Search

Makushin stratovolcano, Alaska, produced three, highly explosive, andesitic eruptions between ~ 9292 and 6215 yBP. Those eruptions are informally named the CFE (“crater-forming eruption”), Nateekin, and Driftwood Pumice, and they deposited significant tephra fallout in the present-day port of Dutch Harbor and City of Unalaska area. The focus of this study is to examine the geochemistry and petrology of those eruptions to better understand Makushin volcano hazards, andesite petrogenesis and eruption triggering by mafic recharge processes. The CFE, Nateekin, and Driftwood Pumice samples range from basaltic andesite to dacite but are predominantly andesitic (SiO2 = 55.6 to 63.5 wt%). The CFE deposits are slightly compositionally stratified, with the top CFE samples slightly more mafic (55 to 60 wt% SiO2) than the basal deposits (58 to 60 wt% SiO2). Disequilibrium mineral compositions and textures in the CFE, Nateekin, and Driftwood Pumice samples, combined with two pyroxene thermometry and An-rich plagioclase microlites (An80) found only in the top of the CFE deposits, provide evidence for repetitive mafic recharge triggering those eruptions, consistent with prior studies. We compare the Makushin geochemical data with data from select satellite vents and cones in the Makushin Volcanic Field (MVF) from prior studies, to examine possible genetic relationships. The geochemical data and Rhyolite-MELTS models run at crustal storage conditions (2 kbar, fO2 = Ni-NiO, and 1.5 and 3.5 wt% H2O) indicate that no single parental magma supplies the MVF satellite cones and Makushin volcano. Instead, two component mixing models better fit the MVF geochemical array. Our Makushin results compare well with models of predominantly andesitic volcanoes that require mafic recharge to mobilize the andesites and trigger eruptions.

Funding: National Science Foundation grants 1925417, 1023672, and 0909541 to J.R.M., and 1824770 to J.G.M. and F.M.L. funded this research, as well as support to A.I.W. and J.R.M. from the John Fell Oxford University Press (OUP) Research Fund and All Souls College, Oxford. Clare Hall, Cambridge provided additional support to J.R.M. through the Sir Nicholas Shackleton fellowship. F.M.L. also acknowledges support from an Irish Research Council Laureate Award (CLICAB project, IRCLA/2017/303). Swiss National Science Foundation grant 18001 funded C.C.R. and W.M.K. European Research Council grant 820047 under the European Union’s Horizon 2020 research and innovation programme supported M.S. A.B. was supported by Marie Curie Career Integration Grant CIG14-631752. The assassination of Julius Caesar in 44 BCE triggered a power struggle that ultimately ended the Roman Republic and, eventually, the Ptolemaic Kingdom, leading to the rise of the Roman Empire. Climate proxies and written documents indicate that this struggle occurred during a period of unusually inclement weather, famine, and disease in the Mediterranean region; historians have previously speculated that a large volcanic eruption of unknown origin was the most likely cause. Here we show using well-dated volcanic fallout records in six Arctic ice cores that one of the largest volcanic eruptions of the past 2,500 y occurred in early 43 BCE, with distinct geochemistry of tephra deposited during the event identifying the Okmok volcano in Alaska as the source. Climate proxy records show that 43 and 42 BCE were among the coldest years of recent millennia in the Northern Hemisphere at the start of one of the coldest decades. Earth system modeling suggests that radiative forcing from this massive, high-latitude eruption led to pronounced changes in hydroclimate, including seasonal temperatures in specific Mediterranean regions as much as 7 °C below normal during the 2 y period following the eruption and unusually wet conditions. While it is difficult to establish direct causal linkages to thinly documented historical events, the wet and very cold conditions from this massive eruption on the opposite side of Earth probably resulted in crop failures, famine, and disease, exacerbating social unrest and contributing to political realignments throughout the Mediterranean region at this critical juncture of Western civilization. Postprint

During the 2005 summer field season, geologists from the University of Alaska, Fairbanks and the U.S. Geological Survey, Alaska Volcano Observatory (AVO) conducted fieldwork in the western Aleutians, on Little Sitkin Island, Alaska. The primary purpose of the fieldwork was to install geophysical networks for volcano monitoring. As part of this effort, AVO geologists conducted reconnaissance fieldwork focused primarily on sample collection for geochemistry. The analytical data table associated with this report is available in digital format as .csv and is also available in .html and .csv from the AVO Geochemical Database (https://avo.alaska.edu/geochem/). Sample descriptions, locations, and sample types are included in the analytical data table. Files can also be downloaded from the DGGS website (http://doi.org/10.14509/30440).

Bogoslof volcano is a shallow submarine/subaerial volcano in the southern Bering Sea about 100 km west of the community of Dutch Harbor, Alaska. The subaerial parts of the volcano consist of two small islands, Bogoslof Island and Fire Island, that together have a total area of about 1.6 km2. Bogoslof was first depicted on a Russian map in 1772 and since then has been observed and visited occasionally. The volcano has had at least nine periods of eruptive activity since 1796 and all of its historical eruptions have been similar in style. Historical Bogoslof eruptions involved the effusion of basalt, trachybasalt, basaltic trachyandesite, and trachyandesite lava domes with above sea level relief of 100–200 m. Many of the eruptions are accompanied by the formation of tuff rings and ejection of ballistic particles. Historical observations suggest that eruption clouds are relatively ash-poor. Minor ash fallout has typically occurred within about 100 km of the volcano. Many of the historical eruptions began at vents that were below sea level, and thus, seawater has played an important role in the style of eruptive activity exhibited by the volcano. At times, eruptive activity has been characterized by Surtseyan style eruptions and magma interaction with wet vent-fill deposits. At other times, the eruptive style has been more magmatically driven and has resulted in the formation of pyroclastic flows and small ash clouds. Preliminary studies of the deposits produced during the 2016–2017 eruption indicate vertical sequences of coarse-grained, horizontally bedded pyroclastic flow and fall deposits with numerous blocks, bombs, and lapilli of dense juvenile and accidental lithic material. These deposits were emplaced by near-vent pyroclastic flows, surges, and explosions some of which originated from shallow, highly crystalline cryptodomes.

The 2016–2017 eruption of Bogoslof primarily produced crystal-rich amphibole basalts. The dominant juvenile tephra were highly microlitic with diktytaxitic vesicles, and amphiboles had large reaction rims. Both observations support a magma history of slow ascent and/or shallow stalling prior to eruption. Plagioclase-amphibole-clinopyroxene mineralogy are also suggestive of shallow magma crystallization. Lavas were emplaced as shallow submarine lava domes and cryptodomes that produced 70 relatively short-lived and water-rich explosions over the course of the 9-month-long eruption. The explosions ejected older trachyandesite lavas that were likely uplifted by cryptodome emplacement that began in December 2016 and continued for many months. Trachyte pumice, similar in composition to a 1796 lava dome, was entrained in basalts by the end of the eruption. The pumice appears to be a largely crystalline magma that was rejuvenated, entrained in the basalt, and heated to ~ 1000 °C. The composition of trachytes require differentiation through stronger amphibole control than the apparent shallow crustal evolution implied for the basalt. This suggests that they are magmas derived from a mid-crustal zone of amphibole crystallization. Nearby arc-front volcanoes that notably lack amphibole have strikingly similar compositional trends. Trace element signatures of the Bogoslof basalts, however, suggest derivation from a mantle source with residual garnet and lower-degree melting than basalts from nearby arc-front volcanoes. The diversity of magmas erupted at Bogoslof thus provides an opportunity not only to probe rare backarc compositions from the Aleutian arc but also to examine the apparent role of amphibole in generating evolved compositions more broadly in arc environments.

