Putting faults in the northern Chilean subduction margin into motion: evidence for remote dynamic earthquake triggering on the plate interface and within the forearc

Dynamic stresses on the order of ~1 kPa from passing waves of mainshock earthquakes can trigger aftershocks at remote distances. Here, we investigate the prevalence of remote earthquake triggering in northern Chile, where aseismic-slip triggering has been documented. Our twofold approach to quantify triggerability includes a statistical difference-of-means test to quantify seismicity-rate changes bracketing candidate mainshock times, and a waveform-based approach to look for triggered earthquakes missing from the local catalog. We find no persistent, statistically-significant seismicity-rate increases associated with any of the candidate mainshocks when considering the local catalog in aggregate. However, catalog statistics reveal evidence for localized triggering both on the subduction interface and within the shallower forearc faults. Waveforms reveal local, uncataloged earthquakes only visible using a high-pass filter that removes the mainshock signal that otherwise overprints the local signals. Based on Japan mainshocks, we cannot rule out antipodal triggering. Areas showing higher triggerability are consistent with regions of low locking inferred from GNSS models and regions of observed aseismic slip. The spatial coincidence of triggering and low-locking, combined with the absence of a stress-triggering threshold, requires non-linear triggering mechanisms, such as altered frictional strength or aseismic-slip triggering, to be consistent with the observations.