Explosive or effusive? Combining experimental petrology with monitoring data to understand the magmatic system at Kelud volcano, Indonesia
One of the challenges in understanding volcanoes is to explain why eruptive style changes from one eruption to the next, or even within the same eruptive period. I will review the different ways of producing both explosive and effusive eruptions from silicic magmas. Eruptive style depends on a set of interrelated magmatic properties, such as viscosity, processes such as gas loss and external properties such as conduit geometry, which altogether initiate various feedbacks. Ultimately, these control the speed at which magmas ascend, decompress and the extent of outgassing on route to the surface, this determines the eruptive style and how it evolves. One such volcano that displays this dichotomy of eruptive behaviour is Kelud volcano in East Java, Indonesia. Eruptions from Kelud are difficult to forecast, because the eruptive style varies considerably, from effusive eruptions e.g. 1920 & 2007 to explosive eruptions in 1990 and 2014, despite near identical bulk rock compositions. Experiments were undertaken to constrain the magma storage conditions such as pressure, temperature and H2O-CO2 ratios prior to both explosive and effusive eruptions at Kelud and compared to monitoring data before the eruptions. The experiments suggest the last phase of magma storage for explosive eruptions is shallow (2-3 km), cold (~1000°C) and water-saturated, whereas CO2 fluxing during recharge prior to effusive eruptions, heats the magma, suppresses water contents, and lowers explosivity. No significant pre-eruptive deformation was recorded via InSAR prior to the 2014 explosive eruption, showing that explosive eruptions can occur without recent magma injections or inflation akin to the eruption of Calbuco volcano in 2015. Instead, the chemical degassing signature may be key to understanding the state of magma underneath the volcano.