Compositional and thermal convection in magma chambers

Magma chambers cool and crystallize at a rate determined by the heat flux from the chamber. The heat is lost predominantly through the roof, whereas crystallization takes place mainly at the floor. Both processes provide destabilizing buoyancy fluxes which drive highly unsteady, chaotic convection in the magma. Even at the lowest cooling rates the thermal Rayleigh number Ra is found to be extremely large for both mafic and granitic magmas. Since the compositional and thermal buoyancy fluxes are directly related it can be shown that the compositional Rayleigh number Rs (and therefore a total Rayleigh number) is very much greater than Ra. In the case of basaltic melt crystallizing olivine Rs is up to 10⁶ times greater than Ra. However compositional and thermal buoyancy fluxes are roughly equal. Therefore thermal and compositional density gradients contribute equally to convection velocities in the interior of the magma. Effects of thermal buoyancy generated by latent heat release at the floor are included.The latent heat boundary layer at the floor of a basaltic chamber is shown to be of the order of 1 m thick with very low thermal gradients whereas the compositional boundary layer is about 1 cm thick with large compositional gradients. As a consequence, the variation in the degree of supercooling in front of the crystal-liquid interface is dominated by compositional effects. The habit and composition of the growing crystals is also controlled by the nature of the compositional boundary layer. Elongate crystals are predicted to form when the thickness of the compositional boundary layer is small compared with the crystal size (as in laboratory experiments with aqueous solutions). In contrast, equant crystals form when the boundary layer is thicker than the crystals (as in most magma chambers). Instability of the boundary layer in the latter case gives rise to zoning within crystals. Diffusion of compatible trace elements through the boundary layer can also explain an inverse correlation, observed in layered intrusions, between Ni concentration in olivine and the proportion of Ni-bearing phases in the crystallizing assemblage.     

Convection in Aid of Adcumulus Growth

Sheet cooling occurs for most large magma bodies emplaced intothe earth's crust The convection that ensues is driven by thestove effect of the feeder, by cooling principally at the roof,and to an important degree by two-phase flow of crystal-liquidsuspensions accelerated by rapid crystal growth upon compression.Significant floor cooling through the cumulate substrate islimited to the early history of an intrusion. Thereafter a thermalmaximum in the T-Z profile (a latent-heat hump) always occursat the cumulus interface. If crystals grow, the latent heatis carried away chiefly by the magma. The solute rejected bythe growing interface is buoyant for mafic and ultramafic cumulatesbut dense for all felsic and feldspar-cotectic cumulates notcontaining cumulus iron oxide minerals. The rejected solute(RS) is light for calc-alkaline cumulates in general, a factprobably germane to the origin of rhyolite by sidewall crystallization. Near the cumulate interface, the ratio R_p of compositional tothermal effects on liquid density lies in the range 10–10⁶,indicating a strong compositional control on density that cannotbe overcome by available thermal contrasts. The solidificationof the cumulate is enhanced by upward removal of light RS butimpeded by stagnation of dense RS. The existence of felsic flatfloor adcumulates proves that adcumulus growth can occur largelyby diffusion, as all other options are precluded. This resultis entirely consistent with known diffusivities of heat andmass, for accumulation rates near ? cm yr^–1. When higheraccumulation rates occur, some residual liquid is trapped. Theflux ratio of heat to matter during adcumulus growth is about300 cal g^–1. From the flux ratios and the one-dimensionalestimates of cooling rates, typical values of the thermal andcompositional gradients during adcumulus growth can be obtained.After a new batch of supercooled magma arrives the thermal gradientdrops from near infinity to > 5000? km^–1 during nucleationand then to a steady state near 40?C km^–1 at the interfaceduring adcumulus growth. The latent heat hump is paired with another thermal maximumabove the boundary layer, which is therefore a heat sink. Byconsideration of the thermal structure and the buoyancy of rejectedsolute it is determined that double diffusive convection isinconsistent with the adcumulus growth of any type of floorcumulate, and with any growth of an ultramafic cumulate. Infiltrationmetasomatism is an orthocumulus process rather than an adcumulusone, and it produces mesocumulates. Multiple stratification of magma will not arise from growthat the floor, but storage of buoyant RS in a polymerized layerrich in plagioclase component appears to occur on a rapid timescalein snowflake troctolite and may have led to anorthosite formationon a slow timescale in the Stillwater Complex. Magma stratificationby multiple injection is likely to be unstable to two-phaseconvection. Compaction of cumulates is limited to cases wheresufficiently thick crystal mush can be shown to exist, and suchthicknesses are rare in large intrusions. Compaction is thereforenot a general alternative to adcumulus growth. Cumulate theoryis very much alive and able to predict testable ideas aboutsilicate diffusivities and convection.     

