Search

Deciphering the architecture of the plumbing system beneath active volcanoes and the pre-eruptive magma dynamic is of key importance to discuss about the eruptive style and petrological warning signal. Here Dominica in the Lesser Antilles arc is chosen as a key case study to illustrate the scientific approach. The first step is a field study in order to well identify the characteristics of the deposits to well constrain in particular the style, number and age of eruptions, and to well sample them for petrological studies. On Dominica, we proposed a new chronostratigraphy of the explosive eruptions along the island in the last 50 kyrs, based on stratigraphic correlations, lithology, 14C dating and glass chemistry. To reconstruct the magma plumbing architecture, melt inclusions composition entrapped in minerals from the key eruptions were investigated. Their volatile content highlights that two magma ponding zones may be identified: a deepest one located at ~ 4 kbars, giving birth to the large pumiceous eruptions, formely recognized as the ignimbritic eruptions, and a shallowest, at 1.5-2 kbars, one leading to Plinian-type eruptions. These results are similar to those obtained by experimental petrology. Such data allows us to conclude that the magma plumbing system is organized as a transcrustal magma system, with the magma ponding zones linked to the structure of the crust (lower and middle crust), as derived from geophysical studies. Melt inclusions also help us to constrain magma source at depth, and the influence of slab-derived fluid respect to sediment melting in a subduction zone. To go further, the petrological processes occuring within the magma reservoirs may be investigated. Tracking the pre-eruptive history of magma storage and ascent is a key challenge of modern volcanology, in particular to gain insight into the timescale of pre-eruptive processes at active volcanoes. By focusing now on crystal composition, the pre-eruptive magma dynamics may be identified adopting the « Crystal System Analysis » approach, and their timescales estimated using the diffusion chronometry (Fe-Mg interdiffusion in orthopyroxens). On active and monitored volcanoes, the correlation between the petrological clock and the pre-eruptive warnings given by the monitoring network (seismic but not only) is now established on different volcanoes, using various petrological clock. In Dominica, we showed that a mixing processes between different magma batches systematically occures a decade prior the eruption within the deepest reservoir giving birth to the oldest ignimbritic eruptions. For the recent Plinian eruptions, similar mixing process begins ~ 10–30 years prior eruption, with more sustained mixing in the last decade, accelerated in the last 2 years. On active and monitored volcanoes, the correlation between the petrological clock and the pre-eruptive warnings given by the monitoring network (seismic but not only) is now established on different volcanoes, using various petrological clock. Thus, on Dominica, that has no monitored eruption, time constrain may help authorities in volcanic risk mitigation in case of volcanic reactivation. Such integrated study may be conducted on any volcanic target and lead to a more comprehensive understanding on the behaviour of the magma plumbing system, with strong implications in volcanic risks.

A transcrustal mush system has been recognized beneath Dominica (Lesser Antilles) with different magma ponding zones that generated a series of pumiceous eruptions from Morne Trois Pitons–Micotrin volcano. Here, the latest, large, pumiceous eruption (Grand Fond - 24 kyrs cal BP) and four, smaller, Plinian eruptions (18-9 kyrs cal BP) are investigated. Pre-eruptive magma dynamics within the mush are unraveled through orthopyroxene phenocrysts by combining a Crystal System Analysis approach (on unzoned and zoned orthopyroxenes) and timescale estimates derived by intracrystalline Fe-Mg interdiffusion modeling. Two magmatic environments are recognized in the mush and have mixed, more or less vigorously, before the successive eruptions. Few interactions between the two magmas began 15-34 years prior to the small Plinian eruptions, but the sustained mixing occurred in the last 2 years. This contrasts with longer timescales (2-80 years) obtained for the larger eruption of Grand Fond with magmas stored deeper. These magma mixing timescales have significant implications for volcanic risk mitigation, with a growing reactivation signal that could be registered at the surface few years prior to the eruptions.

The increase of number and intensity of earthquakes during a pre-eruptive crisis is the main basis of seismic volcano monitoring. However, the exact relationship between the seismic activity and the volcano-magmatic processes remains unclear. Here we present a direct comparison between characteristics of a seismo-volcanic crisis recorded prior to the 2010-2013 eruption of Kizimen volcano (Kamchatka, Russia) and the timescales of processes in the magma plumbing system. These timescales are inferred from the modelling of Fe-Mg intracrystalline interdiffusion in 88 zoned orthopyroxene crystals from dacites and silica-rich andesites samples collected after the eruption. We show that the eruptible magmas were assembled rapidly during a magma mixing episode ~1.5 years before the eruption, which is well correlated with the onset of a seismic crisis. We conclude that the observed seismic re-activation marks the onset of magma mixing leading to destabilization of the reservoir followed by the eruption after ~1.5 years.

