A model for the development of stable isotopic water signatures of tephra deposited on ice following subglacial caldera collapse

Abstract

Large-scale silicic volcanism has occurred frequently in regions mantled by thin to very thick glacial ice cover, with several notable examples during the Pleistocene (Yellowstone, Long Valley, Iceland), including high altitude activity during the most recent glacial (e.g. Cascades, Kamchatka, Andes). More ancient caldera-forming events must have occurred during episodes of long-lasting (ca. 10–50 Ma) snowball Earth glaciation in the Neo- and Paleo-proterozoic, and many extraterrestrial eruptions on other ice-covered planets and moons are also of this general form. Recent work suggests that the process of caldera collapse typically lasts hours to weeks, during which the caldera floor drops by several hundreds to thousands of meters and is covered by hot tephra. With this scenario in mind, we investigate glaciovolcanic interactions immediately following the deposition of thick, hot ash layers on ice and consider the destiny of buried tephra and melting ice inside calderas. Our focus is drawn in particular to the post-emplacement hydration of volcanic glasses, with the goal of assessing whether the δ18O, δ17O and δD signatures in ancient deposits might be used to infer the syn-eruptive climate state. Scaling arguments, augmented by an idealized 1D model, suggest that ice should often survive for several decades or centuries as an active meltwater source to the overlying cooling intracaldera tuff. As liberated glacial water (both liquid and vapor) infiltrates and interacts with the tephra layer, volcanic glasses can become fully hydrated to water saturations of several wt.%. Our theoretical treatment is motivated in part by our recent measurements of lower than modern δD values in products of several Pleistocene eruptions in the western U.S. occupying regions that were likely glaciated immediately prior to the emplacement of volcanic products. We discuss how δ18O–δD and δ18O–Δ17O systematics can be used to recognize syn-glacially hydrated intracaldera tephra, potentially including samples that have been buried, altered and subsequently exposed either by fault uplift and erosional exposure or by drilling operations, such as those being performed currently in the Central Snake River Plain near Yellowstone.