A Fluid-Dynamical Study of Crystal Settling in Convecting Magmas

Thermal convection in magma chambers is believed to be almostalways highly time-dependent, or ‘turbulent’, andpredicted convective velocities are commonly orders of magnitudelarger than settling velocities for typical crystals calculatedfrom Stokes' Law. To understand crystal settling in magma chamberswe have therefore undertaken a theoretical and experimentalstudy of particle settling in a turbulently convecting fluid. The regime of interest is where the ratio, S, of the Stokes'Law settling velocity, v_s, to the root mean square verticalcomponent of convective velocity, W, at mid-depth in the fluidis less than unity. Although v_s < W, settling is still possiblebecause convective velocities are height-dependent and mustdecrease to zero at the boundaries of the fluid. Particles immediatelyadjacent to the bottom boundary settle out with their full Stokes'settling velocities. At the same time, convection is vigorousenough to ensure that the distribution of particles in the fluidis uniform. It follows that the number of particles in suspensiondecays with time according to an exponential law, and the decayconstant is simply the ratio of v_s to h, the depth of the fluid.Experiments confirm this relationship, at least for low particleconcentrations, provided S < 0.5 and there is no re-entrainmentof particles from the floor of the tank. We apply this relationship to crystals in magma chambers andso calculate residence times for typical crystals. We find thatfor basaltic magmas the predicted residence times are smallcompared with the many thousands of years that a chamber takesto solidify if cooling is dominated by conduction through thecountry rock. We therefore conclude that crystal settling maybe an efficient differentiation mechanism. Significant magmaticevolution may, however, take place on time-scales that are competitivewith these residence times. If the settling of crystals is the rate-limiting step duringthe crystallization of a magma chamber it is expected that asteady state will be achieved at which the rate of supply ofcrystals into the convecting magma by crystallization balancesthe rate at which crystals settle out. We show how this ideacan explain both the lack of hydraulic equivalence in cumulaterocks and the commonly observed discrepancy between the relativeproportions of phenocrysts of various phases in fractionatedbasaltic lavas and the calculated relative proportions of thesemineral phases in the fractionating assemblage. Finally, anattempt is made to calculate the steady-state crystal contentof convecting magma chambers. Comparison of the predicted crystalcontents with the observed phenocryst contents of typical basalticlavas suggests that magma chambers may often cool more rapidlythan would be expected for conduction through the country rockalone.     

Assimilation and Fractional Crystallization Controlled by Transport Process of Crustal Melt: Implications from an Alkali Basalt-Dacite Suite from Rishiri Volcano, Japan

Mechanisms of fractional crystallization with simultaneous crustalassimilation (AFC) are examined for the Kutsugata and Tanetomilavas, an alkali basalt–dacite suite erupted sequentiallyfrom Rishiri Volcano, northern Japan. The major element variationswithin the suite can be explained by boundary layer fractionation;that is, mixing of a magma in the main part of the magma bodywith a fractionated interstitial melt transported from the mushyboundary layer at the floor. Systematic variations in SiO₂ correlatewith variations in the Pb, Sr and Nd isotopic compositions ofthe lavas. The geochemical variations of the lavas are explainedby a constant and relatively low ratio of assimilated mass tocrystallized mass (‘r value’). In the magma chamberin which the Kutsugata and Tanetomi magmas evolved, a strongthermal gradient was present and it is suggested that the marginalpart of the reservoir was completely solidified. The assimilantwas transported by crack flow from the partially fused floorcrust to the partially crystallized floor mush zone throughfractures in the solidified margin, formed mainly by thermalstresses resulting from cooling of the solidified margin andheating of the crust. The crustal melt was then mixed with thefractionated interstitial melt in the mushy zone, and the mixedmelt was further transported by compositional convection tothe main magma, causing its geochemical evolution to be characteristicof AFC. The volume flux of the assimilant from the crust tothe magma chamber is suggested to have decreased progressivelywith time (proportional to t^–1/2), and was about 3 x 10^–2m/year at t = 10 years and 1 x 10^–2 m/year at t = 100years. It has been commonly considered that the heat balancebetween magmas and the surrounding crust controls the couplingof assimilation and fractional crystallization processes (i.e.absolute value of r). However, it is inferred from this studythat the ratio of assimilated mass to crystallized mass canbe controlled by the transport process of the assimilant fromthe crust to magma chambers. KEY WORDS: assimilation and fractional crystallization; mass balance model; magma chamber; melt transport; Pb isotope     

Petrology of Calc-Alkaline Lavas at Volcn Ollage and the Origin of Compositional Diversity at Central Andean Stratovolcanoes

