Tales from supereruptions: Combining pumice and mineral textures with phase equilibria to constrain the evolution of giant silicic magma bodies in the crust

Friday, 19 December 2014
Guilherme A R Gualda1, Ayla S Pamukcu2, Kylie A Wright1, Mark S Ghiorso3 and Calvin F Miller1, (1)Vanderbilt University, Earth and Environmental Sciences, Nashville, TN, United States, (2)Brown University, Earth, Environmental, and Planetary Sciences, Providence, RI, United States, (3)OFM Research, Redmond, CA, United States
Supereruption deposits demonstrate that giant magma bodies sporadically exist within the Earth’s crust. We rely on study of such deposits to better understand the underlying magma bodies and their eruptions.

We are studying several deposits: Bishop Tuff (BT, CA USA), Peach Spring Tuff (PST, SW USA), and Oruanui Tuff (OT, NZ). We combine quantitative textural characterization in 3D via x-ray tomography (XRT), focusing particularly on CSDs of major and accessory minerals; characterization of mineral zoning, particularly of Ti and CL in quartz, including inferences from diffusion chronometry; documentation of glass inclusion textures in 3D via XRT, with implications to crystallization timescales; and phase equilibria modeling (rhyolite-MELTS), including glass (inclusion and matrix) composition geobarometry, to constrain crystallization conditions.

CSDs ubiquitously record a growth-dominated regime, characterized by limited nucleation, consistent with pre-eruptive crystallization under low supersaturation. Phenocryst interiors are largely unzoned, consistent with phase equilibria predictions of nearly invariant, effectively isothermal crystallization. Glass compositions record storage over a large spread of depths (~125-250 MPa) for early and late-erupted BT, while the OT represents multiple magma batches evacuated from different depths. Diffusion chronometry and melt inclusion faceting suggest pre-eruptive crystallization over centennial timescales.

CSDs and mineral textures also record syn-eruptive crystallization, which results in huge numbers of small crystals, revealing extensive nucleation prior to eruption. Crystal rims develop on pre-existing phenocrysts, and they can be obvious if compositionally distinct from interiors (BT and PST). In PST, evidence for rim crystallization from hotter magma is very strong. BT contrasts with PST in many ways; evidence for heating is ambiguous, and pumice properties are difficult to reconcile with magma mixing, while the effects of growth rate and decompression are underappreciated. Diffusion chronometry indicates rim crystallization over the final year prior to eruption, under high supersaturation, during eruptive decompression; this reveals the timeframe over which signs of an impending supereruption may be felt at the surface.