A laboratory model of a replenished magma chamber

It has recently been suggested that periodic influxes of hot but heavy magma into the base of a basaltic magma chamber can remain isolated from the rest of the chamber while the new magma cools and crystallization proceeds. When thermal equilibrium is almost complete, the suspended crystals settle out and the residual, less dense liquid can then mix with the fluid above. In the present paper the basic fluid-dynamical processes underlying this model have been investigated in laboratory experiments using aqueous solutions. The lower layer was hot KNO₃ solution, for which saturated solutions become less dense as the temperature decreases. With a cold, deeper layer of less dense NaNO₃ or K₂CO₃ above the lower layer, there was strong convective transfer of heat through a sharp interface separating the layers, at a rate which is predicted here drawing on previous studies carried out with oceanographic applications in mind. Once crystallization began, non-equilibrium effects became important and the observed temperatures differ somewhat from those predicted. In the experiments crystals grew mainly from the bottom rather than while in suspension, but this is not an essential aspect of the model. The important fact is that the density of the residual liquid in the lower layer decreased until it became equal to that of the upper layer, and then the interface broke down so that the two layers mixed thoroughly together, leaving a layer of KNO₃ crystals at the base. No crystallization at all occurred when the hot input liquid was forced to mix initially with the cold solution already in the chamber.     

Silicic magma entering a basaltic magma chamber: eruptive dynamics and magma mixing — an example from Salina (Aeolian islands,Southern Tyrrhenian Sea)

The Pollara tuff-ring resulted from two explosive eruptions whose deposits are separated by a paleosol 13 Ka old. The oldest deposits (LPP, about 0.2 km³) consist of three main fall units (A, B, C) deposited from a subplinian column whose height (7–14 km) increased with time from A to C, as a consequence of the increased magma discharge rate during the eruption (1–8x10⁶ kg/s). A highly variable juvenile population characterizes the eruption. Black, dense, highly porphyritic, mafic ejecta (SiO₂=50–55%) almost exclusively form A deposits, whereas grey, mildly vesiculated, mildly porphyritic pumice (SiO₂=56–67%) and white, highly vesiculated, nearly aphyric pumice (SiO₂=66–71%) predominate in B and C respectively. Mafic cumulates are abundant in A, while crystalline lithic ejecta first appear in B and increase upward. The LPP result from the emptying of an unusual and unstable, compositionally zoned, shallow magma chamber in which high density mafic melts capped low density salic ones. Evidence of the existence of a short crystal fractionation series is found in the mafic rocks; the andesitic pumice results from complete blending between rhyolitic and variously fractionated mafic melts (salic component up to 60 wt%), whereas bulk dacitic compositions mainly result from the presence of mafic xenocrysts within rhyolitic glasses. Viscosity and composition-mixing diagrams show that blended liquids formed when the visosities of the two end members had close values. The following model is suggested: 1. A rhyolitic magma rising through the metamorphic basement enterrd a mafic magma chamber whose souter portions were occupied by a highly viscous, mafic crystal mush. 2. Under the pressure of the rhyolitic body the nearly rigid mush was pushed upwards and mafic melts were squeezed against the walls of the chamber, beginning roof fracturing and mingling with silicic melts. 3. When the equilibrium temperature was reached between mafic and silicic melts, blended liquids rapidly formed. 4. When fractures reached the surface, the eruption began by the ejection of the mafic melts and crystal mush (A), followed by the emission of variously mingled and blended magmas (B) and ended by the ejection of nearly unmixed rhyolitic magma (C).     

