eruption style, although sulfur emissions play a critical role in determining the climactic ...... Eichelberger, J. C., C. R. Carrigan, H. R. Westrich, and R. H. Price,.
Volatile Controls on Magma Ascent and Eruption Katharine V. Cashman Department of Geological Sciences, University of Oregon, Eugene, Oregon
“Gas is the active agent, and the magma is its vehicle” F. A. Perret
Volatiles provide the primary driving force for volcanic eruptions, thus understanding magma degassing is fundamental to understanding volcanic activity in volatile-rich arc environments. A complete picture of volatile behavior requires knowledge of not only (1) how, when and where volatiles saturate, exsolve, and accumulate in magma reservoirs and (2) how they nucleate, expand and coalesce during magma ascent; but also (3) how volatiles affect the phase relations, rheology, and fluid dynamics of the magma, and (4) how volatiles escape to and interact with surrounding wall rocks and hydrothermal systems. Together these interactions determine degassing conditions, rates of magma ascent to the Earth’s surface, and, ultimately, the style and intensity of volcanic eruptions. For example, rapid closed-system degassing provides the explosivity of silicic plinian eruptions, while open-system gas escape permits passive effusion of lava domes. Variations in the details of magma decompression rates and paths create the rich variability in eruptive style that characterizes arc volcanism. Development of a fully integrated perspective on the role of volatiles in volcanic systems requires both better constraints on the time scales and dynamics of processes occurring within volcanic conduits and coupling of conduit processes to other parts of subvolcanic systems.
1. INTRODUCTION
[e.g., Lacroix, 1904; Mercalli 1907; Walker, 1973; Pyle, 1989; Fig. 1], implicit assumptions about the role of volatiles underlie most interpretations of eruptive activity. These assumptions are most obvious in the Volcanic Explosivity Index [Newhall and Self, 1983], which assumes that eruption magnitude and intensity are correlated and together describe eruptive style. However, subsequent work has shown that these parameters are linked only for steady explosive eruptions, and must be decoupled for descriptions of unsteady and effusive activity [Pyle, 2000]. Here I show that much of this decoupling arises from complex feedbacks that develop as a result of competing time scales of decompression-driven vesiculation, crystallization, magma ascent, and volatile loss. Weaving these complexities into new descriptions of eruptive behavior and predictive models of volcanic activity presents the main challenge for the future.
Magma produced in arc environments is typically rich in volatile elements, particularly H2O. Volatile elements are sonamed because they dissolve in silicate melts at high pressures but form a gas at lower pressures. Formation of a free vapor phase creates overpressures in magma reservoirs and adds buoyancy that drives magma ascent; rapid near-surface gas expansion provides the kinetic energy of explosive eruptions. Thus although volatile processes have not been explicitly incorporated into common eruption classification schemes Book Title Book Series Copyright 2004 by the American Geophysical Union 10.1029/Series#LettersChapter# 1
2 VOLATILE CONTROLS
Figure 1. Walker’s [1973] classification of volcanic eruptions based on deposit characteristics.
2. MAGMA STORAGE Most magma resides in shallow (upper crustal) storage regions before beginning the final ascent toward the surface. Here it may saturate in volatiles by cooling and crystallizing, by adding volatiles from a deep source, or by decompression. In this section I review evidence from melt inclusions, volatiles released during eruption, and phase equilibria that many arc magmas are not only saturated in volatiles prior to eruption, but contain several wt% of an exsolved C-O-H-S vapor phase. Understanding the temporal development and spatial distribution of this vapor phase is central to questions of eruption triggering and transfer of magmatic volatiles to ore-forming systems. 2.1. Magmatic Volatiles Reviews of magmatic volatiles are provided in Carroll and Holloway [1994] and Wallace and Anderson [2000] and will be only briefly summarized here. Water and carbon dioxide are the two most common volatile components in natural magmas, followed by sulfur, chlorine and fluorine. Water is much more
soluble in silicate melts than is carbon dioxide, although the solubility of both species decreases with decreasing pressure. Solubility is also a function of melt composition, with both H2O and CO2 being somewhat less soluble in basaltic melts than in rhyolitic melts (Fig. 2a,b). In a vapor-saturated melt, the solubility of a given volatile species is determined by its equilibrium partial pressure in the gas phase. For this reason, addition of H2O to a CO2-bearing melt at constant pressure increases PH2O, which decreases both PCO2 and dissolved CO2 (Fig. 2c). The low abundance of S, Cl and F suggests that exsolution of these components exerts only a minor control on eruption style, although sulfur emissions play a critical role in determining the climactic effects of large eruptions. Magmatic volatiles may be measured directly when they are quenched in melt pockets (inclusions) trapped in phenocrysts (Fig. 3a; inset). Melt inclusions are best preserved in crystals with poor cleavage, such as olivine and quartz, thus detailed melt inclusion studies are limited to end member basaltic or rhyolitic compositions [Lowenstern, 1994;1995; Metrich et al., 2001; Roggensack et al., 1997; Schmitt, 2001; Sisson and Layne, 1993; Wallace et al., 1995;1999]. These studies show that both arc-related basalts and large rhyolitic magma bodies may initially contain > 6 wt% H2O. However, melt inclusions in individual phenocrysts within a given eruptive deposit often contain variable amounts of H2O and CO2, reflecting crystallization over a wide range of magmatic pressures. This range is particularly striking in studies of large ignimbrite units such as the Bishop and Pine Grove Tuffs (Fig. 3a). Correlations between CO2 content and incompatible trace element abundance in these inclusions indicate that crystallization occurred under gas-saturated conditions and created an exsolved volatile phase that would have comprised up to 6 wt% (30 vol%) of the magma reservoir. Measurements of SO2 emissions associated with explosive eruptions support this interpretation. Volcanic plumes often contain “excess” S, that is, sulphur volumes in excess of the amount dissolved in the erupted magma. This excess S must have been exsolved in a C-O-H-S vapor phase prior to eruption [Fig. 3b]. Rec-
Figure 2. (a) H2O solubility in rhyolite (850°C) and basalt (1200°C); (b) CO2 solubility in rhyolite (850°C) and basalt (1200°C); H2O and CO2 solubility in volatile-saturated rhyolitic melt at 850°C. Redrafted from Wallace and Anderson [2000].
CASHMAN
3
Figure 3. (a) H2O and CO2 concentrations in quartz melt inclusions from two large silicic systems: Pine Grove Tuff [outlined field, Lowenstern, 1994]; Bishop Tuff [points, Wallace et al., 1995]. (b) Anticipated (lines) and measured (points) relationship between erupted magma volume and SO2 emissions for andesitic (filled circles; solid lines) and silicic (open diamonds, dashed lines) compositions. Calculated amounts of excess sulfur required to match observed emissions are labeled in wt% next to the appropriate line. Redrafted from Wallace [2003].
onciliation of measured and dissolved S abundances also requires 1–5 wt% vapor [Gerlach and McGee, 1994; Luhr, 1990; Scaillet et al., 1998; Wallace, 2003].
1993]. Evidence of extensive volatile-saturated crystallization in some systems supports SO2-emission data that require a substantial (= 4 wt%) free volatile phase for most recent eruptions of andesitic arc volcanoes (Fig. 3b).
