks of the Lebombo. In: Erlank, A. K. (ed.) Petrogenesis of the Volcanic Rocks of the Karoo Province. Geological Society of South Africa, Special Publication 13, 149–169.
Dick, H. J. B. & Bullen, T. (1984). Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology 86, 54–76.[Web of Science]
Donaldson, C. H. (1976). An experimental investigation of olivine morphology. Contributions to Mineralogy and Petrology 57, 187–213.[Web of Science]
Donaldson, C. H. & Brown, R. W. (1977). Refractory megacrysts and magnesium-rich melt inclusions within spinel in oceanic tholeiites; indicators of magma mixing and parental magma composition. Earth and Planetary Science Letters 37, 81–89.[Web of Science]
Drever, H. I. & Johnston, R. (1957). Crystal growth of forsteritic olivine in magmas and melts. Transactions of the Royal Society of Edinburgh 63, 289–315.
Echeverria, L. M. (1980). Tertiary or Mesozoic komatiites from Gorgona Island, Colombia: field relations and geochemistry. Contributions to Mineralogy and Petrology 73, 253–266.
Eggins, S. M. (1993). Origin and differentiation of picritic arc magmas, Ambae (Aoba), Vanuatu. Contributions to Mineralogy and Petrology 114, 79–100.
Elliott, T. R., Hawkesworth, C. J. & Grönvold, K. (1991). Dynamic melting of the Iceland plume. Nature 351, 201–206.
Elthon, D. & Ridley, W. I. (1979). Comments on: ‘The partitioning of nickel between olivine and silicate melt’ by S. R. Hart and K. E. Davis. Earth and Planetary Science Letters 44, 162–164.
Escher, J. C. & Pulvertaft, T. C. R. (1995). Geological Map of Greenland 1:2 500 000. Copenhagen: Geological Survey of Greenland.
Ford, C. E., Russell, D. G., Craven, J. A. & Fisk, M. R. (1983). Olivine–liquid equilibria: temperature, pressure, and composition dependence of the crystal/liquid cation partition coefficients for Mg, Fe2+, Ca and Mn. Journal of Petrology 24, 256–265.
Fram, M. S. & Lesher, C. E. (1997). Generation and polybaric differentiation of East Greenland Early Tertiary flood basalts. Journal of Petrology 38, 231–275.
Francis, D. (1985). The Baffin Bay lavas and the value of picrites as analogues of primary magmas. Contributions to Mineralogy and Petrology 89, 144–154.
Francis, D. (1995). The implications of picritic lavas for the mantle sources of terrestrial volcanism. Lithos 34, 89–105.
Frost, B. R. (1991). Introduction to oxygen fugacity and its petrological significance. In: Lindsley, D. H. (ed.) Oxide Minerals: Petrologic and Magnetic Significance. Mineralogical Society of America, Reviews in Mineralogy 25, 1–9.
Garcia, M. O. (1996). Petrography and olivine and glass chemistry of lavas from the Hawaii Scientific Drilling Project. Journal of Geophysical Research 101(B), 11701–11713.
Gill, R. C. O., Pedersen, A. K. & Larsen, J. G., (1992). Tertiary picrites in West Greenland: melting at the periphery of a plume? In: Storey, B. C., Alabaster, T. & Pankhurst, R. J. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publication 68, 335–348.
Goodrich, C. A. & Patchett, P. J. (1991). Nd and Sr isotope chemistry of metallic iron-bearing, sediment-contaminated Tertiary volcanics from Disko Island, Greenland. Lithos 27, 13–27.
Gurenko, A. A., Hansteen, T. H. & Schmincke, H.-U. (1996). Evolution of parental magmas of Miocene shield basalts of Gran Canaria (Canary Islands): constraints from crystal, melt and fluid inclusions in minerals. Contributions to Mineralogy and Petrology 89, 422–435.
Haggerty, S. E. (1976). Opaque mineral oxides in terrestrial igneous rocks. In: Rumble, D. (ed.) Oxide Minerals. Mineralogical Society of America, Reviews in Mineralogy 3, Hg101–Hg300.
Haggerty, S. E. (1991). Oxide mineralogy of the upper mantle. In: Lindsley, D. H. (ed.) Oxide Minerals: Petrologic and Magnetic Significance. Mineralogical Society of America, Reviews in Mineralogy 25, 355–416.