Convection and mixing in magma chambers

This paper reviews advances made during the last seven years in the application of fluid dynamics to problems of igneous petrology, with emphasis on the laboratory work with which the authors have been particularly involved. Attention is focused on processes in magma chambers which produce diversity in igneous rocks, such as fractional crystallization, assimilation and magma mixing. Chamber geometry, and variations in the density and viscosity of the magma within it, are shown to play a major role in determining the dynamical behaviour and the composition of the erupted or solidified products.Various convective processes are first reviewed, and in particular the phenomenon of double-diffusive convection. Two types of double-diffusive interfaces between layers of different composition and temperature are likely to occur in magma chambers. A diffusive interface forms when a layer of hot dense magma is overlain by cooler less dense magma. Heat is transported between the layers faster than composition, driving convection in both layers and maintaining a sharp interface between them. If a layer of hot slightly less dense magma overlies a layer of cooler, denser but compositionally lighter magma, a finger interface forms between them, and compositional differences are transported downwards faster than heat (when each is expressed in terms of the corresponding density changes).Processes leading to the establishment of density, compositional and thermal gradients or steps during the filling of a magma chamber are considered next. The stratification produced, and the extent of mixing between the inflowing and resident magmas, are shown to depend on the flow rate and on the relation between the densities and viscosities of the two components. Slow dense inputs of magma may mix very little with resident magma of comparable viscosity as they spread across the floor of the chamber. A similar pulse injected with high upward momentum forms a turbulent “fountain”, which is a very efficient mechanism for magma mixing, as is a turbulent plume of less dense magma rising through the host magma to the top of the chamber.The form of convection in the filled magma chamber is controlled by the shape and size of the chamber, the viscosity of the magma (through the Rayleigh number which is usually high in the early stages of cooling), and by processes at the boundary which produce lighter or denser fluid than that in the interior of the chamber. Compositional convection due to fluid released by crystallization often dominates over thermal convection. If crystallization at the bottom of a funnel-shaped chamber releases a light magma, this convects away from the floor, causing turbulent convection which tends to homogenize the overlying melt. If the magma released is dense, it flows down the sloping floor and stratifies the magma at the base of the chamber. Convection driven by crystallization in an inverted funnel has the reverse effect, e.g. dense fluid released at the sloping roof now has a homogenizing influence. Assimilation of wall rocks can also lead to identical dynamical effects and thus to zoning in magma chambers. Melting of a light roof, for instance, can produce a layer of cool felsic magma overlying the hotter more basic magma in the lower part of the chamber, with a diffusive interface between them. Assimilation has also been discussed for other geometries: assimilation of the walls of dykes, sills and lava flows can occur when the flow is hot and turbulent, whereas if the flow is laminar the magma will chill against the adjacent rocks and protect them from assimilation.When the magma in a chamber is layered, crystallization can cause the composition and density to change in several ways which may lead to mixing. A crystallizing lower layer of hot dense magma can evolve till it has the density of the magma above it, causing sudden overturning and thorough mixing. On the other hand, with a much more viscous layer above, light fluid is released continuously during crystallization and rises to the top of the chamber with little mixing. Overturning of a gas-rich mafic lower layer into a cooler silicic layer can cause a sudden quenching, with the rapid release of gas which could trigger an explosive eruption. Mixing can also occur during eruption, as two layers are drawn up simultaneously from a stratified chamber when a critical flow velocity is exceeded, and they then mix in the outlet vent. Laboratory experiments suggest, however, that magma mixing is inhibited by large viscosity differences, both during the filling and emptying of a magma chamber. Scaling these results to magmas indicates that a basaltic magma can flow into the bottom of a chamber containing rhyolite with little or no mixing between them, and that these two magma types can also flow out through the same exit vent with limited mixing.Each of the phenomena discussed in this review has been studied, at least in a qualitative way, using laboratory experiments to identify and understand a significant physical process occurring in magma chambers. The field of geological fluid mechanics and its application to these problems is still very new, and further advances seem assured as new phenomena are identified and more detailed and quantitative analogue experiments are developed to study them.     