Montagne Pelee is one of the most active volcanoes of the Lesser Antilles arc, with two to three magmatic eruptions per millennium and an estimated magmatic production rate in the order of 0.7 km3/1000 years. Montagne Pelee is also infamous for the very large number of people (30000) killed by an eruptive phenomenon at the onset of the 1902–1905 dome-forming eruption. Active for ~550 kyrs, Montagne Pelee has undergone two major flank collapses that influenced its volcanological as well as magmatic evolution. The last one occurred at around 36 ka. Due to changes in the threshold effect following the decrease in load of the volcanic edifice due to flank collapse, there was a switch in emitted magma from generally andesitic to basaltic andesites. After 10 kyrs of intense activity, the load exerted by the new edifice once again prevented dense basaltic andesite magma from reaching the surface, whereas andesitic magmas, similar to the initial ones, continued to be emitted. All the magmas come from a common magma ponding zone at 200 ± 50 MPa, 875 ± 25 °C, an oxygen fugacity (fO2) between 0.4 and 0.8 log unit above the nickel‑nickel oxide (NNO) oxygen buffer, and melt H2O contents of 5.3–6.3 wt%. Based on comparative on-land and marine tephrochronological studies, we have reconstructed a detailed eruptive history of the volcano over the last 25 kyrs. The volcano produced a succession of Plinian-SubPlinian and dome-forming eruptions, making it a textbook case for studying this duality, which sometimes occurred during a single eruption. We identified more than 55 magmatic eruptions, with a ratio of 2/3 for dome-forming vs. Plinian eruptions. An unusual feature of this volcano is that dome-forming eruptions often start with violent, superficial and laterally directed explosions. These generate highly devastating dilute and turbulent pyroclastic density currents on the southwestern and southern flanks of the volcano, as illustrated by the seven events of this type during the first months of the 1902–1905 eruption. On the basis of the past eruptions over the last millennia, a series of scenarios can be proposed in the event of reactivation, including no magmatic eruption, a phreatic event or a magmatic eruption (Plinian or dome-forming eruption, with or without an explosive phase).

A Ba- and Ti-rich mica (up to 14.0 wt% BaO and 13.1 wt% TiO2) occurs in nephelinites from the Middle Atlas Volcanic Province, Morocco. The rocks show a porphyritic texture composed of olivine, Ti-rich augite, nepheline and Ti-rich magnetite. Oxyphlogopite is found in secondary assemblages consisting of zeolites, clinopyroxene and Ti-oxides. This mineral occurs mostly as euhedral to subhedral crystals lining the walls of zeolitic veinlets and cavities in the nephelinites. The coupled substitutions of (1) Ba for K in the interlayer site and Si for Al in the tetrahedral site and (2) Ti4+ for Mg in the octahedral site and O2− in the (OH, F) site are major mechanisms of Ba and Ti incorporation in these micas. The average structural formula, calculated on the basis of 12 (O, OH, F, Cl), is (Si2.515-Al1.482-Fe3+0.003)4.000 (Al0.050-Ti0.661-Fe2+0.798-Mn0.012-Mg1.500)3.021 (Ca0.014-Ba0.398-Na0.066-K0.502)0.980 O10 (F0.308-Cl0.009-O1.319-OH0.365)2.001, suggesting a combination of oxyphlogopite and oxykinoshitalite, with contribution from the hypothetical end-member “Ti-eastonite”. Ba- and Ti-rich oxymica in the nephelinites may have crystallized under the influence of Ba-rich fluids derived from nephelinitic melts during late magmatic stages. The Ba enrichment in the host nephelinites indicates a lithospheric mantle source enriched by the circulation of metasomatic fluids.

