Magma Genesis, Degassing, and Mixing in Rifts and Arcs.

Abstract

Our understanding of igneous petrogenesis is important for a variety of reasons, such as our ability to identify and understand volcanic hazards, ore deposits, and fundamental tectonic processes. This dissertation investigates two aspects of igneous petrogenesis in two distinct tectonic environments: (i) the relationship between mantle volatiles and alkaline lavas in rift environments and (ii) the formation of intermediate magmas in arc environments. Chapter II presents some of the first volatile concentrations from olivine-hosted melt inclusions in the western branch of the East African Rift System. The melt inclusion H2O and CO2 concentrations measured range from ~0.3 to 2.5 wt% and ~30 to 9,950 ppm, respectively, and have elevated Li concentrations and B/Be ratios relative to MORB. Elevated Li and B concentrations have been used as a fluid tracer to investigate the role of fluid additions to the mantle wedge. As such, volatiles subducted during the ~650 Ma Pan-African orogeny are a plausible source for the elevated volatiles. In Chapter III we investigate the plausibility that mixing of basaltic and dacitic magmas with significantly different initial temperatures and viscosities can mix to produce basaltic andesite erupted from Mutnovsky Volcano, Kamchatka. Plagioclase compositions for Mutnovsky basalts, basaltic andesites, and dacites fall into two distinct populations, An80±10 and An50±15. Basaltic andesites contain both plagioclase populations with a distinct gap between populations. Sodic plagioclase in the basaltic andesites show dissolution/resorption textures, indicating disequilibrium. These new data are consistent with mixing of a basaltic and dacitic component to generate the erupted intermediate lavas at Mutnovsky. Chapter IV presents a quantitative model to describe the evolving viscosity of different magmas (melt + crystals, including water content) and determine if these magmas can physically mix for any proportion of each. We test this model with the samples from Mutnovsky as well as published data from Mt. Hood, Oregon, where mixing of basalt and rhyolite is proposed to generate andesites. These results demonstrate that both Mutnovsky and Mt. Hood magmas satisfy all requirements to generate intermediate magmas by magma mixing and that this model can be applied to any volcanic system to determine the likelihood of mixing.