Volcn Ollage (2117'S) is a large stratovolcano that liesslightly east of the main axis of Quaternary Volcanoes in theAndean Central Volcanic Zone (CVZ). Euptive products range frombasaltic andesite to dacite and define a high-K, calc-alkalinesuite. This compositional range is similar to the collectivecompositional range of the other stratovolcanoes in the CVZ,and it provides a record of both early and late-stage differentiationprocesses operating at the stratovolcanoes. The volumetrically dominant andesitic and dacitic lavas aredivided into four eruptive series on the basis of vent locationsand petrography. In ascending stratigraphic order they are:the Vinta Loma, Chasca Orkho, post-collapse, and La Celosa series.Whole-rock compositions of the lavas are remarkably similarregardless of eruptive series. Variations in phenocryst assemblagesand magmatic f_o2 however, suggest differences in subliquidusvolatile contents for magma chambers developed beneath the summitof the volcano versus those developed beneath the flanks. Basalticandesite magmas are principally preserved as quenched inclusionswithin the andesitic and dacitie lava flows. Large ranges inisotopic ratios over a narrow compositional range indicate thatthe basaltic andesites were derived by crystal fractionationcoupled with large amounts of crustal assimilation. IncreasingCe/Yb ratios with decreasing Yb contents further suggest thatthis initial stage of differentiation occurred at deep crustallevels where garnet was stable. Additional supporting evidencefor differentiation in the deep crust includes isotopic andtrace element compositions that indicate assimilation by thebasaltic andesite magmas of a crust different from upper-crustalrocks exposed at present in the region. Whole-rock major and trace element trends of the dacitic lavascan be simulated largely by fractional crystallization of parentalandesitic magma. The fractionating assemblages for the differenteruptive series are consistent with the observed modes of theparent magmas. Small increases in Sr isotope ratios with increasingRb contents indicate that the fractionating magmas also assimilatedsmall amounts of wall rocks similar in composition to the upper-crustalbasement to the volcano. Consideration of the chemical trends, mineral compositions,and eruptive history of Ollage rocks permits construction ofa model for the evolution of shallow crustal magma chambersbeneath the stratovolcanoes in the CVZ. At a relatively maturestage, the magma chambers may be compositionally, thermally,and density stratified. Temperatures estimated from Fe-Ti oxideand pyroxene thermometry for the chambers beneath Ollage rangefrom 1000 to 790C with increasing SiO₂ from 59 to 67 wt.% inthe upper reaches, and from 1150 to 1020C with increasing SiO₂from 53 to 59 wt.% in the lower reaches. The occurrence of basalticandesite magmatic inclusions within the intermediate lavas andthe repeated eruption of monotonous composition andesitic magmasindicate that the shallow chambers are periodically replenishedwith parental basaltic andesite magmas. Ubiquitous, reversely zoned plagioclase and pyroxene phenocrystsin the lavas at Ollage suggest that convective cooling of thebasaltic andesite releases buoyant derivative liquid that mixeswith the overlying intermediate-composition body of the chambers.Further crystallization and differentiation of the intermediatemagmas may take place in solidification zones at the boundariesof the magma chambers. If so, the return of residual liquidfrom the crystallizing margins and mixing with the interiorare highly efficient such that magma differentiation can bemodeled as a simple, homogeneous, fractional crystallizationprocess.     

The Petrology and Geochemistry of Oto-Zan Composite Lava Flow on Shodo-Shima Island, SW Japan: Remelting of a Solidified High-Mg Andesite Magma

The Oto-Zan lava in the Setouchi volcanic belt is composed ofphenocryst-poor, sparsely plagioclase-phyric andesites (sanukitoids)and forms a composite lava flow. The phenocryst assemblagesand element abundances change but Sr–Nd–Pb isotopiccompositions are constant throughout the lava flow. The sanukitoidat the base is a high-Mg andesite (HMA) and contains Mg- andNi-rich olivine and Cr-rich chromite, suggesting the emplacementof a mantle-derived hydrous (7 wt % H₂O) HMA magma. However,Oto-Zan sanukitoids contain little H₂O and are phenocryst-poor.The liquid lines of descent obtained for an Oto-Zan HMA at 0·3GPa in the presence of 0·7–2·1 wt % H₂Osuggest that mixing of an HMA magma with a differentiated felsicmelt can reasonably explain the petrographical and chemicalcharacteristics of Oto-Zan sanukitoids. We propose a model wherebya hydrous HMA magma crystallizes extensively within the crust,resulting in the formation of an HMA pluton and causing liberationof H₂O from the magma system. The HMA pluton, in which interstitialrhyolitic melts still remain, is then heated from the base byintrusion of a high-T basalt magma, forming an H₂O-deficientHMA magma at the base of the pluton. During ascent, this secondaryHMA magma entrains the overlying interstitial rhyolitic melt,resulting in variable self-mixing and formation of a zoned magmareservoir, comprising more felsic magmas upwards. More effectiveupwelling of more mafic, and hence less viscous, magmas througha propagated vent finally results in the emplacement of thecomposite lava flow. KEY WORDS: high-Mg andesite; sanukitoid; composite lava; solidification; remelting     

Permeable convection above magma bodies

Thermal convection above large shallow magma bodies in the crust is treated as a one-dimensional bottom-heated convection process in permeable media. Solutions for single-phase convection are briefly reviewed and a solution is developed for two-phase permeable convection in bottom-heated media. Heat flow measurement techniques are discussed for permeable geologic zones above magma bodies and these techniques give consistent results for solidifying lava lakes in Hawaii (Kilauea Iki, q=257 W/m²) and Iceland (Heimaey, q = 465 W/m²).The heat loss from a magma body is a strong function of the permeability when a two-phase convection zone occurs above the magma body, and the heat loss is independent of the thickness of the two-phase convection zone. In steady-state two-phase convection zones, where permeability does not vary appreciably with depth, convective heat flow restrictions tend to limit the maximum saturation temperatures at depth to around 250°C—an effect observed in many geothermal steam fields. A conduction-dominated transition zone tends to occur between the two-phase zone and the magma body and the thickness of this transition zone may easily range from a few meters to several kilometers, depending on the permeability.     