2.2. Experimental Phase Equilibria 2.3. Porphyry Ore Deposits The volatile content of intermediate composition magmas is more difficult to constrain by melt inclusion analysis because the common phenocryst phases (plagioclase, pyroxene and hornblende) tend to leak volatiles. As a consequence, the volatile contents of these magmas are best determined through phase equilibria studies [e.g., Costa and Scaillet, 2003; Moore and Carmichael, 1998; Rutherford and Devine, 1988;1996], as illustrated in Figure 4a. The phases most indicative of volatile abundance are hornblende, which generally requires a minimum of 3–4 wt% H2O in the melt, and plagioclase, the stability (liquidus temperature) of which is extremely sensitive to PH2O. As plagioclase abundance is easily measured, it can be used in conjunction with experimental data to infer the water content and PH2O at which magma last resided. For example, an isothermal (950°C) section through the phase diagram in Figure 4a shows that volatile-saturated decompression from 200 to 100 MPa would reduce the H2O content of the liquid from 5.5 to 3.5 wt%, causing an increase in plagioclase abundance from 10 to 25% (Fig. 4b). Thus the phase assemblage of plagioclase, hornblende, pyroxene and Fe-Ti oxides common to intermediate composition calc-alkaline magmas provides an estimate of both the minimum equilibration pressures (from hornblende stability) and the extent of volatile-saturated decompression and crystallization [Blundy and Cashman, 2001; Carmichael, 2002; Sisson and Grove,
Further evidence of volatile concentration in the upper parts of magma reservoirs lies in granite-hosted porphyry Cu and Mo ore deposits [Burnham, 1997; Cloos, 2001; Lowenstern, 1994; Shinohara and Kazahaya, 1995]. Porphyry ore deposits form in arc environments in association with shallow intrusions of intermediate to evolved compositions. Solidified granite porphyries have low bulk volatile contents that contrast with the high volatile content of most silicic melts. This mismatch requires that silicic magma chambers exsolve and discharge most of their volatile components during solidification, either to the atmosphere (via volcanic eruptions) or to hydrothermal systems. Evidence for high volatile concentrations above magma reservoirs can be found in the key stratigraphic components of porphyry systems: a volatile-enriched ‘cupola’ that typically overlies an elongate porphyry intrusion and is tapped by radial and concentric fluid-rich veins that feed the ore deposit. Although this physical picture is consistent with inferred volatile gradients in active volcanic systems, questions remain about the temporal and spatial relationship between ore formation and volcanism. Traditionally viewed as separate, growing evidence for high temperature magmatic fluids involved in ore formation and extensive volatile transfer accompanying volcanic eruptions suggests that links between these two environments need to be re-evaluated.
4 VOLATILE CONTROLS
[Menard and Tait, 2001]. Finally, free volatiles contained within the magma reservoir at the time of eruption will substantially increase the volatile contribution to the atmosphere. For this reason, petrologic estimates of volatile contributions from explosive eruptions are minima, and independent estimates of the free volatile phase are required to assess the full impact of past eruptions on global climate. Future research challenges include (1) identifying the origin of excess volatiles; (2) describing the spatial and temporal evolution of the volatile phase in magma reservoirs of varying depths, compositions, and tectonic environments; (3) linking the volatile evolution of magmatic systems directly to the evolution of ore-forming systems; and (4) determining potential trigger mechanisms for magma ascent and eruption provided by a free volatile phase. 3. MAGMA ASCENT
Figure 4. H2O-saturated andesite phase relations [redrafted from Moore and Carmichael, 1998]. (a) Phase fields and equilibrium H2O contents. Gray line is liquidus; gray field shows PH2O-T range of andesitic magmas; dashed line shows approximate quartz phase boundary [estimated from Barclay et al., 1998; Brugger et al., 2003; Hammer and Rutherford, 2002a; Martel and Schmidt, 2003]. (b) Isothermal section through (a) shows co-variation of melt and plagioclase phase proportions as a function of PH2O (gray field).
2.4. Implications The presence of a free volatile phase in regions of magma storage has enormous implications for magma ascent and eruption. Extensive decompression-induced crystallization explains the plagioclase-phyric nature of many arc magmas, and, perhaps, the common association of porphyry intrusives with magmatic hydrothermal ore deposits. Overpressures generated by exsolution of a free volatile phase may be sufficient to fracture wall rock and permit both gas escape and magma ascent toward the surface. Volatile accumulation and overpressure generation at the roofs of magma reservoirs may trigger rapidly propagating gas-driven fractures [Carrigan, 2000; Lister, 1990; Rubin, 1993] and transfer of volatiles into overlying hydrothermal systems. Such gas-driven fractures may precede magma ascent, causing phreatic (groundwaterdriven) activity before the onset of a full-scale eruption
Prior to eruption, magma must be transferred from the storage reservoir to the surface, a process now recognized as critical to the study of eruptive processes [Jaupart, 2000]. Early models calculated the velocity of a homogeneous fluid moving steadily up a vertical pipe, with frictional losses dictated by wall shear stresses and volatile exsolution/expansion controlled by equilibrium solubility [Wilson et al., 1980]. More recent models allow disequilibrium between the gas and liquid phases [Jaupart and Allegre, 1991; Papale et al., 1998; Woods and Koyaguchi, 1994], compressible fluid flow [Massol and Jaupart, 1999; Massol et al., 2001], coupling with thermodynamic models to calculate phase changes [Mastin and Ghiorso, 2000], open system degassing and crystallization [Melnik and Sparks, 1999;2002a;b] and variable bubble nucleation kinetics [Mangan et al., 2004; Papale, 2001]. These models demonstrate the sensitivity of magma ascent to the kinetics of vesiculation and crystallization, and to related changes in the rheology of bubble- and crystal-bearing melt, the dynamics of magma flow through conduits, and the explosivity of resulting eruptions. Here I review field, experimental and theoretical constraints on both the kinetics of phase changes during decompression and the rheological changes induced by the addition of bubbles and/or crystals to silicic melts. 3.1. Phase Changes Resulting from Decompression Ascent of magma through volcanic conduits decompresses the melt and, as a consequence, decreases the volatile solubility (Fig. 2) and changes the stability of crystalline phases (Fig. 4). In response, volatiles exsolve from the melt by either diffusion into existing gas bubbles or nucleation of new bubbles, hydrous minerals break down, and anhydrous minerals crystallize. The kinetics of these phase changes are controlled by melt com-
CASHMAN
position, volatile content, viscosity and rate of magma decompression. 3.1.1. Bubble nucleation. Nucleation of bubbles within a melt may occur by one of two mechanisms: nucleation may be heterogeneous on existing crystal phases or homogeneous within the melt itself. Experiments on silicic melts show that heterogeneous nucleation occurs at low supersaturations (∆P < 20 MPa) and volatiles exsolve under near-equilibrium conditions. Heterogeneous nucleation produces a single population of bubbles whose number is controlled by the number of crystal sites [Gardner et al., 1999; Hurwitz and Navon, 1994]. Further decompression causes these bubbles to grow, creating a Gaussian size distribution with a mean size that increases with decreasing pressure (Fig. 5a). In the absence of sufficient nucleation sites or at very high rates of decompression, nucleation is homogeneous and continuous [Mangan et al., 2004; Mangan and Sisson, 2000; Mourtada-Bonnefoi and Laporte, 1999;2002], and bubble number decrease exponentially with increasing bubble size (Fig. 5b). Homogeneous nucleation requires high supersaturations (∆P = ~100MPa) that increase with decreasing abundance of H2O and CO2 in the melt. Degassing paths are far from equilibrium, and continuous nucleation creates high bubble number densities (> 108/cm3; Fig. 5c), particularly when the exsolving phase is a mixture of H2O and CO2. In natural systems, bubble nucleation mechanisms may be inferred from the bubble populations preserved in eruptive products. Vesicular clasts of rhyolitic pumice produced by high energy Plinian eruptions have bubble number densities of 108–1010/cm3, equalling or exceeding those produced by homogeneous nucleation of H2O–CO2 vapor mixtures. Bubble numbers decrease exponentially with increasing bubble size [Klug et al., 2002], and groundmass crystals that might serve as nucleation sites are rare or absent. These textural characteristics indicate that Plinian eruptions are driven by homogeneous bubble nucleation at high supersaturation. Initial bubble number densities, in turn, control (1) inter-bubble diffusion distances [Gardner et al., 1999; Mangan and Sisson, 2000], (2) subsequent rates of bubble growth [Proussevitch and Sahagian, 1998], and (3) times required for local volatile depletion and brittle fragmentation [Zhang, 1999]. Conditions of bubble nucleation are more difficult to constrain for other eruption styles because of complications introduced by syn-eruptive crystallization and variable amounts of passive gas loss. However, experimental data suggest that heterogeneous, near-equilibrium, volatile exsolution should increase in importance as magma decompression rates decline. 3.1.2. Degassing-induced crystallization. First suggested early in the 20th century to explain the famous spine of Mt.