Explosive eruptions of the Plinian type inject large amounts of particles (pumice, ash, aerosols) and volatile species into the atmosphere. They result from the rapid discharge of a magma chamber and involve large volumes of magma (from a km3 to hundreds of km3). Such eruptions correspond to a rapid ascent of magma in the conduit driven by the exsolution of volatile species. If the magma supply is continuous, this jet produces a convective eruptive column that can reach tens of km in height and transports gas and particles (pumice, ash, aerosols) directly into the stratosphere. Depending on the latitude of the volcano, the volume of implied magma, the height of the eruptive plume and the composition of the released gaseous and particulate mixture, these events can strongly affect the environment at the local or even at a global scale. Almost all studies on global impacts of volcanic eruptions have largely focused on the sulfur component. Volcanoes are also responsible for the emission of halogens which have a crucial impact on the ozone layer and therefore the climate.The objective of our project is to revisit the issue of the impact of volcanism on the atmosphere and climate by considering not only the sulfur component but also the halogen component. We will provide field work-based constraints on the strength of halogen (Cl and Br) emissions and on degassing processes for key eruptions, we will characterise the dynamics of volcanic plumes, notably the vertical distribution of emissions and we will explore and quantify the respective impacts of sulfur and halogen emissions on the ozone layer and climate. Here we will shed light on the methodology that will combine field campaign, laboratory analysis of collected samples and a hierarchy of modelling tools to study. We use an approach combining field studies, petrological characterization, geochemical measurements including isotopic data, estimation of the volume of involved magma and the height of injection of gases and particles by modelling the eruptive plume dynamic and numerical simulation of the impacts at the plume scale and at the global scale. The first halogen budget will also be presented.

Dominica island experienced the largest explosive eruptions (ignimbrites) of the Lesser Antilles arc. The recent revised chronostratigraphy of the Morne Trois Pitons – Micotrin eruptive activity evidenced a series of plinian eruptions that occurred between 18 ka and 9 ka BP. Here we focus on these recent eruptions in order to determine the magma storage conditions at depth and volatile degassing budget. Volatile concentrations (H2O, CO2) in melt inclusions indicate storage conditions of 200 MPa (~6-8 km deep) and 860-880°C in agreement with experimental constraints from phase-equilibrium data. The magmas were thus stored shallower than those involved during the ignimbritic eruptions (~16 km deep). Magma composition and halogen ratios suggest a common magma origin for all eruptions of Morne Trois Pitons Micotrin volcano in the last 60 kyrs. In addition, for the first time, a complete degassing budget including H2O, CO2, SO2, F, Cl, and Br has been established for all these explosive eruptions. The estimation of their eruptive fluxes towards the atmosphere supports the potential important role of halogen elements in the modification of atmosphere chemistry. Br degassing budget was the same order of magnitude as S whereas F and Cl budgets were 1 and 2 orders of magnitude higher than these two species.

International audience; Ignimbritic eruptions represent catastrophic events due to the magma volume involved and the related consequences on Earth's environment in relation with the released gases and the tephra dispersal. Dominica has been recognized as hosting one of the major ignimbritic eruptions of the last 200 ky in the Lesser Antilles arc, called the Roseau Tuff. But more recent works have evidenced several pumiceous events instead of a single large one. Here we propose a revised chronostratigraphy of the explosive activity that occurred in the last tens of thousands years based on three field trips, new 14C ages, detailed lithological and geochemical investigations, in particular a precise characterization of trace element glass chemistry. This eruptive history reconstruction is mainly based on outcrops along the coast and in the valley, since the luxury vegetation in the center of the island mostly precludes sections close to the central volcanic centers. We thus confirm that the Roseau event has been overestimated and that we may recognize five main ignimbritic events: Grande Savane, Layou (~ 51 ka), Grand Bay, Roseau (~ 33 ka), Grand Fond (~ 24 ka). We discuss the possible volcanic center at their origin, in addition to correlations with some Plinian events of lower magnitude that were identified in the Roseau valley and in the southern part of Dominica. This study may help to better constrain the eruptive history of the most active volcanic island of the Lesser Antilles arc, which has important implications on hazard mitigation.

International audience; The general question of the generation of large-volume silicic eruptions is here addressed through the experimental determination of the storage conditions of the primary magmas that generated ignimbritic eruptions at Dominica Island (Lesser Antilles) during the 24–51 ka period of time. The basal plinian fallouts and pumice pyroclastic flows from the large-volume (~ 5 km3 DRE) events of Layou, Roseau and Grand Fond were investigated, together with the smaller ignimbritic eruptions of Grand Bay and Grande Savane. All samples are dacitic (63–66 wt% SiO2) and contain ~ 30 vol% phenocrysts of plagioclase (~ 21 vol%), orthopyroxene (~ 5 vol%) and Fe-Ti oxides (

