Submarine Basalts Research Articles

Dating crystalline groundmass separates of altered Cretaceous seamount basalts by the [formula omitted] incremental heating technique

Alteration of submarine basalts compromises geochronology using conventional K/Ar and 40 Ar/ 39Ar techniques. To help overcome these problems, we re-evaluated the potential of groundmass dating techniques. Incremental heating on acid-leached groundmass samples, following an overnight bakeout at 200°C and using high-resolution heating schedules, eliminated most of the low temperature alteration effects in the submarine basalts studied. More than 75% of the groundmass analyses ( n=32) display accurate age plateaus consisting of 30–70% of the total amount of 39 Ar K released. More than 50% of the analyses have plateau ages concordant with their total fusion ages implying minor or proportional loss of radiogenic 40 Ar * and 39 Ar K . Overall, we could show a high degree of coherence between ages of groundmass separates and comagmatic phenocrysts. This suggests that the dating of aphyric basalts, which previously has proven problematic, can be accomplished with increasing confidence as well. Adding these rock types to the list of datable submarine basalts significantly enhances our ability to understand the eruptive history of linear submarine volcanic chains.

Thermodynamic analysis of Fe and Mg partitioning between plagioclase and silicate liquid

Thermodynamic analysis of Fe- and Mg-bearing plagioclase and silicate liquid was carried out based on reported element partitioning data between plagioclase and silicate liquid in reduced conditions, solution properties of ternary feldspar, standard state properties of plagioclase endmembers and solution properties of multicomponent silicate liquid. Derived mixing properties of Fe- and Mg-bearing plagioclase are in harmony with estimated results from synthetic experiments in the systems CaAl2Si2O8-CaFeSi3O8 and CaAl2Si2O8-CaMgSi3O8. Based on the determined solution properties of the plagioclase, a computer program to calculate the element partition relationships between Fe- and Mg-bearing plagioclase and multicomponent silicate liquid was developed. The FeO, MgO and MgO/(MgO + FeO) in plagioclase predicted from known liquid compositions and pressure are in agreement with measurements within 0.2 wt%, 0.1 wt% and 0.1 (mol ratio), respectively. The Fe3+ content in plagioclase crystallized at high oxygen fugacity can be estimated with this program. The Fe3+/total Fe ratio in plagioclase crystallized near the quartz-fayalite-magnetite buffer ranges from 0 to 0.5, which is consistent with previous study on natural plagioclase in submarine basalt. Derived solution properties of the Fe- and Mg-bearing plagioclase are also used to calculate equilibrium composition relationship between olivine and plagioclase. Change of XFo in olivine coexisting with plagioclase affects MgO and FeO contents in plagioclase greatly. The present model predicts XFo of coexisting olivine from the chemical composition of plagioclase to ±0.1 accuracy at given pressure and temperature.

The magnetism and microstructure of pulverized titanomagnetite, Fe 2.4Ti 0.6O 4: the effect of annealing, maghemitization and inversion

Titanomagnetite, Fe 2.4Ti 0.6O 4, pulverized in a ball-mill to make the material readily maghemitize on the laboratory time scale, has been used as an analogue for the magnetic mineral of submarine basalts. The ball-milling, in addition to reducing particle size, produces an internal nanocrystalline structure. Annealing at high enough temperature (600°C–800°C) removes the nanocrystalline structure but the titanomagnetite crystals carry a relic of ball-milling induced non-stoichiometry which contributes to magnetic strain anisotropy. Despite this complication, the coercive force of the annealed material is close to the average coercive force of pillow basalts and should provide an improved synthetic analogue for laboratory studies inspired by the magnetism and mineralogy of the submarine crust. The nanocrystalline structure persists in maghemitized pulverized titanomagnetite, and the magnetization process parameters are consistent with an increasing importance of thermal fluctuations as maghemitization proceeds. In the transformations produced by annealing, maghemitization and inversion, the spinel component immediately post-transformation has composition and/or concentration inherited from the pre-transformation spinel, and does not correspond to equilibrium. Later re-equilibration of the coexisting phases, accelerated in the laboratory by elevated temperature, leads to diminution of the spinel component in favour of more stable phases. Unlike the inversion of maghemite to haematite, the inversion product of titanomaghemite is not a unique assemblage of phases.

Why was flood volcanism on submerged continental platforms so common in the Precambrian?