The porosity at which a magma becomes permeable (i.e., the percolation threshold; ϕc) is important for magma degassing; it is also poorly constrained in crystal-bearing systems. To address this, we conduct high pressure-temperature decompression experiments on water-saturated rhyolitic melts with variable crystal contents. We find that crystal-bearing run products become permeable at ~55-vol.% vesicularity (crystal free), a value that is similar to that found in decompression-crystallization experiments using basaltic andesite compositions. Our results provide insight into controls on the eruption styles of hydrous, crystal-bearing magmas in general and controls on pulsatory Vulcanian behavior, in particular.

The magmatic systems feeding arc volcanoes are complex, leading to a rich diversity in eruptive products and eruption styles. This review focuses on examples from the Aleutian subduction zone, encompassed within the state of Alaska, USA because it exhibits a rich diversity in arc structure and tectonics, sediment and volatile influx feeding primary magma generation, crustal magma differentiation processes, with the resulting outcome the production of a complete range in eruption styles from its diverse volcanic centers. Recent and ongoing investigations along the arc reveal controls on magma production that result in diversity of eruptive products, from crystal-rich intermediate andesites to phenocryst-poor, melt-rich silicic and mafic magmas and a spectrum in between. Thus, deep to shallow crustal “processing” of arc magmas likely greatly influences the physical and chemical character of the magmas as they accumulate in the shallow crust, the flow physics of the magmas as they rise in the conduit, and eruption style through differences in degassing kinetics of the bubbly magmas. The broad spectrum of resulting eruption styles thus depends on the bulk magma composition, melt phase composition, and the bubble and crystal content (phenocrysts and/or microlites) of the magma. Those fundamental magma characteristics are in turn largely determined by the crustal differentiation pathway traversed by the magma as a function of tectonic location in the arc, and/or the water content and composition of the primary magmas. The physical and chemical character of the magma, set by the arc differentiation pathway, as it ascends towards eruption determines the kinetic efficiency of degassing versus the increasing internal gas bubble overpressure. The balance between degassing rate and the rate at which gas bubble overpressure builds then determines the conditions of fragmentation, and ultimately eruption intensity.

Eruptions of similar explosivity can have divergent effects on the surroundings due to differences in the behavior of the tephra in the eruption column and atmosphere. Okmok volcano, located on Umnak Island in the eastern Aleutian Islands, erupted explosively between 12 July and 19 August 2008. The basaltic andesitic eruption ejected ∼0.24 km3 dense rock equivalent (DRE) of tephra, primarily directed to the northeast of the vent area. The first 4 h of the eruption produced dominantly coarse-grained tephra, but the following 5 wk of the eruption deposited almost exclusively ash, much of it very fine and deposited as ash pellets and ashy rain and mist. Meteorological storms combined with abundant plume water to efficiently scrub ash from the eruption column, with a rapid decrease in deposit thickness with distance from the vent. Grain-size analysis shows that the modes (although not their relative proportions) are very constant throughout the deposit, implying that the fragmentation mechanisms did not vary much. Grain-shape features consistent with molten fuel-coolant interaction are common. Surface and groundwater drainage into the vents provided the water for phreatomagmatic fragmentation. The available water (water that could reach the vent area during the eruption) was ∼2.8 × 1010 kg, and the erupted magma totaled ∼7 × 1011 kg, which yield an overall water:magma mass ratio of ∼0.04, but much of the water was not interactive. Although magma flux dropped from 1 × 107 kg/s during the initial 4 h to 1.8 × 105 kg/s for the remainder of the eruption, most of the erupted material was ejected during the lower-mass-flux period due to its much greater length, and this tephra was dominantly deposited within 10 km downwind of the vent. This highlights the importance of ash scrubbing in the evaluation of hazards from explosive eruptions.

Magmatic minerals record the pre-eruptive timescales of magma ascent and mixing in crustal reservoirs and conduits. Investigations of the mineral records of magmatic processes are fundamental to our understanding of what controls eruption style, as ascent rates and magma mixing processes are well known to control and/or trigger potentially hazardous explosive eruptions. Thus, amphibole reaction rims are often used to infer pre-eruptive magma dynamics, and in particular to estimate magma ascent rates. However, while several experimental studies have investigated amphibole destabilization during decompression, only two investigated thermal destabilization relevant to magma mixing processes. This study examines amphibole decomposition experimentally through isobaric heating of magnesio-hornblende phenocrysts within a natural high-silica andesite glass. The experiments first equilibrated for 24 h at 870 °C and 140 MPa at H2O-saturated conditions and ƒO2 ∼ Re–ReO prior to rapid heating to 880, 900, or 920 °C and hold times of 3–48 h. At 920 °C, rim thicknesses increased from 17 μm after 3 h, to 55 μm after 12 h, and became pseudomorphs after longer durations. At 900 °C, rim thicknesses increased from 7 μm after 3 h, to 80 μm after 24 h, to pseudomorphs after longer durations. At 880 °C, rim thicknesses increased from 7 μm after 3 h, to 18 μm after 36 h, to pseudomorphs after 48 h. Reaction rim microlites vary from 5–16 μm in size, with no systematic relationship between crystal size and the duration or magnitude of heating. Time-averaged rim microlite growth rates decrease steadily with increasing experimental duration (from 3.97 × 10 − 7 mm s − 1 to 3.1 to 3.5 × 10 − 8 mm s − 1 ). Time-averaged microlite nucleation rates also decrease with increasing experimental duration (from 1.2 × 10 3 mm − 3 s − 1 to 5.3 mm−3 s−1). There is no systematic relationship between time-averaged growth or nucleation rates and the magnitude of the heating step. Ortho- and clinopyroxene together constitute 57–90 modal % mineralogy in each reaction rim. At constant temperature, clinopyroxene abundances decrease with increasing experimental duration, from 72 modal % (3 h at 900 °C) to 0% (48 h at 880 °C, and 36 h at 900 and 920 °C). Fe–Ti oxides increase from 6–12 modal % (after 3–6 h) to 26–34 modal % (after 36–48 h). Plagioclase occurs in relatively minor amounts ( 500 μm from reacted amphibole) are consistent with inter-microlite rim glasses (71.3–77.7 wt.% SiO2) within a given experiment and there is a weakly positive correlation between increasing run duration and inter-microlite melt SiO2 (68.9–78.5 wt.%). Our results indicate that experimental heating-induced amphibole reaction rims have thicknesses, textures, and mineralogies consistent with many of the natural reaction rims seen at arc-andesite volcanoes. They are also texturally consistent with experimental decompression reaction rims. On this basis it may be challenging to distinguish between decompression and heating mechanisms in nature.

