Full waveform modelling of aftershock seismicity in the Chilean subduction zone using the VERCE platform

Wednesday, 17 December 2014
Thomas Garth1, Stephen P Hicks2, Amaya Fuenzalida Velasco1, Emanuele Casarotti3, Alessandro Spinuso4 and Andreas Rietbrock1, (1)University of Liverpool, Liverpool, United Kingdom, (2)University of Liverpool, Liverpool, L69, United Kingdom, (3)National Institute of Geophysics and Volcanology, Rome, Italy, (4)Royal Netherlands Meteorological Institute, De Bilt, Netherlands
The VERCE platform allows high resolution waveforms to be simulated through an interactive web-based portal. The platform runs on a variety of HPC clusters, and waveforms are calculated using SPECFEM3D. We use the full waveform modelling techniques supported on the VERCE platform to test the validity of a number of subduction zone velocity models from the Chilean subduction zone. Waveforms are calculated for aftershocks of the 2010 Mw 8.8 Maule (central Chile) and the Mw 8.1 2014 Pisagua (Northern Chile) earthquakes. For the Maule region, we use a 2D tomographic model of the rupture area (Hicks et al., 2012), and the focal mechanisms of Agurto et al., (2012). For the Pisagua earthquake, we use a 2.5D composite velocity model based on tomographic studies of the region (e.g. Husen et al., 2000, Contreras-Reyes et al., 2012) and Slab1.0 (Hayes et al., 2012). Focal mechanisms for the Pisagua aftershock sequence are produced from waveforms recorded on the IPOC network using the program ISOLA (Sokos and Zahradnik, 2008). We also test a number of synthetic velocity models.

The simulated waveforms are directly compared to waveforms recorded on the temporary deployment for the Maule earthquake aftershocks, and waveforms recorded on the IPOC network for the Pisagua earthquake aftershocks. The waveforms produced by the 3D full waveform simulations are also compared to the waveforms produced by the focal mechanism inversion, which assume a 1D velocity model. The VERCE platform allows waveforms from the full 3D model to be produced easily, and allows us to quantifiably assess the validity of both the velocity model and the source mechanisms. In particular the dependence of the dip of the focal mechanism on the velocity model used is explored, in order to assess the reliability of current models of the plate interface geometry in the Chilean subduction zone.