Plinian eruptions are characterized by high intensity and an overall steady character, and result in a stable convective column. The main processes controlling the dynamics of such steady and stable plume systems have been extensively investigated. Conversely, sub-Plinian eruptions are unsteady, as recorded by the large variability of the products and deposits. Our knowledge of the processes creating this unsteadiness on various timescales remains limited, and still requires more observations as well as theoretical and experimental investigation. Here, we focus on the sub-Plinian eruption of the Greenish Pumice (GP, 19,265 ± 105 BP), Mt. Somma-Vesuvius (Italy). On the basis of coupled geochemical and textural analyses of samples from the well-established stratigraphy of the GP deposits, we investigate volatiles (H2O, CO2, F, Cl) to better constrain the unsteady sub-Plinian eruptive style. This allows us to carry out a detailed study of the degassing processes in relation to the eruption dynamics. We find that degassing by open-system processes generally dominates throughout the entire eruption, but alternates with episodes of closed-system degassing. The fluctuating degassing regimes, responsible for the variable magma ascent rate within the conduit, are also responsible for the eruptive column instability. Volatile behavior is well correlated with textural heterogeneities of the eruptive products. Both reflect higher conduit heterogeneity than for Plinian eruptions, where we find a higher horizontal gradient in magma ascent velocity due to a smaller conduit diameter.

The cycling of the halogens in sub-volcanic systems may play an important role in magmatic ore deposit formation and the emission of climate-altering gases via volcanic outgassing. However, the mobile and reactive nature of the halogens, as well as their typically low abundances in most intrusive igneous rocks, means that their behaviour in shallow crustal level magmatic systems is difficult to constrain. We present data from the ~60 Ma Rum Layered Suite, NW Scotland, utilising the neutron irradiation-noble gas mass spectrometry (NI-NGMS) technique, to provide insight into the provenance and behaviour of Cl, Br and I in magma chamber processes and precious metal enrichment. As a well-characterised and superbly exposed example of an open system sub-volcanic magmatic body, the Rum intrusion is an ideal locality to investigate halogen behaviour in this context. Our data reveal concentrations of

International audience; Gaua, a little known volcano in the northern part of Vanuatu archipelago, went through a long term eruptive activity between September 2009 and July 2010. The eruption started by a phreatic to phreatomagmatic activity which progressively shifted into a magmatic discharge. The first eruptive phase involved the hydrothermal system in place. The latter was likely influenced by seawater seepage, leading to the formation of anhydrite. Magma involved hereafter this opening phase is of basaltic andesite and basaltic trachyandesite composition (high K calc-alkaline series), typical of the northern part of the Vanuatu archipelago. The 2009–2010 activity discharged at least 184 kt of SO2 and a significant amount of halogens (72 kt Cl and 217 kt F). Such halogen releases indicate that Gaua is a strong source of halogens into the atmosphere. High and sustained amount of F discharges are known to induce health issues and should not be ignored on Gaua island.During this eruption the quiescent and voluminous Lake Letas was slightly affected by the eruption. Nevertheless the hydrothermal discharge point into the lake, situated on the southeastern part of Mt. Garet appeared to be relatively active. At this particular location rock forming elements, leached out from volcanic rock by acid fluids released by the new intrusion of magma, were discharged along with anions into Lake Letas. This release has triggered localized chemistry changes in the lake. We speculate that this discharge has also disturbed the bottom water in a limited perimeter, remobilizing reduced Fe to the surface and subsequently triggering the change in the water color by Fe oxidation.

Defining the magma storage conditions of a volcanic system is a major goal in modern volcanology due to its direct implications for the style of a possible eruption and thus on the associated risk of any crisis and the necessary management and mitigation strategies. Below 200 MPa and at equivalent depths, the strongly non-ideal behaviour of the H-C-O-S-Cl-F system in the silicate melt causes unmixing of the fluid phase to form an H2O-rich vapour and a hydrosaline phase in equilibrium with the silicate melt, both responsible for buffering the chlorine (Cl) concentration. Following this equilibrium, the Cl concentration in melts may be used as a geobarometer for alkaline magmas. Systematic application of this method to the main explosive eruptions of Mount Somma-Vesuvius highlights two main magma ponding zones, at ~180–200 and ~100 MPa. At these pressures, the maximum pre-eruptive H2O contents for the different magma compositions can be estimated; the results obtained, largely in agreement with the current literature, therefore confirm the validity of the method. The Cl geobarometer may help scientists to define the variation of the magmatic reservoir location through time and thus provide strong constraints on pre-eruptive conditions, which are of utmost importance for volcanic crisis management.

