Melt inclusions record valuable information about pre-eruptive melt volatile concentrations. However, a vapor bubble commonly forms in inclusions after trapping, and this decreases the dissolved CO2 concentration in the trapped melt. To quantify CO2 loss to bubbles, Raman spectroscopic analysis was used to determine the densities of CO2 vapor in the bubbles. The samples analyzed in this study are from two Cascade cinder cones near Mt. Lassen and two Mexican cinder cones (Jorullo, Paricutin). Using analyses of dissolved CO2 and H2O in the glass in the inclusions, the measured CO2 vapor densities were used to reconstruct the original dissolved CO2 contents of the melt inclusions at the time of trapping. The Raman-restored CO2 values are similar to restored CO2 values calculated using a model of cooling and olivine crystallization in the trapped melts. This thesis includes unpublished co-authored material.

The amount of volatile elements dissolved in silicate melts is a controlling factor in a range of geologic processes, which include hazardous volcanic eruptions, economically-significant ore-forming systems, and global-scale volatile fluxes, which contribute to planetary evolution. While melt volatile contents are important, estimating the origin and fate of volatiles distributed within magmas is challenging because volatiles exsolve from the melt during eruption and are transferred into the atmosphere. Therefore, the stratigraphic record of volcanic and intrusive deposits does not contain direct information regarding the pre-eruptive volatile content of the melt. However, melt inclusions trapped within growing phenocrysts present an opportunity to sample the melt before it has completely degassed. Analysis of melt inclusions is challenging owing to a range of processes which occur after the melt inclusion is trapped and which overprint the original texture and composition of the inclusion at the time of entrapment. Thus, efforts to accurately determine the current composition of the melt inclusion sample and then infer the original composition of the trapped melt which that inclusion represents require a combination of microanalytical, numerical, and/or experimental methods. In Chapter 1, we present a pedagogical approach for estimating the processes that affect the CO2 content of a magma from its origin during melting a C-bearing source material to its exsolution into a free fluid phase during crystallization and degassing. In Chapter 2, we explore different experimental, microanalytical, and numerical methods which may be used to estimate the CO2 contents of melt inclusions that contain fluid bubbles and describe the advantages and disadvantages of each approach. In Chapter 3, we apply some of the methods discussed in the previous chapters to estimate the pre-eruptive volatile content of Haleakala Volcano (Maui) and assess different melting mechanisms that may be active in the Hawaiian plume. / Doctor of Philosophy / Volcanoes are features which form on the Earth’s surface and are located above regions where material melts tens of kilometers (or more) below the surface. The process of melting is studied through laboratory experimentation, and therefore it is possible to estimate the composition of deep subsurface material based on the compositions of volcanic rocks which can be sampled on the Earth's surface. This sub-discipline of geologic research is called "igneous petrology." A fundamental problem in igneous petrology is estimating the volatile content of the Earth's deep interior. Volatile elements are those elements such as hydrogen and carbon, which are stable as gasses in the atmosphere rather than in the mineral components of a rock. It is thought that the gasses produced from volcanic vents, of which the compositions are well known, represent volatile elements which were originally present as dissolved components in the melt. Experiments performed on volcanic rocks have demonstrated that volatile elements can be dissolved in melts at high pressures corresponding to depths within the Earth's crust, and these elements exsolve from the melt when it approaches the surface -- similar to how CO2 can be dissolved in a carbonated beverage, which bubbles out when the beverage is opened. The only geologically-persistent features which preserves the pre-eruptive volatile content of a melt (i.e. how much gas was dissolved before eruption) are droplets of melt which are accidentally trapped within crystals that grow from the melt as it cools near the Earth's surface -- these are called "melt inclusions." While melt inclusions are useful in this regard, they are challenging to apply to geologic problems because they undergo a range of physical and chemical changes after they are trapped, which can alter their composition from the original composition of the melt that was trapped. This dissertation concerns the theory used to infer how volatile elements are distributed within the deep Earth, analytical and numerical methods used to gather relevant information from melt inclusion samples, and an application of these methods to investigate the volatile content of the mantle below Hawaii. Chapter 1 describes a framework for systematically determining the amount of CO2 distrubuted within a given volcanic setting. Chapter 2 compares different methods used to estimate the original volatile content of melt inclusions from Kamchatka, which have formed fluid bubbles -- a common feature present in melt inclusions. Chapter 3 applies the methods described in the first two chapters to estimate how volatile elements are distributed within the Earth's mantle below Hawaii, and how the process of melting transfers them to the Earth's atmosphere.