Many Archean and Proterozoic greenstone belts contain volcanic rocks that erupted subaqueously but interacted with continental crust on their way to the surface. This points to emplacement in a tectonic setting that was rare to absent throughout the Phanerozoic — the eruption of flood basalts onto submerged continental platforms. To explain the situation requires that sea levels were high during periods of flood volcanism, perhaps because the ocean basins were partially filled by extensive, active and high-standing oceanic ridges or by oceanic plateaus. The common preservation of crust-contaminated submarine basalts, even in the oldest supracrustal sequences, may indicate that continental crust was voluminous even at an early stage in Earth history.

Calibration of magnetic granulometric trends in oceanic basalts

The validity of magnetic granulometric estimates relies heavily on the ability to distinguish ultrafine particles from coarser grains. For example, populations with dominantly superparamagnetic (SP) or multidomain (MD) grains both are characterized by low remanence and coercivity, and distinguishing these endmembers may provide valuable clues to the origin of magnetization in the intervening stable single domain (SD) size range. The natural grain size variations associated with variable cooling rates in submarine lavas provide a rare opportunity for examining progressive changes in average magnetic grain size, from SP–SD mixtures in submarine basaltic glass to SD–MD mixtures in flow interiors. Based on microanalysis and rock magnetic measurements on pillow basalt samples dredged from the flanks of the Mid-Atlantic Ridge (ages <1 Ma to 70 Ma), a model of preferential dissolution with time of the finest-grained titanomagnetites has recently been suggested as the major process contributing to long-term temporal changes in remanent intensity of mid-ocean ridge basalts. We evaluated the local and long-term temporal trends in effective magnetic grain size predicted by this model using hysteresis data from a large number of submarine basalt samples which span a range of ages from ∼0 to ∼122 Ma. Specimens were systematically taken along transects perpendicular to the chilled margin of each sample. The large number of data (∼750 loops) and the inferred progressive change in grain size approaching the chilled margin allow recognition of mixing trends between MD and SD grains and between SD and SP grains on a Day-plot. These trends in hysteresis parameters are crucial to resolving the inherent, but frequently overlooked, ambiguity in inferring grain size from hysteresis parameters. We illustrate that two additional rock magnetic tests (warming of a low-temperature isothermal remanence and hysteresis loop shapes) often used to address these ambiguities are inconclusive, requiring some independent knowledge of whether SP or MD grains are likely to be present. Even with a considerably larger data set the substantial intrasample variability in oceanic basalts precludes recognition of any systematic trend in magnetic grain size with age.

Long submarine lava flows: Observations and results from numerical modeling

Long (&gt;100 km) lava flows are relatively common on Mars and Venus and have been identified on the Moon, but they are rarely documented on Earth. However, although ∼75% of the Earth's surface is covered by water, only a small percentage of the ocean floor has been investigated at a resolution sufficient to unequivocally identify the boundaries of long submarine lava flows. Even so, basaltic lava flows as long as 110 km have been identified on the deep (&gt;1500 m) seafloor near Hawaii and the East Pacific Rise. Ambient conditions on the deep ocean floor may favor the development of long lava flows for the following reasons. First, high pressures (&gt;15 MPa) keep volatiles dissolved in basaltic lavas, preventing viscosity increases associated with exsolution and vesiculation. Second, seawater rapidly quenches the surface of submarine basalt flows so that an insulating glass layer, 1–5 cm thick, encases submarine flows within seconds after their emplacement. This glass rind effectively insulates the molten flow interior from additional heat loss, making submarine basalt flows behave as well‐insulated, subaerial tube‐fed flows. Thus, for identical basalt flows emplaced on the deep seafloor and subaerially, a submarine flow could advance farther before stopping. Results of numerical modeling indicate that thin (≤1 m) submarine basalt flows behave similarly to identical subaerial flows, but thicker submarine flows may advance significantly farther than their subaerial counterparts.