International audience; Modern volcanology studies have demonstrated that the crystal "cargo" of magmas can be used to unravel the dynamics of magmatic plumbing systems and constrain timescales of magmatic processes. Specifically, several studies focusing on large silicic eruptions (10 to 100 km3 DRE/eruption) have shown that crystal populations may be rapidly re-mobilized over short timescales of decades to centuries prior to eruption, making these systems more dynamic than previously believed. We present a petrological study of three, major, pumiceous, dacitic eruptions produced on Dominica (Lesser Antilles Arc) in the last 50ky: Layou (~50ka), Roseau (~30ka) and Rosalie (~20ka). We combine crystal texture and chemical analyses with modeling of diffusional relaxation of chemical gradients in orthopyroxene crystals. A systems analysis approach is used to decipher remobilization processes of crystals before each eruption, and to map the plumbing system evolution over time. Timescales are calculated by Fe2+-Mg interdiffusion modeling on selected orthopyroxenes, by inter-calibration of high resolution BSE-images and EMPA analysis profiles. Textural analyses show that for each of the three eruptions about 80-85% of orthopyroxenes are unzoned, while about 15-20% exhibit normal, reverse and multiple zoning. Systems analysis of the zoned crystals reveals the presence of four different magmatic environments (ME): En46-51(ME1), En52-53(ME2), En54-58(ME3) and En59-63(ME4). During the eruptive history, the main magmatic environment, represented by the 80% of unzoned orthopyroxenes, moves from ME2 (for Layou and Roseau) to ME3 (for Rosalie), indicating a shift of the system to less-evolved compositions. For all eruptions, a main crystal transfer pathway can be identified between ME2-ME3, describing a progressive sinking of crystals from the main environment En52-53(ME2) to the less-evolved En54-58 (ME3), which receives crystals, and, by the Rosalie eruption, becomes the volumetrically most significant magmatic environment. Occurrence of ME4 (En59-63) in the Roseau and Rosalie eruptions further suggests a progression toward a less-evolved composition. To assess timescales of these remobilization events, intra-crystalline diffusion has been modeled along the a- and b-axes of 22 zoned orthopyroxenes per eruption at the magmatic temperature of 850°C. Independently of the orthopyroxene zoning patterns, timescales for Layou and Roseau are of the same order of magnitude, showing that the crystal mobilization event occurred tens of years before each eruption. We propose that orthopyroxene crystals of ME2 (~80%, En52-53) represent the main crystallized mush body. About ten years before the Layou and Roseau eruptions, the intrusion in the system of a less evolved magma re-heats the system, extracting eruptible melt plus crystals from the mush and producing the zoning patterns in pyroxenes (En54-56 and En58-63 rims).

article i nfo Article history: Fluorine and chlorine diffusion were measured in two natural phonolitic melts, from Vesuvius (Italy) and from Laacher See (Germany), at 0.5 and 1.0 GPa, between 1250 and 1450 °C at anhydrous conditions and with about 2 and 5 wt.%of dissolvedwater. Thetwo differentstarting materialsallow us to investigate the alkalieffect,Navs.K, onhalogendiffusion.OnecompositionwasaK-rich(~10wt.%)phonoliticmeltcorrespondingtothewhitepumice phaseofthe79ADeruptionofVesuvius,andtheotheraNa-rich(~10wt.%)phonoliticmeltcorrespondingtomost differentiated melt of the 12,000 BC eruption of Laacher See. The diffusion-couple technique in a piston cylinder was used for the experiments. Experiments were performed with only one halogen diffusing and with the simultaneous diffusion of a halogen mixture (F, Cl, Br) in order to evaluate the interactions between the halogens

Diffusion of halogens has the potential to influence petrogenetic processes in magma chambers and conduit degassing processes. This contribution reviews our current state of knowledge concerning halogen diffusion and the influence of halogens on the diffusion of major elements in silicate melts. The addition of halogens to silicate melts at common, natural concentration levels will have little effect on the diffusion of major elements. However, the differences between the diffusivity of water, the diffusivities of halogens, and the diffusivity of sulfur are significant enough that during melt inclusion entrapment, or during rapid bubble or crystal growth, diffusive fractionation between water and the halogens, and between halogens and sulfur, are expected to occur and can influence the compositions of melt inclusions, crystals and volcanic gases.

Dome-forming eruption is a frequent eruptive style and a major hazard on numerous volcanoes worldwide. Lava domes are built by slow extrusion of degassed, viscous magma and may be destroyed by gravitational collapse or explosion. The triggering of lava dome explosions is poorly understood: here we propose a new model of superficial lava-dome explosivity based upon a textural and geochemical study (vesicularity, microcrystallinity, cristobalite distribution, residual water contents, crystal transit times) of clasts produced by key eruptions. Superficial explosion of a growing lava dome may be promoted through porosity reduction caused by both vesicle flattening due to gas escape and syn-eruptive cristobalite precipitation. Both processes generate an impermeable and rigid carapace allowing overpressurisation of the inner parts of the lava dome by the rapid input of vesiculated magma batches. The relative thickness of the cristobalite-rich carapace is an inverse function of the external lava dome surface area. Explosive activity is thus more likely to occur at the onset of lava dome extrusion, in agreement with observations, as the likelihood of superficial lava dome explosions depends inversely on lava dome volume. This new result is of interest for the whole volcanological community and for risk management.