No description available.

Campi Flegrei is a large volcanic complex located west of the city of Naples, Italy. The area has been the site of volcanic activity for more than 60 ka and represents a potential volcanic hazard owing to the large local population. In this study, the geochemistry of the magma associated with two different eruptions at Campi Flegrei has been characterized, with the aim to identify geochemical trends that may help to predict the style and nature of future eruptions. Two eruptions of different age and eruptive style have been selected for study, Fondo Riccio (9.5 ka) and Minopoli 1 (11.1 ka). A scoria (CF-FR-C1) and a bomb (CF-FR-C2) were collected from the Fondo Riccio eruption, and two scoria samples were collected from Minopoli 1 (CF-Mi1-C1 and C2) eruption. The pre-eruptive volatile content of magma plays an important role in the style of eruption and can be assessed from studies of melt inclusions (MI) contained in phenocrysts. Major and trace elements in Fondo Riccio MI show a wider variation compared to those in Minopoli 1 MI suggesting that the Fondo Riccio magma residence time was longer compared to the Minopoli 1 magma. Analyses of volatile contents in MI suggest that Fondo Riccio magma may have been more water-rich than Minopoli 1 magma, consistent with the more explosive character of this eruption compared to Minopoli 1. Trace element data suggest a combination of arc volcanic and upper continental crust magma as the source for the Fondo Riccio and Minopoli 1 eruptions. / Master of Science

A plethora of magmatic processing occurs in magma reservoirs, where melts are stored prior to eruption. Magma reservoirs are complex, open systems, and often multiple reservoirs are partially inter-connected from source to surface, giving rise to the term 'volcanic plumbing system'. Parental melts feeding these reservoirs can have diverse and distinct geochemical and petrological characteristics, and be variably evolved or enriched. These melts can also bring with them a crystal cargo that may remain in equilibrium in the magma reservoir, but may also be modified by reaction, resorption, crystallisation and diffusion. Melts and crystals can be transported between reservoirs, from the upper mantle and through the crust, leading to melt mixing, reactions and volatile exsolution. Basaltic volcanic systems are fed by primitive melts, and due to the rapid ascent of melts and short magma storage times, these volcanoes provide the best means of unravelling the mantle and crustal contribution to geochemical heterogeneity observed in erupted samples. Despite the potential chemical complexity of a magma reservoir, evidence for magma processing and reaction can be preserved in melt inclusion suites and the compositional structure of their host crystals. Magmatic processes during storage and transport at two basaltic volcanoes are investigated using two carefully selected eruptions: the 1669 eruption at Mt. Etna, and the 2007 Father's Day eruption at Kīlauea. A suite of diverse geochemical, petrological and petrographical observations, made at a range of length-scales, are combined and interpreted in tandem with geophysical monitoring data. The conclusions of these studies shed light on the architecture of each volcano's plumbing systems and basaltic plumbing systems in general. This thesis is divided into two parts. The first study unravels the crustal and mantle processes controlling melt geochemical heterogeneity at Mt. Etna, Sicily, during the 1669 eruption, the largest eruption in historical times. The 1669 melt inclusion suite arises from the mixing of two basaltic melts with similar major element compositions but very different trace and volatile element compositions. The melt geochemistry suggests that at least one end-member melt has been heavily influenced by assimilation of carbonate in the crust. The elevation in alkalis, caused by carbonate assimilation, enhances carbon and sulfur solubility in one end member. The melt inclusion suite indicates that mixing of these melts occurred in the shallow crust shortly before eruption and this mixing may be the cause of the enhanced $CO_{2}$ fluxes prior to eruptions at Mt. Etna. The second study is split into two parts. Each uses the eruptive products of the Father's Day eruption at Kīlauea and aims to unravel the connectivity of the plumbing system between the summit and East Rift Zone, with a focus on timescales of storage and transport. The first part investigates the melt geochemistry in terms of heterogeneity and volatile composition, and the second investigates the crystal cargo in terms of features of the macro-scale crystal cargo distribution and the micro-scale geochemical zoning of individual crystals. The integration of observations and models from these two studies constrains the pressure, temperature and composition of source magma feeding the Father's Day eruption. The eruption is investigated in the context of the "magma surge'' event that preceded the intrusion, as well as within the context of the longer-term trends in Kīlauea geochemistry at the summit and East Rift Zone. Melt inclusion and matrix glass volatile systematics provide insights into the degassing path of the magma and the duration of magma transport to the surface is constrained by diffusion modelling. Estimated timescales for ascent by diffusion modelling of macrocryst major element composition, melt inclusion water content and the melt Fe$^{3+}$/Fe$_{tot}$ ratio are in agreement with timescales observed from the geophysical data of $< $8 hours from reservoir depth to eruption. Both studies emphasise how petrological observations can supplement geophysical monitoring datasets collected at the surface to aid our interpretation of volcanic behaviour and eruption forecasting.