The August 7–8, 2008, eruption of Kasatochi Island volcano, located in the central Aleutians Islands, Alaska, produced abundant, compositionally heterogeneous basaltic andesite (52–55 wt% SiO2) that has been interpreted to result from pre-eruptive magma mixing. The basaltic andesite contains two populations of plagioclase phenocrysts. The first, volumetrically dominant population consists of oscillatory-zoned phenocrysts with an overall normal zonation trend toward comparatively sodic rims (An55–65), interrupted by dissolution features and spikes in calcium content (up to ~An85). The second population consists of phenocrysts with highly calcic compositions (~An90). These phenocrysts contain sharp decreases in calcium content close to their rims (reaching as low as ~An60), but are otherwise texturally and compositionally homogeneous. Groundmass plagioclase microlites are generally much more calcic than rims of the first phenocryst population, with more than 50% of measured microlites containing >An80. Major, minor, and trace element concentrations of plagioclase microlites and phenocrysts indicate that oscillatory-zoned phenocrysts derived from cooler (800–950 °C), more silicic mixing magma, while unzoned, calcic phenocrysts were associated with hotter (900–1050 °C), mafic magma. The mixing of these magmas just prior to eruption, followed by decompression during the eruption itself created high effective undercoolings in the mafic end-member, and lead to the nucleation of high-An microlites. MgO and FeO concentrations of plagioclase microlites and high-An phenocryst rims (up to ~0.4 and ~1.3 wt%, respectively) provide further evidence for high mixing- and eruption-induced effective undercoolings.

The Pu‘u Wa‘awa‘a pyroclastic cone and Pu‘u Anahulu lava flow are two prominent monogenetic eruptive features assumed to result from a single eruption during the trachyte-dominated early post-shield stage of Hualālai volcano (Hawaiʻi). Puʻu Wa‘awa‘a is composed of complex repetitions of crudely cross-stratified units rich in dark dense clasts, which reversely grade into coarser pumice-rich units. Pyroclasts from the cone are extremely diverse texturally, ranging from glassy obsidian to vesicular scoria or pumice, in addition to fully crystalline end-members. The >100-m thick Pu‘u Anahulu flow is, in contrast, entirely holocrystalline. Using field observations coupled with whole rock analyses, this study aimed to test whether the Pu‘u Wa‘awa‘a tephra and Pu‘u Anahulu lava flows originated from the same eruption, as had been previously assumed. Crystal and vesicle textures are characterized along with the volatile contents of interstitial glasses to determine the origin of textural variability within Pu‘u Waʻawaʻa trachytes (e.g., magma mixing vs. degassing origin). We find that (1) the two eruptions likely originated from distinct vents and magma reservoirs, despite their proximity and similar age, (2) the textural diversity of pyroclasts forming Pu‘u Wa‘awa‘a can be fully explained by variable magma degassing and outgassing within the conduit, (3) the Pu‘u Wa‘awa‘a cone was constructed during explosions transitional in style between violent Strombolian and Vulcanian, involving the formation of a large cone and with repeated disruption of conduit plugs, but without production of large pyroclastic density currents (PDCs), and (4) the contrasting eruption styles of Hawaiian trachytes (flow-, cone-, and PDC-forming) are probably related to differences in the outgassing capacity of the magmas prior to reaching the surface and not in intrinsic compositional or temperature properties. These results further highlight that trachytes are “kinetically faster” magmas compared to dacites or rhyolites, likely degassing and crystallizing more rapidly.

We conducted melting experiments on a low MgO (3.29 wt.%) basaltic andesite (54.63 wt.% SiO2) from Westdahl volcano, Alaska, at XH2O = 0.7–1 and fO2 ~ Ni–NiO, at pressures = 0.1–180 MPa and temperatures = 900–1,200 °C. We examine the evolution of the melt along a liquid line of descent during equilibrium crystallization at high H2O and fO2 conditions, starting from a high FeOt/MgO, low MgO basaltic andesite. Ti-magnetite formed on the liquidus regardless of XH2O, followed by clinopyroxene, plagioclase, amphibole, and orthopyroxene. We observe slight but significant differences in the phase stability curves between the XH2O = 1 and 0.7 experiments. Early crystallization of Ti-magnetite and suppression of plagioclase at higher pressures and temperatures resulted in strongly decreasing melt FeOt/MgO with increasing SiO2, consistent with a “calc-alkaline” compositional trend, in agreement with prior phase equilibria studies on basalt at similar H2O and fO2. Our study helps quantify the impact of small amounts of CO2 and high fO2 on the evolution of melts formed during crystallization of a low MgO basaltic andesite magma stored at mid- to shallow crustal conditions. Like the prior studies, we conclude that H2O strongly influences melt evolution trends, through stabilization of Ti-magnetite on the liquidus and suppression of plagioclase at high P–T conditions.

Volatile contents of geologic glasses are used to model magma chamber and degassing processes, thus, there is considerable interest in small-scale analytical techniques for analyzing volatiles in glasses. Infrared (IR) spectroscopy has the advantage of determining volatile speciation in glasses (e.g., OH − , molecular H 2 O, molecular CO 2 , and CO 3 2− ). However, sample preparation for the most common IR method used, micro-transmission IR spectroscopy, is complicated because glasses must be prepared as thin, parallel-sided wafers. Raman analysis, while valuable for Fe-poor samples, can be difficult to use for Fe-rich glasses. We have calibrated a micro-reflectance infrared method for determining volatile species using calculated Kramers-Kronig absorbance (KK-Abs.) spectra that requires that only one side of a glass be polished. The method is easier to use than other reflectance methods where it is difficult to determine the baseline for the IR bands. Total H 2 O wt% = m ·(3600 cm −1 KK-Abs.), where m , is the slope of the calibration line that is obtained from a fit to the data. The m value is related to the calculated refractive index, n , for a range of aluminosilicate glass compositions allowing the technique to be applied to samples with unknown calibration slopes. For calc-alkaline andesite glasses we determined calibration slopes for micro-reflectance IR measurements of molecular H 2 O, molecular CO 2 , and CO 3 2− . The method has been calibrated for glasses with up to 6.76 wt% total H 2 O (but is useful for glasses with more than 20 wt% total H 2 O) and has been calibrated for glasses with up to 0.575 wt% total CO 2 . This technique provides a means to analyze volatile abundances in samples that are not possible to analyze or prepare for analysis with transmission micro-IR techniques. We have determined volatile contents in fragile samples such as cracked, vesicular, or crystal-bearing glasses formed by volcanic or impact processes or in high-pressure bubble nucleation experiments and H diffusion experiments. We have monitored H uptake during weathering of basaltic glasses that cannot be polished and determined volatiles in melt inclusions and pumice.