International audience; Petrological data have been acquired on the natural trachytes from the Chaîne des Puys, French Massif Central, and on experimental products from phase equilibria in order to (i) constrain the storage conditions of the trachytic magmas that lead to explosive eruptions (dome destructions as block-and-ash flows or pumice-and-ash flows) and (ii) provide phase relationships and chemical compositions for differentiated alkaline liquids in intraplate continental context. Phase assemblages, proportions, and compositions have been determined on six trachytes with SiO2 contents varying from 62 to 69 wt % and alkali contents of 10.5-12.0 wt %. The samples contain up to 30 % of phenocrysts, mainly consisting of feldspar (15-17 %; plagioclase and/or alkali-feldspar), biotite (2-6 %; except in the SiO2-poorest sample), Fe-Ti oxides (1-3 %) ± amphibole (< 5 %), ± clinopyroxene (~1 %). All samples have apatite and zircon as minor phases and titanite has been found in one sample. Pristine glasses (melt inclusions or residual glasses) in pumice from explosive events are trachytic to rhyolitic (65-73 wt % SiO2 and 10.5-13.0 wt % alkali). H2O dissolved in melt inclusions and the biotite+alkali feldspar+magnetite hygrobarometer both suggest pre-eruptive H2O contents up to 8 wt %. These are so far the highest H2O contents ever reported for alkaline liquids in an intraplate continental context. Melt inclusions also contain ~3400 ppm chlorine, ~700 ppm fluorine, and ~300 ppm sulphur. Crystallisation experiments of the six trachytes have been performed between 200 and 400 MPa, 700 and 900°C, H2O saturation, and oxygen fugacity of NNO +1. The comparison between the natural and experimental phase assemblage, proportion, and composition suggests magma storage conditions at a pressure of 300-350 MPa (~10-12 km deep), melt H2O content ~8 wt % (close to saturation), an oxygen fugacity close to NNO~0.5, and temperatures increasing from 700 to 825°C with decreasing bulk SiO2 of the trachyte. The high H2O contents of the trachytes show that wet conditions may prevail during the differentiation of continental alkaline series. Regardless of the size of the magma reservoir assumed to have fed the trachyte eruptions, calculation of the thermal relaxation timescales indicates that the tapped reservoir(s) are likely to be still partially molten nowadays. The four northernmost edifices may correspond to a single large reservoir with a lateral extension of up to 10 km, which could be possibly reactivated in weeks to months if intercepted by new rising basalt batches.

International audience; Trachytes from the Chaîne des Puys, French Massif Central, have been studied by performing phase equilibria in order to (i) constrain the storage conditions of the trachytic magmas that lead to explosive eruptions (either dome destructions as concentrated or diluted pyroclastic density currents or highly explosive events) and (ii) provide phase relationships and compositions for differentiated alkaline magmas. Phase assemblage, proportion, and compositions have been determined on six trachytes (62-69 wt % SiO2 and 10.5-12.0 wt % alkali) mostly coming from the actual domes dated from 9.2 to 15 Ka and aligned along a 10-km distance in the Chaîne des Puys. All samples contain ~23-30 % of phenocrysts, mainly consisting of plagioclase (15-17 % and K-feldspars for the two SiO2-richest samples), biotites (2-6 % except in the SiO2-poorest sample, where it is absent), and Fe-Ti oxides (1-3 %). The three SiO2-poorest samples also contain ~2 % of amphibole and the SiO2-richest one has 1 % of clinopyroxene. All samples have apatite and zircon as minor phases and sphene for the SiO2-richest one. Glasses (melt inclusions and residual glasses) analysed in pumices resulting from highly explosive events, show trachytic to rhyolitic compositions (65-73 wt % SiO2 and 10.5-13.0 wt % alkali). Analyses of melt inclusions (EMP by-difference method) and the biotite+K-feldspar+magnetite hygrobarometer both suggest pre-eruptive H2O contents up to 7-8 wt %, which are so far the highest contents ever reported for alkaline liquids. The melt inclusions also contain ~3400 ppm chlorine, ~700 ppm fluorine, and ~300 ppm sulphur (EMP analyses). Phase equilibria of six representative trachytes have been performed between 200 and 400 MPa, 700-900°C, H2O saturation, and oxygen fugacity from NNO-1 to NNO+1. The comparison between the natural and experimental products suggests magma storage conditions at pressures of 300-350 MPa, temperatures increasing from 700 to 825°C with decreasing bulk SiO2 contents, oxygen fugacity from NNO to NNO+1, and melt H2O contents close to saturation conditions (~8 wt. %). The high H2O contents of the trachytes show that wet conditions prevail during the differentiation of continental alkaline series.