5
Pelee [Shepherd and Merwin, 1927; Williams, 1932] and quantified during the 1980–1986 eruption of Mount St. Helens, WA [Cashman, 1992], decompression-induced crystallization has now been documented during effusive and intermittent explosive phases of numerous recent andesitic and dacitic eruptions [e.g., Hammer et al., 1999;2000; Martel et al., 2000; Nakada et al., 1995;1999]. These observations have led to new models of cyclical dome growth and explosive activity controlled by complex inter-relations among processes of magma ascent, degassing, and crystallization [Melnik and Sparks, 1999; Sparks, 1997; Voight et al., 1999]. Important constraints on crystallization rates are provided by H2O-saturated decompression experiments on rhyolitic melts [Couch et al., 2003; Geschwind and Rutherford, 1995; Hammer and Rutherford, 2002; Martel and Schmidt, 2003]. As anticipated from phase equilibria studies, plagioclase is the most abundant crystallizing phase. The effective undercooling (∆T) driving plagioclase crystallization is determined by both the final pressure (Pf) and the decompression rate. The extent of crystallization generally increases with
Figure 5. Results of bubble nucleation experiments: (a,b) bubble size distributions for heterogeneous and homogeneous nucleation; (c) log bubble number density (no./cm3) vs. vesicle volume fraction (φb) for homogeneous (open symbols) and heterogeneous (solid symbols) nucleation [data from Gardner et al., 1999; Mourtada-Bonnefoi and Laporte, 1999; 2002; Mangan and Sisson, 2000]. Arrows show general trend for homogeneous nucleation of H2O- and H2OCO2-bearing melts.
6 VOLATILE CONTROLS
increasing ∆T, as does the mechanism of phase change, with crystal nucleation dominant at high ∆Ts (Pf = 5–15 MPa) and crystal growth important at intermediate ∆Ts (Pf ~ 100 MPa). Long times (slow decompression) permit extensive crystallization while short times (rapid decompression) limit plagioclase precipitation. In the extreme, very rapid magma transit to the Earth’s surface will inhibit crystallization altogether. Groundmass textures preserved in erupted samples show evidence of variable decompression rates and magma equilibration depths. When magma rises quickly but stalls at shallow levels, as is common in pulsatory (subplinian) eruptive sequences, crystallization occurs primarily by nucleation, and resulting pyroclasts contain abundant small skeletal or acicular plagioclase crystals (Fig. 6b). When magma ascent is slow (as in many effusive eruptions) or when magma stalls at intermediate depths, crystal growth dominates and the groundmass contains abundant plagioclase microphenocrysts (Fig. 6c). When magma ascent from magma storage depths is sufficiently rapid to permit vesiculation but not crystallization, the result is vesicular pumice (Fig. 6d). The same textural features may be seen in the anhydrous breakdown products of hornblende: rapid ascent to shallow levels produces thin, finegrained breakdown rims; slow ascent permits extensive breakdown and growth of coarse-grained rims; and very rapid ascent prevents hornblende breakdown Thus both hornblende and groundmass textures provide important constraints on magma decompression histories.
Figure 6. (a) Experimentally determined rates of plagioclase growth and nucleation in rhyolitic melt [data from Hammer and Rutherford, 2002]. (b–d) Groundmass textures representative of melts last equilibrated at (b) shallow (~ 10 MPa; Pichincha, Ecuador, breadcrust bomb), (c) intermediate (~ 100 MPa; MSH dome), and (d) deep (> ~ 150 MPa, MSH pumice) pressures.
3.2. Degassing-induced Changes in Magma Rheology All of the physical changes experienced by magma during ascent and degassing translate into dramatic changes in the bulk rheology of bubble- and/or crystal-bearing melts, rheological changes that modulate the timing and explosivity of ensuing eruptions. There are three main controls on magma rheology in volcanic conduits: (1) exsolution of H2O from the melt causes an increase in the melt viscosity, particularly at low water contents (low pressures); (2) addition of bubbles to a fluid creates shear-thinning rheologies; and (3) addition of crystals increases the magma viscosity and, when crystals interact, generates non-Newtonian responses to imposed shear stresses. Together these effects create complex feedback mechanisms that control rates of magma ascent and eruption. Loss of H2O has little effect on melt viscosity over much of the anticipated H2O range of silicic melts: viscosity increases by only an order of magnitude as the water content of a rhyolite melt drops from 7 to 3 wt% (Fig. 7a). However, melt viscosity changes rapidly as H2O decreases to low levels, increasing four orders of magnitude from 1.5 to 0.1 wt% H2O (a pressure change of < 50 MPa). Thus shallow degassing is an effective means of producing large rheological contrasts over small pressure drops. The effect of bubbles on magma rheology depends not only on bubble volume but also on bubble size (radius r), melt viscosity (µ) and shear rate (G; Fig. 7b). Shear-rate sensitivity is expressed by the capillary number Ca = rGµ/Γ, a ratio of competing stresses imposed by shearing that deforms and surface tension (Γ) that restores a bubble to a spherical shape. At high shear rates, bubbles deform and reduce the shear viscosity; at low shear rates bubbles remain close to spherical and increase the shear viscosity [Manga et al., 1998; Pal, 2003; Rust and Manga, 2002]. Although the effect of bubbles on magma viscosity is relatively small (less than a factor of 10 at either Ca limit), the shear-thinning behavior of bubbly magma concentrates shear near conduit walls to create large horizontal pressure and velocity gradients [Massol and Jaupart, 1999]. High shear rates may stretch bubbles to form tube pumice [Klug et al. 2002; Polacci et al. 2001], enhance coalescence and gas escape to produce dense obsidian [Rust et al., 2003] or facilitate shear-induced fragmentation [Papale, 2001]. Adding crystals to a melt increases the suspension viscosity and, at moderate crystal concentrations, changes its rheology. The relationship between viscosity and particle volume fraction is controlled by the maximum packing fraction φm (Fig. 7c), commonly taken as 0.6 [Marsh, 1981]. However, the particle concentration at maximum packing changes with variations in both crystal shape and imposed stress [Zhou et al., 1995]. Additionally, well before maximum packing is
CASHMAN
7
Figure 7. Viscosity variations in bubble- and crystal-bearing magmas. (a) Melt viscosity as a function of dissolved H2O [Hess and Dingwell, 1996]. (b) Variation in relative viscosity (µr) with capillary number Ca for different bubble volume fractions (φb; from Rust and Manga [2002] and Pal [2003]). (c) Variation in log µr with increasing crystal volume fraction (φ) according to the simple Roscoe-Einstein relation for φm = 0.6 [e.g., Marsh, 1981].
achieved, touching crystal networks create complex (nonNewtonian) responses to shear [Kerr and Lister, 1991]. Near the packing limit, large imposed shear stresses at conduit walls causes individual crystals to break; when accompanied by viscous heating this process may help to lubricate the ascent of crystal-rich magmas through volcanic conduits [Polacci et al., 2001; Rosi et al., 2003]. When the crystal concentration exceeds the packing limit, magma may dilate under shear [Smith, 2000]. As high crystallinities are common in very shallow magma instrusions [e.g., Cashman and Hoblitt, 2004], shear-dilatancy may contribute to both deformation and seismic precursors to some eruptions. 3.4. Modeling Volatiles and Magma Ascent Rates of magma ascent thus depend on complex feedbacks determined largely by decompression rate. Rapid homogeneous bubble nucleation induced by very high rates of decompression causes rapid magma expansion and large horizontal velocity gradients because of shear-thinning rheologies. At more modest decompression rates, ascent rates will be controlled by the relative rates of bubble and crystal nucleation. This coupled problem of vesiculation and crystallization during decompression is poorly constrained. Thus future challenges for improving models of magma ascent include (1) improving our understanding of the kinetics of decompressing magmatic systems sufficiently to predict both the bubble- and crystal- content of the magma as a function of location in the conduit, (2) determining the effect of changing phase proportions on magma rheology, and (3) predicting the effect of changing rheologies on continued magma flow. To date, experimental and textural studies have been largely confined to rhyolitic melts, thus need to be extended to more mafic compositions. Additionally, while the effects of water content, temperature and bubble concentration on melt rheology are fairly well constrained, existing rheological models
of crystal-melt suspensions are fairly crude, a reflection of difficulties posed by experimental and numerical studies. 4. DEGASSING AND ERUPTION The physical changes that accompany magma ascent and decompression translate into a wide range of eruptive styles. For example, high rates of gas expansion that accompany rapid decompression and vesiculation cause explosive disruption of magma, generating large and sustained Plinian eruption columns. Alternatively, slow magma ascent accompanied by extensive crystallization and gas loss produces effusion of viscous lava domes. Between these end members lies a continuous range of eruptive styles characterized by variable decompression and degassing trajectories. Here I briefly review degassing styles, physical mechanisms by which gas may be lost during magma ascent, and the translation of degassing style into eruptive behavior. 4.1. Styles of Degassing Volatile exsolution may occur under equilibrium or nonequilibrium conditions, may precede or accompany volcanic eruptions, and may involve variable amounts of degassing (gas escape). When bubbles remain within the melt during magma ascent, the mixture expands as a ‘closed-system’ and the gas volume fraction is a direct measure of the extent of degassing. In contrast, when volatiles segregate from the melt prior to eruption, the system is considered ‘open’ (Fig. 8a). Here vesiculation and eruption may be decoupled, and the gas volume fraction of the erupted material no longer indicates the extent or final equilibration pressure of degassing. Taking a broader perspective, an entire eruptive sequence may be viewed as closed when the rise and decompression of a single magma batch causes both magma and gas emission rates to decrease exponentially with time [Gerlach and McGee, 1994;
8 VOLATILE CONTROLS
Figure 8. (a) Schematic illustration of open- and closed-system degassing. (b) CO2-H2O solubility diagram illustrating schematic open- and closed-system degassing paths. (c) Calculated change in vesicle volume fraction with pressure under closed-system conditions for a range of initial volatile contents (labeled on curves in wt%). Highlighted are example open- and closed-system degassing paths for an initial H2O content of 5 wt%.
