Small Magma Bodies Research Articles

Formation of fast-spreading lower oceanic crust as revealed by a new Mg–REE coupled geospeedometer

A new geospeedometer is developed based on the differential closures of Mg and rare earth element (REE) bulk-diffusion between coexisting plagioclase and clinopyroxene. By coupling the two elements with distinct bulk closure temperatures, this speedometer can numerically solve the initial temperatures and cooling rates for individual rock samples. As the existing Mg-exchange thermometer was calibrated for a narrow temperature range and strongly relies on model-dependent silica activities, a new thermometer is developed using literature experimental data. When the bulk closure temperatures of Mg and REE are determined, respectively, using this new Mg-exchange thermometer and the existing REE-exchange thermometer, this speedometer can be implemented for a wide range of compositions, mineral modes, and grain sizes.Applications of this new geospeedometer to oceanic gabbros from the fast-spreading East Pacific Rise at Hess Deep reveal that the lower oceanic crust crystallized at temperatures of 998–1353 °C with cooling rates of 0.003–10.2 °C/yr. Stratigraphic variations of the cooling rates and crystallization temperatures support deep hydrothermal circulations and in situ solidification of various replenished magma bodies. Together with existing petrological, geochemical and geophysical evidence, results from this new speedometry suggest that the lower crust formation at fast-spreading mid-ocean ridges involves emplacement of primary mantle melts in the deep section of the crystal mush zone coupled with efficient heat removal by crustal-scale hydrothermal circulations. The replenished melts become chemically and thermally evolved, accumulate as small magma bodies at various depths, feed the shallow axial magma chamber, and may also escape from the mush zone to generate off-axial magma lenses.

Petrochemical features of Miocene volcanism around the Çubukludağ graben and Karaburun peninsula, western Turkey: Implications for crustal melting related silicic volcanism

Widespread Neogene volcanism, mainly intermediate and rarely mafic and felsic in composition, was controlled by the extensional tectonic regime in western Turkey. The Karaburun and Cumaovası volcanics are the cases for understanding the magma source(s) and petrological processes, producing the extension-related mafic and felsic volcanism. The Karaburun volcanics (KV) are mainly oriented north to south in the Karaburun peninsula and span a wide spectrum from basalt (20Ma) to rhyolite (16Ma), and younger trachyte and trachydacites (13Ma). The products of the subaerial silicic volcanism (the Cumaovası volcanics, CV; 17Ma) which are represented by cluster of rhyolite domes, related pyroclastics occur within the NE–SW trending Çubukludağ graben, and intermediate and mafic volcanic rocks are lack in this area. The lavas of the Cumaovası volcanics are high silica rhyolites and rare dacites which are calc alkaline, peralumious and enriched significantly in LILE. Extremely low Sr, Ba values, extremely Eu depletions and very low LaN/YbN ratios are typical for the rhyolites of CV, similar to the topaz rhyolites. The Karaburun volcanics, with the exception of the minor alkaline basaltic and trachytic lavas, are mainly calc alkaline and metaluminous intermediate lavas. 87Sr/86Sr ratios of the KV and dacitic samples of CV are close to each other and range from 0.708 to 0.709; while Sr isotopic ratios of the rhyolites are significantly high and variable (0.724–0.786). 143Nd/144Nd ratios of the CV and KV, except for the alkaline samples, are similar for both sequences vary from 0.51230 to 0.51242.Geological, geochemical, isotopic and radiochronologic data reveal that the KV and CV were formed in extensional tectonic setting, but evolved by different petrological processes in different magma chambers. During the Neogene, underplated mafic magma was injected into the crust and hybridized by mantle and crustal derived materials. Geochemical features and trace element modeling for the mafic members of the KV indicate that they were derived from enriched lithospheric mantle and modified by fractional crystallization from basalt to rhyolite (Helvacı et al., 2009). Unexpectedly, the felsic lavas from Cumaovası region have a unique chemical composition, and similar to the extension related rhyolites formed from small magma bodies. Our data reveal that extension related mafic inputs caused crustal anatectic melting and formed felsic melts that rapidly ascended into the upper crust. The Cumaovası felsic rocks were differentiated into the highly evolved silica-rich melts within the magma chambers trapped near the surface.