Hansen, H. & Nielsen, T. F. D. (1999). Crustal contamination in Palaeogene East Greenland flood basalts: plumbing system evolution during continental rifting. Chemical Geology 157, 89–118.
Hanson, B. & Jones, J. H. (1998). The systematics of Cr3+ and Cr2+ partitioning between olivine and liquid in the presence of spinel. American Mineralogist 83, 669–684.[Abstract]
Hart, S. R. & Davis, K. E. (1978). Nickel partitioning between olivine and silicate melt. Earth and Planetary Science Letters 40, 203–219.[Web of Science]
Hart, S. R. & Davis, K. E. (1979). Reply to D. B. Clarke and M. J. O’Hara, ‘Nickel and the existence of high-MgO liquids in nature’. Earth and Planetary Science Letters 44, 159–161.
Helz, R. T. (1987). Diverse olivine types in lava of the 1959 eruption of Kilauea volcano and their bearing on eruption dynamics. US Geological Survey, Professional Paper 1350, 691–722.
Herzberg, C. (1995). Generation of plume magmas through time: an experimental perspective. Chemical Geology 126, 1–16.
Herzberg, C. & O’Hara, M. J. (1998). Phase equilibrium constraints on the origin of basalts, picrites, and komatiites. Earth-Science Reviews 44, 39–79.
Herzberg, C. & Zhang, J. (1996). Melting experiments on anhydrous peridotite KLB-1: compositions of magmas in the upper mantle and transition zone. Journal of Geophysical Research 101(B), 8271–8295.
Hill, R. I. (1991). Starting plumes and continental break-up. Earth and Planetary Science Letters 104, 398–416.
Holm, P. M., Gill, R. C. O., Pedersen, A. K., Larsen, J. G., Hald, N., Nielsen, T. F. D. & Thirlwall, M. F. (1993). The Tertiary picrites of West Greenland: contributions from ‘Icelandic’ and other sources. Earth and Planetary Science Letters 115, 227–244.[Web of Science]
Jurewich, A. J. G. & Watson, E. B. (1988). Cations in olivine, part 1: Calcium partitioning and calcium–magnesium distribution between olivines and coexisting melts, with petrological applications. Contributions to Mineralogy and Petrology 99, 176–185.
Kamenetsky, V., Métrich, N. & Cioni, R. (1995a). Potassic primary melts of Vulsini (Roman Province): evidence from mineralogy and melt inclusions. Contributions to Mineralogy and Petrology 120, 186–196.
Kamenetsky, V., Sobolev, A. V., Joron, J.-L. & Semet, M. P. (1995b). Petrology and geochemistry of Cretaceous ultramafic volcanics from eastern Kamchatka. Journal of Petrology 36, 637–662.
Kamenetsky, V. S., Eggins, S. M., Crawford, A. J., Green, D. H., Gasparon, M. & Falloon, T. J. (1998). Calcic melt inclusions in primitive olivine at 43°N MAR: evidence for melt–rock reaction/melting involving clinopyroxene-rich lithologies during MORB generation. Earth and Planetary Science Letters 160, 115–132.[Web of Science]
Kent, A. J. R., Stolper, E. M., Woodhead, J., Hutcheon, I. D. & Francis, D. (1998). Using glass inclusions to investigate a heterogeneous mantle: an example from N- and EMORB-like lavas from Baffin Island. Mineralogical Magazine 62A(Goldschmidt Conference), 765–766.
Kilinc, A., Carmichael, I. S. E., Rivers, M. L. & Sack, R. O. (1983). The ferric–ferrous ratio of natural silicate liquids equilibrated in air. Contributions to Mineralogy and Petrology 83, 136–140.[Web of Science]
Krishnamurthy, P. & Cox, K. G. (1977). Picrite basalts and related lavas from the Deccan traps of Western India. Contributions to Mineralogy and Petrology 62, 53–75.
Langmuir, C. H., Klein, E. M. & Plank, T. (1992). Petrological systematics of mid-ocean ridge basalts: constraints on melt generation beneath ocean ridges. In: Phipps Morgan, J., Blackman, D. K. & Sinton, J. M. (eds) Mantle Flow and Melt Generation at Mid-ocean Ridges. Geophysical Monograph, American Geophysical Union 71, 183–280.