Crystallization and Layering of the Skaergaard Intrusion

Solidification of large slowly cooled intrusions is a complexprocess entailing progressive changes of rheological propertiesas the crystallizing magma passes through successive stagesbetween a viscous Newtonian fluid and a brittle solid rock.Studies of this transition in the Skaergaard intrusion indicatethat most crystallization took place in an advancing front ofsolidification against the floor, walls, and roof where crystalsnucleated and grew in a static boundary layer, much in the mannerproposed by Jackson in 1961. The non-Newtonian properties ofthe crystallizing magma account for the fact that plagioclase,which was lighter than the liquid, is a major component of rockson the floor, while mafic minerals that were heavier than theliquid accumulated under the roof. Crystals that nucleated andgrew in these zones were trapped by an increasingly rigid zonethat advanced more rapidly than the crystals sank or floated.If any crystals escaped entrapment, they were those of the largestsize and density contrast. The rates of accumulation in different parts of the intrusionwere not governed by rates of gravitational accumulation somuch as by the nature of convection and heat transfer. Cumulatetextures, preferred orientations of crystals, and layering,all of which have been taken as evidence of sedimentation, canbe explained in terms of in situ crystallization. Layering cannothave been caused by density currents sweeping across the floor;it is well developed on the walls and under the roof, lacksthe size and density grading and mineralogical compositionsthat would be expected, and shows no evidence of having beenaffected by obstructions in the paths of the currents. We propose an alternative origin of layering that is based onprocesses governed by the relative rates of chemical and thermaldiffusion during cooling. Intermittent layering resulted fromgravitational stratification of the liquid, and cyclic layeringwas produced by an oscillatory process of nucleation and crystalgrowth. The effects of differentiation during in situ crystallizationare strongly dependent on relative rates of diffusion of individualcomponents, and some of the compositional variations in differentparts of the intrusion can be explained in terms of these differences.     

Petrology of the Marginal Border Series of the Skaergaard Intrusion

The Marginal Border Series (MBS) of the Skaergaard intrusionconsists of rocks formed by in situ crystallization againstthe walls of the intrusion. Most of these rocks are productsof fractional crystallization, though samples believed to representchilled liquid occur locally at the intrusive contact. The MBScomprises only 5% of the exposed volume of the intrusion, butwithin its thickness, the order of crystallization and the compositionsof fractionated rocks and minerals vary systematically withdistance inward from the intrusive contact in largely the samemanner as rocks and minerals upward through the Layered Series(LS). Earliest differentiates are cumulates of olivine and plagioclase.The most basic compositions of cumulus plagioclase (An₇₂) andolivine (Fo₈₄) in these rocks indicate that the amount of fractionationpreceding formation of the exposed LS was substantially lessthat previously believed. Field and compositional data indicatethat picritic blocks are xenoliths rather than cumulates ofthe Skaergaard magma. Xenoliths of gneiss in all stages of reactionare locally abundant; however, there is no evidence that uppercrustal material contaminated the magma from which the MBS cumulatesformed. Compositions of cumulus minerals in the MBS differ fromthose in comparable LS rocks. Cumulates in the lower marginscontain more calcic plagioclase, more magnesian augite in allbut the late differentiates, and more iron-rich olivine. Thecompositions of cumulus olivine and to a lesser degree thoseof other mafic silicates, were modified to more iron-rich compositionsby re-equilibration with relatively large amounts of interstitialliquid. The lower MBS and LS crystallized from the same magma, but fractionationoccurred at different rates on the walls and floor of the intrusion.The upper margin may have crystallized from a magma of modifiedcomposition and fractionated at rates different from that inthe lower margin and Upper Border Series (UBS). Crystals onthe floor and roof of the intrusion accumulated faster or moreefficiently than on the walls. At any given stage of fractionation,crystals also accumulated against all sides of the magma chamberat about the same rate. Either the rates of cooling, crystallization,and crystal retention affected accumulation rates locally asfunctions of rock type and geometry of the walls, or these rateswere largely independent of wall rock owing to buffering ofconductive heat loss possibly to an envelope of hydrothermalfluid circulating around the crystallizing magma. The appearanceor disappearance of cumulus minerals in the lower MBS occursat higher structural levels than in the LS and at lower structurallevels than in the UBS. These relationships together with cumulusmineral compositions indicate that magma at the margins wasalways somewhat less fractionated than that at the floor androof of the chamber. It is proposed that these relationshipsreflect the combined effects of liquid and crystal fractionationof the magma within largely independent convection systems inthe lower and upper parts of the chamber.     

The stratigraphy of western Australia

Abstract

Measurements of intensity of magnetization and susceptibility have been made at intervals of 5 ft. on some bore cores from Tasmanian tholeiites of Jurassic age. In all cases the direction of magnetization was nearly vertical. One sill appears to have its upper 700 ft. magnetized normally and its lower part reversed.