Platinum group elements zoning and mineralogy of chromitites from the cumulate sequence of the Nurali massif (Southern Urals, Russia)

The Nurali lherzolite massif is one of the dismembered ophiolite bodies associated with the Main Uralian Fault (Southern Urals, Russia). It comprises a mainly lherzolitic mantle section, an ultramafic clinopyroxene-rich cumulate sequence (Transition Zone), and an amphibole gabbro unit. The cumulate section hosts small chromitite bodies at different stratigraphic heights within the sequence. Chromitite bodies from three different levels along a full section of the cumulate sequence and two from other localities were investigated. They differ in the host lithology, chromitite texture and composition, and PGE content and mineralogy. Chromitites at the lowest level, which are hosted by clinopyroxenite, form cm-scale flattened lenses. They have high Cr# and low Mg# chromites and are enriched in Pt and Pd relative to Os and Ir. At a higher, intermediate level, the chromitites are hosted by dunite. They form meter thick lenses, contain low Cr# and high Mg# chromites, have high PGE contents (up to 26,700 ppb), and are enriched in Os, Ir and Ru relative to Pt and Pd, reflecting a mineralogy dominated by laurite–erlichmanite and PGE–Fe alloys. At the highest level are chromitites hosted by olivine–enstatite rocks. These chromitites have high Cr# and relatively low Mg# chromites and very low PGE content, with laurite as the dominant PGE mineral. The platinum group minerals (PGMs) show extreme zoning, with compositions ranging from erlichmanite to almost pure laurite and from Os-rich to Ru-rich alloys, with variable and irregular zoning patterns. Two chromitite bodies up to 6 km from the main sequence can be correlated with the latter based on geochemistry and mineralogy, implying that the variations in chromitite geochemistry are due to processes that operated on the scale of the massif rather than those that operated on the scale of the outcrop. Pertsev et al. [Pertsev, A.N., Spadea, P., Savelieva, G.N., Gaggero, L., 1997. Nature of the transition zone in the Nurali ophiolite, Southern Urals. Tectonophysics 276, 163–180.] propose that the Transition Zone formed by solidification of a series of small magma bodies that partially overlapped in time and space. The magmas formed by successive partial melting of the underlying mantle. We suggest that this process determined the changing PGE geochemistry of the successive batches of magma. The PGE distribution fits a model of selected extraction from the mantle, where monosulphide solid solution–sulphide liquid equilibrium was attained until complete melting of the monosulphide solid solution. Later and localized variations in fS 2 resulted in the formation of different PGMs with complex zoning patterns.

Slow cooling of middle and lower oceanic crust inferred from multicomponent magnetizations of gabbroic rocks from the Mid‐Atlantic Ridge south of the Kane fracture zone (MARK) area

The remanent magnetization of gabbroic material of the Mid‐Atlantic Ridge south of the Kane fracture zone (MARK) area provides constraints on both the thermal structure and tectonic history of the lower crust in this slow spreading environment. The remanence of these gabbroic samples is often complex, with the juxtaposition of intervals of apparently normal and reversed polarity rocks over small spatial scales (tens of centimeters to a few meters). Moreover, several samples when thermally demagnetized have a reversed polarity magnetization component between higher and lower stability normal polarity components. Given the nominal age (∼1 Ma) of the crust, we suggest that this pattern of normal/reversed‐normal polarity most plausibly reflects emplacement and/or cooling through three successive polarity intervals, Jaramillo normal polarity interval (1.07–0.99 Ma), a portion of the Matuyama reversed polarity interval (0.99–0.78 Ma), and the Brunhes normal polarity interval (0.78 Ma to present). A small number of samples with three well‐defined magnetization components have magnetic characteristics compatible with a remanence carried by fine‐grained, possibly single domain, magnetite. Laboratory unblocking temperatures in these samples therefore allow estimation of lower crustal temperatures at the time of the Jaramillo/Matuyama (0.99 Ma) and Matuyama/Brunhes (0.78 Ma) polarity transitions. Together with depth estimates derived from fluid inclusion studies these results suggest that middle and lower crustal temperatures remained as high as ∼350°–475°C for a minimum of 0.21 m.y. after emplacement. We suggest that continued injection of liquid, in the form of sills or small magma bodies, over a broad region (half width of ∼3 km) is responsible for this slow cooling. In addition, inclinations of the highest stability component from these drill sites are remarkably similar to that expected from an axial geocentric dipole, suggesting that little, if any, resolvable tilt occurred during uplift of these rocks to the seafloor.