Zapata et al., 1997]. Volcanic systems may be considered open when gases are either lost from the conduit prior to eruption, or when volatiles fed from a deep source flux through the magmatic system. In the latter case, eruptive activity is modulated by the ability of the system to act as a conduit for the volatile phase [Edmonds et al., 2003], and ultimately controlled by deep inputs to the magmatic reservoir [Sparks and Young, 2002]. Volatile exsolution under closed-system conditions causes a decrease in H2O and CO2 over much of the decompression range [Fig. 8b; e.g., Newman and Lowenstern, 2002]. Equilibrium calculations show that even modest amounts of H2O produce final gas volume fractions > 0.99 if closed-system conditions are maintained to atmospheric pressure (Fig. 8c). In contrast, early gas loss during open-system degassing causes rapid CO2 depletion of the melt at near-constant H2O values (Fig. 8b). Once CO2 is lost, continued decompression depletes H2O. The gas volume fraction of the magma will be less (by an amount equal to the gas loss from the system) than predicted by equilibrium models. 4.2. Physical Controls on Open-system Degassing Gas segregation in open systems may cause violent explosions if those gases are contained (and pressurized) within the conduit, or effusion of bubble-poor magma if gases are removed from the conduit [e.g., Jaupart, 1998]. Degassing conditions depend on both magma ascent rate and the relative
permeabilities of magma and wall rock. Magma permeability may develop by brittle fracture of viscous magma or development of connected bubble networks within expanding bubbly melt. Magma will fracture when either the melt viscosity or the strain rate is sufficiently high, a condition most likely to be met along conduit walls [Gonnerman and Manga, 2003]. Evidence for syn-eruptive magma fracture includes brecciated obsidian clasts in subplinian pyroclastic deposits [Rust et al., 2004], tuffisite veins found within exposed magma conduits [Tuffen et al., 2003] and the occurrence of hybrid earthquakes interpreted to result from some combination of fracture and fluid flow [Chouet, 1996; White et al., 1998]. Formation of connected bubble networks requires both a sufficiently high volume fraction of the gas phase and rupture of adjacent bubble walls (coalescence). Most models of permeability development in bubble-melt suspensions rely on percolation theory, which predicts development of touching networks of spherical bubbles at a bubble volume fraction ~ 30% [Sahimi, 1994; Garboczi et al., 1995; Saar and Manga, 1999]. Coalescence is controlled by rates of fluid drainage and capillary pressures in stationary low viscosity melts. In rapidly decompressing melts, bubble expansion is likely to play a major role in film thinning, as are local pressure differences resulting from adjacent bubbles of different sizes [Klug and Cashman, 1996]. Silicic pyroclasts have high permeabilities over a range of porosities, and show no evidence of a threshold in gas volume fraction for the development of permeable networks (Fig. 9). Early and rapid expansion-driven coales-
CASHMAN
9
gas with dispersed liquid drops or isolated solid particles. Fragmentation provides the explosive force of magmatic eruptions by converting the potential energy of expanding magma to the kinetic energy of individual fragments [Cashman et al., 2000], and may be viewed as either dynamic or static. Dynamic fragmentation occurs when rapidly expanding bubbly melt disintegrates as the result of fluid instabilities when melt viscosity is low [e.g., Mader, 1998], or by exceedance of the tensile strength (brittle failure) when melt viscosity is high (Fig. 10a). In contrast, fragmentation is static when rapid decompression of previously vesiculated melt causes successive brittle fracture of the porous material [e.g., Dingwell, 1998; Fig. 10b]. Figure 9. Permeability-φb relations measured in volcanic rocks; line is empirical fit to data of Klug and Cashman [1996].
cence observed in recent vesiculation experiments [Mourtada-Bonnefoi and Laporte, 2002], and high SO2 fluxes from Soufriere Hills Volcano immediately following dome collapse [Edmonds et al., 2003], provide additional evidence for rapid and extensive development of permeable bubble networks. 4.3. Degassing and Explosive Eruptions By definition, explosive eruptions involve fragmentation of bubbly magma, where fragmentation refers to the transformation of a liquid or solid with dispersed gas bubbles to a
4.3.1. Explosive eruptions involving rapid closed-system degassing. For magma to degas as a closed system, rates of bubble nucleation, expansion, and fragmentation must exceed those of bubble coalescence and gas migration. Eruptions driven by closed-system degassing are typically explosive, with eruption intensity and magnitude correlated, volatile emissions proportional to erupted volume, and highly expanded pyroclasts. Of the range of eruptive activity exhibited by andesitic to rhyolitic volcanoes, plinian eruptions most closely meet these criteria. Plinian eruptions produce large volumes of magma that are emitted rapidly to produce sustained eruption columns and thick deposits of pumice and ash [Cioni et al., 2000]. Eruption intensity is typically > 107 kg/s and correlates approximately with erupted volume [Pyle, 2000]. Pumice clasts are
Figure 10. (a, b) Cartoons illustrating dynamic (syn-eruptive vesiculation) and “static” (pre-eruptive vesiculation) fragmentation mechanisms. (c) “Normal” and (d) “Tube” pumice samples from the 1991 eruption of Pinatubo Volcano, Philippines [from Polacci et al., 2001]. (e) Breadcrust bomb from the 1999 eruption of Guagua Pichincha Volcano, Ecuador; note the angularity of the original fracture surfaces. (f) Smooth and scoriaceous surface textures on the 1980 Mount St. Helens lava dome.