The magnetic properties vary very rapidly and show a structure in which narrow regions with high values are superposed on a general increase to a maximum at about 300 ft. from the top of the sill. This coincides with a minimum of the density and a maximum of normative magnetite and total iron and also in the size of the areas of mesostasis.

It is suggested that this behaviour is due to differentiation and that magnetic measurements form an extremely sensitive tool for the study of this process. An attempt is made to show that differentiation by the settling of individual crystals through the sill is impossible in thick sills in which convection does not occur. and that the mechanism must be one of roof stoping in which large blocks of mixed crystals and residual magma fall from the roof to the floor of the sill. The regions with exceptionally high magnetic values which have been observed are regarded as composed of magma trapped between these blocks.     

Crystal settling and convection in the Shiant Isles Main Sill

The 168 m-thick Shiant Isles Main Sill is a composite body, dominated by an early, 24 m-thick, picrite sill formed by the intrusion of a highly olivine-phyric magma, and a later 135 m-thick intrusion of olivine-phyric magma that split the earlier picrite into a 22 m-thick lower part and a 2 m-thick upper part, forming the picrodolerite/crinanite unit (PCU). The high crystal load in the early picrite prevented effective settling of the olivine crystals, which retain their initial stratigraphic distribution. In contrast, the position of the most evolved rocks of the PCU at a level ~80% of its total height point to significant accumulation of crystals on the floor, as evident by the high olivine mode at the base of the PCU. Crystal accumulation on the PCU floor occurred in two stages. During the first, most of the crystal load settled to the floor to form a modally and size-sorted accumulation dominated by olivine, leaving only the very smallest olivine grains still in suspension. The second stage is recorded by the coarsening-upwards of individual olivine grains in the picrodolerite, and their amalgamation into clusters which become both larger and better sintered with increasing stratigraphic height. Large clusters of olivine are present at the roof, forming a foreshortened mirror image of the coarsening-upwards component of the floor accumulation. The coarsening-upwards sequence records the growth of olivine crystals while in suspension in a convecting magma, and their aggregation into clusters, followed by settling over a prolonged period (with limited trapping at the roof). As olivine was progressively lost from the convecting magma, crystal accumulation on the (contemporaneous) floor of the PCU was increasingly dominated by plagioclase, most likely forming clusters and aggregates with augite and olivine, both of which form large poikilitic grains in the crinanite. While the PCU is unusual in being underlain by an earlier, still hot, intrusion that would have enhanced any driving force for convection, we conclude from comparison with microstructures in other sills that convection is likely in tabular bodies >100 m thickness.     

Chemical and thermal zonation in a mildly alkaline magma system Infiernito Caldera,Trans-Pecos Texas

Postcollapse lavas of the Infiernito caldera grade stratigraphically upward from nearly aphyric, high-silica rhyolite (76% SiO₂) to highly prophyritic trachyte (62% SiO₂). Plagioclase, clinopyroxene, orthopyroxene, magnetite, ilmenite, and apatite occur as phenocrysts throughout the sequence. Sanidine, biotite, and zircon are present in rocks with more than about 67% SiO₂. Major and trace elements show continuous variations from 62 to 76% SiO₂. Modeling supports fractional crystallization of the observed phenocrysts as the dominant process in generating the chemical variation.Temperatures calculated from coexisting feldspars, pyroxenes, and Fe-Ti oxides agree and indicate crystallization from slightly more than 1100° C in the most mafic trachyte to 800° C in high-silica rhyolite. The compositional zonation probably arose through crystallization against the chilled margin of the magma chamber and consequent rise of more evolved and therefore less dense liquid.Mineral compositions vary regularly with rock composition, but also suggest minor mixing and assimilation of wall rock or fluids derived from wall rock. Mixing between liquids of slightly different compositions is indicated by different compositions of individual pyroxene phenocrysts in single samples. Liquid-solid mixing is indicated by mineral compositions of glomerocrysts and some phenocrysts that apparently crystallized in generally more evolved liquids at lower temperature and higher oxygen fugacity than represented by the rocks in which they now reside. Glomerocrysts probably crystallized against the chilled margin of the magma chamber and were torn from the wall as the liquid rose during progressive stages of eruption. Assimilation is indicated by rise of oxygen fugacity relative to a buffer from more mafic to more silicic rocks.Calculation of density and viscosity from the compositional and mineralogical data indicates that the magma chamber was stably stratified; lower temperature but more evolved, thus less dense, rhyolite overlay higher temperature, less evolved, and therefore more dense, progressively more mafic liquids. The continuity in rock and mineral compositions and calculated temperature, viscosity, and density indicate that compositional gradation in the magma chamber was smoothly continuous; any compositional gaps must have been no greater than about 2% SiO₂.     