Surface heat flow density at the phlegrean fields caldera (SOUTHERN ITALY)

The Phlegrean Fields area is a Holocene caldera located west of Naples, southern Italy. The recent post-caldera activity is characterized by several eruptive centers inside the collapsed area. In order to investigate the still active volcanic processes, surface heat flow measurements were carried out in 1995 in 30 sites of the Phlegrean Fields, and a heat flow map compiled. Filtering of the map reveals some well-defined anomalies superimposed on a general southward-increasing trend. Local anomalies are related to small magma bodies, whereas the observed general trend has been attributed to the effect of groundwater flow. This effect was calculated and removed. The undisturbed mean value of the surface heat flow density in the eastern sector is 149 mW\\m 2, which is above the regional value of 85 mW\\m 2 assigned to the eastern part of the Tyrrhenian Sea, and which is probably influenced by a very large, deep magmatic body.

Petrology and geochemistry of the ultrapotassic rocks from the Sabatini Volcanic District, central Italy: the role of evolutionary processes in the genesis of variably enriched alkaline magmas

The Sabatini Volcanic District (SVD) is a large volcanic field characterised by the lack of any major volcanic center. Its activity, spread over a wide area, started at 0.6 Ma and developed through five main phases, during which several calderas and the Bracciano lake volcano-tectonic depression were formed. All the volcanic rocks belong to the Roman-type ultrapotassic series (HKS), ranging from leucite tephrites to leucite and haüyne phonolites. Although the major- and compatible-element contents indicate a single series of evolution, there are differences in the incompatible trace-element abundances. A high-Ba series (HBaS) has been distinguished from a low-Ba series (LBaS), with the former also enriched in all other incompatible elements (e.g., REE, Nb, Zr, Th) except Rb. The HBaS rocks are plagioclase-free, leucite-bearing lavas and were abundantly outpoured from the Bracciano volcanoes during the third and fifth phase of activity. Plagioclase- and phlogopite-bearing rocks constitute the LBaS and were erupted during the other phases generally from smaller and eccentric volcanic centers. The initial 87Sr/ 86Sr values are higher in the HBaS rocks and do not vary significantly with magma evolution (0.71047–0.71080), but cover a wider range in the LBaS rocks (0.70944–0.71038), with the lowest Sr isotope ratios occurring in the least evolved lavas. The higher Ca content in the olivine and Ti and Al IV in the clinopyroxene, and the lower ulvöspinel content of the Ti-magnetites of the HBaS rocks suggest an evolution at lower pressure and higher temperature for this magma. The observed petrologic characteristics suggest that the HBaS magma evolved at lower depths by processes of refilling, tapping, fractionation and probably assimilation (RTFA), where the crystallisation rate of clinopyroxene+leucite±olivine dominates over the input rate of the fresh magma. The LBaS magma evolved at slightly higher pressure, in separate and small magma bodies, by fractional crystallisation of clinopyroxene+plagioclase±phlogopite±olivine that was often associated with crustal assimilation (AFC). It has been suggested that RTFA processes with high input rate/crystallisation rate ratios could also be responsible for the differentiation between the HBaS and LBaS. The different processes of evolution undergone by the HBaS and LBaS could have been related to the different volumes of magma rising from the source.

The role of seamount volcanism in crustal construction at the Mid‐Atlantic Ridge (24°–30°N)

An analysis of approximately 6000 km2 of Sea Beam swaths indicates that the floor of the median valley of the Mid‐Atlantic Ridge (MAR) between 24°–30°N is strewn with near‐circular volcanoes. We have identified 481 seamounts in the height range 50–650 m. Because many of the features which appear to be seamounts do not meet our criteria fully, they have not been counted, and our numbers are likely a minimum estimate of the population. The large abundance of seamounts on the median valley floor indicates that seamount volcanism plays an important role in the accretionary processes at this section of the MAR. The summit height distribution of the MAR seamount population is consistent with the exponential frequency‐size distribution model for off‐axis eastern Pacific seamounts (Jordan et al., 1983; Smith and Jordan, 1987). The model parameters for the MAR population yield an average of about 195 seamounts per one thousand square kilometers, and a characteristic height of about 60 m. Shape statistics compiled from the 481 seamounts give parameters that are exactly the same, or cover the same range in values as those obtained from a study of Pacific seamounts in the height range 140–3800 m (Smith, 1988) implying a universal control on seamount construction. Based on the volcanic morphology, we identify 18 spreading segments along the ridge. All except two contain a prominent axial volcanic ridge. Many of the seamounts identified are associated with the axial ridges. From the Sea Beam swaths, we infer that the ridges are composed of piled up seamounts and hummocky flows, and interpret them to be the primary sites of crustal construction. This style of volcanism contrasts strongly with that observed at the East Pacific Rise where seamounts are virtually absent at the spreading axis. The construction of seamounts and hummocky flows may be directly related to the fact that the MAR is fed by magma chambers that are limited in size and frequency. We suggest that small magma pockets with slow eruption rates produce seamounts either from pipes or initial fissure eruptions that collapse to construct a single edifice; small magma bodies with somewhat higher eruption rates produce hummocky fissure fed flows. Using buoyancy arguments, the heights of the MAR seamounts are related to the depth of the magma bodies. We hypothesize that small magma bodies rise buoyantly to the base of the hydrothermally cooled brittle lid where they are trapped, and then erupt. If this hypothesis is correct, the thickness of the brittle lid controls seamount height, and the exponential distribution of seamount summit heights implies that the thickness of the brittle Hd follows an exponential distribution in time and space with characteristic thickness 1.4–1.9 km. If seamounts and hummocky flows are fed each from their own discrete magma body trapped at the brittle/ductile transition, then the lower crust of the MAR will be made up of the products of a multitude of small plutons.