Bubble growth in visco-elastic magma: implications to magma fragmentation and bubble nucleation

We present a visco-elastic bubble growth model, accounting for viscous and elastic deformations and for volatile mass transfer between bubbles and melt. We define the borders between previous bubble growth models accounting for incompressible viscous melt, and our new model accounting also for elastic deformation; this is done by a set of end-member analytical solutions and numerical simulations. Elastic deformation is most prominent for magma of small vesicularity, where the growth regime depends on the shear modulus. For high shear modulus, bubble growth is slow and follows an exponential law in a viscous growth regime, while for low shear modulus bubbles quickly follow a square-root diffusive solution. Our model provides all the elastic components (stresses, strains and strain rates) required for defining criteria for failure and magma fragmentation. We suggest two failure criteria, a stress related one based on the internal friction and the Mohr-Coulomb failure theory, and a strain related one based on fibre elongation experiments. We argue that both criteria are equivalent if we consider their shear modulus dependency and its effect on magma rheology. Last, we apply our model to the process of bubble nucleation. In the incompressible case, following nucleation, growth is slow and leads to long incubation times during which bubbles may be dissolved back into the melt. The elastic response in magmas with low shear modulus results in a short incubation time, increasing the probability of survival. The above effects emphasize the significance of visco-elasticity for the dynamic processes occurring in magmas during volcanic activity.     

Replenishment of a mafic magma in a zoned felsic magma chamber beneath Rishiri Volcano, Japan

The trachytic Tanetomi lava from Rishiri Volcano, northern Japan, provides useful information concerning how a replenished mafic magma mixes with a compositionally zoned felsic magma in a magma chamber. The Tanetomi lava was erupted in the order of Lower lava 1 (LL1, 59.2-59.8 wt.% in SiO₂), Lower lava 2 (LL2, 58.4-59.1 wt.%), and Upper lava (UL, 59.9-65.1 wt.%). Evidence for mixing with a mafic magma is observed only in the LL2, in which a greater amount of crystals derived from the mafic magma occurs in rocks with higher SiO₂ content. The whole-rock compositional trend of the Tanetomi lavas is fairly smooth except for the LL2 lava composition, which scatter along the main composition trend. There is no reasonable composition of basaltic magma on the extrapolation of the LL2 composition trend, and the trend cannot be explained by a simple two-component magma mixing. Before the replenishment, the felsic magma was zoned in composition (58-65 wt.% in SiO₂) and temperature (1030-920°C) in the magma chamber located at the pressure of ~2 kbar. The compositional variation of the main felsic magma was produced by extraction of a fractionated interstitial melt from mush zones along the chamber walls and its subsequent mixing with the main magma (boundary layer fractionation). The LL1 magma tapped the magma chamber soon after the replenishment, before the mafic magma mixed with the overall felsic magma. Then the basalt magma mixed heterogeneously with the upper part of the felsic magma by forced convection as a fountain during injection. The mixing of the basalt magma with compositionally zoned felsic magma resulted in the characteristic composition trend of the LL2. The fraction of basaltic magma in the LL2 magma is estimated to be at most 10%. Despite such a small proportion, the basalt magma was mixed completely with the felsic magma, probably because the crystallinity of undercooled basalt magma was low enough to behave as a liquid.     

Evolution of double diffusion convection in a felsic magma chamber

The onset of double diffusion convection (DDC) is modeled in a two-dimensional case in respect to magma chambers. The viscosity model for the melt takes into account the effects of temperature and concentration of the dissolved component (H₂O). The upper boundary of the convecting magma chamber is assumed to be anhydrous and at constant temperature, whereas the lower boundary is treated as being hydrous permeable with a temperature greater than that within the upper boundary. The case of positive compositional and thermal buoyancy of melt is studied assuming a H₂O diffusion coefficient small in comparison with thermal diffusivity. The DDC has been modeled using a system of equations solved by the finite difference method on a square grid. The convective pattern evolution has been studied for fixed boundary conditions as well as for cooling and degassing. Due to the higher viscosity in the upper zone, the upper boundary layer is thicker than the lower one. The variation of water concentration in this zone of the convective cell can be significant. In nature, the high gradient of water concentration can be responsible for the observed variations of water content in minerals crystallized from a granite melt (e.g., biotite). Because of a high Lewis number (= 100), temperature variations in the magma chamber decay much faster than the water concentration. In this case the intensive convection can continue at a constant temperature due to the non-zero water content in the chamber. In principle, the effect can be applied to the formation of magmatic bodies. If the cooling and degassing system reaches a uniform temperature distribution prior to the crystallization temperature, water content throughout the body may still remain variable.     