Density changes during the fractional crystallization of basaltic magmas: fluid dynamic implications

The dynamical behaviour of basaltic magma chambers is fundamentally controlled by the changes that occur in the density of magma as it crystallizes. In this paper the term fractionation density is introduced and defined as the ratio of the gram formula weight to molar volume of the chemical components in the liquid phase that are being removed by fractional crystallization. Removal of olivine and pyroxene, whose values of fractionation density are larger than the density of the magma, causes the density of residual liquid to decrease. Removal of plagioclase, with fractionation density less than the magma density, can cause the density of residual liquid to increase. During the progressive differentiation of basaltic magma, density decreases during fractionation of olivine, olivine-pyroxene, and pyroxene assemblages. When plagioclase joins these mafic phases magma density can sometimes increase leading to a density minimum. Calculations of melt density changes during fractionation show that compositional effects on density are usually greater than associated thermal effects.In the closed-system evolution of basaltic magma, several stages of distinctive fluid dynamical behaviour can be recognised that depend on the density changes which accompany crystallization, as well as on the geometry of the chamber. In an early stage of the evolution, where olivine and/or pyroxenes are the fractionating phases, compositional stratification can occur due to side-wall crystallization and replenishment by new magma, with the most differentiated magma tending to accumulate at the roof of the chamber. When plagioclase becomes a fractionating phase a zone of well-mixed magma with a composition close to the density minimum of the system can form in the chamber. The growth of a zone of constant composition destroys the stratification in the chamber. A chamber of well-mixed magma is maintained while further differentiation occurs, unless the walls of the chamber slope inwards, in which case dense boundary layer flows can lead to stable stratification of cool, differentiated magma at the floor of the chamber.In a basaltic magma chamber replenished by primitive magma, the new magma ponds at the base and evolves until it reaches the same density and composition as overlying magma. Successive cycles of replenishment of primitive magma can also form compositional zonation if successive cycles occur before internal thermal equilibrium is reached in a chamber. In a chamber containing well-mixed, plagioclase — saturated magma, the primitive magma can be either denser or lighter than the resident magma. In the first case, the new magma ponds at the base and fractionates until it reaches the same density as the evolved magma. Mixing then occurs between magmas of different temperatures and compositions. In the second case a turbulent plume is generated that causes the new magma to mix immediately with the resident magma.     

地下综合管廊穿越地裂缝变形与受力特征研究

以西安市昆明路地下综合管廊穿越f3地裂缝为研究对象,基于有限元数值模拟分析了地裂缝错动作用下分段地下综合管廊的变形与受力特征。结果表明:地裂缝错动作用下地下管廊顶板竖向沉降变形整体上呈现反“S”形特征,其变形量随地裂缝错动量的增大而增大;管廊结构纵向变形大致可划分3个变形段即下盘翘曲变形段、不均匀沉降段和上盘整体沉降段;在管廊设计使用寿期100 a内地裂缝错动量为50 cm时,管廊接头部位顶板的水平位移在地裂缝带处达到峰值,为4.1 cm,而底板水平位移为3.2 cm,管廊接头部位易发生张开、错位破坏现象,应予以加固;在地裂缝带附近,上盘管廊底板的接触压力减低至0,存在底板脱空现象,应预留注浆孔便于必要时进行注浆加固处理,而下盘管廊底板的接触压力则有明显增大的趋势;当地裂缝错动量超过20 cm时管廊结构顶、底板的拉应变超过了混凝土的极限拉应变,管廊变形破坏模式主要为拉张破坏。研究结果可以为西安市及其他地裂缝发育区地下综合管廊穿越地裂缝带的结构设计提供科学依据。     

Fractional crystallization and magma mixing in the Tigalak layered intrusion,the Nain anorthosite complex,Labrador

