Item

Abstract

Combining multiphysics numerical models with observations of volcanic unrest is critical for evaluating the evolution of magma systems, their potential for eruption, and eruption triggering mechanisms. Thermomechanical finite element method (FEM) models provide estimates of the magma system stress evolution and stability through time, while data assimilation provides a statistical framework for evaluating model results and linking them with observations. Here we present recent advancements in statistical data assimilation for combining ground deformation with thermomechanical FEMs to provide forecasts of magma chamber evolution and stability. We illustrate how the Ensemble Kalman Filter (EnKF) has been adapted to assimilate geodetic observations of surface deformation into multiphysics FEMs and track the evolution of stress in volcanic systems. Stability and eruption potential is evaluated using model estimates of magma reservoir overpressure, tensile stress and failure along the calculated magma reservoir boundary, and Mohr-Coulomb failure in the host rock. The EnKF approach not only provides a robust framework for evaluating near-real time volcanic unrest, it also allows for investigating historical eruptions via hindcasts to determine precursors and triggering mechanism. We discuss recent successes implementing the approach to hindcast the 2008 eruption of Okmok Volcano, Alaska (Albright et al., GRL, 2019), and forecast the 2018 eruption of Sierra Negra Volcano, Galápagos (Gregg et al., Sci. Adv., 2022). Our findings demonstrate the potential for utilizing the EnKF – FEM approach to investigate eruption potential and triggering mechanisms at restless volcanoes worldwide.