Crystal-rich andesites that erupted from Redoubt Volcano in 2009 range from 57.5 to 62.5 wt.% SiO 2 and have phenocryst and phenocryst-melt relations consistent with staging in the upper crust. Early explosive products are low-silica andesites (LSA, 2 ) that ascended from deeper crustal levels during or before the 6 months of precursory activity, but a broad subsequent succession to more evolved and cooler products, and predominantly effusive dome growth, are interpreted to result from progressive mobilization and mixing with differentiated magmas tapped from pre-2009 Redoubt intrusions at ~ 3–6 km depth. Initial explosions on March 23–28 ejected predominantly LSA with a uniform phenocryst assemblage of high-Al amphibole, ~ An 70 plagioclase, ortho- and clinopyroxene, FeTi oxides (890 to 960 °C), and traces of magmatic sulfide. Melt in the dominant microlite-poor LSA was compositionally uniform dacite (67–68 wt.% SiO 2 ) but ranged to rhyolite with greater microlite growth. Minor amounts of intermediate- to high-silica andesite (ISA, HSA; 59–62.5 wt.% SiO 2 ) also erupted during the early explosions and most carried rhyolitic melt (72–74 wt.% SiO 2 ). A lava dome grew following the initial tephra-producing events but was destroyed by an explosion on April 4. Ejecta from the April 4 explosion consists entirely of ISA and HSA, as does a subsequent lava dome that grew April 4–July 1; LSA was absent. Andesites from the April 4 event and from the final dome had pre-eruptive temperatures of 725–840 °C (FeTi oxides) and highly evolved matrix liquids (77–80 wt.% SiO 2 ), including in rare microlite-free pyroclasts. ISA has mixed populations of phenocrysts suggesting it is a hybrid between HSA and LSA. The last lavas from the 2009 eruption, effused May 1–July 1, are distinctly depleted in P 2 O 5 , consistent with low temperatures and high degrees of crystallization including apatite. Plagioclase–melt hygrometry and comparison to phase equilibrium experiments are consistent with pre-eruptive storage of all three magma types at 100–160 MPa (4–6 km depth), if they were close to H 2 O-saturation, coincident with the locus of shallow syn-eruptive seismicity. Deeper storage would be indicated if the magmas were CO 2 -rich. Relatively coarse-grained clinopyroxene-rich reaction rims on many LSA amphibole phenocrysts may result from slow ascent to, or storage at, depths shallow enough for the onset of appreciable H 2 O exsolution, consistent with pre-eruptive staging in the uppermost crust. We interpret that the 2009 LSA ascended from depth during the 8 or more months prior to the first eruption, but that the magma stalled and accumulated in the upper crust where its phenocryst rim and melt compositions were established. Ascent of LSA through stagnant mushy intrusions residual from earlier Redoubt activity mobilized differentiated magma pockets and interstitial liquids represented by HSA, and as LSA–HSA hybrids represented by ISA, that fed the subsequently erupted lava domes.

Permeability development in magmas controls gas escape and, as a consequence, modulates eruptive activity. To date, there are few experimental controls on bubble growth and permeability development, particularly in low viscosity melts. To address this knowledge gap, we have run controlled decompression experiments on crystal-free rhyolite (76 wt.% SiO2), rhyodacite (70 wt.% SiO2), K-phonolite (55 wt.% SiO2) and basaltic andesite (54 wt.% SiO2) melts. This suite of experiments allows us to examine controls on the critical porosity at which vesiculating melts become permeable. As starting materials we used both fine powders and solid slabs of pumice, obsidian and annealed starting materials with viscosities of ∼ 10 2 to ∼ 10 6 Pa s . We saturated the experiments with water at 900° (rhyolite, rhyodacite, and phonolite) and 1025 °C (basaltic andesite) at 150 MPa for 2–72 hrs and decompressed samples isothermally to final pressures of 125 to 10 MPa at rates of 0.25–4.11 MPa/s. Sample porosity was calculated from reflected light images of polished charges and permeability was measured using a bench-top gas permeameter and application of the Forchheimer equation to estimate both viscous ( k 1 ) and inertial ( k 2 ) permeabilities. Degassing conditions were assessed by measuring dissolved water contents using micro-Fourier-Transform Infrared (μ-FTIR) techniques. All experiment charges are impermeable below a critical porosity ( ϕ c ) that varies among melt compositions. For experiments decompressed at 0.25 MPa/s, we find the percolation threshold for rhyolite is 68.3 ± 2.2 vol.% ; for rhyodacite is 77.3 ± 3.8 vol.% ; and for K-phonolite is 75.6 ± 1.9 vol.% . Rhyolite decompressed at 3–4 MPa/s has a percolation threshold of 74 ± 1.8 vol.% . These results are similar to previous experiments on silicic melts and to high permeability thresholds inferred for silicic pumice. All basaltic andesite melts decompressed at 0.25 MPa/s, in contrast, have permeabilities below the detection limit ( ∼ 10 − 15 m 2 ), and a maximum porosity of 63 vol.%. Additionally, although the measured porosities of basaltic andesite experiments are ∼10–35 vol.% lower than calculated equilibrium porosities, μ-FTIR analyses confirm the basaltic andesite melts remained in equilibrium during degassing. We show that the low porosities and permeabilities are a consequence of short melt relaxation timescales during syn- and post-decompression degassing. Our results suggest that basaltic andesite melts reached ϕ c > 63 vol.% and subsequently degassed; loss of internal bubble pressure caused the bubbles to shrink and their connecting apertures to seal before quench, closing the connected pathways between bubbles. Our results challenge the hypothesis that low viscosity melts have a permeability threshold of ∼30 vol.%, and instead support the high permeability thresholds observed in analogue experiments on low viscosity materials. Importantly, however, these low viscosity melts are unable to maintain high porosities once the percolation threshold is exceeded because of rapid outgassing and collapse of the permeable network. We conclude, therefore, that melt viscosity has little effect on percolation threshold development, but does influence outgassing.

The origin of tholeiitic (TH) versus calc-alkaline (CA) magmatic trends has long been debated. Part of the problem stems from the lack of a quantitative measure for the way in which a magma evolves. Recognizing that the salient feature in manyTH^CA discrimination diagrams is enrichment in Fe during magma evolution, we have developed a quantitative index of Fe enrichment, the Tholeiitic Index (THI):THI1⁄4Fe4·0/Fe8·0, where Fe4·0 is the average FeO* concentration of samples with 4 1wt % MgO, and Fe8·0 is the average FeO* at 8 1wt % MgO. Magmas withTHI41 have enriched in FeO* during differentiation from basalts to andesites and are tholeiitic; magmas with THI51 are calc-alkaline. Most subduction zone volcanism is CA, but to varying extents; the THI expresses the continuum of Fe enrichment observed in magmatic suites in all tectonic settings.To test various controls on the development of CA trends, we present new magmatic water measurements in melt inclusions from eight volcanoes from the Aleutian volcanic arc (Augustine, Emmons, Shishaldin, Akutan, Unalaska, Okmok, Seguam, and Korovin). Least degassed H2O contents vary from 2 wt % (Shishaldin) to47 wt % (Augustine), spanning the global range in arc mafic magmas.Within the Aleutian data, H2O correlates negatively with THI, from strongly calc-alkaline (Augustine, THI1⁄4 0·65) to moderately tholeiitic (Shishaldin, THI1⁄41·16). The relationship between THI and magmatic water is maintained when data are included from additional arc volcanoes, back-arc basins, ocean islands, and mid-ocean ridge basalts (MORBs), supporting a dominant role of magmatic water in generating CA trends. An effective break between TH and CA trends occurs at 2 wt % H2O. Both pMELTs calculations and laboratory experiments demonstrate that the observed co-variation of H2O and THI in arcs can be generated by the effect of H2O on the

Experimental bubble nucleation studies are used for determining the nucleation mechanism as a function of experimental conditions, the resulting bubble number density, and can also yield estimates of the melt-vapor surface tension. This provides important information on gas exsolution in silicate melts, which can be applied towards understanding magmatic degassing in volcanic conduits. At present, determination of nucleation processes in tiny experimental samples relies upon visual observations. To improve the characterization of the spatial distribution of bubbles, we present a new application of spatial point pattern analysis. This technique allows the quantitative description of the spatial distribution of nucleation sites and has the potential to distinguish between homogeneous, heterogeneous, and multiple nucleation events. Since point pattern analysis highlights clustering or spatial regularity among objects, it may improve our understanding of the melt structure underlying the spatial distribution of nucleation sites, as well as interactions between bubble populations resulting from different nucleation pulses within a single experimental sample.

