Silicic caldera volcanoes are often spatially associated with hydrothermal systems that are economically important for geothermal power and the localisation of ore deposits, and also influence their restless behaviour and associated hazards. However, despite their potential importance, the influence that caldera-related structures, lithologies and magmatism have on controlling hydrothermal fluid pathways, and the physiochemical conditions of the fluid is not fully understood. Ancient, exhumed calderas provide an opportunity to examine fossil fluid pathways in a more complete structural, lithological and magmatic context than is possible in poorly-exposed modern calderas. In this thesis I use a combination of field mapping, scanline transects, reflectance spectroscopy, alteration mineralogy, vein textures and fluid inclusion microthermometry to reconstruct and better understand the volcanotectonic controls on hydrothermal fluid flow at the 22.9 Ma Lake City caldera in Colorado, U.S.A.

Field mapping, scanline transects and petrographic analyses are used in Chapter 2 to characterise the architecture of the caldera margin and the structural and lithological controls on the distribution of veins and alteration at Lake City. The caldera margin consists of relatively straight segments linked by more structurally complex intersections; these structural intricacies produce a zone of deformation that can reach >300 m wide. Structural analyses show that the wide (up to ~60 m) fault core of the ring fault contains abundant subparallel veins, with orientations similar to that of the caldera margin. Smaller displacement faults inside the caldera generally have narrow (<1 m), hydrothermally cemented fault cores with more variably oriented veins in the surrounding damage zone. These field data suggest that fluid flow is controlled by fault connectivity, the location and displacement of faults, and lithology. I propose a conceptual model where permeability is enhanced by: 1) the presence of permeable lithologies, 2) a high density of faults and fractures, and 3) orientations of faults and fractures that promote the formation of permeable discontinuity intersections.

Secondary alteration minerals form due to interaction between country rock and hydrothermal fluid. Mineral assemblages and compositions can be used to estimate physiochemical parameters of the hydrothermal system (e.g. pH, temperature, water/rock ratio) that are important for geothermal and ore exploration. The composition of hydrothermal white mica is sensitive to changes in physiochemical conditions. Short wave infrared (SWIR) reflectance spectroscopy is a well-established tool for investigating alteration mineralogy, including white mica composition. However, the newest high spatial resolution, automated systems, such as the Corescan HCI-3, are poorly represented in the literature compared to older low spatial resolution systems, such as the ASD TerraSpec. In Chapter 3, I compare the performance of Corescan and TerraSpec SWIR systems in measuring the composition of white mica, as estimated using the wavelength position of the ~2200 nm AlOH absorption feature (lAlOH). The Corescan and TerraSpec correlate well with each other, although there are small absolute differences that should be taken into account if combining data from both systems. The Corescan results for lAlOH correlate slightly better than TerraSpec with the aluminium content of white mica as determined by scanning electron microscope (SEM) energy-dispersive x-ray spectroscopy (EDS). The spatial distribution of white mica composition at Lake City caldera suggests that high-Al white mica generally correlates with quartz-sericite-pyrite alteration and low d18O compositions in the centre of the caldera, although there are significant deviations from this pattern on the western caldera margin. These results confirm the usefulness of SWIR reflectance spectroscopy and white mica composition as a tool for studying hydrothermal alteration.

infrared (SWIR) reflectance spectroscopy is a well-established tool for investigating alteration mineralogy, including white mica composition. However, the newest high spatial resolution, automated systems, such as the Corescan HCI-3, are poorly represented in the literature compared to older low spatial resolution systems, such as the ASD TerraSpec. In Chapter 3, I compare the performance of Corescan and TerraSpec SWIR systems in measuring the composition of white mica, as estimated using the wavelength position of the ~2200 nm AlOH absorption feature (lAlOH). The Corescan and TerraSpec correlate well with each other, although there are small absolute differences that should be taken into account if combining data from both systems. The Corescan results for lAlOH correlate slightly better than TerraSpec with the aluminium content of white mica as determined by scanning electron microscope (SEM) energy-dispersive x-ray spectroscopy (EDS). The spatial distribution of white mica composition at Lake City caldera suggests that high-Al white mica generally correlates with quartz-sericite-pyrite alteration and low d18O compositions in the centre of the caldera, although there are significant deviations from this pattern on the western caldera margin. These results confirm the usefulness of SWIR reflectance spectroscopy and white mica composition as a tool for studying hydrothermal alteration.