The first section of this work examines melt inclusions in phenocrysts from Volcán Popocatépetl and Volcán de Colima within the Trans Mexican Volcanic Belt (TMVB). These inclusions are dacitic to rhyolitic. Trends in melt inclusion major element and water concentrations form the evolved extension of other Mexican volcanics including those presumably derived directly from primitive melts. Water concentrations in Popocatépetl and Colima melt inclusions are similar (0.3 to 3.4 weight percent Hsub2O). Melt-vapor equilibration pressures calculated from dissolved Hsub2O and COsub2 (Popocatépetl) or Hsub2O (Colima) in melt inclusions correspond to depths of entrapment of 12 km or less. Water and carbon dioxide concentrations correlate negatively with SiOsub2 and potassium. Normalized olivine-augite-quartz compositions are consistent with near cotectic crystallization under vapor-saturated conditions at pressures of 1.5 kb or less. Our results show that Popocatépetl and Colima magmas have undergone vapor-saturated crystallization during ascent in conjunction with varying degrees of mixing between degassed rhyo-dacitic and less degassed, mafic melts in the upper portions of the crust. These data suggest melt evolution occurred in conduits or inter-fingered dikes rather than a large stratified magma chamber. Part II looks at the Masaya caldera in Nicaragua. This volcano has erupted frequently in recorded history, producing lava lakes and very high gas emissions. Melt inclusions from Masaya are basaltic, with low Hsub2O (below 0.5 wt. %), low S (less than 300 ppm) and high COsub2 concentrations (up to approximately 6000 ppm). Relationships between water, sulfur, Cl and F in combination with Masaya's high COsub2 and Ba/Zr and Ba/Nb ratios suggest that Masaya has undergone a multi stage degassing process involving 1) shallow degassing, 2) recycling of magma into a deeper reservoir, and 3) fluxing of previously degassed magma with a nearly pure COsub2 vapor. Trace element signatures of melt inclusions are consistent with contributions that have been variably metasomatized by fluids generated by dehydration of subducted sediments and/or altered oceanic crust.