The intensity of the Earth's magnetic field over the past 160 million years

In contrast to our detailed knowledge of the directional behaviour of the Earth's magnetic field during geological and historical times1,2, data constraining the past intensity of the field remain relatively scarce. This is mainly due to the difficulty in obtaining reliable palaeointensity measurements, a problem that is intrinsic to the geological materials which record the Earth's magnetic field. Although the palaeointensity database has grown modestly over recent years3,4,5 these data are restricted to a few geographical locations and more than one-third of the data record the field over only the past 5?Myr—the most recent database5 covering the time interval from 5 to 160?Myr contains only about 100 palaeointensity measurements. Here we present 21 new data points from the interval 5–160?Myr obtained from submarine basalt glasses collected from locations throughout the world's oceans. Whereas previous estimates for the average dipole moment were comparable to that of the Earth's present field6, the new data suggest an average dipole moment of (4.2 ± 2.3) × 1022?A?m2, or approximately half the present magnetic-field intensity. This lower average value should provide an important constraint for future efforts to model the convective processes in the Earth's core which have been responsible for generating the magnetic field.

Velocities of a natural mid-ocean ridge basalt glass

We report, for the first time, ultrasonic velocity values for a pure (>95%), natural, submarine basalt glass of mid-ocean ridge basalt composition, from 10 to 1000 MPa at room temperature. These new data show that basalt glass, abundant in the upper oceanic crust, has the lowest velocity of any primary solid component of the oceanic crust. In addition, natural basalt glass has a steeper pressure-dependence of velocity than previously measured in more crystalline samples, indicating that cracks in natural basalt glass are weaker than in more crystalline rocks. To obtain values for the pure glass phase, we correct the natural glass data for the low-pressure closure of cracks, and the presence of minor mineralogic components and vesicles. These new data provide a baseline for evaluating the effect of abundant basalt glass and glassy mesostasis in oceanic upper crust on in situ seismic velocities. In addition, data on the elastic and seismic properties of natural glasses are useful for a better understanding of glass structure, and glass relaxation, with potential applications to submarine volcanology.

Potential links between ocean plateau volcanism and global ocean anoxia at the Cenomanian-Turonian boundary

The Cenomanian-Turonian boundary is marked globally by the extinction of marine invertebrates and an increase in the accumulation and preservation of organic carbon-rich sediments (black shales). The timing of this boundary also coincides with large-scale volcanism from at least three large igneous provinces: the Caribbean and Ontong-Java oceanic plateaus and the Madagascar flood basalts. In this paper, we assess the possibility that hydrothermalism associated with large-scale submarine magmatism was responsible for the reduction of dissolved O 2 in the oceans. We investigate two potential mechanisms: the oxidation of reduced material in hydrothermal effluents and the stimulation of primary productivity in the water column due to the injection of hydrothermal Fe into surface waters. With the first mechanism, we find that a 10,000-km 3 submarine basalt eruption (three orders of magnitude larger than recent midocean ridge eruptions) could release enough reduced hydrothermal material to consume at least 6 percent of the dissolved O 2 in seawater in a well-ventilated ocean (such as the present ocean). With the second mechanism, we calculate that even a small percentage of the hydrothermal Fe released from a single large flow could have led to a significant increase in primary productivity in areas where Fe was the biolimiting nutrient (such as open-ocean areas in the Pacific). The potential impact of both of these O 2 -reducing mechanisms at the Cenomanian-Turonian boundary would have been significantly greater if Cretaceous oceans were less well ventilated than the present oceans.

Rare gas systematics in Red Sea Ridge basalts

From south to north, the 4He/³He ratios of submarine basalts from the Red Sea and Gulf of Tadjoura increase from 45,000 (R/Ra=15) in the Gulf of Tadjoura to 84,000 (R/Ra=8.6) in the Mabahiss deep. Our helium isotopic results indicate that the mantle plume proposed in the north of the Red Sea, on the basis of the Sr and Pb isotopic ratios, does not posess a high ³He signature. The neon and Ar isotopic ratios of the Red Sea glasses are explained by mixing between AIR, MORB and a plume with a solar like isotopic Ne signature and low 40Ar/36Ar, similar to that of Loihi.

Petrogenesis of alkaline basalts from Socorro Island, Mexico: Trace element evidence for contamination of ocean island basalt in the shallow ocean crust