Heat diffusion and size reduction of a spherical magma chamber

An analytical solution is derived for the size reduction of a spherical magma chamber cooling by conduction. The use of moving boundary conditions and the constraint of a spherical symmetry allow one to ignore the details of the heat redistribution processes which take place within the magma chamber. The dependence of the solution on the initial conditions is investigated. A simple solution is found for short time, which is shown to be valid for times long enough to make it useful in the volcanological context. Moreover, the general solution confirms that the hydrothermal contribution to heat transfer in Phlegraean Fields cannot be extremely important.     

Incipient magma chamber formation as a result of repetitive intrusions

An analytical solution for periodic magma intrusions in conduits was developed to study the onset of shallow magma chamber formation. The solution is based on determining when a repetitive series of intrusions can cause the wall rock of a conduit to reach its melt temperature. The results show that magma chamber formation in conduits is a strong function of the volume rate of intrusion and that magma chamber formation is likely when the intrusion rate exceeds 10^?3 km³/ yr. which agrees with observations by other investigators. Once this critical value of intrusion rate is reached, magma chambers are likely to begin forming after only a few intrusive pulses (less than ten). Results for both cylindrical conduits and dikes show cylindrical conduits are more favourable for the formation of shallow magma chambers.     

The messina earthquake of 1908 and the magma chamber of Etna

Comparison between observed and theoretical seismic intensities of the Messina earthquake of 1908 shows an area of negative anomalies, which seems to correspond to the «shadow» of the Etna magma chamber. This magma chamber has probably an elongated form, being 40 km wide and about 100 km long; the mean depth appears to be close to 5 km, but the thickness cannot be estimated from the present data.     

采用有限元等效体力方法数值分析火山区岩浆压力变形源—以长白山火山为例

火山区岩浆压力变形源的反演计算采用解析方法存在难以考虑地形的限制,采用传统有限元方法则存在网格依赖和计算量大的问题,反演过程中每一次正演由于岩浆房位置和大小变化都需要重新生成一次网格,耗费巨大的计算量和网格生成时间.为了克服上述问题,首次在长白山火山区使用"有限元等效体力"方法考虑地形影响反演地下岩浆压力变形源,计算岩浆应力扰动对周边断层稳定性的影响.在火山区地下压力变形源引起的地表形变计算中,地表地形影响不可忽略.埋深越浅,地表最大径向位移ur所在的位置越靠近岩浆囊中心.当坡度达到30°时,最大垂向位移uz所在位置不再位于岩浆囊正上方.椭球状岩浆囊压力源可以较好地模拟长白山火山地区2002—2003年间的GPS和水准测量.岩浆房扰动应力场和区域构造应力场的叠加有可能造成天池西部近EW向,天池北部以NW-NNW向为主的现今应力方向.岩浆房压力源引起的库仑应力变化有利于天池火山口NW向震群在空间上主要分布于火山口的西南和东北部.     

Rhythmic layering in the magma chamber of basic-ultrabasic intrusions due to diffusion and intermittent convection

The following facts have been established in this study. (1) Fine rhythmic layering can form due to diffusion. For example, the behavior of the elements Fe, Mg, Ca, and Al with a high diffusion coefficient (D ₁ ? 1.1 × 10^?5 cm²/s) can lead to the formation of fine rhythmic layering in rocks, whereas elements with low diffusion coefficients (D ₁ ? 2.7 × 10^?10 cm²/s) do not produce rhythmic layering and the distribution of their concentrations is monotonic. Equations derived to describe layering are used to calculate the distribution of magnesium (MgO) concentration in the Kivakka layered intrusion. Comparison of the theoretical distribution curve with in situ data showed their fairly good agreement, confirming the possible role of diffusion in the formation of fine rhythmic layering. (2) Conditions favorable for intermittent convection can occur in a magma chamber. The equation derived to describe the intermittent convection has a periodic solution. A distinctive feature of this type of convection is that zones of active mixing alternate with zones of relative stagnation. Conditions suitable for diffusion can arise in the latter, leading to the formation of fine (with layers up to 5 cm thick) rhythmic layering. An example of possible development of intermittent convection is the layered Akanvaara massif, where zones with a relatively smooth distribution of chemical elements alternate with zones of contrasting layering of rocks.     