The eruptive history of Mt. Somma–Vesuvius is characterised by large explosive events: Pomici di Base eruption (22,030 ± 175 yr cal BP), Mercato (8890 ± 90 yr cal BP), Avellino (3945 ± 10 yr cal BP) and Pompeii (79 AD). Pre-eruptive conditions and sin-eruptive degassing processes of the Avellino eruption, the highest-magnitude Plinian event, have been investigated, using volatile contents (F, Cl, H 2 O) in melt inclusions and residual glass, and textural characteristics of pumice clasts of the 9 fallout layers sampled in detail in a representative sequence. The sequence displays an up-section sharp colour change from white to grey, corresponding to variations in both magma composition and textural characteristics. The pre-eruptive conditions have been constrained by systematic measurements of Cl content in both melt inclusions and matrix glass of pumice clasts. The pumice glass composition varies from Na-rich phonolite (white pumice) to K-rich phonolite (grey pumice). The measured Cl values constantly cluster at 5200 ± 400 ppm (buffer value), whatever the composition of the melt, suggesting that the entire magma body was saturated with sub-critical fluids. This Cl saturation constrains the pre-eruptive pressures and maximum H 2 O contents at 200 ± 10 MPa and 6.3 ± 0.2 wt.% H 2 O for the white pumice melt and 195 ± 15 MPa and 5.2 ± 0.2 wt.% H 2 O for the grey pumice melt. The fluid phase, mainly composed of a H 2 O-rich vapour phase and brine, probably accumulated at the top of the reservoir and generated an overpressure able to trigger the onset of the eruption. Magma degassing was rather homogeneous for the white and grey eruptive units, mostly occurring through closed-system processes, leading to a typical Plinian eruptive style. A steady-state withdrawal of an H 2 O-saturated magma may explain the establishment of a sustained Plinian column. Variation from white to grey pumice is accompanied by decrease of mean vesicularity and increase of mean microcrystallinity and permeability related to significant vesicle coalescence. Despite this, the ascending magma column still evolves under closed-system degassing, without significant gas loss through conduit walls. The Avellino eruption shows numerous similarities with the 79 AD Pompeii eruption in pre-eruptive conditions, degassing processes and eruptive style which are discussed here.

The 79 ad Plinian eruption of Vesuvius produced first a white pumice fallout from a high steady eruptive column, and then a grey pumice fallout originating from an oscillatory eruptive column with several partial column collapse events after which there was a total column collapse. This first total collapse was followed by renewed Plinian activity and produced the last grey pumice (GP) fallout deposit of the eruption. Textural characteristics (vesicularity and microcrystallinity) of a complete sequence of the pumice fallout deposits are presented along with the major element compositions and residual volatile contents (H2O, Cl) to constrain the degassing processes and the eruptive dynamics. Large variations in residual volatile contents exist between the different eruptive units. Textural features also strongly differ between white and grey pumices, but also within the grey pumices. The degassing processes were thus highly heterogeneous. We propose a new model of the 79 ad eruption in which pre-eruptive conditions (H2O saturation, magma temperature and viscosity) are the critical controls on the diversity of the syn-eruptive degassing processes and hence the eruptive dynamics. Cl contents measured in melt inclusions show that only the white pumice and the upper part of the grey pumice magma were H2O saturated prior to eruption. The white pumice eruptive units represent a typical closed-system degassing evolution, whereas the first grey pumice one, stored under similar pre-eruptive saturation conditions, follows a particular open-system degassing evolution. We suggest that the oscillatory regime that dominated the grey pumice eruptive phase is linked to pre-eruptive water undersaturation of most of the grey magma, and the associated time delays necessary for H2O exsolution. We also suggest that the high residual H2O content of the last grey pumice, deposited after the renewal of Plinian activity following the first total column collapse event, is due to syn-eruptive saturation of GP magma and reduced H2O exsolution efficiency resulting from speciation of dissolved H2O in the melt.

