No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

This work is based on gas monitoring data: COSPEC-derived sulphur dioxide emission rates and Open-path FTIR-derived HCl:SO<SUB>2</SUB> mass ratios in the volcanic plume at Soufrière Hills Volcano, Montserrat, West Indies, collected during the 1995-2000 eruption, as well as geochemical analyses of phenocryst-hosted melt inclusions and matrix glass of the erupted products. The following chapters deal with various aspects of the behaviour of the elements sulphur, chlorine and water in the pre-, syn- and post-eruption melt and the gas species sulphur dioxide and hydrogen chloride in the volcanic plume. Melt inclusion and matrix glass compositions (major, volatile and trace elements) have been used to build a picture of melt evolution with respect to sulphur and chlorine. Concentrations of chlorine and sulphur in the melt; both pre- and syn-eruption, indicate that these species are derived from different sources and from different depths. Surface emission data are used to constrain these models (chapter 3). Open-path FTIR measurements have yielded HCl emissions throughout the eruption of Soufrière Hills Volcano, with the most recent measurements being obtained during dome building in November and December 1999 and January, August, September and October 2000. These data have corroborated theories on chlorine degassing from the melt on ascent derived from analytical results and we have recognised clear dome building and "residual period" HCl emission signatures.

At Mt. Etna, Italy, vigorous gas-rich eruptions in 2001, 2002 and 2003 were followed by gas-poor eruptions in 2004, 2006 and 2007. Contrary to expectation, melt inclusion compositions indicate that magmas erupted between 2004 to 2007 did not follow similar degassing paths as recorded in 2001 and 2002 and that are expected from the solubility laws of CO₂ and H₂O. Instead melts stored in the plumbing system since 2002 reequilibrated with CO₂-rich gases from depth. Sustained gas percolation caused loss of water and enhancement of CO₂ in the evolving melt. At Piton de la Fournaise melt inclusions trapped in olivines record degassing of various batches of magma and the fractionation of olivines at various depths within the plumbing system. The host melt which carries these olivines to the surface represents an infiltration of new magma which erupts rapidly incorporating olivines along the way. The host melt also records processes of diffusive fractionation during groundmass crystallisation. Semi-volatile trace metals and Li have been found to behave in a volatile fashion at both Mt. Etna and Piton de la Fournaise. At Mt. Etna, CO₂gas fluxing may have been important for causing the transfer of Cu from magmas at depth, to magmas stored in the shallow plumbing system. At Piton de la Fournaise trace metals are not simply behaving as incompatible elements but rather show the influence S and H₂O loss during degassing. However, when investigating trace metal concentrations in volcanic products this thesis shows that the affect of sulphide immiscibility should not be neglected. Loss of chalcophile trace metals to sulphide melts prevents partitioning of the element into a gas phase. Destabilisation of these melts on the other hand, may release enriched concentrations of trace metals to a gas phase.

The Baikal Rift is a zone of active lithospheric extension, located in the Sayan-Baikal fold belt between the Siberian Craton and the Aldan Shield. Rifting was initiated in this region ~35 Mya. The ~20 to 3 Ma Vitim Volcanic Field (VVF) lies on the eastern shoulder of the rift and consists of ~5000 km³ of alkali basalts and basanites, with minor nephelinites and tholeiitic basalts. 140 drill core samples have been used to study temporal and spatial variations in the volcanic pile, which is up to 700 m thick in places. The VVF lavas have been analysed by XRF and ICP-MS for their whole-rock chemistry. Variations in major element abundances (e.g. MgO, which varies from 14.6-3.3 wt.%) reflect fractional crystallisation of olivine, clinopyroxene and to a lesser extent, plagioclase. Prior to eruption, the nephelinitic magmas appear to have crystallised in chambers in the upper crust (at 1 kbar) whereas the tholeiitic basalts appear to have undergone fractionation at the base of the crust (9 kbar). At a given MgO content, Na₂O, P₂O₅ and K₂O and incompatible trace elements decrease systematically from nephelinites through to tholeiitic basalts, suggesting that partial melting is also responsible for variations in whole-rock chemistry. ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd vary from 0.70391 to 0.70487 and 0.51272 to 0.51287 respectively, and indicate that the Vitim magmas have not assimilated significant quantifies of continental crust. Fractionation-corrected compositions may therefore be used to study mantle processes operating beneath the VVF. Nephelinites, basanites, alkali basalts and tholeiites display a wide range in [La/Yb]_n ratios (5-33) which is interpreted as evidence of variation in the degree of partial melting. Trace element ratios, e.g. Ba/Sr and La/K, suggest that phlogopite and possibly amphibole are present as residual mantle phases during generation of the nephelinites and basanites but not the tholeiites.

This study uses the strontium isotope tracer system to constrain hydrothermal flux estimates and investigate the hydrothermal contribution to the oceanic strontium budget. This is achieved through a multi-directional approach. Firstly, strontium isotope data for ocean and ophiolite crust are compiled to investigate the degree of isotopic alteration displayed by modern and ancient crustal profiles; and calculate the cumulative hydrothermal contribution to the oceanic strontium budget. This reveals an apparent imbalance in the oceanic strontium budget, because the hydrothermal contribution calculated is only a third of that required to balance a revised riverine input and the buffering effect of carbonate digenesis. Secondly, a multi-phase hydrothermal system in the northern Oman ophiolite is investigated as an analogue for understanding modern oceanic process and the magnitude of hydrothermal fluxes in arc-related environments. Field mapping and trace elements discrimination methods are combined to identify three hydrothermal regimes which correlate with the complex magmatic-tectonic evolution of the area. The first regime is associated with formation and cooling of the crustal sequence within an oceanic spreading environment; the second and third regimes are associated with later magmatism in an off-axis environment. Strontium isotope geochemistry is used to investigate the character of each regime and predict a time-integrated high-temperature fluid flux of ~4.5±1.1 x 10⁷ kg m^-2 for the initial spreading related hydrothermal event and lower estimates for the later regimes. The flux calculated is significantly higher than a similar prediction made for mid-ocean ridge systems (eg. Teagle et al, 2003), supporting previous suggestion that oceanic spreading systems in supra-subduction settings support greater hydrothermal fluxes than normal oceanic environments.