Fountains in Magma Chambers

Cyclic layering is a common feature of the ultramafic zone oflayered intrusions and is usually attributed to the entry ofnew pulses of dense magma into the chamber. Since the crystallizationof olivine and bronzite lowers the density of the magma, a newpulse of the parent magma will be denser than the fractionatedmagma in the chamber. If the new pulse enters with excess momentumit will initially rise up into the host magma to form a fountain,then fall back around the feeder when negative buoyancy forcesovercome the initial momentum of the pulse. Laboratory experimentsusing aqueous solutions with both point and line sources havebeen conducted to obtain a quantitative understanding of thefluid-dynamical processes that are important in fountains. Itis observed that convection within the fountain is highly turbulent,resulting in appreciable entrainment of the host magma. A gravity-stratifiedhybrid layer develops at the floor and this breaks up into aseries of double-diffusive convecting layers if the new pulseis hotter than the host magma. The number of layers that formdepends on a number of factors, especially R, the ratio of thecontributions of composition and heat to the total density differencebetween the host magma and the new pulse. Raising the valueof R, results in the formation of more, thinner layers. The thickness of the hybrid layer at any time t is given byH = h₀+(V₀/A)t where V₀ is the volume flux through the feederand A is the horizontal area of the chamber. h₀ is related tothe initial steady-state height of the fountain and, for a linesource, is given by h₀=CU₀^4/3 d^–1(g/)^–2/3 whereU₀ is the volume flux per unit length, g is the accelerationdue to gravity, d is the width of the feeder, is the densityof the host magma, is the density difference between the magmasand C is a constant. Calculations based on these results and the consideration ofthe flow in the feeder dykes below the chamber indicate thata fountain will rise at least 350 m in a continental magma chamberif the feeder width is greater than 10 m. This will lead toextensive mixing between the new pulse and the fractionatedmagma in the chamber, producing a zoned hybrid layer at thefloor that is commonly over 1000 m thick. If the chamber receivesmany pulses of dense magma, the resulting zoning may persistthroughout much of the life of the chamber, especially if thefirst pulse to enter becomes contaminated by light magma releasedby melting at the margins. The highest Mg/Fe ratio for olivineand pyroxenes from cyclic units from the ultramafic zones oflayered intrusions is often well below the value expected forminerals crystallizing from a melt derived directly from themantle, supporting the hypothesis that new pulses of dense magmacan mix extensively with the fractionated magma in the chamber. The feeder dykes to some oceanic magma chambers, such as theBay of Islands Ophiolite, are believed to be narrower, so thatfountains do not rise more than a few metres above the floorof the chamber. This restricts mixing between the input magmaand the host magma and can result in the formation of a hybridzone that is only a few metres thick.     

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.     

Pyroxenes of the Dufek Intrusion, Antarctica

The Dufek intrusion is a stratiform mafic body, 24,000 to 34,000km² in area and 8 to 9 km thick, in the Pensacola Mountainsof Antarctica. Textures, structures, magmatic stratigraphy,and chemical variation indicate that layered gabbros and relatedrocks of this body developed by accumulation of crystals thatsettled on the floor of a magma chamber. The major cumulus phasesin the exposed part of the intrusion are plagioclase, pyroxene,and iron-titanium oxides. The base of the Dufek intrusion is not exposed, and both Ca-richand Ca-poor pyroxene coexist as cumulus phases in the lowerexposed rocks. The Ca-rich pyroxenes belong to an augite-ferroaugiteseries (Ca_36.4Mg_48.7Fe_14.9-Ca_30.0Mg_23.5Fe_46.5) that extendsup through the 300 m thick capping granophyre. The Ca-poor pyroxenesbelong to a bronzite-inverted pigeonite series (Ca_3.5Mg_69.1Fe_27.4-Ca_11.4Mg_34.0Fe_54.6)that extends only to about 200 m below the granophyre layer.In addition to the cumulus pyroxenes some rocks contain post-cumulusgreen calcic augite and ferrohypersthene. The compositional change of the cumulus pyroxenes with stratigraphicheight is one of general iron enrichment. Superimposed on thistrend are (1) a 1 km thick section in the lower part of thebody that shows slight to no iron enrichment and (2) a markedreversal in the Fe/(Fe+Mg) ratio about 1 km below the top ofthe body. The variations from the general trend are associatedwith cyclic units and are best explained by convective overturnof the magma. In general, the pyroxene compositional trends are similar tothose of the Skaergaard and Bushveld intrusions. One significantdifference in the Dufek intrusion is the limited iron enrichmentof its Ca-rich pyroxenes, that may relate to a slower decreaseof P_O2 during crystallization of the Dufek magma.     

