Scientists analysing seismic pre- and aftershocks in Chile

Last year, in April 2014, Northern Chile was struck by a magnitude 8.1 earthquake offshore the city of Iquique. The quake occurred in a subduction zone, where an oceanic tectonic plate collides with a continental plate and is forced down into the Earth's mantle. At their interface the plates are locked. Hence, with time, large stresses build up which are released as earthquakes. During such an event, the plates suddenly move against each other by the amount of accumulated strain. This cyclic process shows some regularities. Estimations for this region based on observations of the last 500 years suggest that it takes roughly 100 years to build up stresses able to cause a large earthquake. Since the last such event happened in 1877, and since the northerly and southerly bordering plate segments broke in 1995 and 2001 respectively, the interspace forms a seismic gap.

In order to better understand physical processes of earthquakes as well as the seismic gap, the installation of the international monitoring network IPOC (Integrated Plate boundary Observatory Chile) was initiated in 2006. The German Research Centre for Geosciences (GFZ) plays a leading role within this cooperation. The network consists of a variety of sensors: broadband seismometers record the velocity of the ground motion, accelerometers the acceleration. Using GNSS receivers (Global Navigation Satellite System) - a high-tech version of usual navigation systems - displacements in the millimeter range can be detected. Hereby, the strain occurring between earthquakes (interseismic), as well as the offset during (co-seismic) and the dislocation after events (post-seismic) are of interest. Creepmeters and tiltmeters monitor the activity at known crustal faults and magnetotelluric instruments are able to measure variations of electrical resistance in the ground which might be associated with fluid relocations at fault zones. These land-based measurements are complemented with satellite-based radar observations (InSAR) of topographic changes.

The abundance of data generated by IPOC from before-, during and after the earthquake has allowed to investigate the tectonic processes with unprecedented resolution in their temporal development. Since July 2013 three series of foreshocks with increasing maximum magnitude (but still much smaller than the main event) have been recorded. During this time, the so-called b-value, describing the ratio between smaller and larger events was reduced. Such a behaviour is often interpreted as a consequence of increasing stresses. Shortly before the main event the b-value increased again. Comparison of the location of the foreshock clusters with results based on interseismic GPS-data revealed that they happened in a region upward the zone of strongest locking between the plates. The large number of smaller events increased the stress at the most strongly coupled area until the forces were large enough to cause failure of this asperity leading to the main shock. The analysis of the seismic data enabled the reconstruction of the spatial and temporal evolution of the event. It nucleated in the area of the foreshocks and then propagated downwards along the plate interface into the area of strongest coupling. The rupture process had a duration of almost three minutes and the maximum displacement between the plates was about five meters.

Fortunately, the damages and casualties caused by the earthquake were rather moderate for an event of this size, due to the distance to the coast and low population density. However, simulations based  on the current locking state between the plates and historical earthquake activity indicate that only parts of the accumulated stresses were released, and that the potential for an event significantly larger than the Iquique earthquake of 1st of April 2014 is still present. It is possible that the remaining energy is released in a single earthquake, but since part of the stress was already released in the central region of the seismic gap, it is considered more likely that several smaller events will occur. To the present day it is impossible even with the best measurements to predict earthquakes in a sense that location, time and magnitude could be announced. But it is possible to give probabilities for the occurrence of events within a period of time. Such information can for instance be used to develop adequate building codes.