The Middle Scoria deposit represents an explosive eruption of basaltic andesite magma (54 wt. % SiO2) from Okmok volcano during mid-Holocene time. The pattern of dispersal and characteristics of the ejecta indicate that the eruption opened explosively, with ash textural evidence for a limited degree of phreatomagmatism. The second phase of the eruption produced thick vesicular scoria deposits with grain texture, size and dispersal characteristics that indicate it was violent strombolian to subplinian in style. The third eruptive phase produced deposits with a shift towards grain shapes that are dense, blocky, and poorly vesicular, and intermittent surge layers, indicating later transitions between magmatic (violent strombolian) to phreatomagmatic (vulcanian) eruptive styles. Isopach maps yield bulk volume estimates that range from 0.06 to 0.43 km3, with ~ 0.04 to 0.25 km3 total DRE. The associated column heights and mass discharge values calculated from isopleth maps of individual Middle Scoria layers are 8.5 – 14 km and 0.4 to 45 × 106 kg/s. The Middle Scoria tephras are enriched in plagioclase microlites that have the textural characteristics of rapid magma ascent and relatively high degrees of effective undercooling. Those textures probably reflect the rapid magma ascent accompanying the violent strombolian and subplinian phases of the eruption. In the later stages of the eruption, the plagioclase microlite number densities decrease and textures include more tabular plagioclase, indicating a slowing of the ascent rate. The findings on the Middle Scoria are consistent with other explosive mafic eruptions, and show that outside of the two large caldera-forming eruptions, Okmok is also capable of producing violent mafic eruptions, marked by varying degrees of phreatomagmatism.

article Crystals in volcanic rocks are sensitive records of magma chamber and conduit conditions under volcanoes. Plagioclase is an invaluable tool to identify ascent rates for calc-alkaline magmas, but may be absent in alkaline melts. In contrast, leucite is common in alkaline magmas and is potentially useful to investigate storage and ascent conditions prior to volcanic eruptions. Leucite microphenocrysts are ubiquitous within the products from all phases of the 79AD eruption of Vesuvius. Steady-state (isobaric-isothermal) and dynamic (decompression) experiments on white phonolitic pumice from the opening (EU1) and lower Plinian (EU2) phases of the eruption were performed at temperature conditions ranging from 800 to 850 °C to test the possibility that leucites within this 'white' magma formed during ascent. However, multiple-step

This study focuses on constraining bubble nucleation and H 2 O exsolution processes in alkalic K-phonolite melts, using “white pumice” of the 79 AD eruption of Vesuvius as starting material. The first set of experiments consisted of H 2 O solubility runs at 1153 to 1250 K and pressures between 50 and 200 MPa, to constrain equilibrium water concentrations along the decompression pathways. The decompression experiments were equilibrated with H 2 O at 150 MPa and 1173 and 1223 K, and then decompressed at 3 to 17 MPa/s before rapid quenching. Experiments nucleated bubbles within the first 50 MPa pressure drop, producing maximum bubble number densities (N V ), corrected to melt volume, of 3.8 × 10 14 m − 3 at 1173 K and 4.3 × 10 13 m − 3 at 1223 K. Most bubbles were not visibly attached to crystals, except for a subset attached to pyroxenes primarily in the 1173 K experiments. When compared with prior bubble nucleation studies, the reduced nucleation ΔP and relatively low N V observed indicate predominantly a heterogeneous nucleation mechanism. Melt–vapor–crystal wetting angles measured in 1173 K experiments from bubbles attached to pyroxene crystals are 36 to 69°, which are similar to those measured on titanomagnetite crystals in calc-alkaline dacite melts. The 1223 K experiments have porosities and water concentrations that largely track equilibrium, despite the rapid decompression rate. The 1173 K experiments deviate strongly from equilibrium trends in both porosity and water concentration, and slower H 2 O diffusion rates are likely the cause of the inhibited bubble growth. Bubble number densities from 79 AD Vesuvius natural EU2 pumice are relatively high (2 to 4 × 10 15 m − 3 ; [Gurioli, L., Houghton, B.F., Cashman, K.V., Cioni, R., 2005. Complex changes in eruption dynamics during the 79 AD eruption of Vesuvius. Bull. Volcanol. 67: 144–159.]) when corrected to vesicularity. In comparison, corrected N V 's from homogeneous and heterogeneous bubble nucleation experiments from this study and prior work are at least factor of 5 lower, indicating perhaps that the natural magmas initially nucleated bubbles in the presence of CO 2 . The disequilibrium H 2 O exsolution seen in the 1173 K experiments indicates that inhibited bubble growth could lead to delayed exsolution in the conduit in cooler K-phonolite magmas.

Core samples from the conduit of Unzen Volcano, obtained only 9 years after cessation of the 1991–1995 eruption, exhibit important differences in physical characteristics and mineralogy, and subtle differences in bulk chemistry from erupted samples. These differences in the conduit samples reflect emplacement under a confining pressure where about half of the original magmatic water was retained in the melt phase, maintenance at hypersolidus temperature for some unknown but significant time span, and subsequent subsolidus hydrothermal alteration. In contrast, magma that extruded as lava underwent decompression to 1 atm with nearly complete loss of magmatic water and cooling at a sufficiently rapid rate to produce glass. The resulting hypabyssal texture of the conduit samples, while clearly distinct from eruptive rocks, is also distinct from plutonic suites. Given the already low temperature of the conduit (less than 200 °C, [Nakada, S., Uto, K., Yoshimoto, M., Eichelberger, J.C., Shimizu, H., 2005. Scientific Results of Conduit Drilling in the Unzen Scientific Drilling Project (USDP), Sci. Drill., 1, 18–22]) when it was sampled by drilling, this texture must have developed within a decade, and perhaps within a much shorter time, after emplacement. The fact that all trace-element concentrations of the conduit and the last-emplaced lava of the spine, 1300 m above it, are identical to within analytical uncertainty provides strong evidence that both were produced during the same eruption sequence. Changes in conduit magma that occurred between emplacement and cooling to the solidus were collapse of vesicles from less than or equal to the equilibrium value of about 50 vol.% to about 0.1 vol.%; continued resorption of quartz and reaction of biotite phenocrysts due to heating of magma prior to ascent by intruding mafic magma; breakdown of hornblende; and micro-crystallization of rhyolitic melt to feldspar and quartz. Subsolidus changes were deposition of calcite and pyrite, growth of sericite in anorthite-rich zones of plagioclase, and development of montmorillonite as an alteration product. Significant changes in bulk composition were depletion of Mg, Fe and Na and enrichment in C and S. These changes were due mainly to the breakdown of hornblende and plagioclase, and addition of carbonate and pyrite, respectively. The identical concentrations of REEs in the conduit and surface lava are consistent with low water to rock ratios during alteration. This suggests to us that despite convective hydrothermal removal of heat from the conduit, chemical open-system effects were limited to early loss of magmatic water and later addition of magmatic CO2 and SO2 and/or H2S streaming up the conduit from deeper levels.