Petrology of Rhyolitic and Mixed Magma Ejecta from the 1875 Eruption of Askja, Iceland

Volcanic activity in Askja central volcano and its fissure swarmin 1875 occurred in response to a crustal rifting episode inIceland, resulting in up to 70 km lateral flow of magma withinthe crust, caldera collapse and a plinian eruption of acid magma(0·2 km³ dense-rock equivalent). Petrologic studies ofthe predominantly rhyolitic and crystal-poor ejecta reveal thata complex array of other liquid compositions was also present,including icelandite (0.75 per cent) and basalt (1.9 per cent),as well as leucocratic xenoliths of trondhjemite type. Mineralgeothermometers indicate that the rhyolite evolved at 990 to1010 °C and 0·5 Kb P_H2O, the icelandite at 1005 to1020 °C and at f_O2 10^–10 atm. and the basalt at 1140to 1170 °C. A petrologic model of Askja in 1875 consists of a density-stratifiedmagma chamber with a rhyolitic upper part and a lower part offerrobasalt, with an intervening layer of icelandite. The modelcalculations show that the icelandite can be derived from ferrobasaltby 50 per cent fractional crystallization, but one-stage fractionalcrystallization models cannot account for generation of theacid magma. Simple partial or complete fusion of the field-associatedtrondhjemite xenoliths cannot produce the acid magma. Instead,a more complex fusion, hybridization and fractional crystallizationmodel is presented, which is consistent with the available petrologicevidence. This model involves large-scale fusion of pre-existingtrondhjemite intrusions or reactivation of previously consolidatedroof-rock in the magma chamber followed by hybridization ofthe acid magma with 7 to 14 per cent basaltic magma. Finally,10 to 11 per cent fractional crystallization of the dacite hybridis required to produce the observed compositional range withinthe rhyolite ejecta. The 1875 explosive eruption was causedby the ascent of tholeiitic basalt magma from depth during crustalrifting. Influx of new basalt magma in 1874–75 triggeredconvective mixing and hybridization in the compositionally zonedmagma chamber.     

Pleistocene Underplating and Metasomatism of the Lower Continental Crust: a Xenolith Study

Xenoliths from Engeln–Kempenich in the East Eifel volcanicfield (Germany) comprise gabbroic to ultramafic cumulates, andmeta-igneous and meta-sedimentary granulite- to amphibolite-facieslithologies. They provide evidence for Pleistocene heating andmetasomatism of the lower continental crust by mafic magmas.The metamorphic xenoliths were divided into three types: (1)primitive type P, which are little affected by metasomatic replacementstructures; (2) enriched type E1 defined by metasomatic replacementof primary pyroxene and garnet by pargasitic amphibole and biotite;(3) enriched type E2 defined by breakdown of hydrous phases.Type E rocks are geochemically related to type P and cumulatexenoliths by compositional trends. During modal metasomatism,type E rocks were oxidized. Type E1 rocks were typically enrichedin Rb, Th, U, Nb, K, light rare earth elements (LREE) and Zr,and E2 enriched in Rb, Th, U, Nb, K, REE, Zr, Ti and Y, relativeto type P rocks. Formation of the hydrous, chlorine-bearingphases amphibole and scapolite containing glass and fluid inclusionsin the E1 rocks provides evidence for a water and Cl-bearingfluid phase coexisting with silicate melt. Accordingly, we calculated10 mol % H₂O back into the CO₂-dominated fluid inclusions, inagreement with experimental data on the composition of a fluidphase coexisting with mafic alkaline melts at elevated pressure.Primary CO₂-dominated fluid inclusions coexisting with glassinclusions in metamorphic corona phases and neoblasts, and incumulate xenoliths, have overlapping densities. Fluid inclusionbarometry using the corrected densities indicates that bothcumulates and metamorphic xenoliths originated from the samedepth at 22–25 km (650 ± 50 MPa). This is interpretedas being a main magma reservoir level within the upper partof the lower crust close to the Conrad discontinuity, wherethe xenoliths represent wall-rocks. The Conrad discontinuityseparates an upper-crustal layer, consisting of preferentiallyductile granodioritic and tonalitic gneisses, and more brittlelower-crustal mafic granulites. The brittle–ductile transitionappears to be a preferred level of magma stagnation. KEY WORDS: continental lower crust; fluids; magma chamber; metasomatism; xenoliths     

Remobilization of Andesite Magma by Intrusion of Mafic Magma at the Soufriere Hills Volcano, Montserrat, West Indies