The mechanisms and timescales associated with the triggering of caldera-forming eruptions remain ambiguous and poorly constrained. Do such eruptions start vigorously, then escalate, or can there be episodicity? Are they triggered through internal processes (e.g. recharge, buoyancy), or can external modulations play an important role? Key to answering these questions is the ability to reconstruct the state of the magma body immediately prior to eruption. My dissertation research seeks to answer these questions through detailed investigation of four voluminous caldera-forming eruptions: (1) 650 km3, 0.767 Ma Bishop Tuff, Long Valley, (2) 530 km3, 25.4 ka Oruanui eruption, Taupo, (3) 2,500 km3, 2.08 Ma Huckleberry Ridge Tuff, Yellowstone and (4) 250 km3, 26.91 Ma Cebolla Creek Tuff, Colorado. The main techniques I applied integrated glass geochemistry (major, trace and volatile), diffusion modeling, and detailed field sampling. In chapters two, three, and four these methods are applied to the initial fall deposits of three supereruptions (Bishop, Oruanui and Huckleberry Ridge) that preserve field-evidence for different opening behaviors. These behaviors range from continuous deposition of fall deposits and ignimbrite (Bishop), to repetitive start/stop behavior, with time breaks between eruptive episodes on the order of weeks to months (Oruanui, Huckleberry Ridge). To reconstruct the timescales of opening activity and relate this to conduit processes, I used two methods that exploit diffusion of volatiles through minerals and melt, providing estimates for the rate at which magmas ascended to the surface. This knowledge is then integrated with the pre-eruptive configuration of the magma body, based on melt inclusion chemistry, to interpret what triggered these systems into unrest. Finally, in chapter five I take a different approach by integrating geochemical data for melt inclusions and phenocryst minerals to test whether the mechanism of heat and volatile recharge often called upon to trigger crystal-rich dacitic magmas (the so-called monotonous intermediates), is applicable to the Cebolla Creek Tuff. This dissertation includes both previously published and unpublished co-authored material, and three online supplementary excel files.

Two magmatic systems were investigated using different petrological tools: 1) Origin of Ponza trachyte was studied combining data from MI with trends predicted by thermodynamic modeling. MI data were compared with known phase relations in the ternary feldspar and anorthite-diopside-albite systems to constrain the parameters used in the modeling. MI data are consistent with melt evolution from a basaltic parent via a fractional crystallization mainly of pyroxene and feldspars. These data and the results from the modeling, suggest a genetic link between the Ponza trachyte and coeval alkali olivine basalts on the nearby Ventotene Island. 2) We evaluated the range of magmatic temperatures within the crystallization interval for a basanite with different olivine-spinel geothermometers. While olivine spinel pair records the evolution of the basanite during crystallization, low temperatures calculated with the geothermometers are unrealistic. This is likely due to the presence of significant amounts of Ti in our magmatic spinels. Indeed Ti is not taken into account in the geothermometers. We tested the possibility of accounting for the presence and effects of Ti using a linear correction for the Fe+2 content in our spinels. While this generated more realistic temperatures at the low end of the range, it also increased the dispersion in the data, suggesting that spinel behavior is more complex and that the presence of Ti affects content and site occupancy of other elements as well. / Ph. D.

