Comparing the InSAR observations of deformation for Andean retroarc earthquakes triggered by the Mw 8.8 2010 Maule earthquake with seismic fault plane solutions and geologic constraints

Following the February 27, 2010 Maule earthquake in Chile (Mw = 8.8), heightened levels of seismicity were observed in South America. We focus on whether stress changes due to the Maule earthquake or local structures controlled fault orientations for these triggered events. We compare the fault plane orientations and slip vectors inferred from InSAR (Interferometric Synthetic Aperture Radar) and seismic data with stress changes due to the Maule earthquake and the orientations of pre-existing weaknesses near the triggered events. First, we estimate the focal mechanism using first motion polarities from teleseismic and regional seismic data. Second, we use InSAR to constrain possible fault planes consistent with the data, instead of merely focusing on a single "best" model. We use two metrics to evaluate models with different fault plane geometries: a numerical residual that quantifies the fit of the synthetic model, and a visual inspection aimed at selecting models that generate similar deformation signals and approximately random residuals. We prefer a distribution of models chosen using the latter metric, because the set of models better reflects the uncertainty in fault orientation due to spatially correlated noise caused by atmospheric water vapor and ionospheric variations. In addition, we compare these results with constraints on fault plane orientations and slip directions of active or recently active faults from published data on nearby field studies. The 2010 (Mw = 6.2) Salta, NW Argentina earthquake is one of several triggered events to which we apply the above methods. The Salta earthquake occurred 15 hours after the Maule earthquake at an epicentral distance of 1450 km from the main event. The timing and location of the Salta event suggest that it was dynamically triggered by the Maule earthquake. The first motion seismic arrivals are consistent with several of the fault planes preferred by the InSAR observations. Further, the distribution of fault planes from the InSAR study is generally consistent with the variation in local fault plane orientations representing the most recent deformation sequence. The shortening and extension directions inferred from each one of the three datasets, seismic, InSAR and geologic, are consistent with the other two datasets. Since the seismic deformation is consistent with geologic fault orientations and kinematic axes, we infer that while the Maule earthquake likely brought the Salta region closer to failure, the Salta earthquake fault geometry is primarily controlled by local kinematics and pre-existing crustal weaknesses. We also present work on three other potentially triggered earthquakes with moment magnitude ranges of 4.8-6.0, and at epicentral distances of 400-1300 km from the Maule earthquake.