The 1995–1999 eruption of the Soufriere Hills volcano,Montserrat, has produced a crystal-rich andesite containingquench-textured mafic inclusions, which show evidence of havingbeen molten when incorporated into the host magma. Individualcrystals in the andesite record diverse histories. Amphibolephenocrysts vary from pristine and unaltered to strongly oxidizedand pseudomorphed by anhydrous reaction products. Plagioclasephenocrysts are commonly reverse zoned, often with dusty sievetextures. Reverse zoned rims are also common on orthopyroxenephenocrysts. Pyroxene geothermometry gives an average temperatureof 858 ± 20°C for orthopyroxene phenocryst cores,whereas reverse zoned rims record temperatures from about 880to 1050°C. The heterogeneity in mineral rim compositions,zoning patterns and textures is interpreted as reflecting non-uniformreheating and remobilization of the resident magma body by intrusionof hotter mafic magma. Convective remobilization results inmixing together of phenocrysts that have experienced differentthermal histories, depending on proximity to the intruding maficmagma. The low temperature and high crystallinity are interpretedas reflecting the presence of a cool, highly crystalline magmabody beneath the Soufriere Hills volcano. The petrological observations,in combination with data on seismicity, extrusion rate and SO₂fluxes, indicate that the current eruption was triggered byrecent influx of hot mafic magma. KEY WORDS: Montserrat; eruption; magma mixing; mafic inclusion; sieve texture     

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.     

Open-System Magma Chamber Evolution: an Energy-constrained Geochemical Model Incorporating the Effects of Concurrent Eruption, Recharge, Variable Assimilation and Fractional Crystallization (EC-E'RA{chi}FC)

Significant petrogenetic processes governing the geochemicalevolution of magma bodies include magma Recharge (includingformation of ‘quenched inclusions’ or enclaves),heating and concomitant partial melting of country rock withpossible ‘contamination’ of the evolving magma body(Assimilation), and formation and separation of cumulates byFractional Crystallization (RAFC). Although the importance ofmodeling such open-system magma chambers subject to energy conservationhas been demonstrated, the effects of concurrent removal ofmagma by eruption and/or variable assimilation (involving imperfectextraction of anatectic melt from wall rock) have not been considered.In this study, we extend the EC-RAFC model to include the effectsof Eruption and variable amounts of assimilation, A. This model,called EC-E'RAFC, tracks the compositions (trace elements andisotopes), temperatures, and masses of magma body liquid (melt),eruptive magma, cumulates and enclaves within a composite magmaticsystem undergoing simultaneous eruption, recharge, assimilationand fractional crystallization. The model is formulated as aset of 4 + t + i + s coupled nonlinear differential equations,where the number of trace elements, radiogenic and stable isotoperatios modeled are t, i and s, respectively. Solution of theEC-E'RAFC equations provides values for the average temperatureof wall rock (T_a), mass of melt within the magma body (M_m),masses of cumulates (M_ct), enclaves (M_en) and wall rock () and the masses of anatectic melt generated () and assimilated (). In addition, t trace element concentrations and i + s isotopic ratios inmelt and eruptive magma (C_m, _m, _m), cumulates (C_ct, _m, _m), enclaves(C_en, , ) and anatectic melt (C_a, , ) as a function of magma temperature (T_m) are also computed. Input parametersinclude the (user-defined) equilibration temperature (T_eq),a factor describing the efficiency of addition of anatecticmelt () from country rock to host magma, the initial temperatureand composition of pristine host melt (, , , ), recharge melt (, , , ) and wall rock (, , , ), distribution coefficients (D_m, D_r, D_a) and their temperaturedependences (H_m, H_r, H_a), latent heats of transition (meltingor crystallization) for wall rock (h_a), pristine magma (h_m)and recharge magma (h_r) as well as the isobaric specific heatcapacity of assimilant (C_p,a), pristine (C_p,m) and recharge(C_p,r) melts. The magma recharge mass and eruptive magma massfunctions, M_r(T_m) and M_e(T_m), respectively, are specified apriori. M_r(T_m) and M_e(T_m) are modeled as either continuous orepisodic (step-like) processes. Melt productivity functions,which prescribe the relationship between melt mass fractionand temperature, are defined for end-member bulk compositionscharacterizing the local geologic site. EC-E'RAFC has potentialfor addressing fundamental questions in igneous petrology suchas: What are intrusive to extrusive ratios (I/E) for particularmagmatic systems, and how does this factor relate to rates ofcrustal growth? How does I/E vary temporally at single, long-livedmagmatic centers? What system characteristics are most profoundlyinfluenced by eruption? What is the quantitative relationshipbetween recharge and assimilation? In cases where the extractionefficiency can be shown to be less than unity, what geologiccriteria are important and can these criteria be linked to fieldobservations? A critical aspect of the energy-constrained approachis that it requires integration of field, geochronological,petrologic, and geochemical data, and, thus, the EC-ERAFC ‘systems’approach provides a means for answering broad questions whileunifying observations from a number of disciplines relevantto the study of igneous rocks. KEY WORDS: assimilation; energy conservation; eruption; open system; recharge     

Liquid Lines of Descent of Island Arc Magmas and Genesis of Rhyolites: Evidence from the Shirahama Group, Japan

