Magmatic volatile content and the overpressure ‘sweet spot’: Implications for volcanic eruption triggering and style

Volatile exsolution is widely considered to be capable of generating magmatic overpressure and triggering volcanic eruptions. Despite its role as an eruption trigger, exsolution-driven overpressurisation is relatively poorly understood. Part of the problem is that thermodynamic models do not consider how the behaviour of small quantities of magma scales up to reservoir level – where variations in temperature and crystallinity become important. In contrast, many thermomechanical models focus only on magma injection, and do not consider how overpressure evolves spatially or temporally, when related to crystallisation and volatile exsolution. Here, we use Rhyolite-MELTS to track exsolution-driven overpressure during cooling for a variety of natural compositions, storage pressures, initial volatile contents and magmatic XH2O (molar H2O/ (CO2 + H2O)). We then couple these outputs to a thermal model to determine the timescales and spatial extent of overpressurisation with varying volatile content.

We find that the highest overpressures occur in magmas which are initially at their H2O solubility limit, with the addition or removal of H2O resulting in a decrease in peak overpressure. We also find that maximum overpressure decreases with the addition of CO2 (decreasing XH2O) at typical magma storage pressures of 100–230 MPa. The higher overpressures generated at the volatile ‘sweet spot’ have a greater potential to trigger eruptions – or to favour their initiation by making the system more susceptible to other triggers, such as magma injection. The reduction in overpressure with increasing or decreasing initial H2O suggests that triggering by volatile exsolution is less likely for these magmas. Peak overpressure at the volatile sweet spot also coincides with an increased incidence of explosive eruptions at water contents ∼4–5.5 wt%. This suggests that higher magmatic overpressures may produce more explosive eruptions, by driving faster initial ascent rates and decreasing outgassing efficiency in the conduit. Our thermal modelling demonstrates that, for small magmatic systems, exsolution-driven overpressurisation can operate on timescales which are much shorter than the crustal relaxation timescale. In these cases, overpressure cannot be dissipated by a visco-elastic crustal response, and therefore has the potential to trigger a volcanic eruption.