Co-seismic deformation of the 2010 Maule, Chile earthquake: Validating the least squares collocation interpolation method

Least squares collocation (LSC) has been successfully applied to develop the Velocity Model for SIRGAS (VEMOS) (Drewes and Heidbach, 2012) used to predict the velocities in the Geocentric Reference System for the Americas (Sistema de Referencia Geocentrico para las Americas, SIRGAS) GNSS reference frame. After the 2010 (Mw 8.8) Maule, Chile earthquake, the co-seismic and ongoing post-seismic deformation changed both the coordinates and velocities of geodetic benchmarks and continuous operating GPS reference stations (CORS) within the region affected (latitude -28 to -40). This deformation made VEMOS invalid for the estimation of velocities of the reference frame. To correctly obtain coordinates in the pre-seismic frame using post-seismic coordinates, it is necessary to estimate the deformation produced by the earthquake, both co- and post-seismic. Since neither the Argentine nor the Chilean CORS GPS networks are sufficiently dense to directly determine the deformation at arbitrary locations (by using the closest station), a densification of the observations of the deformation field using LSC was recently proposed. In this paper, we used a finite element model (FEM) to simulate the co-seismic deformation of the 2010 Maule earthquake. The FEM was then used to test the LSC of the co-seismic deformation field. We found that LSC cannot be used to correctly predict the behavior of the deformation in the near field due to the complexity of the elastic response of the earth’s crust. Nevertheless, the method correctly interpolates the far field deformation. As an alternative to the LSC method, the authors propose to use a finite element geophysical model that allows for a correct approximation of the co-seismic deformation, both in the near and far fields.