We performed three sets of decompression experiments on a hydrated K-phonolite melt at 880 °C in order to study the exsolution and degassing behavior of this low viscosity melt. The range of decompression rates studied was 0.01 MPa/s to 0.25 MPa/s. The pressure range was 200 MPa to 10 MPa. We determined that the melt exsolved water in equilibrium at all pressures at the decompression rates studied using micro-reflectance FTIR, a new method which is particularly well suited to determining the dissolved water concentration in high porosity experimental samples. Below 40–50 MPa at all decompression rates, the samples exsolved water in equilibrium, but the sample porosities deviated from values derived from equilibrium calculations/experiments and instead porosities remained constant or even decreased at lower pressures. The bubbles in all samples were predominantly deformed with features characteristic of partially collapsed bubble textures apparent at low pressures. Analysis of the bubble size distributions and deformation parameters indicate that melt shearing due to bubble growth and attendant sample expansion may cause the bubbles to deform, leading to the formation of bubble chains. The deformation and alignment of bubbles increases connectivity and lowers the porosity at which extensive permeability develops. In this situation, many of the connected bubbles are in contact with external vapor leading to significant degassing at 40–50 MPa. This degassing reduces the sample porosity and creates collapsed bubble textures at low pressures. This behavior differs significantly from that of hydrated rhyolite melts.

Okmok volcano is situated on oceanic crust in the central Aleutian arc and experienced large ( 15 km) caldera-forming eruptions at 12 000 years BP and 2050 years BP. Each caldera-forming eruption began with a small Plinian rhyodacite event followed by the emplacement of a dominantly andesitic ash-flow unit, whereas effusive interandpost-caldera lavashavebeenmorebasaltic.Phenocryst assemblages are composed ofolivineþ pyroxeneþ plagioclase Fe^Ti oxides and indicate crystallization at1000^11008Cat 0 1^0 2GPa in thepresence of 0^4% H2O.The erupted products follow a tholeiitic evolutionary trend and calculated liquid compositions range from 52 to 68 wt % SiO2 with 0 8^3 3 wt%K2O.Major and trace element models suggest that the more evolved magmas were produced by 50^60% in situ fractional crystallization around the margins of the shallow magma chamber. Oxygen and strontium isotope data ( O 4 4^4 9o, Sr/ Sr 0 7032^0 7034) indicate interaction with a hydrothermally altered crustal component, which led to elevated thorium isotope ratios in some caldera-forming magmas. This compromises the use of uranium^thorium disequilibria [(Th/U)1⁄4 0 849^0 964] to constrain the time scales of magma differentiation but instead suggests that the age of the hydrothermal system is 100 ka. Modelling of the diffusion ofstrontiuminplagioclase indicates thatmanyevolved crystal rims formed less than 200 years prior to eruption.This addition of rim material probably reflects the remobilization of crystals from the chamber margins following replenishment. Basaltic recharge led to the expansion of the magma chamber, which was responsible for the most recent caldera-forming event.

Volcanoes can erupt explosively in Plinian style or effusively as lava flows. Most models for such ranges in activity are based on silicic magma, which may not be appropriate for less viscous basic magma. Although basic magma erupting at Arenal volcano has not varied significantly in bulk composition, the volcano has exhibited a full range in eruptive style, from Plinian activity in 1968 to Strombolian bursts to lava flows. We examined groundmass textures of samples erupted over that range of activity to investigate the controls on the variability. Microlite textures in lavas collected both hot (rapid quenched) and cold show that most samples have textures that are overprinted by crystallization as a result of cooling. Despite that overprint, microlites in the Plinian sample have unique crystal morphologies and vesicles that are much smaller and more spherical than those in the other samples. We interpret those differences as recording a change in degassing style as a result of changing ascent rate in the conduit. To constrain the potential changes in ascent rate, a limited number of decompression experiments were run at rates from 0.0013 to 0.25 MPa/s. Crystal textures and morphologies vary greatly as decompression rates change, and compared to our observed differences in the natural groundmass, it appears that magma erupted in the Plinian event decompressed between 0.0013 and 0.025 MPa/s, whereas magma erupted in non-Plinian events decompressed slower than 0.0013 MPa/s. The change in eruptive style from explosive Plinian to lava effusion thus resulted from an order of magnitude decrease in magma ascent rate. Plinian magma probably rose too quickly to allow bubbles to coalesce and allow the magma to degas efficiently, whereas at other times magma rose more slowly, which allowed bubbles to coalesce and gas to escape leading to less explosive activity.

We experimentally studied the dacitic magma ejected during the first event in the Usu 2000 eruption to investigate the conditions of syneruptive magmatic ascent. Geophysical data revealed that the magma reached under West Nishiyama, the location of the event’s craters, after rising beneath the summit. Prior study of bubble-size distributions of ejecta shows two stages (stage 1 and stage 2) with different magma ascent rates, as the magma accelerated beneath West Nishiyama with the start of the second stage. To simulate ascent of stage 1 from the main reservoir, which was located at a depth of 4–6 km (125 MPa) to 2 km (50 MPa) beneath West Nishiyama, decompression experiments were conducted isothermally at 900°C following two paths. Single step decompression (SSD) samples were decompressed rapidly (0.67 MPa/s) to their final pressure and held for 12 to 144 hours. Multiple step decompression (MSD) samples were decompressed stepwise to their final pressure and quenched instantly. In MSD, the average decompression rates and total experimental durations varied between 0.01389 to 0.00015 MPa/s and 1.5 to 144 hours, respectively. Syneruptive crystallization was confined to stage 1, and the conditions of ascent were determined by documenting similarities in decompression-induced crystallization between ejecta and experiments. Core compositions, number densities, and shapes of experimental microlites indicate that ascent to 2 km depth occurred in less than 1.5 h. Volumes and number densities of experimental microlites from the SSD experiments that best replicate the decompression rate to 2 km indicate that the magma remained at 2 km for approximately 24 h before the eruption. Stagnation at a depth of 2 km corresponds with horizontal transport through a dike from beneath the summit to West Nishiyama, according to geodetic results. The total magma transport timescale including stage 2 is tens of hours and is shorter than the timescale of precursory seismicity (3.5 days), indicating that the erupted magma did not move out of the reservoir for the first 2 days. This is consistent with the temporal change in numbers of earthquakes, which reached a peak after 2 days.