Postcaldera mafic volcanism on Socorro Island, Mexico, an island located in the eastern Pacific Ocean on a segment of a mid‐ocean ridge spreading center abandoned at ∼3.5 Ma, dominantly comprises alkalic basalt, hawaiite, and mugearite with subordinate mildly alkalic‐transitional basalt. Most major element data are consistent with differentiation of alkalic basalt to hawaiite and mugearite by up to 50% fractionation of plagioclase+clinopyroxene+olivine±Fe−Ti oxides. High‐field strength element abundances are consistent with this interpretation, and the narrow range in Zr/Nb suggests that parental magmas formed by similar degrees of partial melting of a relatively homogeneous mantle source. Sr and Nd isotopic ratios exhibit relatively narrow ranges (0.7031–0.7032, 0.5128–0.5130, respectively). Despite this, a subset of the alkalic basalts has negative Ce anomalies and abundances of P2O5, Ba, Y, and some rare earth elements in excess of those predicted by fractional crystallization models; these chemical features require open‐system processes. The observed heterogeneities are perhaps best explained by assimilation of ocean crustal components by basalt in a shallow magma reservoir. Regional mantle heterogeneity is evident from elemental and isotopic data for mildly alkalic‐transitional basalt from Socorro Island, mugearite from San Benedicto Island, and submarine basalts from near Socorro Island. Evidence presented here indicates that compositional variations in ocean island basalts can be introduced by ocean crustal contamination and underscores the importance of considering this source of chemical heterogeneity in basalts before characterizing associated mantle.

An Experimental Study of Water and Carbon Dioxide Solubilities in Mid-Ocean Ridge Basaltic Liquids. Part II: Applications to Degassing

Degassing processes in basaltic magmas rich in both water and carbon dioxide can be modeled using the solubilities of the endmember systems and the assumption of Henry's law. Suites of vapor-saturated basaltic melts having a range of initial CO_2/H_2O ratios and erupted over a narrow depth interval will define negatively sloped arrays on an H_2O vs CO_2 plot. It is important that all of the major volatile species be considered simultaneously when interpreting trends in dissolved volatile species concentrations in magmas. Based on measured concentrations of water and carbon dioxide in basaltic glasses, the composition of the vapor phase at 1200°C that could coexist with a basaltic melt and the pressure at which it would be vapor saturated can be calculated. The range in vapor compositions in equilibrium with submarine basalts reflects the range in water contents in the melts characteristic of each environment. The ranges in the molar proportion of CO_2 in vapor phases (X^ν_(CO2)) calculated to be in equilibrium with submarine tholeiitic glasses are 0⋅93–1⋅00 for mid-ocean ridge basalts (MORB), 0⋅60–0⋅99 for glasses from Kilauea [representative of ocean island basalts (OIB)] and 0–0⋅94 for glasses from back-arc basins (BABB). MORB glasses from spreading centers ranging from slow (e.g. the Mid-Atlantic Ridge) to fast (e.g. East Pacific Rise, 9–13°N) are commonly supersaturated with respect to CO_2-rich vapor, resulting from magma ascent rates so rapid that magmas erupt on the sea-floor without having been fully degassed by bubble nucleation and growth during ascent. In contrast to the MORB glasses, volatile contents in submarine glasses from Kilauea are consistent with having been in equilibrium with a vapor phase containing 60–100 mol% CO_2 at the pressure of eruption, reflecting differences in average magma transport rates during eruptions at mid-ocean ridges and hotspot volcanoes. Degassing during decompression of tholeiitic basaltic magma is characterized by strong partitioning of CO_2 into the vapor phase. During open system degassing, CO_2 is rapidly removed from the melt with negligible loss of water, until a pressure is reached at which the melt is in equilibrium with nearly pure water vapor. From this pressure downward, the water content of the melt follows the water solubility curve. During closed system degassing, water and CO_2 contents in vapor-saturated basaltic magmas will depend strongly on the vapor composition as determined by the initial volatile concentrations. Deviation from open system behavior, toward lower dissolved H_2O and CO_2 saturation concentrations at a given pressure, will be greatest in melts having high total volatile concentrations and high CO_2:H_2O ratios. Closed system degassing of basaltic melts having the low initial H_2O and CO_2 contents typical of MORB and OIB, however, are similar to the open system case.

Magnetic properties of zero‐age oceanic crust; A new submarine lava flow on the Juan de Fuca Ridge

A volcanic eruption that occurred in October, 1993 on the Juan de Fuca Ridge shows unusually high magnetization when compared with older samples from the off‐axis Juan de Fuca Ridge and other Pacific sites. These data, plus data from other young oceanic basalts, indicate that oceanic crustal magnetization may decay rapidly, by 10% in the first year. The considerably slower decrease in magnetization for rocks from older, off‐axis sites agrees with previous models of progressive low temperature oxidation of the mineral titanomagnetite. The magnetization for newly erupted submarine basalts, however, is not consistent with the exponential decay predicted from older sites, indicating that oxidation may not be the sole process responsible for the short‐term decay. Similarly, weak field magnetic susceptibility of the younger rocks does not show an obvious corresponding decrease with increasing age, as expected for the oxidation of titanomagnetite to titanomaghemite. These preliminary results indicate that decay of magnetization of oceanic crustal rocks with time may have two distinct stages, with only the later stage being due to the progressive oxidation of the magnetic minerals.