10 VOLATILE CONTROLS
vesicular, with high bubble number densities indicating homogeneous nucleation, rapid decompression and degassing under largely closed-system conditions. In detail, mean clast vesicularities of ~ 75% indicate fragmentation after near-spherical bubbles achieve close packing, which inhibits continued expansion [Sparks, 1978]. Fragmentation is dynamic, with isolated bubbles rupturing individually to form ash and permeable melt regions preserved as pumice [Klug and Cashman, 1996]. The extent of permeability development prior to fragmentation thus controls the volumetric proportion of pumice, while the extent of bubble deformation, expansion and collapse preserved in individual pyroclasts records variations in strain rate within the conduit (Fig. 10c,d). 4.3.2. Explosive eruptions involving pre-eruptive degassing. Magma that ascends slowly maintains near-equilibrium melt volatile concentrations and may experience variable amounts of degassing prior to eruption. Eruption intensity is proportional to the pre-eruptive overpressure achieved in the conduit, fragmentation is (initially) static, and emitted gas volumes may be either greater or less than amounts initially dissolved in the erupted magma. Pre-eruptive degassing and partial open-system behavior are responsible for the range of explosive activity classified in andesitic-to-rhyolitic volcanoes as vulcanian, pelean, or subplinian. Vulcanian explosions are defined as small to moderate volcanic outbursts that last seconds to minutes [Morrissey and Mastin, 2000]. Vulcanian eruptions are powered by overpressures ≤ 10–15 MPa and produce discrete violent explosions, ballistic blocks, and atmospheric shock waves. The juvenile products of vulcanian eruptions range from dense blocks to vesicular pumice and include ubiquitous breadcrusted bombs. Prismatic breadcrusted blocks (Fig. 10e) and angular pumice clasts indicate extensive brittle fragmentation of previously vesiculated and variably degassed material from the uppermost conduit or capping dome. Suggested source models for magmatic vulcanian eruptions include pressurization of a lava cap or dome at shallow levels, and closedsystem degassing at depth followed by episodic expulsion of gas into the conduit. The first type of behavior is exemplified by eruptions of Soufriere Hills Volcano, Montserrat, in 1997 [Clarke et al., 2002; Druitt and al., 2002; Formenti et al., 2003], where volatile exsolution precedes eruption, the extent of degassing varies with position in the conduit, and fragmentation occurs by brittle fracture in response to a suddenly imposed decompression wave. Pelean (dome explosion) eruptions probably occur by a similar mechanism, but with a pressure source that is within the dome rather than within the conduit. The second model arises from seismic and deformation signals from volcanoes such as Popocatepetl, Mexico, that indicate gas accumulation at depths of > 2–3 km
prior to episodic release [e.g., Arcinega-Ceballos et al., 2003]. The mechanism by which this occurs and the structure of the conduit system that permits rapid gas transfer to shallow levels is not known. Subplinian eruptions may be viewed as transitional between vulcanian and plinian, with lower magnitudes and intensities than plinian eruptions but longer durations than vulcanian explosions. Like vulcanian eruptions, however, subplinian eruptions are often characterized by unsteady (pulsatory) convective columns [Cioni et al., 2000].. Subplinian pyroclastic deposits are dominated by vesicular pumice but include clasts with a wide range in density and groundmass crystallinity [e.g., Cashman and Blundy, 2000]. Similarities in the instantaneous eruption rates and pyroclast characteristics (vesicularity and texture) of plinian and subplinian eruptions suggest similar fragmentation mechanisms. However, the wide density range of subplinian deposits indicates a component of preeruptive volatile exsolution and gas escape during temporary magma arrest in shallow conduits. Thus it is not surprising that many subplinian eruptions are preceded by vulcanian explosions. 4.4. Degassing and Effusion of Lava Flows and Domes Effusive eruptions produce lava flows or domes, with eruptive activity that may be continuous or episodic and may occur by extrusion of lava flow lobes (exogenous growth; Fig. 10f) or by intrusion of magma into an existing dome (endogenous growth). Continuous exogenous growth occurs when eruption rates are high and the system is sufficiently open to volatiles to prevent explosive disruption of the magma. Decreasing the rate of magma supply causes activity to become episodic until sufficiently low rates of magma ascent permit crustal thickening and endogenous growth styles [Fink and Griffiths, 1998; Kaneko et al., 2002; Swanson and Holcomb, 1990]. Very low rates of magma ascent produce highly crystalline volcanic spines that often (but not always) represent the end of eruptive activity. Delayed degassing caused by late-stage crystallization of anhydrous phases within silicic domes or shallow conduits may create sufficiently high pore pressures to drive explosive dome-collapse events [Sparks, 1997]. Specifically, as quartz appears on the liquidus at very low pressures (~ 5–10 MPa; Fig. 4a), rapid cotectic crystallization of quartz and feldspar at shallow depths could create explosive overpressures when capped by a sufficiently rigid and impermeable plug [Blundy and Cashman, 2001]. Evidence for high internal pressures within growing domes includes impulsive seismic signals, emitted jets of high-pressure gas, and high SO2 emissions from domes immediately after collapse events [Sparks and Young, 2002].
CASHMAN
4.5. Degassing Determined by Rates of Magma Ascent As illustrated above, the rate of magma decompression controls the kinetics of vesiculation and crystallization, which in turn determine a magma’s degassing history. This history is preserved (imperfectly) in the textures (vesicle and crystal content) of the erupted material (Fig. 11a). In general, closedsystem degassing during plinian eruptions generates abundant vesicular juvenile clasts, while partial gas loss from conduits prior to subplinian and vulcanian eruptions produces variably degassed and crystallized juvenile material along with low-density pumice. More extensive degassing prior to effusive eruption of lava domes and flows creates crystal-rich and vesicle-poor lava. This comparison suggests that variations in the relative abundance of different clast density populations should provide a measure of pre-eruptive gas loss from the system. Using this interpretation, Figure 11a presents a general picture of volcanic systems that become increasingly open to volatiles as eruption styles become less energetic (change from explosive to effusive). As one goal of volcano monitoring is prediction of transitions in eruptive styles, it would be useful to quantify this general observation. A measure of eruptive energy is the intensity, or mass eruption rate (MER, in kg/s). In sustained eruptions, MER is a direct measure of magma supply rate (MSR) from depth. In pulsatory (subplinian, vulcanian) eruptions, however, the instantaneous MER is a more a measure of overpressure and fragmentation conditions than of MSR. In fact, the condition
11
MSR < MER may define pulsatory activity [Scandone and Malone, 1985]. Under these conditions, processes of magma ascent and degassing can be considered separately from those of fragmentation and eruption. Separation of these processes suggests a classification framework that both incorporates variable degassing scenarios and provides a direct link to recent decompression experiments. The relationship between decompression rate (MSR) and degassing history is shown schematically in Figure 11b. The vertical axis uses clast density ranges shown in Figure 11a to approximate the extent of pre-eruptive degassing. Limiting magma supply rates are estimated at >107 kg/s for plinian eruptions, 105 to 107 kg/s for subplinian eruptions, 104 to 105 kg/s for vulcanian eruptions, and < ~104 for dome effusion, based on direct observations of recent activity [e.g., Druitt et al., 2002; Geschwind and Rutherford, 1995; Nakada and Motomura, 1999; Pyle, 2000; Scandone and Malone, 1985]. Exact values will depend on the material properties of both the erupting magma and the wall rock. To estimate decompression rates corresponding to bounding MSRs, I assume a minimum conduit diameter of 10m (Fig. 11b). This yields maximum decompression rates of ~ 1 MPa/s and 0.01 MPa/s for the subplinian/plinian and subplinian/vulcanian boundaries, respectively. Similarly, magma ascent rates > 0.02 m/s required to preserve pristine hornblende phenocrysts [Rutherford and Gardner, 2000] correspond to MSR > 4000 kg/s (0.0005 MPa/s). Although approximate, these boundaries illustrate the utility of using MSR as a framework for linking eruption style to
Figure 11. (a) Clast vesicularity range of large Plinian eruptions [Houghton and Wilson, 1989; Polacci et al. 2003], moderate Plinian eruptions [Gardner et al., 1996; Houghton and Wilson, 1989; Polacci et al., 2001], vulcanian and subplinian eruptions [Gardner et al., 1998; Formenti et al., 2003], and dome effusion [Eichelberger et al., 1986; Fink et al, 1992]. (b) Alternative classification scheme based on magma supply rate and exsolution/degassing style. Limiting MSR values are provided by estimated eruption intensities for sustained eruptions [Pyle, 2000], subplinian and vulcanian activity [Formenti et al., 2003; Scandone and Malone, 1985], and rates of dome growth [Geschwind and Rutherford, 1995; Nakada et al., 1995; Pyle, 2000].
12 VOLATILE CONTROLS
degassing conditions. Decompression rates in excess of ~ 1 MPa/s inhibit crystal nucleation [Martel and Schmidt, 2003] and promote rapid, homogeneous bubble nucleation [Mangan and Sisson, 2004]. These rates correspond to MSR > 107 kg/s, consistent with the estimated minimum MSR for plinian activity. When nucleation is heterogeneous (microlites are present) decompression rates < 0.025 MPa/s are required for maintenance of melt-vapor equilibrium over a large pressure range [Gardner et al., 1999]. This value is approximately equivalent to an MSR of ~ 105 kg/s, and may define limiting conditions for vulcanian eruptions (which require extensive pre-eruptive degassing). For 107 > MSR > 105 kg/s, decompression rates are sufficiently slow to allow limited crystallization and degassing, but sufficiently fast that eruptions are probably driven by syn-, rather than pre-eruptive vesiculation. Finally, decompression rates of 0.005 MPa/s required for hornblende breakdown lie below the limiting MSR for lava dome effusion (~ 104 kg/s), consistent with observations of dome lavas with and without hornblende breakdown rims [Rutherford and Gardner, 2000]. In summary, Figure 11b, although schematic, provides a framework that (1) links explosive and effusive eruptive styles, (2) describes the control exerted by volatiles in determining eruptive style, and (3) can be calibrated using a combination of decompression experiments, numerical models, and observations of active volcanoes. This approach could easily be extended to include MSR limits to eruptive styles characteristic of hydrous mafic volcanism, which is also an important component of arc environments. 5. SUMMARY Volatiles play a critical role in all aspects of volcanic activity in arc environments, from exsolution and accumulation in magma reservoirs, to degassing that occurs during magma ascent to the surface, to the eruptions themselves. Most importantly, the rates, styles and timing of volatile exsolution and degassing determine conditions of eruption. Recent advances in methods of measuring volatiles in melts, fumaroles and eruptive plumes provide excellent data on the behavior of the gases themselves. New experimental work constrains the degassing and crystallization behavior of silicic melts, although these studies have yet to be extended to more mafic and alkalic compositions. Of particular interest are the potential interactions between degassing and crystallization in melts where crystals nucleate easily, and may thus directly influence the evolution of the gas phase. Also important are the implications of strain-rate dependent rheology of bubble- and crystal-rich melts for both permeability development and for the flow of magma through volcanic conduits.