Le Bas, M. J. (ed.) (2000). A Classification of Igneous Rocks and Glossary of Terms. Recommendations of the IUGS Subcommission on the Systematics of Igneous Rocks, 2nd edn. Oxford: Blackwell Scientific (in press).
Lesher, C. M. (1989). Komatiite-associated nickel sulfide deposits. In: Whitney, J. A. & Naldrett, A. J. (eds) Ore Deposition Associated with Magmas. Reviews in Economic Geology 4, 45–101.
Li, J.-P., O’Neil, H. St C. & Seifert, F. (1995). Subsolidus phase relations in the system MgO–SiO2–Cr–O in equilibrium with metallic Cr, and their significance for the petrochemistry of chromium. Journal of Petrology 36, 107–132.
Lightfoot, P. C., Hawkesworth, C. J., Olshevsky, K., Green, A., Doherty, W. & Keays, R. R. (1997). Geochemistry of Tertiary tholeiites and picrites from Qeqertarssuaq (Disko Island) and Nuussuaq, West Greenland with implications for the mineral potential of comagmatic intrusions. Contributions to Mineralogy and Petrology 128, 139–163.
Maaløe, S. (1973). Temperature and pressure relations of ascending primary magmas. Journal of Geophysical Research 78, 6877–6886.
Maaløe, S. (1979). Compositional range of primary tholeiitic magmas evaluated from major-element trends. Lithos 12, 59–72.
Maaløe, S. & Hansen, B. (1982). Olivine phenocrysts of Hawaiian olivine tholeiite and oceanite. Contributions to Mineralogy and Petrology 81, 203–211.
Maaløe, S., Tumyr, O. & James, D. (1989). Population density and zoning of olivine phenocrysts in tholeiites from Kauai, Hawaii. Contributions to Mineralogy and Petrology 101, 176–186.
Maurel, C. & Maurel, P. (1982). Étude expérimentale de l’équilibre Fe2+–Fe3+ dans les spinelles chromifères et les liquides silicatés basiques coexistants, à 1 atm. Comptes Rendus de l’Académie des Sciences 295, 209–212.
McKenzie, D. & Bickle, M. J. (1988). The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology 29, 625–679.
Melluso, L., Beccaluva, L., Brotzu, P., Gregnanin, A., Gupta, A. K., Morbidelli, L. & Traversa, G. (1995). Constraints on the mantle sources of the Deccan Traps from the petrology and geochemistry of the basalts of Gujarat State (western India). Journal of Petrology 36, 1393–1432.
Murck, B. W. & Campbell, I. H. (1986). The effects of temperature, oxygen fugacity and melt composition on the behaviour of chromium in basic and ultrabasic melts. Geochimica et Cosmochimica Acta 50, 1871–1887.
Nielsen, T. F. D., Soper, N. J., Brooks, C. K., Faller, A. M., Higgins, A. C. & Matthews, D. W. (1981). The pre-basaltic sediments and the Lower Basalts at Kangerdlugssuaq, East Greenland: their stratigraphy, lithology, palaeomagnetism and petrology. Meddelelser om Grønland, Geoscience 6, 28 pp.
Nisbet, E. G., Cheadle, M. J., Arndt, N. T. & Bickle, M. J. (1993). Constraining the potential temperature of the Archaean mantle: a review of the evidence from komatiites. Lithos 30, 291–307.[Web of Science]
Nisbet, E. G., Bickle, M. J. & Martin, A. (1977). The mafic and ultramafic lavas of the Belingwe greenstone belt, Rhodesia. Journal of Petrology 18, 521–566.
Pearce, T. H. (1984). The analysis of zoning in magmatic crystals with emphasis on olivine. Contributions to Mineralogy and Petrology 86, 149–154.
Pedersen, A. K. (1979a). A shale buchite xenolith with armalcolite and native iron in a lava from Asuk, Disko, central West Greenland. Contributions to Mineralogy and Petrology 69, 83–94.
Pedersen, A. K. (1979b). Basaltic glass with high-temperature equilibrated immiscible sulphide bodies with native iron from Disko, central West Greenland. Contributions to Mineralogy and Petrology 69, 397–407.
Pedersen, A. K. (1985). Reaction between picrite magma and continental crust: early Tertiary silicic basalts and magnesian andesites from Disko, West Greenland. Bulletin Grønlands Geologiske Undersøgelse 152, 126 pp.
