Date of Award

Fall 1-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Geology and Geophysics

First Advisor

Ague, Jay

Abstract

The production and transport of metamorphic fluids are key to a range of Earth processes, including seismogenesis, global geochemical cycling, and the maintenance of planetary habitability over geologic time. It is therefore of great importance to understand the distribution of fluids in the lithosphere and how their interactions with rocks along deep flow pathways affect the transport of volatiles and dissolved species in metamorphic systems. This dissertation aims to trace metamorphic fluid sources and pathways in various petro tectonic settings, and to assess how fluid-rock interactions modify fluid compositions, with the goal of better understanding the processes underlying the global cycling of volatiles and elements of societal and geologic interest. Following the introductory Chapter 1, Chapter 2 uses anisotropy-aware receiver function analysis of seismic data to identify the presence of water in an active subduction zone. In particular, we compare results from a typically dipping segment of the Alaska subduction zone beneath the southwestern Kenai peninsula with the adjacent, nearly flat slab segment associated with the subduction of the Yakutat terrane. Our results provide evidence for hydrous conditions above the slab in both segments. This suggests that metamorphic dehydration and the transfer of slab-derived fluids to the narrow “mantle wedge” can be observed above flat slabs, and sheds further light on the distribution of volatiles in flat slab settings. Chapter 3 uses a numerical model to link the formation of a distinctive, spike-shaped geochemical profile across a partially altered marble layer in an exhumed subduction zone section on Syros, Greece, to carbonate dissolution during subduction. Previous work has suggested that carbonate dissolution may make considerable contributions to subduction zone CO2 fluxes, so it is important to identify where it has occurred and to understand the fluid flow processes driving it. The modeled profile is characterized by a spike in Sr concentration at the reaction front between a precursor marble and the resulting altered calc-silicate. Sr concentration is uniform on the altered side behind the reaction front, and tapers off to the baseline precursor marble value ahead of the reaction front. Model results show that this spike shape can be formed by carbonate dissolution processes. Furthermore, we suggest that the primary fluid flow direction must have been orthogonal to the propagation of the reaction front, and was likely directed by layering and/or folding. Chapters 4–6 are focused on a mélange in northeastern CT, USA, that is cut by a large network of amphibolite facies silicate-carbonate veins. Although the origin of the mélange itself is unclear, the veins likely formed during collisional orogenesis, and thus record fluid transport and fluid-rock interaction in this setting. These chapters address the transport and mineralization of REEs and CO2 in this metamorphic fluid flow system, and also place constraints on the geochemistry of the throughgoing fluid and the timing of fluid infiltration. Chapter 4 examines the role of garnet in sequestering heavy REEs (HREEs) from deep fluids during fluid-rock interaction. In this mélange, garnet growth in silicate-carbonate veins is widespread in some metanorite host blocks, but is uncommon in metagabbro or metaorthopyroxenite host blocks. Mass balance calculations reveal that garnet growth in some metanorite blocks is associated with net HREE gain, suggesting that the garnet sequestered these elements from the throughgoing fluid. These net gains are absent in garnet-free veins hosted in metagabbro and metaorthopyroxenite host blocks. Additionally, some garnet-bearing veins in metanorite host blocks preserve evidence for local redistribution of HREEs, implying that garnet may prevent locally mobilized HREEs from leaving the system. If HREE sequestration in garnet is widespread in the lithosphere, garnet may play a major role in the lithospheric HREE budget and could negatively impact HREE availability for ore-forming processes. Chapter 5 aims to constrain the processes driving CO2-bearing fluid infiltration and carbonate mineralization in veins and selvages, as well as the amount of CO2 sequestered in the vein system. Fluids from the matrix were introduced to block interiors by fracturing. Petrographic observations suggest that fluid infiltration outward from veins into host rocks, and associated exchange of material between veins and developing altered selvages, was facilitated by fluid migration along grain boundaries and development of nanoporosity during mineral dissolution-reprecipitation reactions. Thermodynamic modeling of vein and selvage mineral assemblages indicates that the XCO2 of the vein-forming fluid must have been ~0.5 or greater. Mass balance calculations reveal that veins and selvages sequestered ~150–450 kg CO2 per m3 rock, indicating that these rocks had the potential to be efficient CO2 traps. However, fluid infiltration in the mélange blocks was primarily localized in the veins + selvages, which are only ~10% of the blocks by volume. These rocks, therefore, did not reach their full carbonation potential, emphasizing that the extent of CO2 mineralization is also influenced by fluid infiltration mechanisms and how pervasively the fluid can access the host rock. Chapter 6 combines a range of isotopic tracers (?34S in pyrrhotite and pyrite, ?18O and ?13C in carbonate minerals, and ??18O and ??17O in quartz and kyanite) in vein and selvage minerals to better constrain the source of the mélange fluid. The stable isotope signatures of these vein and selvage phases are generally independent of host block type, suggesting that these signatures are largely controlled by the fluid composition. Furthermore, these isotopic signatures reflect a mixture of sources, and are consistent with incorporation of mélange block material into the matrix during deformation and fluid-rock interaction. This is supported by ?Hfi values obtained for zircons from a highly altered matrix-block rind sample, which generally fall between the values obtained for host rock zircons. U-Pb zircon geochronology indicates that peak mélange fluid activity occurred c. 350 Ma, providing novel evidence for extensive fluid infiltration and fluid-rock interaction in the mélange during Neoacadian orogenesis.

Download

Share

COinS