[ Back ]


Lattice-Boltzmann model applied to volcanology

Author Andrea Parmigiani
Director of thesis
Co-director of thesis
Summary of thesis Large volcanic eruptions are spectacular natural events, which display power outputs exceeded only by asteroid impacts. Despite their destructiveness, the risk perceived by modern-day societies is strongly underevaluated due to the relatively long recurrence intervals between eruptions when compared to the typical human memory timescale; the last major eruption that had a global impact on climate and killed more than 100.000 people occurred in 1815 (Tambora, Sumbawa, Indonesia). Such eruptions occur on average every 150-200 years, and the next one will have devastating effects on human infrastructures and population. A better understanding of the physics of volcanic eruptions and more predictive capabilities are urgently needed to reduce damages in the future. Since the advent of powerful computers a few decades ago, numerical simulations of natural phenomena have been one of the prime tools to help human societies protecting itself against natural disasters (hurricanes, tsunamis, floods, earthquakes, volcanic eruptions, ...) and gain a deeper knowledge about complex phenomena involving a larger number of free parameters. In weather forecasting, such numerical simulations are becoming extremely sophisticated and useful. In volcanology, complex models are also becoming available, but due to the greater difficulty of visualizing the internal plumbing of volcanic edifices and obtaining accurate measures of physical properties in magmas, the degree of sophistication has not reached the level of physical models for surface and atmospheric processes. However, geological data for volcanic systems is rapidly becoming available at present and using the knowledge acquired by computer scientists over the last decades, we appear to be at a particularly favourable time to generate such models. Our plan is to develop incrementally, over the next three years, Lattice-Boltzmann codes that we allow us to tackle multiple issues related to volcanology. In order to get results soon, we will start by validating and using a 2D code simulating a gas phase flow through a complex porous medium determined on natural volcanic rocks by X-ray tomography (application to gas escape from permeable magmatic foam). In parallel, we will develop a 2D code with variable temperature that can handle phase changes (crystallization, melting) in order to better understand how magma behave in crustal reservoirs (magma chambers). par For both of these applications, literature review and code foundation have already been laid out. Finally, we will construct a 3D code that will couple variable temperature fields with phase changes and multi-phase flow in porous media to better understand complex multiphase flow in magma chambers and provide predicting capacities on how volcanoes reawaken after periods of quiessence.
Administrative delay for the defence