Near-real-time tracking of volcanic systems using ground deformation data

In order to forecast volcanic eruptions, it is important to track not only the state of the volcanic system in near-real-time, but also the conditions that could possibly trigger an eruption. Here we present a simple and computationally efficient framework to track the pressure state of a shallow magmatic system and the parameters (e.g. source bottom pressure, magma influx rate, and damage state) that may accelerate or trigger an eruption. Our approach is a model-data fusion technique where we combine process-based volcano dynamical models (i.e. Got et. al., 2017; Reverso et. al., 2014; Lengliné et. al., 2008), Bayesian inversion, and data assimilation. Bayesian inversion is used to constrain non-evolving model parameters, whereas data assimilation through the Ensemble Kalman Filter (EnKF) is utilized to track the state of the volcanic system in near-real-time. We will introduce the Caltech/JPL-developed AlTar open-source Bayesian inversion code, which was initially developed for earthquake kinematic rupture models, and has now been extended to model volcano deformation sources. As AlTar is designed to sample high-dimensional problems in parallel (Minson et. al., 2013) and the EnKF only uses incoming and available data, our framework can be used to accommodate and model large InSAR and GNSS data sets in real-time. We present synthetic performance tests which track the rapid acceleration of deformation prior to eruption and provide insights to possible mechanisms catalyzing an eruption, such as for the 2005 and 2018 eruptions of Sierra Negra volcano, Galapagos, Ecuador.

References:

[1] Got JL, Carrier A, Marsan D, Jouanne F, Vogfjörd K, Villemin T. An analysis of the nonlinear magma-edifice coupling at Grimsvötn volcano (Iceland). Journal of Geophysical Research: Solid Earth. 2017 Feb;122(2):826-43.

[2] Lengliné O, Marsan D, Got JL, Pinel V, Ferrazzini V, Okubo PG. Seismicity and deformation induced by magma accumulation at three basaltic volcanoes. Journal of Geophysical Research: Solid Earth. 2008 Dec;113(B12).

[3] Minson SE, Simons M, Beck JL. Bayesian inversion for finite fault earthquake source models I—Theory and algorithm. Geophysical Journal International. 2013 Jun 20;194(3):1701-26.

[4] Reverso T, Vandemeulebrouck J, Jouanne F, Pinel V, Villemin T, Sturkell E, Bascou P. A two-magma chamber model as a source of deformation at Grímsvötn Volcano, Iceland. Journal of Geophysical Research: Solid Earth. 2014 Jun;119(6):4666-83.