Deformation and stress changes before, during and after the 2014 Pisagua earthquake inferred from RADARSAT-2 data

Abstract:

The Mw 8.1 Pisagua earthquake occurred off the coast of northern Chile, in a seismic gap that hadn’t ruptured in a major earthquake since 1877. It soon became clear from the magnitude of the earthquake, and from the distribution of aftershocks, that this event, large though it was, had by no means ruptured the entire seismic gap. The question immediately raised was, therefore: what is the remaining earthquake potential in this area of northern Chile? To address this question, we use RADARSAT-2 data to investigate aseismic and major seismic activity over the entire sequence of foreshocks, mainshock, and the first four postseismic months. Seismic event locations are used throughout the study to inform our interpretations. Multiple interferograms are inverted for slip distributions on the subduction interface and other contributing faults during the three phases. GPS displacements published by Ruiz et al. (2014) are used alongside the InSAR data to help constrain the preseismic fault slip. The subduction zone interface is simulated using the Slab1.0 model, meshed with triangular elements. We find that in the mainshock, slip occurred between about 15 and 40 km depth, with a maximum of ~7 m of slip at a depth of ~25 km. We also resolve up to 2.5 m of slip in the Mw 7.6 aftershock that occurred two days after the mainshock. The postseismic InSAR dataset, which is still being acquired and developed, is also inverted to identify the main zones of afterslip. We spatially compare the mainshock slip patches with zones of aseismic slip and also seismicity clusters during the foreshock and aftershock sequences. Stress changes on the interface due to the largest foreshock (Mw 6.7) are calculated, which indicates that stress was significantly enhanced in the area of the mainshock. In addition, the stressing effect of the possible slow slip event inferred by Ruiz et al. (2014) is investigated. The results from the various datasets are integrated to map out areas that (a) still have rupture potential and (b) are now zones of enhanced stress.