采用有限元等效体力方法数值分析火山区岩浆压力变形源——以长白山火山为例

火山区岩浆压力变形源的反演计算采用解析方法存在难以考虑地形的限制,采用传统有限元方法则存在网格依赖和计算量大的问题,反演过程中每一次正演由于岩浆房位置和大小变化都需要重新生成一次网格,耗费巨大的计算量和网格生成时间.为了克服上述问题,首次在长白山火山区使用"有限元等效体力"方法考虑地形影响反演地下岩浆压力变形源,计算岩浆应力扰动对周边断层稳定性的影响.在火山区地下压力变形源引起的地表形变计算中,地表地形影响不可忽略.埋深越浅,地表最大径向位移u_r所在的位置越靠近岩浆囊中心.当坡度达到30°时,最大垂向位移u_z所在位置不再位于岩浆囊正上方.椭球状岩浆囊压力源可以较好地模拟长白山火山地区2002—2003年间的GPS和水准测量.岩浆房扰动应力场和区域构造应力场的叠加有可能造成天池西部近EW向,天池北部以NW-NNW向为主的现今应力方向.岩浆房压力源引起的库仑应力变化有利于天池火山口NW向震群在空间上主要分布于火山口的西南和东北部.     

Particle dynamics and aggregation of crystals in a magma chamber with application to Kilauea Iki olivines

Crystals settling in relative proximity to each other will settle faster than their respective Stokes' velocities (U_S) due to their mutual interactions. Pairs of equant spheres (within 100 diameters of each other) are accelerated to 1.55U_S. Modification of Han and Lawler's [Han, M., Lawler, D.F., 1991. Interactions of two settling spheres: settling rates and collision efficiency. Journal of Hydraulic Engineering 117 (10) 1269–1289] equations to include inertia shows that the velocities of a pair of spheres differ and change during descent, allowing them to come into contact. In quiescent settling experiments with olivine-shaped clay prisms, the pairs of prisms usually settled faster than their U_S. The clay prisms aggregated 80% of the time, with the pairs of prisms rotating and tumbling as they fall, coming into and out of contact. Size variations in Kilauea Iki olivines control aggregation during quiescent settling. The settling velocities of individual crystals are variable, increasing when they are incorporated into aggregates. A suspension of olivine crystals (more than 0.01 vol.%) in a magma will settle slower than the average Stokes' velocities, except when a buoyant phase is present. Addition of a buoyant phase, such as CO₂ bubbles, can increase the settling velocity of the olivines when the total concentration (olivine plus bubbles) (ε) is <35%, or carry the olivines in a packed bed when ε>35%.     

Refractory megacrysts and magnesium-rich melt inclusions within spinel in oceanic tholeiites: Indicators of magma mixing and parental magma composition

Refractory megacrysts of olivine, plagioclase, chromian diopside and Cr-Al spinel, which were not in equilibrium with the host oceanic tholeiite on eruption, are present in samples from several dredge sites and DSDP drill sites in the Atlantic and Pacific Oceans. They have multiple origins: (1) cognate or accidental mantle fragments; (2) relict fragments from fractional crystallization of parental liquids considerably more primitive than oceanic tholeiite; and most commonly (3) the fractional crystallization products of such liquids mixed with oceanic tholeiite magma. Melt inclusions in chrome-spinel phenocrysts provide evidence for this postulated Mg- and Ca-rich magma which has counterparts in the Scottish Tertiary Province and in west Greenland.     