High-magnetization zones near spreading center discontinuities

Three-dimensional inversions of the magnetic field reveal the presence of zones of high magnetization over the propagating limb of large spreading center discontinuities. The presence of high-magnetization zones is confirmed by rock magnetic measurements where available. Magnetization highs are associated with basalts which tend to have high Fe contents and low Mg numbers. These data suggest that high-magnetization zones observed over large ridge axis discontinuities are associated with highly differentiated basalts enriched in iron. Following Christie and Sinton [1], such highly evolved basalts may be the result of shallow-level crystal fractionation in small magma bodies with a low supply rate. These small magma bodies are postulated to correspond to the first stages in the development of a sub-axial magmatic system as a result of the propagation of one of the limbs of the offset into older lithosphere. Because high-magnetization zones at large ridge axis discontinuities often correspond to gaps in the along-axis extent of a seismically detectable magma chamber, these magma bodies may be smaller than a few hundred meters. Rock magnetic measurements suggest that the enrichment in iron associated with increased differentiation may be accompanied, in a few cases, by an increase in the concentration of titanomagnetite within the basalts and in the magnetization of the rocks. However, the exact relationship between high magnetization intensities and iron enrichment is complex and unclear, and may be significantly affected by factors such as magnetic mineralogy and crystallization history.

Mineral relationships in gabbroic rocks from fracture zones of Indian Ocean ridges: evidence for extensive fractionation, parental diversity and boundary-layer recrystallization

Summary Evolved two-pyroxene gabbros and ferrogabbros predominate in gabbroic suites dredged from five fracture zones on the SW and Central Indian Ridges. Compositions of olivines (Fo 56 -Fo 85 ) and plagioclases (An 10 -An 80 ) and the generally low magnesium numbers of orthopyroxene (0.58–0.84) and clinopyroxene (0.50–0.89) indicate that most gabbros crystallized from liquids more fractionated than those represented by basalts from adjacent ridge segments and the fracture zones themselves. This disparity, the paucity of diabases and gabbros, the absence of more magnesian gabbros and olivine-spinel cumulates, and the abundance of serpentinite in the dredge collections suggest that the fracture-zone gabbros crystallized in small magma bodies such as dykes or sills. These were emplaced laterally from large central magma chambers along ridge axes towards the fracture zones beneath a carapace of less fractionated basalt. Rare olivine gabbros, however, contain magnesian clinopyroxene (magnesium number >0.84) and orthopyroxene (magnesium number >0.70), compositions which are more magnesian than those of pyroxenes that have been crystallized experimentally from certain ocean-ridge basalts at 1 atm. It has been suggested that such high-magnesium pyroxenes result from moderate- to high-pressure fractionation. However, the magnesian pyroxenes in the Indian Ocean samples follow olivine and plagioclase in the crystallization sequence and have compositions appropriate for crystallization at low pressure. For example, they match compositions of phenocrysts in refractory (low-Na 2 O, low-TiO 2 ) siliceous (51–53% SiO 2 ) basalts from Deep Sea Drilling Project sites in Eocene-Cretaceous portions of the Indian Ocean. Such refractory basalts have not been studied experimentally, but the early crystallization of bronzite at low pressure in one of them is consistent with the siliceous and magnesian composition of the host glass. Alternatively, the more magnesian pyroxenes in the gabbros may have resulted from a process of in situ boundary-layer recrystallization (resorption and recrystallization) of minerals on the walls of small magma bodies as high-temperature magmas were repeatedly injected along them. This mechanism is suggested by phase relationships in simplified basaltic systems at low pressure and by the compositions of rounded (resorbed) mega-crystal minerals which occur in many tholeiites from the Indian Ocean and elsewhere.

