Abstract

Preexisting networks of seismometers and continuous GPS in Northern Chile successfully captured surface motions and seismicity leading up to the 1 April M-w 8.1. Here we compare continuous GPS (cGPS) with predictions of seismic dislocations for the final foreshock swarm, beginning with the 16 March M-w 6.7. Results show that the cumulative cGPS motion can be largely explained by seismic slip because evolutions of cGPS positions for most stations stay within the ranges of seismic predictions (given sensible ranges of assumed source errors). However, cGPS motions between 18-21 and 25-31 March outpace seismic predictions, supporting the existence of aseismic transients that were most probably the afterslip from preceding bursts of seismicity. A parameter search reveals that the 16 March M-w 6.7 cGPS displacements can be recreated with a fault plane significantly rotated anticlockwise from the strike of the plate interface, suggesting that failure was on a structure other than the plate interface.