The seismic detection of a magma chamber beneath Mount Etna,sicily

Macroseismic studies, linear refraction profiles, and a two-dimensional seismic array study on Mount Etna, have all detected anomalous low velocity zones beneath the volcano. Seismic travel time delays together with high frequency attenuation observations confirm the presence of a large volume of partial melt beneath the volcano. A simplified three-dimensional model of the main storage system of Mount Etna is presented.Paper presented at the Symposium Volcanoes of the Earth and Planets, held at the University of Lancaster, March 17, 1981.     

长白山天池火山岩浆系统分析

针对外媒报道长白山天池火山在近2年内有可能喷发的言论,在长白山天池火山区布测了一条长度约为103 km的二维大地电磁测深观测剖面,对火山区深部电性结构进行探测研究.由于研究区内不明来源的电磁干扰非常强,对数据采用了远参考处理、Robust处理、Rhoplus分析、张量阻抗分解和基于一维层状介质电阻率与相位互算方法等先进处理技术,获取到一批在强干扰区质量较为可靠的电磁数据,利用数据计算分析了长白山天池火山区二维构造走向和感应矢量特征,采用NLCG二维反演技术对资料进行了二维反演解释,并将反演结果与前人探测结果进行了对比分析.探测结果表明:在天池火山口下方存在明显的直立型岩浆通道,岩浆通道在下方约5~8 km位置形成关闭;在火山口下方往北方向附近,在埋深位置约7 km深处存在一个明显的低阻异常体,电阻率小于10 Ωm,且与岩浆通道对接,推测其可能是地表浅部发育的岩浆囊;在长白山山门附近C07-C09号测点之间和C04-C05号测点之间,在埋深约7~17 km深处发现近直立型低阻带,低阻带与下方低阻体直接相连,推测低阻带内赋存有活动的岩浆;随着埋深的增加,从天池火山口南部约20 km位置往北方向,在埋深13~30 km之间壳内广泛发育明显的低阻异常体,推测其可能是活动的岩浆囊.反演结果与前人探测结果整体电性特征相似,但又局部不同.     

Injection of additional batches of magma into the formed magma chamber: The case study of Lukkulaisvaara intrusion,North Karelia

The probable implications of the injection of additional batches of melt into the magma chamber and the correlation of ore formation to these processes are considered. The assumed model of hydraulic impact (stamp) explains a number of the structural features of layered basite-hyperbasite intrusions and, in particular, formation of microgranular rocks, whose structure indicates a high rate of crystallization, and, probably, platinum group element (PGE) mineralization in the Lukkulaisvaara layered intrusion, North Karelia. It is shown that intrusion of additional batches of magma can lead to thermodiffusion, which is most effective in the vicinity of the contact between the chamber and the portion of magma. This, in turn, should result in the redistribution of chemical components in this zone and, probably, to anomalous concentrations of productive components at early penetration stages.     

Cretaceous episodic growth of the Japanese Islands

Abstract The Japanese Islands formed rapidly in situ along the eastern Asian continental margin in the Cretaceous due to both tectonic and magmatic processes. In the Early Cretaceous, huge oceanic plateaus created by the mid-Panthalassa super plume accreted with the continental margin. This tectonic interaction of oceanic plateau with continental crust is one of the significant tectonic processes responsible for continental growth in subduction zones. In the Japanese Islands, Late Cretaceous-Early Paleogene continental growth is much more episodic and drastic. At this time the continental margin uplifted regionally, and intra-continent collision tectonics took place in the northern part of the Asian continent. The uplifting event appears to have been caused by the subduction of very young oceanic crust (i.e. the Izanagi-Kula Plate) along the continental margin. Magmatism was also very active, and melting of the young oceanic slab appears to have resulted in ubiquitous plutons in the continental margin. Regional uplift of the continental margin and intra-continent collision tectonics promoted erosion of the uplifted area, and a large amount of terrigenous sediment was abruptly supplied to the trench. As a result of the rapid supply of terrigenous detritus, the accretionary complexes (the Hidaka Belt in Hokkaido and the Shimanto Belt in Southwest Japan) grew rapidly in the subduction zone. The rapid growth of the accretionary complexes and the subduction of very young, buoyant oceanic crust caused the extrusion of a high-P/T metamorphic wedge from the deep levels of the subduction zone. Episodic growth of the Late Cretaceous Japanese Islands suggests that subduction of very young oceanic crust and/or ridge subduction are very significant for the formation of new continental crust in subduction zones.     