White Island, New Zealand, is an active andesitic volcano that is located near the southern end of the Tonga-Kermadec Volcanic Arc at the convergent plate boundary where the Pacific Plate is being subducted beneath the Indian-Australian Plate. The plate tectonic setting, volcanic features and the petrology of White Island are thought to be characteristic of the environment associated with formation of porphyry copper deposits. White Island has only been active for about 10 Ka and, as such, is thought to be an ideal location to study early magmatic processes associated with formation of porphyry copper deposits. In this study, the geochemistry of the silicate melt at White Island has been characterized through detailed studies of silicate melt inclusions, phenocrysts, and matrix glass contained in recent ejecta (1977-1991). Most melt inclusions contained only glass, however, daughter minerals present in multiphase melt inclusions in the 1991 sample indicate a different P-T history compared to the other samples. Samples studied are vesicular porphyritic andesitic dacites containing phenocrysts of plagioclase, orthopyroxene, and clinopyroxene. A glassy matrix containing crystallites surrounds the phenocrysts. Both mineral and silicate melt inclusions occur in all three phenocryst phases. Inclusions of plagioclase occur in pyroxenes and inclusions of orthopyroxene and clinopyroxene occur in plagioclase. Compositions of minerals are independent of mode of occurrence - that is, plagioclase (and orthopyroxene and clinopyroxene) compositions are the same regardless of whether they occur as phenocrysts or as inclusions in another mineral. Moreover, compositions of mineral inclusions and phenocrysts show no systematic variation within individual samples or in samples representing different eruptive events, indicating that the magma chamber is chemically homogenous over the time-space scale being sampled. Various major, trace element and volatile compositional features of economic and non-economic (or barren) porphyry copper systems were compared to the White Island data. The Al2O3/(Na2O+K2O+CaO) ratio observed in economic porphyry copper deposits is always greater than or equal to 1.3, and glass in one phase melt inclusions, as well as glass in unhomogenized (1991) inclusions from White Island equal or exceed this value. The glass in the unhomogenized 1991 melt inclusions is corundum normative, with Si/(Si+Ca+Mg+Fet)>0.91, and K/(K+Ca+Mg+Fet)>0.36, all of which are characteristic of productive systems. Melt inclusions from White Island also show a positive Eu anomaly similar to that found in productive porphyry deposits, whereas non-productive systems show a negative Eu anomaly. Copper concentrations (170-230 ppm) in melt inclusions from White Island are sufficiently high to generate an economic porphyry copper deposit based on theoretical models. High Cl/H2O ratios (0.15) in melt inclusions furthermore indicate that copper will be efficiently partitioned from the melt into the magmatic aqueous phase. The inferred pressure in the magma chamber at depth (1 kbar) is ideal for extracting copper from the melt, and mineral phases (pyrrhotite, biotite or amphibole) which could scavenge copper before it could be partitioned into the magmatic vapor phase are absent. Concentrations of S in the melt are also low, which would prevent pyrrhotite from crystallizing. The tectonic setting and geochemical characteristics of the magma body at White Island are similar to features observed in economic porphyry systems elsewhere. These data suggest that development of economic porphyry copper mineralization at White Island is likely. / Master of Science

Strand et al. (2002) discovered small metal alloy grains rich in Cu, Co, and Sn (maximum size 150 µm) in vesicles of lava from Kilauea Volcano. These alloys are also found in basaltic rocks of several Italian volcanoes. To better understand the origin of these metal-rich grains, bombs from Stromboli Volcano were examined. Two bomb types were collected from Stromboli: pumiceous bombs and scoriaceous bombs. Bulk rock trace element geochemistry indicates that there are no significant differences in Cu, Co, or Sn (the three major components of the metal alloys) between the pumiceous and scoriaceous bombs. Comparison of olivine melt inclusion and matrix glass concentrations from these rocks shows that the pumiceous bombs are more primitive (melt inclusions: MgO 2.7-5.8 wt. %; matrix glass: MgO 5.1-6.50 wt. %) and are more S-rich (melt inclusions: maximum 0.13 wt. %; matrix glass: maximum 0.06 wt. % ) than the scoriaceous bombs. The melt inclusions and matrix glass in the scoriaceous bombs are more evolved (melt inclusions: MgO 3.0-4.3 wt. %; matrix glass: MgO 2.7-3.7 wt. %) and are S-poor (melt inclusions: maximum 0.06 wt. %; matrix glass: b.d.l. ). However, Cl concentrations in melt inclusions and matrix glass are more similar for both bomb types. Metal alloys were counted in thin section for each sample. The crystallized interiors of the bombs contain more metal grains than the glassy exteriors. Pumiceous bombs (from more primitive, S-rich magma) contain more metal grains of a larger size than the scoriaceous bombs (from more fractionated, S-poor magma). This indicates that S (and Cl) are probable transport ligands for the metals in the alloys. As S (and Cl) move through the glass of an erupted cooling bomb, they complex with volatile chalcophile metals (Cu, Co, and Sn). These vapor-phase metal sulfides and chlorides move to inflating vesicles. Here the sulfide and chloride complexes become reduced and metal alloys condense, as S and Cl escape as gas. Non-degassed primitive magma may provide more S (but not necessarily more metals) to create the higher abundance of alloys hosted by the vesicles of the pumiceous bombs.