Magmatic evolution of the Easter microplate-Crough Seamount region (South East Pacific)

The Easter microplate-Crough Seamount region located between 25° S–116° W and 25° S–122° W consists of a chain of seamounts forming isolated volcanoes and elongated (100–200 km in length) en echelon volcanic ridges oriented obliquely NE (N 065°), to the present day general spreading direction (N 100°) of the Pacific-Nazca plates. The extension of this seamount chain into the southwestern edge of the Easter microplate near 26°30′ S–115° W was surveyed and sampled. The southern boundary including the Orongo fracture zone and other shallow ridges ( 0.25) MORBs which are similar in composition to other more recent basalts from the Southwest and East Rifts spreading axes of the Easter microplate. Incompatible element ratios normalized to chondrite values [(Ce/Yb)N = 1−2.5}, {(La/Sm)N = 0.4−1.2} and {(Zr/Y)N = 0.7−2.5} of the basalts are also similar to present day volcanism found in the Easter microplate. The volcanics from the Easter microplate-Crough region are unrelated to other known South Pacific intraplate magmatism (i.e. Society, Pitcairn, and Salas y Gomez Islands). Instead their range in incompatible element ratios is comparable to the submarine basalts from the recently investigated Ahu and Umu volcanic field (Easter hotspot) (Scientific Party SO80, 1993) and centered at about 80 km west of Easter Island. The oblique ridges and their associated seamounts are likely to represent ancient leaky transform faults created during the initial stage of the Easter microplate formation (≈ 5 Ma). It appears that volcanic activity on seamounts overlying the oblique volcanic ridges has continued during their westward drift from the microplate as shown by the presence of relatively fresh lava observed on one of these structures, namely the first Oblique Volcanic Ridge near 25° S–118° W at about 160 km west of the Easter microplate West Rift. Based on a reconstruction of the Easter microplate, it is suggested that the Crough seamount (< 800 m depth) was formed by earlier (7–10 Ma) hotspot magmatic activity which also created Easter Island.

Thermally Activated Magnetic Viscosity In Natural Multidomain Titanomagnetite

SUMMARY Magnetic viscosity and hysteresis versus temperature and field have been studied between room and Curie temperatures on two selected coarse-grained submarine basalts carrying multidomain (MD) low Curie temperature titanomagnetite. These samples exhibit typical characteristics of magnetic viscosity in rocks, including imperfect linearity of the magnetization change with logarithm of time and a large increase in viscosity with temperature, followed by a decrease occurring some 30°-40° below the average Curie point of the rock sample. the field dependence of magnetic viscosity exhibits a maximum located slightly above the sample coercive force. Other experiments, reported elsewhere, show that diffusion plays a negligible role in these rocks. Thus, magnetic viscosity is probably entirely of thermal origin. In spite of the MD characteristics of our samples at room temperature, alternating field (AF) demagnetization of remanence shows, in one of them, the presence of a large component of high coercivity of PSD or SD origin. the remanence of the other sample seems to be wholly of MD origin. the thermal dependences of magnetic viscosity and irreversible susceptibility are only approximately similar, and the field dependences are quite different. This indicates that the proportionality of irreversible susceptibility with magnetic viscosity predicted by the Neel and the Street and Wooley theories of thermal fluctuations in MD grains does not hold for natural titanomagnetite like ours, even though it is verified for synthetic ferromagnetic substances. However, our experimental data are compatible with the fundamentals of the theory of thermal fluctuations if we assume that the pinning of Bloch walls in natural titanomagnetite is due to several kinds of energy barriers with a broad distribution of both activation volumes va and critical fields hc. Thus, the remanent magnetization acquired in a low field (1 Oe) applied for 1 d at room temperature is found to be pinned by two categories of defects with (1) va= 7x10−15cm3 and hc= 3Oe and (2) va= 1x10−15cm3 and hc= 15Oe. Using a (va, hc) diagram to represent each barrier, we show that the ‘blocking’(or ‘unblocking’) curves are quite dissimilar for field and time effects, just as for SD particles. This indicates that the proportionality between magnetic viscosity and irreversible susceptibility requires particular distributions of energy barriers in the (va, hc) plane and cannot be considered as a general characteristic of thermally activated magnetic viscosity.