A model of the Phlegraean Fields magma chamber in the last 10,500 years

Volcanological and petrological data suggest that the Phlegraean Fields volcanic activity has been fed, at least in the last 10,500 years, by a not-refilled magma chamber where trachytic residual liquids were produced by fractionation of a trachybasaltic magma. Using estimated volumes of the erupted products andP–T data obtained through petrological studies, a conductive thermal model of the chamber was built up in order to estimate its past and present size. Results suggest a volume decrease from approximately 14 to 1.4 km³ of the trachybasaltic magma in 10,500 years. Trachytic liquid would also be present in the chamber in a minimum amount of 0.4 km³. The model allowed some insights on the petrogenesis of the Phlegraean trachytes, suggesting that they were erupted as liquids because thermally buffered within the magma chamber.     

Co-existing calcic amphiboles in calc-alkaline andesites: Possible evidence of a zoned magma chamber

Hornblende-biotite andesites erupted from Mount Price and Clinker Peak volcanoes, southwestern British Columbia, contain two texturally and compositionally distinct calcic amphiboles: pargasitic hornblende xenocrysts and magnesio-hornblende microphenocrysts. Disequilibrium relationships exhibited by these amphiboles and associated minerals suggest that the magnesio-hornblendes precipitated under chemical and thermal conditions that were intermediate between those under which pargasitic hornblende and biotite, respectively, crystallized. Experimental studies of crystallization in double-diffusive systems (Chen and Turner, 1980; Turner, 1980; McBirney, 1980) suggest that these varied magmatic environments can be explained as a consequence of progressive crystallization within a zoned magma chamber. Although gravitational settling may have played a role, the observed mineral assemblages probably developed by convective mixing of crystals precipitated at the cooling margins with those crystallized in the interior of the compositionally stratified magma column.     

Gravity evidence for an extinct magma chamber beneath Syrtis Major, Mars: a look at the magmatic plumbing system

Syrtis Major is an ancient basaltic shield volcano on Mars with a basal diameter of 1100 km. The free-air gravity anomaly is 126 mGal at spherical harmonic degree 50 and reaches its maximum amplitude over the 2 km deep topographic caldera. The observed gravity anomaly cannot be explained by flexurally supported surface topography and requires the presence of a buried, high-density load. The geologically most reasonable interpretation of this high-density load is that it represents the magma chamber of Syrtis Major, now solidified and filled at least in part by dense igneous cumulates. Pyroxene is likely to be the dominant cumulate mineral in this system, although olivine may also be present. Gravity models presented here define the structure of the buried load and in essence provide a look at the magmatic plumbing system of this volcano. The preferred model involves a buried load that is approximately 300×600 km across, roughly twice as large as the topographic caldera. Both the buried load and the caldera are elongated in the north-south direction. In the center of the buried load, the minimum thickness is 2.8 km for an olivine-dominated cumulate system or 3.9 km for a pyroxene-dominated system. The best terrestrial analog for this structure is the Bushveld Complex, an igneous cumulate body that is similar in size and thickness to the Syrtis Major structure. Assuming that the mean crustal density is 2600 kg m⁻³ due to impact brecciation, the elastic lithosphere at Syrtis Major was 10-15 km thick at the time when the topographic load was emplaced. This corresponds to a lithospheric thermal gradient of 28-52 K/km and a surface heat flux of 70-130 mW m⁻². Higher resolution gravity data, such as that which is planned for the 2005 Mars Reconnaissance Orbiter, will permit further refinement of the dimensions of this structure.