Pedersen, A. K. & Pedersen, S. (1987). Sr isotope chemistry of contaminated Tertiary volcanic rocks from Disko, central West Greenland. Bulletin of the Geological Society of Denmark 36, 315–336.
Pedersen, A. K., Larsen, L. M. & Dueholm, K. S. (1993). Geological section along the south coast of Nuussuaq, central West Greenland, 1:20 000 Coloured Geological Sheet. Copenhagen: Geological Survey of Greenland.
Pedersen, A. K., Larsen, L. M., Pedersen, G. K. & Dueholm, K. S. (1996). Filling and plugging of a marine basin by volcanic rocks: the Tunoqqu Member of the Lower Tertiary Vaigat Formation on Nuussuaq, central West Greenland. Bulletin Grønlands Geologiske Undersøgelse 171, 5–28.
Pedersen, G. K., Larsen, L. M., Pedersen, A. K. & Hjortkjær, B. F. (1998). The synvolcanic Naajaat lake, Paleocene of West Greenland. Palaeogeography, Palaeoclimatology, Palaeoecology 140, 271–287.
Roeder, P. L. & Emslie, R. F. (1970). Olivine–liquid equilibrium. Contributions to Mineralogy and Petrology 29, 275–289.[Web of Science]
Saunders, A. D., Fitton, J. G., Kerr, A. C., Norry, M. J. & Kent, R. W. (1997). The North Atlantic Igneous Province. In: Mahoney, J. J. & Coffin, M. L. (eds) Large Igneous Provinces. Geophysical Monograph, American Geophysical Union 100, 45–93.
Sinton, J. M. & Detrick, R. S. (1992). Mid-ocean ridge magma chambers. Journal of Geophysical Research 97(B), 197–216.
Sobolev, A. V. & Nikogosian, I. K. (1994). Petrology of long-lived mantle plume magmatism: Hawaii, Pacific, and Reunion Island, Indian Ocean. Petrology 2, 111–144.
Sobolev, A. V. & Shimizu, N. (1994). The origin of typical N-MORB: the evidence from a melt inclusions study. Mineralogical Magazine 58A, 862–863.
Storey, M., Duncan, R. A., Pedersen, A. K., Larsen, L. M. & Larsen, H. C. (1998). 40Ar/39Ar geochronology of the West Greenland Tertiary volcanic province. Earth and Planetary Science Letters 160, 569–586.
Ulmer, P. (1989). The dependence of the Fe2+–Mg cations-partitioning between olivine and basaltic liquid on pressure, temperature and composition. An experimental study to 30 kbars. Contributions to Mineralogy and Petrology 101, 261–273.
Upton, B. G. J. (1988). History of Tertiary igneous activity in the N Atlantic borderlands. In: Morton, A. C. & Parson, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication 39, 429–453.
White, R. & McKenzie, D. (1989). Magmatism at rift zones: the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research 94(B), 7685–7729.
Yang, H.-J., Frey, F., Garcia, M. O. & Clague, D. A. (1994). Submarine lavas from Mauna Kea volcano, Hawaii: implications for Hawaiian shield stage processes. Journal of Geophysical Research 99(B), 15577–15594.