The Tigalak intrusion is a dominantly dioritic layered body, about 80 km² in area, which ranges in composition from norite to granodiorite. Local areas of the layered rocks display upward fractionation from norite to ferrodiorite. Periodic reversals of mineral composition trends record the emplacement of less fractionated dioritic magma. Heterogeneous mixtures of dioritic and granodioritic rocks occur widely in mappable lenses and layers that alternate up section and along the strike with more uniformly layered rocks. In these mixtures, chilled dioritic pillows occur abundantly in a hybrid cumulate matrix of granodiorite to diorite composition. Cross-cutting granodioritic dikes grade upward into stratigraphically-bound lensoid masses of the hybrid cumulates. It appears that the hybrid rocks formed as a result of the emplacement of the granodioritic magma through lower cumulates into the dioritic magma chamber and that the dioritic pillows represent chilled bodies of Ferich dioritic magma that commingled with cooler granodioritic magma and settled to the floor of the Tigalak magma chamber. The restricted distribution of these mixtures of hybrid cumulates and chilled pillows indicates that mixing between granodioritic and dioritic liquids was limited in time and lateral extent. Periodic injections of granodioritic liquids may have collected as a separate layer below the roof of the magma chamber and above dioritic magma.     

深部厚顶煤巷道围岩变形破坏机制模型试验研究

为研究深部厚顶煤巷道围岩变形破坏特性及其机制,以赵楼煤矿千米深井厚顶煤巷道为工程背景,开展了大比尺地质力学模型试验,对让压型锚索箱梁支护系统作用下的巷道围岩位移、应力演化规律进行的研究表明：巷道顶底板围岩竖向应力释放较两帮剧烈,水平应力释放反之,巷道顶板中部围岩是顶板竖向应力释放的主要部位。通过与现场试验结果对比验证,总结出深部厚顶煤巷道围岩变形破坏的3个主要特征：顶板变形破坏较两帮和底板严重、顶板围岩变形破坏主要发生在煤岩交界面以下的煤体中、巷中是顶板变形破坏的关键部位,并进一步分析了相应机制：顶板煤岩松软破碎、自承能力差、顶板及其巷中竖向应力释放相对更为剧烈、矩形巷道顶板受力状态差等因素,导致顶煤所受径向应力低,碎胀变形剧烈,且弯曲变形、离层严重,顶板受力结构恶化,最终导致顶板控制困难。     

The Skaergaard Layered Series: I. Structure and Average Compositions

Re-examination of the Skaergaard Layered Series in the lightof more extensive field work and sampling shows that the lithologiczones vary laterally as well as vertically, in both their bulkchemical composition and their mineralogical assemblages. Themargins of the zones differ from both the central part of theLayered Series and Marginal Border Series in being richer inFeO^*, TiO₂, K₂, P₂O₂, and most excluded elements. Mafic mineralstend to be more abundant and more iron-rich, plagioclase ismore albitic and more strongly zoned, and apatite and biotiteare more abundant near the margins. When the average compositions of successive zones are compared,the abundances of most excluded components are seen to declineupward as far as Middle Zone then reverse their trends and increasethrough Upper Zone. P₂O₅ and K₂O are negatively correlated inUpper Zones B and C, owing, perhaps, to separation of immisciblefelsic liquids from the iron-rich magma. No evidence has beenfound for introduction of a new batch of less differentiatedmagma. Layered rocks have an average composition that is more maficthan that of homogenous rocks at the same level. Blocks thatfell from the roof have the opposite relation; they are greatlyenriched in felsic components compared to the original compositionsof the Upper Border Series from which they came. Although some of the compositional variations may be consistentwith differing degrees of fractionation of trapped liquids,no consistent relation has been found between the degree offractionation and rates of crystal accumulation or cooling atthe walls. Contamination with the metamorphic wall rocks, eitherby assimilation or by hydrothermal fluids, seems to have hadonly local effects and cannot account for the large-scale variations.At least some of the compositional differences must have resultedfrom late-stage processes that redistributed certain componentsafter the intrusion reached advanced stages of solidification.     

Convective crystal dissolution

The effect of free and forced convection on crystal dissolution is examined both theoretically and experimentally. Well-established relationships for heat and mass transfer are applied to obtain approximate expressions for the dissolution velocity and the associated thickness of the compositional boundary layer. These expressions are found to be in good agreement with experimental observations of the dissolution of quartz crystals in basalt and NaCl crystal in water. When applied to light felsic crystals in basaltic magmas, the expressions predict that forced convection will produce a boundary layer thickness of about 100 μm and a dissolution velocity of order 10⁻⁶ cm s⁻¹. These velocities are too slow for xenocrysts to be dissolved significantly during magma ascent in dykes, but are sufficient for cm-size crystals to dissolve in the interior of a convecting magma chamber. Larger crystals are likely to accumulate at the chamber's roof, where free convection is predicted to dissolve them at velocities of order 10⁻⁷ cm s⁻¹. In an Appendix, the dissolution of the chamber's walls is also considered, and a velocity of order 10⁻⁸cm s⁻¹ is predicted. Editorial responsibility: T.L. Grove     