P to S converted wave observations at Long Valley Caldera

We report observations of P to S converted teleseismic body waves recorded with seismic stations located on the resurgent dome of Long Valley caldera. We observe strong conversions occurring at boundaries that could be the tops of small magma bodies at depths of 4–9 kilometers beneath the resurgent dome. This technique could prove to be an extremely powerful tool in the search for magma bodies for geothermal energy development. Our results from a two month deployment of triggered, digital teleseismic recorders are discussed.

Three‐dimensional P velocity structure of the crust below Newberry Volcano, Oregon

Utilizing teleseismic P residuals, we have detected a column of high P velocity material extending from within 10 km of the surface below the summit of Newberry Volcano, Oregon, to midcrustal depths near 25 km. We interpret this column to be the expression of a swarm of predominantly subsolidus gabbroic sills and dikes which were intruded as the volcano was built. The high P velocities observed below the volcano severely limit the size of magma chambers which could presently exist in the crust below Newberry Volcano. Those possible include a few percent of partial melt distributed through large volumes of a mafic intrusion zone in the midcrust; a few, smaller, higher melt fraction zones in the midcrust with dimensions less than 6 km and whose aggregate volume is only a few percent of enclosing volumes of 200 km3; small magma bodies with dimensions of a few kilometers located within the upper 10 km of the crust; or a mafic, crystal‐rich magma of arbitrary dimensions located in the upper few km. The low P velocities detected in the upper 4 km beneath the center of the summit caldera may be partially caused by a magma chamber in the second of these catagories.

The El Cajete Series, Valles Caldera, New Mexico

Three rhyolite eruptive units totaling 3 km3 dense rock equivalent volume are the youngest products from the Valles caldera, New Mexico. These pyroclastic and effusive units, herein called the El Cajete Series, were erupted over an appreciable time period with Plinian and ignimbrite‐forming activity preceding an effusive phase by a lengthy but indeterminable interval. New U‐Th disequilibrium and published fission track dates support an age for the event between 170 and 130 ka. The explosive phases formed the El Cajete Plinian deposit (about 1.3 km3), minor dry pyroclastic surges, and the Battleship Rock tuff (about 1.0 km3), a valley‐confined welded ignimbrite. The Plinian eruption column is estimated to have been 28 km high during deposition of the most widely dispersed pumice fall unit. Slow effusion of the Banco Bonito obsidian lava flow (<1.0 km3) onto a dissected surface cut into the El Cajete and Battleship Rock pyroclastic deposits, accompanied by minor explosive activity, terminated the event. Twenty‐four bulk‐rock samples from all three eruptive stages exhibit little significant compositional variation, and the three units are petrographically identical apart from differences arising from contrasting eruption styles. Very few phenocrysts appear to have been in equilibrium with their enclosing high‐silica rhyolitic liquid. Plagioclase grains are typically strongly resorbed, while biotite and hornblende frequently occur in aggregate grains in which textures characteristic of plutonic rocks can be discerned. These features result from partial melting of preexisting crustal igneous rock, probably of dioritic to granodioritic composition, and subsequent eruption of the resulting liquid plus restite crystals. Rapid generation and eruption of rhyolitic magma during this most recent phase of activity in the Jemez Mountains may imply that the Valles magma system is presently in a state where small magma bodies are transient phenomena.

A magmatic model of Medicine Lake Volcano, California

Medicine Lake volcano is a Pleistocene and Holocene shield volcano of the southern Cascade Range. It is located behind the main Cascade arc in an extensional tectonic setting where high‐alumina basalt is the most commonly erupted lava. This basalt is parental to the higher‐silica calc‐alkaline and tholeiitic lavas that make up the bulk of the shield. The presence of late Holocene, chemically identical rhyolites on opposite sides of the volcano led to hypotheses of a large shallow silicic magma chamber and of a small, deep chamber that fed rhyolites to the surface via cone sheets. Subsequent geophysical work has been unable to identify a large silicic magma body, and instead a small one has apparently been recognized. Some geologic data support the geophysical results. Tectonic control of vent alignments and the dominance of mafic eruptions both in number of events and volume throughout the history of the volcano indicate that no large silicic magma reservoir exists. Instead, a model is proposed that includes numerous dikes, sills, and small magma bodies, most of which are too small to be recognized by present geophysical methods.