The Mio-Pliocene Shirahama Group, Izu Peninsula, Central Japan,a well-exposed submarine volcanic arc complex of lava flows,pyroclastic rocks and associated shallow intrusives, is characterizedby a tholeiitic series (basalt to dacite) and a calc-alkalineseries (andesite to dacite). Chemical variations in the tholeiiticseries and calc-alkaline series are consistent with crystalfractionation from basalt and magnesian andesite (boninite),respectively. Crystal–liquid phase relations of thesemagmas have been investigated by study of sample suites fromthese two series. Compositions of liquids in equilibrium withphenocrysts were determined by microprobe grid analyses, inwhich 49 points were averaged in 03 mm 03 mm groundmassareas. The liquid compositions, coupled with the phenocrystmineralogy of the same samples, define the liquid lines of descentof these volcanic arc magmas. Major findings include the following:(1) Crystallization of the tholeiitic series magma is consistentwith early stage crystallization in the simple system Fo–Di–Silica–H₂O,with olivine having a reaction relation to augite and the tholeiiticliquid. (2) The later stage products of the tholeiitic seriesmagma are, however, crystal-poor (<10%) dacites with no maficminerals, suggesting that tholeiitic liquids, hypersthene andaugite were no longer on the cotectic (3) A characteristic ofthe calc-alkaline series magmas is the development of rhyoliticliquids. Hypersthene, augite, plagioclase and Fe–Ti oxideoccur in most calc-alkaline rocks studied, and hornblende andquartz can be found in about half of these. However, their differentiationpaths show that the cotectic relation between quartz and liquidended at a later stage, resulting in the resorption of quartzphenocrysts and ultimately in the formation of quartz-free magmas.(4) The late-stage liquids of both the tholeiitic and calc-alkalineseries have deviated from their cotectics, which cannot be explainedby fractional crystallization alone. The addition of H₂O froman outside system is probably required to explain the differentiationpaths. (5) The formation of chilled margins, the in situ crystallizationof a magma chamber in the solidification zone, and/or the migrationof groundwater into the magma chamber are thought to be likelyprocesses affecting magmas during their migration and intrusioninto the crust. An extreme effect of H₂O addition would be tolower the liquidus temperatures of all precipitating silicatephases far below their restorable range before eruption, resultingin the production of aphyric magmas. Even when a temperaturedecrease in the magma chamber causes a liquid to intersect theliquidus of a pre-existing phase, the addition of H₂O shiftsthe cotectic toward SiO₂, resulting in quartz being the lastphase to crystallize. The resorption of quartz is interpretedto be the result of a liquidus boundary shift caused by theaddition of H₂O. The genesis of aphyric rhyolites is thereforeinferred to result from fractional crystallization followingaddition of H₂0. KEY WORDS: Shirahama Group; Japan; island arc; rhyolite; magma series     

The Volume and Composition of Melt Generated by Extension of the Lithosphere

Calculation of the volume and composition of magma generatedby lithospheric extension requires an accurate mitial geotherm,and knowledge of the variation and composition of the melt fractionas a function of pressure and temperature. The relevant geophysicalobservations are outlined, and geotherms then obtained fromparameterized convective models. Experimental observations whichconstrain the solidus and liquidus at various pressures aredescribed by simple empirical functions. The variation in meltfraction is then parameterized by requiring a variation from0 on the solidus to 1 on the liquidus. The composition of the melts is principally controlled by themelt fraction, though those of FeO, MgO, and SiO² in additionvary with pressure. Another simple parameterization allows theobserved compositions of major elements in 91 experiments tobe calculated with a mean error of 1.1%, and those of TiO₂ andNa₂O to 0.3%. These expressions are then used to calculate theexpected compositions of magma produced by adiabatic upwelling.The mean major element composition of the most magnesium-richMORB glasses resemble the mean composition calculated for amantle potential temperature T_p of 1280?C. Adiabatic meltingduring upwelling of mantle of this temperature generates a meltthickness of 7 km. The observed variations of the MgO and TiO₂concentrations in a large collection of MORB glass compositionssuggest that extensive low pressure fractional crystallizationoccurs, but that its effect on the concentrations of SiO₂, Al₂O₃,and CaO is small. There is no evidence that normal oceanic crustis produced from magmas containing more than 11% MgO. The mantlepotential temperature within hot rising jets is about 1480?Cand can generate 27 km of magma containing 17% MgO. Extension of the continental lithosphere generates little meltunless ß> 2 and T_p> 1380?C. The melts generatedby larger values of ß or of T_p are alkali basalts,and change to tholeiites as the amount of melting increases.Large quantities of melt can be generated, especially at continentalmargins, where estimates of ß obtained from changesin crustal thickness will in general be too small.     