Strong degassing at ridge crests: The behaviour of dissolved carbon and water in basalt glasses at 14°N, Mid-Atlantic Ridge

The popping rock 2πD43 dredged at 14°N on the Mid-Atlantic Ridge contains the largest gas content so far reported in MORB [11], with most of the gas preserved in vesicles (17 vol%). This gas has chemical ( CO 2 H 2O , C He , C N , He Ar ) and isotopic ( δ 13C) characteristics that are remarkably constant throughout the dredge. We have taken this opportunity to study the behaviour and variability of dissolved species (carbon and water) against such a well-defined fluid. Sixteen new crushing experiments, carried out in steps and on much smaller (38–170 mg) fragments than in our previous study [11], have been conducted to investigate small-scale variations in vesicle carbon composition. The results confirm the overall constancy of δ 13C ( −3.62 ± 0.05‰ vs. −3.68 ± 0.07‰ for the large fragments). However, variations of 0.5–3.5‰ occur in about one-third of the aliquots analyzed, reflecting outgassing effects on the smaller vesicles and a slight chemical evolution. δ 18O variations from 9.29 to 20.89‰ are associated with the δ 13C variations, and correspond to variations in H 2O concentrations in vesicles ranging from ∼ 5 ± 2 vol% for most of the aliquots to high values of 40–85 vol% for small vesicles. The amount of carbon extracted below 900°C decreases with decreasing grain size, and its isotopic composition can be modelled as a variable mixture of CO 2 from small vesicles, precipitated carbonate and amorphous carbon deposits. On samples crushed down to complete exhaustion of vesicle CO 2, carbon deposits on vesicles walls and cracks have been shown to be about 40 ppm C ( ∼ 1% of vesicle CO 2) at δ 13 C = −16 ± 4‰ . The carbon recovered on fusion, which corresponds to carbon dissolved in the glass, is constant within a given batch of samples but varies from 57 to 94 ppm C from batch to batch, with the corresponding δ 13C values varying from −5.6 to −11.8‰. Water contents vary from 4160 to 5300 ppm, with a δD of −64 ± 2‰, except for an atypical, vesicle-poor sample (2620 ppm). The 13C fractionations between gas and dissolved carbon vary widely, between 2 and 8‰, with most of these values occurring in samples of constant vesicle δ 13C (ca. −3.7‰). Dissolved carbon and water concentrations vary from approximate equilibrium solubility at the seafloor up to 100% and 30% supersaturation respectively. Given their respective solubility laws, this corresponds to similar saturation depths of 0.6–1.3 km below the seafloor. 13C fractionations between vesicle CO 2 and dissolved carbon record disequilibrium outgassing features of deeper origin during the eruptive process of the magma, from ∼ 2 to 4 km below the seafloor, which is most likely the depth of the magma chamber. These outgassing distillation effects lead to variable levels of supersaturation and isotopic disequilibrium that depend on the style (i.e., ascent rate) of eruption, until, at the very end of the process, kinetic isotope effects eventually partially reverse the isotopic trend. These results suggest that, in most occurrences, submarine basalt glasses display the characteristics of magma chambers outgassed to more than 90% (i.e., 100–200 ppm total (vesicle + dissolved) carbon), with δ 13C lowered from an initial value of ca. −4‰ to between −6 and −8‰.

