V51B-4741:
What Caused the 2001-2002 Unrest at Cotopaxi Volcano, Ecuador? Insights from a Finite Element Based Geodetic Inversion
Friday, 19 December 2014
James Hickey, University of Bristol, Bristol, BS8, United Kingdom, Joachim Gottsmann, University of Bristol, School of Earth Sciences, Bristol, United Kingdom and Patricia A Mothes, Instituto Geofisico, Quito, Ecuador
Abstract:
A general inflation of the edifice and increased long-period/very-long-period seismicity define the 2001-2002 period of non-eruptive unrest at Cotopaxi volcano, Ecuador. This study focuses on the observed deformation – simultaneous contraction of seven baselines recorded by an electronic distance meter (EDM) network. To determine the cause of this deformation we model the system using Finite Element analysis with COMSOL Multiphysics. Our models incorporate subsurface heterogeneity, real topography and represent the source as a spheroidal cavity. This set up allows the EDM baselines to be modelled in three dimensions and account for the steep relief of the iconic stratovolcano, as opposed to analytical models that are either restricted to two dimensional EDM calculations and/or a flat Earth surface. To further assess the importance of topography, subsurface mechanics, and the 2-or-3D approach, we conduct a sensitivity analysis using both Finite Element and analytical techniques. We solve the Finite Element inverse problem with a least-squares approach, searching for the optimum location (longitude, latitude, depth) and over-pressure of the source to fit the EDM deformation data within its error. This optimization procedure was repeated for each source shape, orientation, size and aspect ratio using a series of nested parameter constraint grids. All source shapes converge on a location beneath the south to south-west of the edifice at a central depth of 0.5 – 2.0 km above sea level (summit at 5897 m). High-eccentricity oblate spheroids generally provide the best-fit to the observed data and may be interpreted as a sill-like intrusion as the cause of the deformation. Finally, additional forward Finite Element models are used to assess the implications of inelastic rheology, failure locations and gravity anomalies associated with the intrusion.