Largest aftershock nucleation driven by afterslip during the 2014 Iquique sequence

Megathrust faults are known to host both seismic and aseismic slip. Laboratory experiments and numerical simulations have demonstrated that seismic-aseismic interaction can be involved in the earthquake source process such as nucleation and termination. However, models of seismic-aseismic interaction regarding the source process are still controversial because their observational evidence is limited to the small number of events among instrumentally recorded earthquakes. This is likely due to the low signal-to-noise ratio of observations and/or short duration of the nucleation and termination processes. In this study, we newly report aseismic slip accompanying intriguing seismicity during the 2014 Iquique earthquake sequence by analysing seismicity and high-rate GPS crustal deformation data.

We document early postseismic deformation during the 3 days following the M 8.1 mainshock and demonstrate that afterslip started immediately after the mainshock and led 27 hours later to the M 7.6 largest aftershock which located ~120 km further south. The interevent afterslip peaks down-dip of the mainshock with decaying moderate aftershock rate, exhibiting typical postseismic megathrust response. A local peak of the afterslip is inferred between the mainshock and the largest aftershock epicentres. This local peak suggests that this area acted as an aseismic barrier to the southward mainshock rupture propagation so that the two big quakes did not occur simultaneously.

The geodetic moment everywhere decreased with time during the 27h interevent stage with different decaying rate. The decay was slower in the afterslip area between the two epicentres than the main down-dip peak. Interestingly, the seismicity rate and associated moment release in this area increased with time during the interevent 27 hours. We propose that the largest aftershock nucleation was driven by the afterslip. Contrary to predictions of some numerical simulation models, our result implies that aseismic slip during the nucleation process does not necessarily accelerate. Our new observational discovery illuminates the mechanical connection between sequential great earthquakes mediated by aseismic slip.