A recent focus on processes occurring along conduit margins is exciting, as it marks the first step toward filling a large gap in volatile studies, that is, coupling of magmatic degassing processes to conditions of volatile loss through conduit walls [e.g., Jaupart and Allegre, 1991]. Newly recognized links between gas emissions, seismic signals and magmatic processes within conduits provide the tantalizing vision that real-time monitoring of gas migration within subvolcanic systems will soon be possible. Also promising are new research avenues arising from rapid developments in remote sensing. Volatile emissions can now be measured directly during large eruptions, greatly enhancing our view of syn- and post-eruptive volatile behavior. Fully integrated volatile studies, however, will require extension of this linked chemical and physical analysis in both space and time. Spatial scales of observation are improving with the use of InSAR (radar interferometry) to document periodic deep replenishments to magma reservoirs, migration of hydrothermal waters in response to changing stress fields, and intrusions of magma into the upper crust. Improved temporal scales of observation will come from integration of different monitoring techniques (to improve resolution of short time scales) with studies conducted on active and exhumed magmatic-hydrothermal systems. Acknowledgements. I would like to thank Maggie Mangan, Bernard Chouet, Hugh Tuffen, Paul Wallace and Helge Gonnerman for preprints of their papers. Stimulating discussions in both the field and the laboratory with Rafaello Cioni, Lucia Gurioli, Laura Pioli, Mauro Rosi, Alison Rust, Paul Wallace and Heather Wright helped to develop many of the ideas presented here. Mauro Rosi, Tony Fowler, Steve Sparks, Claude Jaupart, Chris Hawkesworth, Heather Wright and Rob Nicholson provided helpful comments on the manuscript. This work was supported by the National Science Foundation grant EAR0207362.
REFERENCES Arcinega-Ceballos, A., B. A. Chouet, and P. Dawson, Long-period events and tremor at Popocatepetl volcano (1994–2000) and their broadband characteristics, Bulletin of Volcanology, 65, 124–135, 2003. Blundy, J., and K. V. Cashman, Ascent-driven crystallisation of dacite magmas at Mount St. Helens, 1980–1986, Contributions to Mineralogy and Petrology, 140, 631–651, 2001. Burnham, C. W. (1997) Magmas and hydrothermal fluids. in Geochemistry of Hydrothermal Ore Deposits, ed. H. L. Barnes, 63–125. Carmichael, I. S. E., The andesite aqueduct: perspectives on the evolution of intermediate magmatism in west-central (105–99°W) Mexico, Contributions to Mineralogy and Petrology, 143, 641–663, 2002. Carrigan, C. R., Plumbing systems, in Encylopedia of Volcanoes, edited by H. Sigurdsson, pp. 149–170, Academic Press, San Diego, 2000.
CASHMAN Carroll, M. R., and J. R. Holloway, Volatiles in Magmas, Mineralogical Society of America, 1994. Cashman, K. V., Groundmass crystallization of Mount St. Helens dacite, 1980–1986: A tool for interpreting shallow magmatic processes, Contributions to Mineralogy and Petrology, 109, 431–449, 1992. Cashman, K. V., and J. Blundy, Degassing and crystallization of ascending andesite and dacite, Phil. Trans. R. Soc. Lond., 358, 1487–1513, 2000. Cashman, K. V. and R. P. Hoblitt, Magmatic precursors to the 18 May 1980 eruption of Mount St. Helens, USA, Geology, 32, 141–144, 2004. Cashman, K. V., B. Sturtevant, P. Papale, and O. Navon, Magmatic fragmentation, in Encyclopedia of Volcanoes, edited by H. Sigurdsson, pp. 421–430, Academic Press, New York, 2000. Chouet, B. A., Long-period seismicity: its source and use in eruption forecasting, Nature, 380, 309–316, 1996. Cioni, R., P. Marianelli, R. Santacroce, and A. Sbrana, Plinian and subplinian eruptions, in Encyclopedia of Volcanoes, edited by H. Sigurdsson, pp. 477–494, Academic Press, San Diego, 2000. Clarke, A. B., B. Voight, A. Neri, and G. Macedonio, Transient dynamics of vulcanian explosions and column collapse, Nature, 415, 897–901, 2002. Cloos, M., Bubbling magma chambers, cupolas, and porphyry copper deposits, International Geology Review, 43, 285–311, 2001. Costa, F., and B. Scaillet, Massive atmospheric sulfur loading of the AD 1600 Huaynaputina eruption and implications for petrologic sulfur estimates, Geophysical Research Letters, 30, doi:10.1029/2002GL016402, 2003. Couch, S., R. S. J. Sparks, and M. R. Carroll, The kinetics of degassing-induced crystallization at Soufriere Hills Volcano, Montserrat, Journal of Petrology, 44, 1477–1502, 2003. Dingwell, D. B., Recent experimental progress in the physical description of silicic magma relevant to explosive volcanism, Geological Society of London Special Publication, 145, 9–26, 1998. Druitt, T. H., et al., Episodes of cyclic Vulcanian explosive activity with fountain collapse at Soufriere Hills Volcano, Montserrat, Geological Society of London Memoirs, 21, 281–304, 2002. Edmonds, M. C., C. Oppenheimer, D. M. Pyle, R. A. Herd, and G. Thompson, SO2 emissions from Soufriere Hills Volcanoe and their relationship to conduit permeability, hydrothermal interaction and degassing regime, Journal of Volcanology and Geothermal Research, 124, 23–43, 2003. Eichelberger, J. C., C. R. Carrigan, H. R. Westrich, and R. H. Price, Non-explosive silicic volcanism, Nature, 323, 598–602, 1986. Fink, J. H., S.vW. Anderson, and C. R. Manley, Textural constraints on effusive silicic volcanism: Beyond the permeable foam model, Journal of Geophysical Research, 97, 9073–9083, 1992. Fink, J. H., and R. W. Griffiths, Morphology, eruption rates, and rheology of lava domes: Insights from laboratory models, Journal of Geophysical Research, 103, 527–545, 1998. Formenti, Y., T. H. Druitt, and K. Kelfoun, Characterisation of the 1997 Vulcanian explosions of Soufriere Hills Volcano, Montserrat, by video analysis, Bulletin of Volcanology, 65, 587–605, 2003.
