Syn-eruptive “rehydration” of ash at Redoubt Volcano, Alaska during eruption through a glacier

Abstract:

Magma-water interactions can increase explosivity and are generally thought to aid in rapid quenching, yet even short timescales of cooling in the presence of water may allow for partial rehydration or reequilibration of H isotopes in glass. We analyze tephras from the well-sampled, month-long 2009 eruptive series of Redoubt Volcano to understand 1) to what extent magmatic degassing of water continues in an ash plume and/or 2) if hot, vaporized meteoric H₂O – from Drift glacier in this case – can partially reset magmatic H₂O and δD in tephras. Volcanic degassing drives melt δD values to increasingly low values as H₂O exsolves and decreases in the melt. If degassing continues in an eruptive ash plume, then H₂O and δD in tephras should decrease with increasing distance from the vent. Our sample suite (n=25) instead shows δD increasing away from the vent as H₂O decreases. Tephras recording the highest H₂O (0.57 wt.%) and lowest δD (−127‰) are medial (11 km). Water concentrations decrease by ~50% and increase by >10-15‰ δD in more distal samples (<120 km). This indicates that 2009 Redoubt tephras cannot be degassing with increasing residence time in the plume and/or atmosphere. Medial samples appear to be effected by high temperature, short duration interactions with low δD meteoric water. These tephras may have partially rehydrated on the order of minutes at temperatures of 300-400°C as has been demonstrated in submarine eruptions [1]. Alternatively, the tephras could have reequilibrated with vapor produced from glacial meltwater [2]. Giachetti et al. (2020) demostrate that H isotope reequilibration can occur without increases in total H₂O in obsidian pyroclasts on the walls of shallow volcanic conduits. These results demonstrate that tephras are neither completely nor homogeneously degassed during eruption and prior to deposition. Furthermore, these results demonstrate that H₂O in in glass is sensitive to complex processes in the shallow conduit and during eruption at the vent. [1] Mitchell, McIntosh, Houghton, Carey, and Shea (2018), EPSL, 494, 135-147. [2] Giachetti, Hudak, Shea, Bindeman, and Hoxsie (2020), EPSL, 530: 115909.