Skip to content

Linking volcanic plume dynamics with sedimentation processes using a multi-GPU accelerated Lattice Boltzmann solver.

Jonathan Lemus1,2 , Riccardo Simionato1,2, Christophe Coreixas2,3, Jonas Latt2, Costanza Bonadonna1

  • Affiliations: 1Department of Earth Sciences, University of Geneva, Geneva, Switzerland; 2Department of Computer Science, University of Geneva , Geneva, Switzerland; 3Department of Mechanics and Aerospace Engineering, Southern University of Science and Technology, Shenzhen, China

  • Presentation type: Poster

  • Presentation time: Thursday 16:30 - 18:30, Room Poster Hall

  • Poster Board Number: 91

  • Programme No: 3.12.17

  • Theme 3 > Session 12

Abstract

Explosive eruptions release significant amounts of volcanic ash (tephra) into the atmosphere, posing serious threats to communities. Tephra can harm human health, damage infrastructure, pollute ecosystems, and disrupt economic and transport systems. While numerical dispersion models have advanced considerably, gaps remain in understanding processes linked to premature tephra sedimentation, such as particle aggregation and gravitational instabilities. Tephra dispersion and sedimentation depend heavily on source conditions at the volcano vent, requiring a detailed understanding of complex eruption dynamics. Explosive eruptions involve compressible flows, potential transonic behaviors, buoyancy effects, multiphase interactions, and intense turbulence during plume ascent. To explore the influence of the plume dynamics on the sedimentation, we developed a numerical model simulating volcanic plume dynamics from initial stages to wind-driven large-scale dispersion. The model leverages the Lattice Boltzmann Method (LBM) for accurately simulating complex flows and integrates low-diffusive finite difference schemes for effective transport modeling of species in the fluid. A notable feature of our approach is its multi-GPU computing capability, enabling efficient, high-performance 3D simulations on large scales. Model validation against experimental data, including turbulent and thermal jets, demonstrates strong stability and accuracy across various configurations. These results enhance our understanding of volcanic plume dynamics and improve parameterization in operational models, aiding better prediction and mitigation of volcanic hazards.