Zhou, M.-F. & Kerrich, R. (1992). Morphology and composition of chromite in komatiites from the Belingwe greenstone belt, Zimbabwe. Canadian Mineralogist 30, 303–317.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. M. LARSEN, L. M. HEAMAN, R. A. CREASER, R. A. DUNCAN, R. FREI, and M. HUTCHISON Tectonomagmatic events during stretching and basin formation in the Labrador Sea and the Davis Strait: evidence from age and composition of Mesozoic to Palaeogene dyke swarms in West Greenland Journal of the Geological Society, December 1, 2009; 166(6): 999 - 1012. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Larsen and A. K. Pedersen Petrology of the Paleocene Picrites and Flood Basalts on Disko and Nuussuaq, West Greenland J. Petrology, September 1, 2009; 50(9): 1667 - 1711. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Beccaluva, G. Bianchini, C. Natali, and F. Siena Continental Flood Basalts and Mantle Plumes: a Case Study of the Northern Ethiopian Plateau J. Petrology, July 1, 2009; 50(7): 1377 - 1403. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J.R. Kent Melt Inclusions in Basaltic and Related Volcanic Rocks Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 273 - 331. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Meyer, J. van Wijk, and L. Gernigon The North Atlantic Igneous Province: A review of models for its formation Geological Society of America Special Papers, January 1, 2007; 430(0): 525 - 552. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. ZHANG, J. J. MAHONEY, J. MAO, and F. WANG Geochemistry of Picritic and Associated Basalt Flows of the Western Emeishan Flood Basalt Province, China J. Petrology, October 1, 2006; 47(10): 1997 - 2019. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. HOLM, J. R. WILSON, B. P. CHRISTENSEN, L. HANSEN, S. L. HANSEN, K. M. HEIN, A. K. MORTENSEN, R. PEDERSEN, S. PLESNER, and M. K. RUNGE Sampling the Cape Verde Mantle Plume: Evolution of Melt Compositions on Santo Antao, Cape Verde Islands J. Petrology, January 1, 2006; 47(1): 145 - 189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. THOMPSON, C. J. OTTLEY, P. M. SMITH, D. G. PEARSON, A. P. DICKIN, M. A. MORRISON, P. T. LEAT, and S. A. GIBSON Source of the Quaternary Alkalic Basalts, Picrites and Basanites of the Potrillo Volcanic Field, New Mexico, USA: Lithosphere or Convecting Mantle? J. Petrology, August 1, 2005; 46(8): 1603 - 1643. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. JOHNSON, S. A. GIBSON, R. N. THOMPSON, and G. M. NOWELL Volcanism in the Vitim Volcanic Field, Siberia: Geochemical Evidence for a Mantle Plume Beneath the Baikal Rift Zone J. Petrology, July 1, 2005; 46(7): 1309 - 1344. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. FEDORTCHOUK and D. CANIL Intensive Variables in Kimberlite Magmas, Lac de Gras, Canada and Implications for Diamond Survival J. Petrology, September 1, 2004; 45(9): 1725 - 1745. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Herzberg Partial melting below the Ontong Java Plateau Geological Society, London, Special Publications, January 1, 2004; 229(1): 179 - 183. [Abstract] [PDF]
|
|
|
|
 |
|
 Y. XU, X. HUANG, M. A. MENZIES, and R. WANG Highly magnesian olivines and green-core clinopyroxenes in ultrapotassic lavas from western Yunnan, China: evidence for a complex hybrid origin European Journal of Mineralogy, December 1, 2003; 15(6): 965 - 975. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. W. PEATE, J. A. BAKER, J. BLICHERT-TOFT, D. R. HILTON, M. STOREY, A. J. R. KENT, C. K. BROOKS, H. HANSEN, A. K. PEDERSEN, and R. A. DUNCAN The Prinsen af Wales Bjerge Formation Lavas, East Greenland: the Transition from Tholeiitic to Alkalic Magmatism during Palaeogene Continental Break-up J. Petrology, February 1, 2003; 44(2): 279 - 304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. LARSEN, A. K. PEDERSEN, B. SUNDVOLL, and R. FREI Alkali Picrites Formed by Melting of Old Metasomatized Lithospheric Mantle: Manitdlat Member, Vaigat Formation, Palaeocene of West Greenland J. Petrology, January 1, 2003; 44(1): 3 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. HERZBERG and M. J. O'HARA Plume-Associated Ultramafic Magmas of Phanerozoic Age J. Petrology, October 1, 2002; 43(10): 1857 - 1883. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. G. J. Upton, B. G. J. UPTON, A. C. SKOVGAARD, J. McCLURG, L. KIRSTEIN, M. CHEADLE, C. H. EMELEUS, W. J. WADSWORTH, and A. E. FALLICK Picritic magmas and the Rum ultramafic complex, Scotland Geological Magazine, July 1, 2002; 139(4): 437 - 452. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Pedersen, L. M. Larsen, P. Riisager, and K. S. Dueholm Rates of volcanic deposition, facies changes and movements in a dynamic basin: the Nuussuaq Basin, West Greenland, around the C27n-C26r transition Geological Society, London, Special Publications, January 1, 2002; 197(1): 157 - 181. [Abstract] [PDF]
|
|
|
|
|
|
R. N. THOMPSON, S. A. GIBSON, A. P. DICKIN, and P. M. SMITH Early Cretaceous Basalt and Picrite Dykes of the Southern Etendeka Region, NW Namibia: Windows into the Role of the Tristan Mantle Plume in Parana-Etendeka Magmatism J. Petrology, November 1, 2001; 42(11): 2049 - 2081. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||