Role of olivine cumulates in destabilizing the flanks of Hawaiian volcanoes

The south flank of Kilauea Volcano is unstable and has the structure of a huge landslide; it is one of at least 17 enormous catastrophic landslides shed from the Hawaiian Islands. Mechanisms previously proposed for movement of the south flank invoke slip of the volcanic pile over seafloor sediments. Slip on a low friction decollement alone cannot explain why the thickest and widest sector of the flank moves more rapidly than the rest, or why this section contains a 300 km3 aseismic volume above the seismically defined decollement. It is proposed that this aseismic volume, adjacent to the caldera in the direction of flank slip, consists of olivine cumulates that creep outward, pushing the south flank seawards. Average primary Kilauea tholeiitic magma contains about 16.5 wt.% MgO compared with an average 10 wt.% MgO for erupted subaerial and submarine basalts. This difference requires fractionation of 17 wt.% (14 vol.%) olivine phenocrysts that accumulate near the base of the magma reservoir where they form cumulates. Submarine-erupted Kilauea lavas contain abundant deformed olivine xenocrysts derived from these cumulates. Deformed dunite formed during the tholeiitic shield stage is also erupted as xenoliths in subsequent alkalic lavas. The deformation structures in olivine xenocrysts suggest that the cumulus olivine was densely packed, probably with as little as 5–10 vol.% intercumulus liquid, before entrainment of the xenocrysts. The olivine cumulates were at magmatic temperatures (>1100°C) when the xenocrysts were entrained. Olivine at 1100°C has a rheology similar to ice, and the olivine cumulates should flow down and away from the summit of the volcano. Flow of the olivine cumulates places constant pressure on the unbuttressed seaward flank, leading to an extensional region that localizes deep intrusions behind the flank; these intrusions add to the seaward push. This mechanism ties the source of gravitational instability to the caldera complex and deep rift systems and, therefore, limits catastrophic sector failure of Hawaiian volcanoes to their active growth phase, when the core of olivine cumulates is still hot enough to flow.

Mantle Plume Helium in Submarine Basalts from the Galápagos Platform

Helium-3/helium-4 ratios in submarine basalt glasses from the Galapagos Archipelago range up to 23 times the atmospheric ratio in the west and southwest. These results indicate the presence of a relatively undegassed mantle plume at the Galápagos hot spot and place Galápagos alongside Hawaii, Iceland, and Samoa as the only localities known to have such high helium-3/helium-4 ratios. Lower ratios across the rest of the Galápagos Archipelago reflect systematic variations in the degree of dilution of the plume by entrainment of depleted material from the asthenosphere. These spatial variations reveal the dynamics of the underlying mantle plume and its interaction with the nearby Galápagos Spreading Center.

Origin of magnetic viscosity and estimate of long‐term induced magnetization in coarse‐grained submarine basalts

The characteristics of magnetic viscosity and susceptibility disaccommodation of a submarine basalt containing multidomain (MD) titanium‐rich titanomagnetite are found to be quite dissimilar at room temperature. Thus, the viscous induced magnetization (VIM) of submarine basalts in situ is not due to diffusion but to thermal fluctuations. Experiments above room temperature show that a linear log t extrapolation of laboratory measurements at room temperature yields an underestimate of the Brunhes VIM. A theoretical calculation of this projected VIM is difficult because of the thermal dependence of the activation volume of barriers in MD low‐Curie point minerals.

Temporal and geochemical variability of volcanic products of the Marquesas Hotspot

The Marquesas archipelago is a short, NW‐SE trending cluster of islands and seamounts that formed as a result of volcanic activity over a weak hotspot. This volcanic chain lies at the northern margin of a broad region of warm and compositionally diverse mantle that melts to build several other subparallel volcanic lineaments. Basalts dredged from submerged portions of volcanoes along the Marquesas lineament decrease in age from northwest to southeast. The new sample age distribution yields a volcanic migration rate significantly slower than that expected for Pacific plate motion over a stationary Marquesas hotspot. This and the aberrant orientation of the chain indicate deflection of the plume by westward upper mantle flow. The interaction of this weak plume with upper mantle flow accounts for the temporal and spatial patterns in Marquesan volcanism. The compositions of subaerial and submarine basalts reflect the mixing of at least two mantle sources, distinguished by Sr, Nd, and Pb isotope and trace element compositions. There is a consistent evolutionary pattern at each volcano, from early tholeiitic to later alkalic basalt eruptions. Tholeiitic and transitional lava compositions can be derived by variable degrees of partial melting of a source composed of depleted mid‐ocean ridge basalt mantle (DMM) and mantle characterized by radiogenic Pb (HIMU). Alkalic lava compositions appear to be dominantly the result of smaller degrees of melting of a more radiogenic mantle source (EM II). Large‐scale melting of the lower lithosphere or upper mantle (DMM+HIMU) entrained within a sheared, thermally buoyant plume (EM II) could produce the tholeiitic and transitional basalts found in the main shields of the volcanoes, while alkalic basalts could result from melting of mantle of EM II composition at the edges of the hotspot.