Magmatic Evolution of the La Pacana Caldera System, Central Andes, Chile: Compositional Variation of Two Cogenetic, Large-Volume Felsic Ignimbrites

La Pacana is one of the largest known calderas on Earth, andis the source of at least two major ignimbrite eruptions witha combined volume of some 2700 km³. These ignimbrites have stronglycontrasting compositions, raising the question of whether theyare genetically related. The Toconao ignimbrite is crystal poor,and contains rhyolitic (76–77 wt % SiO₂) tube pumices.The overlying Atana ignimbrite is a homogeneous tuff whose pumiceis dacitic (66–70 wt % SiO₂), dense (40–60% vesicularity)and crystal rich (30–40 % crystals). Phase equilibriaindicate that the Atana magma equilibrated at temperatures of770–790°C with melt water contents of 3·1–4·4wt %. The pre-eruptive Toconao magma was cooler (730–750°C)and its melt more water rich (6·3–6·8 wt% H₂O). A pressure of 200 MPa is inferred from mineral barometryfor the Atana magma chamber. Isotope compositions are variablebut overlapping for both units (⁸⁷Sr/⁸⁶Sr_i 0·7094–0·7131;¹⁴³Nd/¹⁴⁴Nd 0·51222–0·51230) and are consistentwith a dominantly crustal origin. Glass analyses from Atanapumices are similar in composition to those in Toconao tubepumices, demonstrating that the Toconao magma could representa differentiated melt of the Atana magma. Fractional crystallizationmodelling suggests that the Toconao magma can be produced by30% crystallization of the observed Atana mineral phases. Toconaomelt characteristics and intensive parameters are consistentwith a volatile oversaturation-driven eruption. However, thelow H₂O content, high viscosity and high crystal content ofthe Atana magma imply an external eruption trigger. KEY WORDS: Central Andes; crystal-rich dacite; eruption trigger; high-silica rhyolite; zoned magma chamber     

Experimental Petrology of the 1991-1995 Unzen Dacite, Japan. Part II: Cl/OH Partitioning between Hornblende and Melt and its Implications for the Origin of Oscillatory Zoning of Hornblende Phenocrysts

High-temperature–pressure experiments were carried outto determine the chlorine–hydroxyl exchange partitioncoefficient between hornblende and melt in the 1992 Unzen dacite.Cl in hornblende and melt was analyzed by electron microprobe,whereas OH in hornblende and melt was calculated assuming anionstoichiometry of hornblende and utilizing the dissociation reactionconstant for H₂O + O = 2(OH) in water-saturated melt, respectively.The partition coefficient strongly depends on the Mg/(Mg + Fe)ratio of hornblende, and is expressed as ln K₁ = (Cl/OH)_hb/(Cl/OH)_melt= 2·37 – 4·6[Mg/(Mg + Fe)]_hb at 2–3kbar and 800–850°C. The twofold variation in Cl contentin the oscillatory zoned cores of hornblende phenocrysts inthe 1991–1995 dacite cannot be explained by the dependenceof the Cl/OH partition coefficient on the Mg/(Mg + Fe)_hb ratio,and requires c. 80% variation of the Cl/OH ratio of the coexistingmelt. Available experimental data at 200 MPa on Cl/OH fractionationbetween fluid and melt suggest that c. 1·2–1·8wt % degassing of water from the magma can explain the required80% variation in the Cl/OH ratio of the melt. The negative correlationbetween Al content and Mg/(Mg + Fe) ratio in the oscillatoryzoned cores of the hornblende phenocrysts is consistent withrepeated influx and convective degassing of the fluid phasein the magma chamber. KEY WORDS: chlorine; element partitioning; hornblende; oscillatory zoning; Unzen volcano     

The Influence of Viscosity on Fountains in Magma Chambers

Geological observations suggest that basaltic magmas mix readilybut that rhyolites and basalts can erupt through the same volcanicvent without mixing. The implication is that viscosity may havean important influence on the mixing of magmas. This paper examines the influence of viscosity on magma mixingwhen a magma of low viscosity is injected as a turbulent fountaininto a chamber containing a magma of higher viscosity. Threeseries of experiments were carried out. In the first, the lowviscosity fluid was injected into the tank at a Reynolds number(Re₁) of about 1000 to ensure that flow within the fountainwas fully turbulent. Both fluids were maintained at the sametemperature and the viscosity of the host fluid was varied systematically.If the viscosities of the two fluids are similar, turbulencein the fountain leads to extensive mixing and to formation ofa density-stratified layer which collects at the bottom of thetank. However, if the viscosity ratio of the fluids used isgreater than 400, no detectable mixing occurs. Motion withinthe fountain is again fully turbulent but the high momentumof the input fluid is not transmitted to the more viscous hostfluid. Experiments carried out at intermediate viscosity ratiosresulted in intermediate amounts of mixing. Theoretical considerations suggest that the criterion for mixingin turbulent fountains, when a fluid of low viscosity v₁ isinjected into a fluid of much higher viscosity v₂ is wd>kv₂ where w is the flow velocity through an input pipe of diameterd and k is a constant. This relationship is believed to be ageneral one, applying equally to jets and plumes, although thevalues of k may vary with the nature of the flow. The secondseries of experiments was designed to test this relationshipand determine values of k for fountains. The results show that,if wd/v₂ > 70, the inflowing fluid mixes with the host fluidas there was no viscosity difference between them. However,if wd/v₂ < 7, little or no mixing occurs even if motion withinthe fountain is fully turbulent. In the third series of experiments the temperature of the inputfluid was 70?C above that of the host fluid. This led to heatingofa thin boundary layer of the host fluid, lowering the viscosityof the host fluid adjacent to the fountain and thus allowingadditional mixing between the fluids. This effect was most obviousin the experiment with the highest viscosity ratio (1229) wherea small amount of mixing was detected, compared with no mixingin the equivalent experiment in the first series. The criterion for mixing, wd> kv₂ can be applied to naturalmagmas. The results show that if a primitive basaltic magmais injected, as a turbulent fountain or plume, into a chambercontaining fractionated basaltic magma, the two magmas willmix readily. However, if basaltic magma is injected into a chambercontaining granitic melt, little or no mixing occurs.     