Olivine-rich units in the Fongen–Hyllingen Intrusion, Norway: implications for magma chamber processes

The Fongen–Hyllingen Intrusion (FHI) is considered to have crystallised from stratified magma residing in a bowl-shaped magma chamber. Seven olivine-rich units, representing the most primitive cumulates in the central part of the intrusion, are associated with compositional reversals and are interpreted as having formed at the lowest part of the magma chamber floor. Based on phase-relationships, the crystallisation order is explained in terms of magma mixing and fractional crystallisation. Repeated influxes of small volumes of dense, primitive magma at the base of the chamber had a major impact on the crystallising assemblage on the local floor and a decreasing effect towards the flanks of the chamber. This was due to the small volume of replenishing magma, the geometry of the chamber and the consequent restriction of magma mixing to the deepest part of the chamber where the new magma was emplaced. It is estimated that the chamber floor sloped as little as 1–2°, but this was sufficient to give widely different cumulate sequences near the bottom of the chamber and on the flanks.     

The Cougar Point Tuff: Implications for Thermochemical Zonation and Longevity of High-Temperature, Large-Volume Silicic Magmas of the Miocene Yellowstone Hotspot

The 12·7–10·5 Ma Cougar Point Tuff in southernIdaho, USA, consists of 10 large-volume (>10²–10³ km³each), high-temperature (800–1000°C), rhyolitic ash-flowtuffs erupted from the Bruneau–Jarbidge volcanic centerof the Yellowstone hotspot. These tuffs provide evidence forcompositional and thermal zonation in pre-eruptive rhyolitemagma, and suggest the presence of a long-lived reservoir thatwas tapped by numerous large explosive eruptions. Pyroxene compositionsexhibit discrete compositional modes with respect to Fe andMg that define a linear spectrum punctuated by conspicuous gaps.Airfall glass compositions also cluster into modes, and thepresence of multiple modes indicates tapping of different magmavolumes during early phases of eruption. Equilibrium assemblagesof pigeonite and augite are used to reconstruct compositionaland thermal gradients in the pre-eruptive reservoir. The recurrenceof identical compositional modes and of mineral pairs equilibratedat high temperatures in successive eruptive units is consistentwith the persistence of their respective liquids in the magmareservoir. Recurrence intervals of identical modes range from0·3 to 0·9 Myr and suggest possible magma residencetimes of similar duration. Eruption ages, magma temperatures,Nd isotopes, and pyroxene and glass compositions are consistentwith a long-lived, dynamically evolving magma reservoir thatwas chemically and thermally zoned and composed of multiplediscrete magma volumes. KEY WORDS: ash-flow tuff; Bruneau–Jarbidge; rhyolite; Yellowstone hotspot; residence time     

Diffusion coupling between trace and major elements and a model for calculation of magma residence times using plagioclase

Residence times of plagioclase crystals in magma reservoirs can be determined by modeling the compositional zoning of trace elements in these crystals. We present a formulation to model diffusion of trace elements in plagioclase paying special attention to certain thermodynamic and kinetic aspects. In particular, we account for the compositional dependence on anorthite content of the chemical potential and diffusion coefficients of trace elements (e.g., Mg), the choice of suitable boundary conditions and potential effects of diffusion in more than one dimension. We show that contrary to intuition, diffusive fluxes of trace elements may be coupled to major element concentration gradients, and ignoring such coupling can lead to incorrect estimates of timescales. We illustrate application of the model using plagioclase crystals from a suite of gabbroic xenoliths from a Holocene dacitic lava flow of Volcán San Pedro (Chilean Andes, 36°S). The inferred timescale for metasomatism of the xenoliths by evolved liquids is on the order of 100 (30 to 148) yr and serves to illustrate how trace element zoning in plagioclase provides a window into timescales of magmatic processes inaccessible by isotopic or other methods.     

