Abstract
In order to understand the cause of a caldera-forming eruption, it is crucial to study the pre-eruptive processes, from build-up to the collapse. In recent years, the cyclicity of large-scale magmatic systems and their corresponding caldera collapses has been described through various geochemical and petrological studies. Numerical modelling studies have investigated the collapse stage using a pressurised cavity or gas-filled chamber, studying the evolution of the surrounding lithosphere by inflating and deflating the chamber. Another modelling approach involves investigating the displacement effects of moving a piston-shaped object. In this study, we use a multi-physical numerical modelling approach to investigate different aspects of the caldera cycle, including the accumulation of magma, the evolution of stress, the strain rate and the dynamics within the forming magmatic system. To achieve this, we couple an open-source thermal evolution magma intrusion code with a Stokes solver. The coupling allows for the evaluation of thermal and volumetric changes in the magmatic system, taking into account the complex interaction between magma dynamics and the surrounding host rock. The use of nonlinear visco-elasto-plastic rheology enables the assessment of fracturing and thermal effects on the lithosphere, which create downward propagating weak zones that are later utilised as eruption channels. These faults not only provide pathways but are also associated with the caldera collapse stage of the eruption cycle, which leads to geomorphological transformations of the surface. Our approach is also applicable for assessing the evolution of current magmatic systems taking into account their topography. The model is applied to the well-studied area of the Toba caldera in Sumatra. The magma is injected into areas where seismic inversion studies have identified a melt phase. This allows for an assessment of the system's evolution and provides insights into an active magmatic system. Additionally, it serves as a reference for correlating and referencing our model with observational data.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have