This study presents a pre-eruptive magma storage model for the rhyodacite and andesite magmas erupted during the 3430 yBP caldera-forming eruption of Aniakchak volcano, Alaska, derived from phase equilibria experiments and petrological data. The compositions of Fe–Ti oxide pairs from the early erupted Plinian rhyodacite pumice yield core temperatures of 871–900°C, with rims up to ∼942°C, and fO2 from −10.6 to −11.8 log units. Melt inclusions entrapped in plagioclase phenocrysts have H2O contents between 3 and 5 wt%, estimated by FTIR and electron microprobe volatiles by difference methods, with no detectable CO2. Assuming water saturation, this corresponds to entrapment pressures between ∼65 and 150 MPa. Phase equilibria results reproduce the natural phase assemblages at $$P_{\text H_{2}\text O}$$ of 95–150 MPa at 870–880°C, assuming water saturation. A mismatch in experimental versus natural glass SiO2 and Al2O3, and MELTS models for H2O-undersaturated conditions indicate that the rhyodacite may not have been H2O saturated. MELTS models with $$ X_{\text{H}_2\text{O}} =0.8$$ and P total of 125–150 MPa at 870–880°C reproduce the natural groundmass glass Al2O3 composition best, indicating the magma may have been slightly H2O undersaturated. Those pressures correspond to storage at 4.5–5.4 km depth in the crust. MELTS models and VBD estimates from melt inclusions in titanomagnetite grains from the andesite indicate pre-eruptive conditions of ∼1,000°C and > 110 MPa, corresponding to a minimum residence depth of ∼4.1 km assuming water saturation or greater if the magma was H2O undersaturated. Previous geochemical studies indicate separate histories of the two magmas, though they retain some evidence that they are ultimately related through fractional crystallization processes. Analogous to the 1912 Novarupta magmas, the rhyodacite and andesite presumably originated within the same crystal mush zone beneath the edifice, yet were separated laterally and underwent different degrees of crustal assimilation. The andesite must have resided in close proximity, with ascent occurring in response to movement of the rhyodacite, and resulting in extensive syn-eruptive mingling.

Aniakchak caldera, Alaska, produced a compositionally heterogeneous ignimbrite 3400 years ago, which changes from rhyodacitic at the base to andesitic at the top of the eruptive sequence. Interpretations of compositionally heterogeneous ignimbrites typically include either in situ fractional crystallization of mafic magma and generation of a stratified magma body or replenishment of a silicic magma chamber by mafic inputs. Another possibility, silicic replenishment of a more mafic chamber, exists. Geochemical characteristics of the caldera-forming rhyodacite and several late precaldera rhyodacites indicate independent origins for each, within a maximum of 5000 years prior to caldera formation. Isotopic considerations preclude derivation of the caldera-forming rhyodacite from the caldera-forming andesite. However, the caldera-forming rhyodacite can be explained as the residual liquid of a mostly crystallized basalt, with addition of crustal material. The Aniakchak andesite probably formed in a shallow chamber by successive mixing events involving small volumes of basalt and rhyodacite, together with contamination. The pre-caldera rhyodacites represent erupted portions of intruding silicic magma, whereas another portion homogenized with the resident mafic magma. The caldera-forming event reflects a large influx of rhyodacite, which erupted before significant mixing occurred and also triggered draining of much of the andesitic magma from the chamber.

The purpose of this study was to replicate experimentally the growth of new rims around highly anorthitic plagioclase "core" phenocrysts, analogous to the incorporation of xenocrysts into a silicic magma body through magma mixing. Aniakchak rhyodacite forms the bulk starting composition, and phase-equilibria experiments constrained the pre-eruption magma conditions to be ∼110 MPa and 870-880 °C. The experimental runs were seeded with Great Sitkin anorthite (An 9 1 - 9 5 ) crystals. New rim growth of An 2 8 - 3 8 plagioclase occurred at rates between 3.5 (′0.3) x 10 - 1 0 to 60.6 (′20.0) x 10 - 1 0 cm/s at pressures and temperatures from 50 to 150 MPa and 825 to 880 °C. The values in parentheses are ′1σ standard deviation. Microlite crystallization (An 2 7 - 4 1 ) occurred in all experiments within the plagioclase stability field, and their growth rates varied from 4.4 (′1.3) x 10 - 1 0 to 65.7 (′10.1) x 10 - 1 0 cm/s. The rim and microlite growth rates are similar to one another within each experiment, and microlite number density (N v ) is correlated approximately inversely with rim growth rates. Microlite crystallinities increased from 4.2 to 49.7 vol% as a function of increasing ΔT e f f up to 95 °C. The results indicate growth-dominated crystallization at low ΔT e f f , and nucleation dominated crystallization at high ΔT e f f , in agreement with previous studies. Assuming the experiments apply to nature, the rim growth rates can provide a minimum estimate on how fast magma mixing can occur. Rims that are 10 to 100 μm wide can grow in ∼10 days to 4 months, recording fast mixing timescales as long as eruption occurs shortly after mixing. The growth rate estimates presented here generally agree with those derived from sodic rims growing around anorthite cores after mixing between basalt and andesite prior to the 1996 eruption of Karymsky volcano, Kamchatka.

We have experimentally studied degassing of phonolite melts between 50 and 200 MPa and 825 and 850 °C. Our results indicate that decompressed phonolite melts contain equilibrium water contents regardless of decompression rate or final pressure. In addition, the total porosities of our experiments agree with predicted equilibrium values, regardless of final pressure or decompression rate. Combined observations of mean bubble diameters, porosities, and arrangement of bubbles in our experiments indicate that degassing occurs preferentially in the large bubble fraction of the population. Because of this, retardation of growth in the small bubble population occurs. This may lead to bimodal bubble size distributions in completely degassed phonolites. Our results indicate that water saturated phonolites at temperatures >800 °C can degas in equilibrium, while rhyolites follow a disequilibrium trend, at decompression rates of ∼0.5–1.0 MPa/s. Thus, hotter, water-saturated phonolitic magmas are likely to maintain equilibrium magma water contents during ascent at rates up to 20 m/s, which are typical of highly explosive eruptions. These results may apply to eruptions such as the 79 AD eruption of Vesuvius, which is the type location for Plinian-style, explosive eruptions. However, caution must be used when applying these results to lower temperature, more differentiated, and water-undersaturated phonolite magmas. If fragmentation can occur at depths of ∼2 km or greater in the conduit, our results indicate that highly explosive eruptions of phonolitic magma may occur in the absence of water oversaturation in the magma during ascent. Fast ascent rates and high shear stresses in the conduit may provide a better explanation for fragmentation in this case.