International audience; Residual concentrations of halogens (F, Cl, Br, I) and H2O in glass (matrix glass and melt inclusions) have been determined in a series of volcanic clasts (pumice and lava‐dome fragments) of plinian, vulcanian and lava dome‐forming eruptions. Felsic magmas from calc‐alkaline, trachytic and phonolitic systems have been investigated: Montagne Pelée and Soufrière Hills of Montserrat (Lesser Antilles), Santa Maria‐Santiaguito (Guatemala), Fogo (Azores) and Vesuvius (Italy). The behavior of halogens during shallow H2O degassing primarily depends on their incompatible character and their partitioning between melt and exsolved H2O vapor. However, variations in pre‐eruptive conditions, degassing kinetics, and syn‐eruptive melt crystallization induce large variations in the efficiency of halogen extraction. In all systems studied, Cl, Br and I are not fractionated from each other by differentiation or by degassing processes. Cl/Br/I ratios in melt remain almost constant from the magma reservoir to the surface. The ratios measured in erupted clasts are thus characteristic of pre‐eruptive magma compositions and may be used to trace deep magmatic processes. F behaves as an incompatible element and, unlike the other halogens, is never significantly extracted by degassing. Cl, Br and I are efficiently extracted from melts at high pressure by H2O‐rich fluids exsolved from magmas or during slow effusive magma degassing, but not during rapid explosive degassing. Because H2O and halogen mobility depends on their speciation, which strongly varies with pressure in both silicate melts and exsolved fluids, we suggest that the rapid pressure decrease during highly explosive eruptions prevents complete equilibrium between the diverse species of the volatiles and consequently limits their degassing. Conversely, degassing in effusive eruptions is an equilibrium process and leads to significant halogen output in volcanic plumes.

International audience; Pre-eruptive conditions and degassing processes of the AD 79 plinian eruption of Mt. Vesuvius are constrained by systematic F and Cl measurements in melt inclusions and matrix glass of pumice clasts from a complete sequence of the pumice-fallout deposits. The entire ‘white pumice' (WP) magma and the upper part of the ‘grey pumice' (GP) magma were saturated relative to sub-critical fluids (a Cl-rich H2O vapour phase and a brine), with a Cl melt content buffered at ~ 5300 ppm, and a mean H2O content of ~ 5%. The majority of the GP magma was not fluid-saturated. From these results it can be estimated that the WP magma chamber had a low vertical extent (< 500 m) and was located at a depth of ~ 7.5 km while the GP magma reservoir was located just beneath the WP one, but its vertical extent cannot be constrained. This is approximately two times deeper than previous estimates. H2O degassing during the WP eruption followed a typical closed-system evolution, whereas GP clasts followed a more complex degassing path. Contrary to H2O, Cl was not efficiently degassed during the plinian phase of the eruption. This study shows that F and Cl behave as incompatible elements in fluid-undersaturated phonolitic melts. H2O saturation is necessary for a significant partitioning of Cl into the fluid phase. However, Cl cannot be extracted in significant quantity from phonolitic melts during rapid H2O degassing, e.g. during plinian eruptions, due to kinetics effects. Halogen contents are better preserved in volcanic glass (melt inclusions or matrix glass) than H2O, therefore the combined analysis of both volatile species is required for reliable determination of pre-eruptive conditions and syn-eruptive degassing processes in magmas stored at shallow depths.

Supporting information for d'Augustin et al, 2020.-supplementary figures-supplementary tables; International audience; The Morne Trois Pitons‐Micotrin volcanic complex on the island of Dominica (Lesser Antilles) emitted a series of plinian eruptions between 18 ka and 9 ka BP. We studied it to constrain magma storage conditions and volatile degassing balances, by comparison with the three previous ignimbrites (~60‐24 ka BP). Volatile concentrations in glass inclusions and mineral‐melt thermobarometry indicate storage at ≤200 MPa (~6‐8 km) and 860‐880°C. The magmas feeding these plinian eruptions were stored at a shallower depth than those that older ignimbrites from the same volcanic complex (~16 km). Close magma composition and similar halogen ratios, however, suggest a common source for the magmas feeding both the plinian eruptions and the ignimbrites. The large eruptive fluxes of F, Cl and Br to the atmosphere (up to 1.4‐2.8 x 10‐1 Mt/km3, 1.5‐4.0 Mt/km3 and 2‐4 x 10‐2 Mt/km3, respectively), estimated by the petrological method, support the potentially important role of volcanic halogens in modifying the chemistry of the atmosphere, though Cl is underestimated here because of buffering in a fluid phase. The behaviour of S, potentially partitioned in the same fluid phase, prevents here the calculation of an eruptive outgassing budget.