Imaging the magmatic system at the Laguna del Maule Volcanic Field, central Chile (36° S, 70° W) using magnetotellurics and surface wave tomography: Comparisons and limitations to anomaly-matching

The Laguna del Maule volcanic field (LdMVF), central Chile has experienced significant upward ground deformation since 2007 suggesting the presence of a restless magmatic system at depth. Broadband magnetotelluric (MT) data were collected at the LdMVF between 2009 and 2016 and these data were inverted to create a 3-D electrical resistivity model of the subsurface. Two primary conductive features were identified. At 4 km depth, a strong conductor (C3) is located beneath recent vents in the northwest LdMVF and dips to the north. The conductivity of C3 is too high to be explained as partial melt alone and requires the presence of well-connected hydrothermal brine. A deeper conductor (C4) is located north of the lake at >8 km depth and is interpreted as a region of andesitic partial melt (<40% melt).

Seismic surface wave tomography results show a low Vs zone centered beneath the southwest side of the lake. This has been modelled as a large, low-velocity zone (V1) between 2 and 8 km depth which is estimated to be approximately 5% partial melt. The dip of V1 and C3 are similar and there is some overlap, but the centers of the anomalies are offset from one another by approximately 5 km.

The fact that the modelled features do not align may be due to a variety of factors and may highlight important differences in the LdM magmatic system. The deeper feature (C4) is likely beyond the resolution limits of the current seismic array. The low melt fractions estimated for feature V1 would likely result in poorly-connected networks which have little effect on the electrical resistivity. Similarly, interconnected brine lenses could significantly lower the resistivity while having little effect on seismic velocity. Both these explanations suggest that differences could be geological rather than related to shortcomings of either geophysical model.

The plumbing system beneath the LdMVF is very complex and shows a 3-D network which includes both vertical and lateral transport from mid-crustal depth to surface vents. The LdMVF system also highlights the fact that different geophysical methods are sensitive to different geological features and that anomalies from different methods may not always be coincident. Attempts to fit these datasets with correlated geophysical parameters may result in incorrect interpretations and models with higher misfits.