13
Garboczi, E. J., K. A. Snyder, J. F. Douglas, and M. F. Thorpe, Geometrical percolation threshold of overlapping ellipsoids, Physical Review E, 52, 819–828, 1995. Gardner, C. A., K. V. Cashman, and C. A. Neal, Tephra-fall deposits from the 1992 eruption of Crater Peak, Alaska: implicatinos of clast textures for eruptive processes, Bulletin of Volcanology, 59, 537–555, 1998. Gardner, J., M. Hilton, and M. R. Carroll, Experimental constraints on degassing of magma: isothermal bubble growth during continuous decompression from high pressure, Earth and Planetary Science Letters, 168, 201–218, 1999. Gardner, J. E., R. M. E. Thomas, C. Jaupart, and S. Tait, Fragmentation of magma during Plinian volcanic eruptions, Bulletin of Volcanology, 58, 144–162, 1996. Gerlach, T. M., and K. A. McGee, Total sulfur dioxide emissions and pre-eruption vapor-saturated magma at Mount St. Helens, 1980–88, Geophysical Research Letters, 1994. Geschwind, C.-H., and M. J. Rutherford, Crystallization of microlites during magma ascent: the fluid mechanics of 1980–1986 eruptions at Mount St. Helens, Bulletin of Volcanology, 57, 356–370, 1995. Gonnerman, H., and M. Manga, Explosive volcanism may not be an inevitable consequence of magma fragmentation, Nature, 426, 432–435, 2003. Hammer, J., and M. Rutherford, An experimental study of the kinetics of decompression-induced crystallization in silicic melt, Journal of Geophysical Research, 107, doi:10.1029/2001JB000281, 2002. Hammer, J. E., K. V. Cashman, R. Hoblitt, and S. Newman, Degassing and microlite crystallization during the pre-climactic events of the 1991 eruption of the Mt. Pinatubo, Phillipines, Bulletin of Volcanology, 60, 355–380, 1999. Hammer, J. E., K. V. Cashman, and B. Voight, Magmatic processes revealed by textural and compositional trends in Merapi dome lavas, Journal of Volcanology and Geothermal Research, 100, 165–192, 2000. Hess, K.-U. and D. B. Dingwell, Viscosity of hydrous leucogranitic melts: A non-Arrhenian model, American Mineralogist, 81, 1297–1300, 1996. Houghton, B. F., and C. J. N. Wilson, A vesicularity index for pyroclastic deposits, Bulletin of Volcanology, 51, 451–462, 1989. Hurwitz, S., and O. Navon, Bubble nucleation in rhyolitic melts: Experiments at high pressure, temperature, and water content, Earth and Planetary Science Letters, 122, 267–280, 1994. Jaupart, C., Gas loss from magmas through conduit walls during eruption, Geological Society of London Special Publication, 145, 73–90, 1998. Jaupart, C., Magma ascent at shallow levels, in Encyclopedia of Volcanoes, edited by H. Sigurdsson, pp. 237–245, Academic Press, New York, 2000. Jaupart, C., and C. J. Allegre, Gas content, eruption rate and instabilities of eruption regime in silicic volcanoes, Earth and Planetary Science Letters, 102, 413–429, 1991. Kaneko, T., M. J. Wooster, and S. Nakada, Exogenous and endogenous growth of the Unzen lava dome examined by satellite infrared
14 VOLATILE CONTROLS image analysis, Journal of Volcanology and Geothermal Research, 116, 151–160, 2002. Kerr, R. C., and J. R. Lister, The effects of shape on crystal settling and the rheology of magmas, Journal of Geology, 99, 457–467, 1991. Klug, C., and K. V. Cashman, Permeability development in vesiculating magma, Bulletin of Volcanology, 58, 87–100, 1996. Klug, C., K. V. Cashman, and C. R. Bacon, Structure and physical characteristics of pumice from the climactic eruption of Mt. Mazama (Crater Lake), Oregon, Bulletin of Volcanology, 64, 486–501, 2002. Lacroix, A., La Montagne Pelee et ses eruptions, 62 pp., Masson et Cie, Paris, 1904. Lister, J. R., Buoyancy-driven fluid fracture: the effects of material toughness and of low-viscosity precursors, Nature, 411, 678–680, 1990. Lowenstern, J. B., Dissolved volatile concentrations in an ore-forming magma, Geology, 22, 893–896, 1994. Lowenstern, J. B., Applications of silicate melt inclusions to the study of magmatic volatiles, in Magmas, Fluids, and Ore Deposits, edited by J. F. H. Thompson, pp. 71–99, Mineralogical Society of Canada, 1995. Luhr, J. F., Experimental phase relations of water- and sulfur- saturated arc magmas and the 1982 eruptions of El Chichon Volcano, Journal of Petrology, 31, 1071–1114, 1990. Mader, H. M., Conduit flow and fragmentation, Geological Society of London Special Publication, 145, 51–71, 1998. Manga, M., J. Castro, K. V. Cashman, and M. Loewenberg, Rheology of bubble-bearing magmas, Journal of Volcanology and Geothermal Research, 87, 15–28, 1998. Mangan, M. T., L. G. Mastin, and T. Sisson, Gas evolution in eruptive conduits: combining insights from high temperature ad pressure decompression experiments with steady-state flow modeling, Journal of Volcanology and Geothermal Research, 129, 23–36, 2004. Mangan, M. T., and T. Sisson, Delayed, disequilibrium degassing in rhyolite magma: decompression experiments and implications for explosive volcanism, Earth and Planetary Science Letters, 183, 441–455, 2000. Marsh, B. D., On the crystallinity, probability of occurrence, and rheology of lava and magma, Contributions to Mineralogy and Petrology, 78, 85–98, 1981. Martel, C., J.-L. Bourdier, M. Pichavant, and H. Traineau, Textures, water content and degassing of silicic andesites from recent plinian and dome-forming eruptions at Mont Pelee volcano (Martinique, Lesser Antilles arc), Journal of Volcanology and Geothermal Research, 96, 191–206, 2000. Martel, C., and B. C. Schmidt, Decompression experiments as an insight into ascent rates of silicic magmas, Contibutions to Mineralogy and Petrology, 144, 397–415, 2003. Massol, H., and C. Jaupart, The generation of gas overpressure in volcanic eruptions, Earth and Planetary Science Letters, 166, 57–70, 1999. Massol, H., C. Jaupart, and D. W. Pepper, Ascent and decompression of viscous vesicular magma in a volcanic conduit, Journal of Geophysical Research, 106, 16223–16240, 2001.
Mastin, L. G., and M. S. Ghiorso, A numerical program for steadystate flow of magma-gas mixtures through vertical eruptive conduits, U.S. Geological Survey Open-File Report, 00–209, 1–59, 2000. Melnik, O., and R. Sparks, Nonlinear dynamics of lava dome extrusion, Nature, 402, 37–41, 1999. Melnik, O., and R. S. J. Sparks, Dynamics of magma ascent and lava extrusion at Soufriere Hills Volcano, Montserrat, Geological Society of London Memoirs, 21, 153–170, 2002a. Melnik, O., and R. S. J. Sparks, Modeling conduit flow dynamics during explosive eruptions at Soufriere Hills Volcano, Montserrat, Geological Society of London Memoirs, 21, 307–317, 2002b. Menard, T., and S. R. Tait, A phenomenological model for precursor volcanic eruptions, Nature, 411, 678–680, 2001. Mercalli, G., I vulcani attivi della Terra, 421 pp., Ulrico Hoepli, Milan, 1907. Metrich, N., A. Bertagnini, P. Landi, and M. Rosi, Crystallization driven by decompression and water loss at Stromboli Volcano (Aeolian Islands, Italy), Journal of Petrology, 42, 1471–1490, 2001. Moore, G., and I. S. E. Carmichael, The hydrous phase equilibria (to 3 kbar) of an andesite and basaltic andesite from western Mexico: constraints on water content and conditions of phenocryst growth, Contributions to Mineralogy and Petrology, 130, 304–319, 1998. Morrissey, M. M., and L. G. Mastin, Vulcanian eruptions, in Encyclopedia of Volcanoes, edited by H. Sigurdsson, pp. 463–475, Academic Press, San Diego, 2000. Mourtada-Bonnefoi, C. C., and D. Laporte, Experimental study of homogeneous bubble nucleation in rhyolitic magmas, Geophysical Research Letters, 26, 3505–3508, 1999. Mourtada-Bonnefoi, C.C., and D. Laporte, Homogeneous bubble nucleation in rhyolitic magmas: An experimental study of the effect of H2O and CO2, Journal of Geophysical Research, 107, doi:10.1029/2001JB000290, 2002. Nakada, S., and Y. Motomura, Petrology of the 1991–1995 eruption at Unzen: effusion pulsation and groundmass crystallization, Journal of Vocanology and Geothermal Research, 89, 173–196, 1999. Nakada, S., Y. Motomura, and H. Shimizu, Manner of magma ascent at Unzen Volcano (Japan), Geophysical Research Letters, 22, 567–570, 1995. Newhall, C. G., and W. G. Melson, Explosive activity associated with the growth of volcanic domes, Journal of Volcanology and Geothermal Research, 17, 111–131, 1983. Newman, S. and J. B. Lowenstern, VolatileCalc: a silicate melt-H2OCO2 solution model written in Visual Basic for Excel, Computers and Geosciences, 28, 597–604, 2002. Pal, R., Rheological behavior of bubble-bearing magmas, Earth and Planetary Science Letters, 207, 165–179, 2003. Papale, P., Dynamics of magma flow in volcanic conduits with variable fragmentation eff iciency and nonequilibrium pumice degassing, Journal of Geophysical Research, 106, 11043–11066, 2001. Papale, P., A. Neri, and G. Macedonio, The role of magma composition and water content in explosive eruptions I. Conduit ascent