Quantification of Crustal Contamination in Open Magmatic Systems

The DePaolo (1981a, Earth Planet. Sci. Lett. 53, 189–202)equations for assimilation with fractional crystallization (AFC)are extended and solved to yield directly the fraction of crustassimilated during AFC with or without magma replenishment orrecharge. Errors in the computed crust/magma ratio are alsopresented. Using graphical methods and a sufficient number ofelements no assumptions need be made about the value of theratios r (= rate of assimilation/rate of fractional crystallization)and ß (= rate of recharge/rate of assimilation); infact, the method yields independent estimates of these ratiosin addition to the crust/magma ratio. Thus the average relativesizes of the fluxes of wall-rock melt, replenishing magma, andfractionating phases are known and can be incorporated in abudget for the whole magma-crust system. The method for calculatingthe crust/magma ratio is extended to encompass variable valuesof r, ß, and bulk distribution coefficient (D). Simple(bulk) mixing is effectively a special case of AFC where ßor r and/or D = 0. Only for highly incompatible elements (D<01)do bulk mixing calculations approximate the true crust/magmaratio if AFC processes operated. For more compatible elements(e.g., oxygen) bulk mixing calculations considerably overestimatethe crust/magma ratio if r and ß are constant. Thecomplications of compositional stratification and diffusionin magma chambers are considered qualitatively. Two examples illustrate the application of the methods developedhere. For the upper zone of the Lille Kufjord instrusion innorthern Norway, an example of mid-crustal AFC without magmarecharge, either r = 0.04 and the crust/magma ratio is 0.04,or r = 0.08 and the crust/magma ratio is 0.06. For lavas ofthe Andean Central Volcanic Zone, an example of deep AFC withpossible recharge, r = 0.15, ß = 3.2, and the crust/magmaratio is 0.16.     

Experimentally Determined Conditions in the Fish Canyon Tuff, Colorado, Magma Chamber

The Fish Canyon Tuff, Colorado, forms one of the largest (3000km³ known silicic eruptions in Earth history. The tuff is ahomogeneous quartz latite consisting of 40% phenocrysts (plagioclase,sanidine, biotite, hornblende, quartz, magnetite, apatite, sphene,and ilmenite) in equilibrium with a highly evolved rhyoliticmelt now represented by the matrix glass. Melt inclusions trappedin hornblende and quartz phenocrysts are identical to the newlyanalyzed matrix glass composition indicating that hornblendeand quartz crystallized from a highly evolved magma that subsequentlyexperienced little change. This study presents experimentalphase equilibrium data which are used to deduce the conditions(P, T, f_O2, f_H2O, etc.) in the Fish Canyon magma chamber priorto eruption. These new data indicate that sanidine and quartzare not liquidus phases until 780?C temperatures are achieved,consistent with Fe-Ti oxide geothermometry which implies thatthe magmatic temperature prior to eruption was 760?30?C. NaturalFe-Ti oxide pairs also suggest that log f_O2 was -12.4 (intermediatebetween the Ni-NiO and MnO-Mn₃O₄ oxygen buffers) in the magmachamber. This f_O2.102 is supported by the experimentally determinedvariations in hornblende and melt Mg-numbers as functions off_O2 A new geobarometer based on the aluminum content of hornblendesin equilibrium with the magmatic assemblage hornblende, biotite,plagioclase, quartz, sanidine, sphene, ilmenite or magnetite,and melt is calibrated experimentally, and yields pressuresaccurate to ?0.5 kb. Total pressure in the Fish Canyon magmachamber is inferred to have been 2.4 kb (equivalent to a depthof 7.9 km) based on the Al-content of natural Fish Canyon hornblendesand this new calibration. This depth is much shallower thanhas been proposed previously for the Fish Canyon Tuff. Variationsin experimental glass (melt) composition indicate that the magmawas water-undersaturated prior to eruption. X_H2O in the fluidphase that may have coexisted with the Fish Canyon magma isestimated to have been 0.5 by comparing the An-content of naturalplagioclases to experimental plagioclases synthesized at differentX_H2O and P_totals. This ratio corresponds to about 5 wt.% waterin the melt at depth. The matrix glass chemistry is reproducedexperimentally under these conditions: 760?C, 2.4 kb, X_H2O=0.5,and log f_o2=NNO+2 log units. The fugacity of SO₂ (91 b) is calculatedfrom the coexistence of pyrrhotite and magnetite. Maximum CO₂fugacity (2520 b) is inferred assuming the magma was volatilesaturated at 2.4 kb.