Internal triggering of volcanic eruptions: tracking overpressure regimes for giant magma bodies

Understanding silicic eruption triggers is paramount for deciphering explosive volcanism and its potential societal hazards. Here, we use phase equilibria modeling to determine the potential role of internal triggering - wherein magmas naturally evolve to a state in which eruption is inevitable - in rhyolitic magma bodies. Whole-rock compositions from five large to super-sized rhyolitic deposits are modeled using rhyolite-MELTS. By running simulations with varying water contents, we can track crystallization and bubble exsolution during magma solidification. We use simulations with variable enthalpy and fixed pressure for the five compositions. The interplay between bubble exsolution and crystallization can lead to an increase in the system volume, which can lead to magma overpressurization. We find that internal triggering is possible for high-silica rhyolite magmas crystallizing at pressures below 300 MPa (<11 km depth in the crust), revealing a window of eruptibility within the upper crust from which high-silica eruptions emanate. At higher pressures, the critical overpressure threshold for eruption is only reached once crystallinities are high, >50 wt.%, which makes magma immobile - high-silica rhyolite eruptions from such depths would require external triggering, but examples are scarce or entirely absent. Calculated crystallinities at which the critical overpressure threshold is reached compare favorably with observed crystal contents in natural samples for many systems, suggesting that internal evolution plays a critical role in triggering eruptions. Systems in which fluid saturation happens late relative to crystallization or in which degassing is effective can delay or avoid internal triggering. We argue that priming by crystallization and bubble exsolution is critical for magma eruption, and external triggering serves simply as the final blow - rather than being the driving force - of explosive rhyolitic eruptions. (C) 2017 Elsevier B.V. All rights reserved.