We have experimentally studied the process of bubble coalescence in rhyolite and phonolite melts of natural composition. The experiments involved decompression of water-saturated melts equilibrated at pressures and temperatures from 100 to 150 MPa and 775 to 840 °C in vertically oriented, rapid-quench capable, cold seal pressure vessels. One type of experiments (rhyolite MCR-100, 120, 150 and phonolite LSP-120 series') approximates a “static” bubble coalescence case, where we held the decompressed samples for ∼5 seconds to 4320 minutes (3 days) before quenching. The second type (rhyolite LPC-100 series) replicates an “expanding” bubble coalescence environment, where we continually decompressed the experiments at a rate of 0.5 MPa/s, examining samples quenched at ending pressures between 10 and 80 MPa. Our “static” case (MCR-100, 120, and 150, and LSP-120) results show significant increases in the modal bubble sizes and in the sizes of the largest bubbles, corresponding to measurable broadening in the size distributions. Their bubble number densities (NV) decrease as a function of hold time at their final pressures (PF), and can be fit well by power law functions. Our “expanding” case experiments (LPC-100) show a significant drop in NV during the duration of the experiments that can be fit by an exponential equation as NV vs. either time or PF. Average estimates of bulk coalescence rates indicate a ∼1 order of magnitude drop in NV for “static” case rhyolites in a 2–3 day period, and ∼2 orders of magnitude for phonolites within a 3 day period. Despite a ∼2 order of magnitude difference in viscosity, coalescene in the phonolite is only slightly faster than the rhyolite. The “expanding” case experiments show a ∼1 order of magnitude drop in NV over 180 seconds. Thus, NV's decrease 4 orders of magnitude faster in expanding vs. static bubbly rhyolite melts. Our results imply that significant bubble coalescence can occur in rhyolite magmas at relatively fast (∼20 m/s) ascent rates in the conduit. Thus, bubble interconnectivity, leading to high permeability, is possible during ascent. Bubble coalescence may occur during second boiling in magma bodies that are stalled in the crust. The timescales over which this occurs is much faster than the estimated rise rates for bubbles to reach the top of the magma chamber.

We have studied interactions between bubbles of two distinct size classes in rhyolite melts experimentally decompressed between 200 and 80 MPa. The first set of ‘decompression’ bubbles has a size range (Rdec) of 1–11 μm and is formed from nucleation and growth upon isothermal decompression of the melt. The larger populations of ‘hydration’ bubbles are on average 30–40 μm in radius (Rhyd) and are formed from pore spaces present that were filled with water vapor during the saturation runs prior to the decompression experiments. The first type of interaction results in the elongation of decompression bubbles oriented radially around the larger hydration bubbles. The degree of elongation increases both as a function of distance and with increasing ratio of hydration to decompression bubble size (Rhyd/Rdec). The second type of interaction studied results in a reduction of the size of decompression bubbles located within a range of distances from 10 to 65 μm from a hydration bubble surface, relative to the modal size of the unaffected bubbles in the same sample. In addition, within an average distance of 10 μm, melt next to the hydration bubble surface is depleted in decompression bubbles. Our results indicate that concentration gradients in the melt are probably responsible for bubble size reduction and the depleted zones, because the predicted time scales for Ostwald ripening are much longer than those of the experiments. These effects persist even to the lowest ending pressures studied (80 MPa), which indicates that size distributions of small bubbles may be affected by concentration gradients in the depleted melt shell surrounding large bubbles. Large bubbles present in an ascending magma, prior to a subsequent nucleation event, could therefore affect the growth of the smaller bubble population occurring within the depleted melt shell of the larger bubbles, and produce a bimodal vesicle size distribution. Elongated decompression bubbles may be strained as a result of melt flowing away from the much larger hydration bubbles as they grow. Estimates of capillary number (Ca) plotted against deformation (Df) indicate that bubbles in water-rich rhyolite melts are deformable, even at small sizes (1 μm) and small values of Ca. Our results show a different trend of Df with Ca than previous studies in non-geological systems predict, indicating that viscosity effects may be important. The preservation of deformation textures depends strongly on relaxation time, explaining the lack of deformation textures in less viscous natural lavas.

International audience; Vesicle populations in volcanic pumice provide a partial record of shallow magma ascent and degassing. Here we compare pumice textures from the well-characterized 79AD Vesuvius eruption to those generated during isothermal decompression experiments. Three series of experiments were conducted using starting material from the first two phases of the eruption (eruptive units EU1 and EU2). Samples were decompressed from 100 or 150 MPa to final pressures of 10–25 MPa using conditions appropriate for simulating eruption conditions (T = 850 °C, dP/dT = 0.25 MPa/s). The experiments differed not only in starting material but also in temperature at which samples were annealed prior to decompression, which determined the initial number of crystals present in the melt. Results show that experiments approach the vesicle number densities and sizes of pumice samples, but show narrower size distributions. The wider size range of pumice samples suggests continuous, rather than instantaneous nucleation, which may reflect non-linear rates of decompression. All experiments exhibited equilibrium degassing, a process that was probably aided by heterogeneous bubble nucleation on oxide microlites. We conclude that delayed bubble nucleation cannot explain the explosivity of the Vesuvius eruption, which instead appears to require high rates of magma decompression.

Okmok volcano, in the central Aleutian arc, Alaska, produced two caldera-forming eruptions within the last ∼12,000 years. This study describes the stratigraphy, composition, and petrology of those two eruptions. Both eruptions initially produced small volumes of felsic magmas, followed by voluminous andesite and basaltic andesite. The Okmok I eruption produced >30 km 3 DRE of material on Umnak Island, and Okmok II ∼15 km 3 . However, a significant proportion of material not accounted for here was deposited into the oceans during both events. The Okmok I pyroclastic flow deposits contain evidence for interaction with snow/ice, particularly along the northern flanks of the caldera. Although both Okmok I and II eruptions involved a phreatomagmatic component, the accumulation of a large volume (>15km 3 ) of volatile-rich, mafic-intermediate magma in the shallow crust may provide the driving force for the catastrophic eruptions. Agglutinate deposits associated with Okmok II indicate energetic lava fountaining simultaneous with caldera-collapse, similar to other descriptions of mafic-intermediate caldera-forming deposits such as in the New Hebrides.

In the late morning of 12 July 2008, the Alaska Volcano Observatory (AVO) received an unexpected call from the U.S. Coast Guard, reporting an explosive volcanic eruption in the central Aleutians in the vicinity of Okmok volcano, a relatively young (∼2000-year-old) caldera. The Coast Guard had received an emergency call requesting assistance from a family living at a cattle ranch on the flanks of the volcano, who reported loud “thunder,” lightning, and noontime darkness due to ashfall. AVO staff immediately confirmed the report by observing a strong eruption signal recorded on the Okmok seismic network and the presence of a large dark ash cloud above Okmok in satellite imagery. Within 5 minutes of the call, AVO declared the volcano at aviation code red, signifying that a highly explosive, ash-rich eruption was under way.

The North Pacific is ringed by a convergent boundary that is segmented into active volcanic arcs that stretch from Japan through the Kuriles and Kamchatka to the Aleutians and mainland Alaska. The abundant volcanic eruptions and earthquakes and the complex tectonic problems offer an unparalleled opportunity to study arc processes. The North Pacific subduction zones are large, spanning three nations, and the solid Earth problems are widely varied. Studying arc processes in this region requires strong international collaborations across the different scientific disciplines, and a recent international workshop generated recommendations that will facilitate such research. International student and researcher exchange programs, accessible archives of geodetic data, and identification of key scientific problems will help remove the barriers to research imposed by international boundaries and the size and remoteness of the region.