Time-series analysis of geodetic data at the Laguna del Maule volcanic field, Chile, (2007-2017)

Abstract:

The Laguna del Maule (LdM) volcanic field in Chile continues to deform extremely rapidly. The total vertical displacement over the last ten years now exceeds two meters! Since 2007, the maximum vertical velocity has exceeded 200 mm/yr, as estimated from two types of geodetic data: interferometric synthetic aperture radar (InSAR) and the Global Positioning System (GPS) [Le Mével et al., 2015]. The rate of uplift at LdM is among the highest ever observed geodetically for a volcano that is not actively erupting. In this paper, we analyze InSAR data from several satellite missions, including the ENVISAT and SENTINEL-1 missions operated by the European Space Agency (ESA), the ALOS missions operated by the Japanese Space Exploration Agency (JAXA) and the TerraSAR-X and TanDEM-X missions operated by the German Space Agency (DLR). To compute the pair-wise combinations that map the deformation field, we employ a semi-automatic, self-consistent workflow based on the GMTSAR5 software [Sandwell et al., 2011]. To model the observed deformation field, we employ the General Inversion of Phase Technique — GIPhT [Feigl and Thurber, 2009] implemented in open-source software [https://github.com/feigl/gipht]. To evaluate the first and second temporal derivatives of displacement, we perform time-series analysis using a general approach named temporal adjustment that is based on graph theory and implemented in the open-source software GraphTreeTA [Reinisch et al., 2016]. To verify the time series of displacement estimated from the remotely sensed InSAR data, we compare them to time series of relative position estimated from the GPS network at LdM operated by OVDAS (including seven stations occupied intermittently and five stations recording continuously) using the GIPSY software [https://gipsy-oasis.jpl.nasa.gov]. This approach allows rigorous statistical testing of hypotheses to describe the geophysical processes that drive the deformation, including: (1) an empirical “double-exponential” model [Le Mevel et al., 2015]; (2) a dynamic model of viscous magma flowing through -- and interacting with -- a conduit through the crust [Le Mével et al., 2016], and (3) a dynamic model of (magmatic or aqueous) fluids flowing through a poroelastic medium, as proposed by Le Mével, Gregg, and Feigl [this Chapman Conference].