CASHMAN dynamics, Journal of Volcanology and Geothermal Research, 87, 75–93, 1998. Polacci, M., P. Papale, and M. Rosi, Textural heterogeneities in pumices from the climactic eruption of Mount Pinatubo, 15 June 1991, and implications for magma ascent dynamics, Bulletin of Volcanology, 63, 83–97, 2001. Polacci, M., L. Pioli, and M. Rosi, The Plinian phase of the Campanian Ignimbrite eruption (Phlegrean Fields, Italy): evidence from density measurements and textural characterization of pumice, Bulletin of Volcanology, doi 10.1007/s00445-002-0268-4, 2003. Proussevitch, A. A., and D. L. Sahagian, Dynamics and energetics of bubble growth in magmas: analytical formulation and numerical modeling, Journal of Geophysical Research, 103, 18223–18251, 1998. Pyle, D. M. (1989) The thickness, volume and grain size of tephra fall deposits. Bulletin of Volcanology 51, 1–15. Pyle, D. M., Sizes of Volcanic Eruptions, in Encyclopedia of Volcanoes, edited by H. Sigurdsson, pp. 263–269, Academic Press, New York, 2000. Robertson, R.e.a., The explosive eruption of Soufriere Hills Volcano, Montserrat, West Indies, 17 September, 1996, Geophysical Research Letters, 25, 3429–3432, 1998. Roggensack, K., R. L. Hervig, S. B. McKnight, and S. N. Williams, Explosive basaltic volcanism from Cerro Negro Volcano: Influences of volatiles on eruptive style, Science, 277, 1639–1642, 1997. Rosi, M., P. Landi, M. Polacci, A. DiMuro, and D. Zandomeneghi, Role of conduit shear on ascent of the crystal-rich magma feeding the 800-year-BP Plinian eruption of Quilotoa Volcano (Ecuador), Bulletin of Volcanology, DOI:10.1007/s00445-030312-z, 2003. Rubin, A. M., On the thermal viability of dikes leaving magma chambers, Geophysical Research Letters, 20, 257–260, 1993. Rust, A. C., K. V. Cashman, and P. J. Wallace, Magma degassing buffered by vapor flow through brecciated conduit margins, Geology, (in press), 2004. Rust, A. C., and M. Manga, Effects of bubble deformation on the viscosity of dilute suspensions, Journal of nonNewtonian Fluid Mechanics, 104, 53–63, 2002. Rust, A. C., M. Manga, and K. V. Cashman, Determining flow type, shear rate and shear stress in magmas from bubble shapes and orientations, Journal of Volcanology and Geothermal Research, 122, 111–132, 2003. Rutherford, M. D., and J. D. Devine, The May 18, 1980 eruption of Mount St. Helens 3. Stability and chemistry of amphibole in the magma chamber, Journal of Geophysical Research, 93, 11949–11959, 1988. Rutherford, M. J., and J. D. Devine, Pre-eruption pressure-temperature conditions and volatiles in the 1991 dacitic magma of Mount Pinatubo, in Fire and Mud: Eruptions and Lahars of Mount Pinatubo, Phillipines, edited by C. Newhall, and R. Punongbayan, pp. 751–766, University of Washington Press, Seattle, 1996. Rutherford, M. R., and Gardner, J. E., Rates of magma ascent, in Encyclopedia of Volcanoes, edited by H. Sigurdsson, pp. 207–218, Academic Press, New York, 2000.
15
Saar, M. O., and M. Manga, Permeability-porosity relationship in vesicular basalts, Geophysical Research Letters, 26, 111–114, 1999. Sahimi, M., Applications of Percolation Theory, 258 pp., Taylor and Francis, London, 1994. Scaillet, B., B. Clemente, B. W. Evans, and M. Pichavant, Redox control of sulfur degassing in silicic magmas, Journal of Geophysical Research, 103, 23937–23949, 1998. Scandone, R., and S. D. Malone, Magma supply, magma discharge and readjustment of the feeding system of Mount St. Helens during 1980, Journal of Volcanology and Geothermal Research, 23, 239–262, 1985. Schmitt, A. X., Gas-saturated crystallization and degassing in largevolume, crystal-rich dacitic magmas from the Altiplano-Puna, northern Chile, Journal of Geophysical Research, 106, 30561–30578, 2001. Shepherd, E. S., and H. E. Merwin, Gases of the Mt. Pelee lavas of 1902, Journal of Geology, 35, 97–116, 1927. Shinohara, H., and K. Kazahaya, Degassing processes related to magma chamber crystallization, edited by J. Thompson, GACMAC, 1995. Sisson, T., and G. D. Layne, H2O in basalt and basaltic andesite glass inclusions from four subduction-related volcanoes, Earth and Planetary Science Letters, 117, 167–184, 1993. Sisson, T. W., and T. L. Grove, Temperatures and H2O contents of low MgO high-alumina basalts, Contributions to Mineralogy and Petrology, 113, 167–184, 1993. Smith, J. V., Textural evidence for dilatant (shear thickening) rheology of mama at high crystal concentrations, Journal of Volcanology and Geothermal Research, 99, 1–7, 2000. Sparks, R. J. S., The dynamics of bubble formation and growth in magmas: a review and analysis, Journal of Vocanology and Geothermal Research, 3, 1–37, 1978. Sparks, R. S. J., Causes and consequences of pressurization in lava dome eruptions, Earth and Planetary Science Letters, 150, 177–189, 1997. Sparks, R. S. J., and S. R. Young, The eruption of Soufriere hills Volcano, Montserrat (1995–1999): overview of scientific results, Geological Society of London Memoirs, 21, 45–69, 2002. Swanson, D. A., and R. T. Holcomb, Regularities in the growth of the Mount St. Helens dacite dome, 1980–1986, in Lava Flows and Domes, edited by J. H. Fink, pp. 3–24, Springer-Verlag, Berlin, 1990. Tuffen, H., D. B. Dingwell, and H. Pinkerton, Repeated fracture and healing of silicic magma generates flow banding and earthquakes?, Geology, 31, 1089–1092, 2003. Voight, B., et al., Magma flow and cyclic instability at Soufriere Hills volcano, Montserrat, British West Indies, Science, 283, 1138–1142, 1999. Walker, G. P. L., Explosive volcanic eruptions—a new classification scheme, Geologische Rundschau, 62, 431–446, 1973. Wallace P. J. (2003) From mantle to atmosphere: Magma degassing, explosive eruptions, and volcanic volatile budgets. In: Bodnar BJ, DeVivo B (Eds.) Developments in Volcanology, Elsevier Science, pp. 105–127.
16 VOLATILE CONTROLS Wallace, P. J., and A. T. Anderson, Volatiles in Magmas, in Encylopedia of Volcanoes, edited by H. Sigurdsson, pp. 149–170, Academic Press, San Diego, 2000. Wallace, P J., A. T. Anderson, and A. M. Davis, Gradients in H2O, CO2, and exsolved gas in a large-volume silicic magma system: Interpreting the record preserved in melt inclusions from the Bishop Tuff, Journal of Geophysical Research, 104, 20097–20122, 1999. Wallace, P. J., A. T. Anderson Jr., and A. M. Davis, Quantification of pre-eruptive exsolved gas contents in silicic magmas, Nature, 377, 612–616, 1995. White, R. A., A. D. Miller, L. Lynch, and J. A. Power, Observations of hybrid seismic events at Soufriere Hills Volcano, Montserrat: July 1995 to September 1996, Geophysical Research Letter, 25, 3657–3660, 1998. Williams, H., The history and character of volcanic domes, University of California, Bulletin of the Department of Geological Sciences, 21, 51–146, 1932.
Wilson, L., R. S. J. Sparks, and G. P. L. Walker, Explosive volcanic eruptions—IV. The control of magma properties and conduit geometry on eruption column behavior, Geophys J R Astr Soc, 63, 117–148, 1980. Zapata, J. A.et al., SO2 fluxes from Galeras Volcano, Columbia, 1989–1995: Progressive degassing and conduit obstruction of a Decade Volcano, Journal of Volcanology and Geothermal Research, 77, 195–208, 1997. Zhang, Y., A criterion for the fragmentation of bubbly magma based on brittle failure theory, Nature, 402, 648–650, 1999. Zhou, J. Z.Q., T. Fang, G. Luo, and P. H. T. Uhlerr, Yield stress and maximum packing fraction of concentrated suspensions, Rheologica Acta, 34, 544–561, 1995. Katharine V. Cashman, Address TK?, University of Oregon, Eugene, Oregon