The Origin of Marginal Compositional Reversals in Basic-Ultrabasic Sills and Layered Intrusions by Soret Fractionation

Marginal reversals—a common feature of many basic differentiatedigneous bodies regardless of their size and bulk composition—areremarkable in being a mirror of the Layered Series. These aredistinguished by: (1) an apparent lack of mass balance betweenthe lower part of the marginal reversals, including chilledmargins, and the bulk composition of the intrusions; (2) mineralcrystallization sequences and (3) mineral compositional trends,which are both essentially the opposite of those in the LayeredSeries; (4) the cotectic composition of rocks composing themarginal reversals; (5) the capacity to form from both phenocryst-richand phenocryst-free parental magmas; (6) the capability to developalong the floor, subvertical walls and even the roof of magmachambers. None of the current models of magma chamber evolutioncan provide an adequate explanation for the characteristic featuresof the marginal reversals. The problem can be resolved in thecontext of a model combining Soret diffusion in thin liquidboundary layers at the magma chamber margins and vigorous convectionin the main magma body. The key proposal is that the formationof marginal reversals takes place through the non-equilibriumevolution of liquid boundary layers as a result of a temperaturegradient imposed by the cold country rock. The fundamental explanationfor the mirror image of a marginal reversal is that the non-equilibriumSoret fractionation works in a manner opposite to that of theequilibrium crystal–liquid fractionation that producesthe Layered Series. KEY WORDS: marginal compositional reversals; sills; layered intrusions; Soret fractionation     

Efficiency of compaction and compositional convection during mafic crystal mush solidification: the Sept Iles layered intrusion,Canada

Adcumulate formation in mafic layered intrusions is attributed either to gravity-driven compaction, which expels the intercumulus melt out of the crystal matrix, or to compositional convection, which maintains the intercumulus liquid at a constant composition through liquid exchange with the main magma body. These processes are length-scale and time-scale dependent, and application of experimentally derived theoretical formulations to magma chambers is not straightforward. New data from the Sept Iles layered intrusion are presented and constrain the relative efficiency of these processes during solidification of the mafic crystal mush. Troctolites with meso- to ortho-cumulate texture are stratigraphically followed by Fe–Ti oxide-bearing gabbros with adcumulate texture. Calculations of intercumulus liquid fractions based on whole-rock P, Zr, V and Cr contents and detailed plagioclase compositional profiles show that both compaction and compositional convection operate, but their efficiency changes with liquid differentiation. Before saturation of Fe–Ti oxides in the intercumulus liquid, convection is not active due to the stable liquid density distribution within the crystal mush. At this stage, compaction and minor intercumulus liquid crystallization reduce the porosity to 30%. The velocity of liquid expulsion is then too slow compared with the rate of crystal accumulation. Compositional convection starts at Fe–Ti oxide-saturation in the pore melt due to its decreasing density. This process occurs together with crystallization of the intercumulus melt until the residual porosity is less than 10%. Compositional convection is evidenced by external plagioclase rims buffered at An₆₁ owing to continuous exchange between the intercumulus melt and the main liquid body. The change from a channel flow regime that dominates in troctolites to a porous flow regime in gabbros results from the increasing efficiency of compaction with differentiation due to higher density contrast between the cumulus crystal matrix and the equilibrium melts and to the bottom-up decreasing rate of crystal accumulation in the magma chamber.     

The fluid dynamics of a basaltic magma chamber replenished by influx of hot,dense ultrabasic magma

This paper describes a fluid dynamical investigation of the influx of hot, dense ultrabasic magma into a reservoir containing lighter, fractionated basaltic magma. This situation is compared with that which develops when hot salty water is introduced under cold fresh water. Theoretical and empirical models for salt/water systems are adapted to develop a model for magmatic systems. A feature of the model is that the ultrabasic melt does not immediately mix with the basalt, but spreads out over the floor of the chamber, forming an independent layer. A non-turbulent interface forms between this layer and the overlying magma layer across which heat and mass are transferred by the process of molecular diffusion. Both layers convect vigorously as heat is transferred to the upper layer at a rate which greatly exceeds the heat lost to the surrounding country rock. The convection continues until the two layers have almost the same temperature. The compositions of the layers remain distinct due to the low diffusivity of mass compared to heat. The temperatures of the layers as functions of time and their cooling rate depend on their viscosities, their thermal properties, the density difference between the layers and their thicknesses. For a layer of ultrabasic melt (18% MgO) a few tens of metres thick at the base of a basaltic (10% MgO) magma chamber a few kilometres thick, the temperature of the layers will become nearly identical over a period of between a few months and a few years. During this time the turbulent convective velocities in the ultrabasic layer are far larger than the settling velocity of olivines which crystallise within the layer during cooling. Olivines only settle after the two layers have nearly reached thermal equilibrium. At this stage residual basaltic melt segregates as the olivines sediment in the lower layer. Depending on its density, the released basalt can either mix convectively with the overlying basalt layer, or can continue as a separate layer. The model provides an explanation for large-scale cyclic layering in basic and ultrabasic intrusions. The model also suggests reasons for the restriction of erupted basaltic liquids to compositions with MgO<10% and the formation of some quench textures in layered igneous rocks.