Oliver John Higgins, Michael Stock

Melt-based thermobarometers are essential tools to recover pre-eruptive magma storage conditions through their application to bulk rock and liquid chemistry. In active volcanic systems, thermobarometric results can be combined with independent geophysical data during or after an eruption to validate conceptual models. In this contribution, we revisit the thermobarometer for melts equilibrated with the mineral assemblage of olivine + plagioclase + augitic clinopyroxene (OPAM). We first demonstrate that the most widely applied OPAM thermobarometer suffers from both random and systematic uncertainty even for anhydrous melts, and that the uncertainty increases proportionally with melt H2O. To address this issue, we use a modern compilation of anhydrous and hydrous OPAM-saturated experiments to regress a new melt-based OPAM thermometer and barometer. Our new equations recover a validation dataset with a standard error estimate (SEE) of ±1.1 kbar and ±35 ˚C for pressure and temperature respectively, as well as a low systematic uncertainty that does not depend on melt H2O. Additionally, we present a novel statistical approach to determine the probability that a given melt is OPAM-saturated, which can be used alongside petrographic observations. Our thermobarometer and saturation test are presented as a straightforward R script which reads from an input csv file to be populated with natural data. We benchmark the new calibrations on the products of the 2015 eruption of Wolf Volcano (Isabela Island, Galápagos) and the 2014-2015 Holuhraun eruption (Iceland), both of which have independent geophysical estimates of magma storage that agree well with our thermobarometric results.