We consider various cases of seismicity, induced by artificial fluid injections in boreholes. Like many other authors, we support the hypothesis that to a large extent the triggering of this seismicity is caused by a diffusive process of the pore pressure relaxation in porous (or fractured), saturated rocks. We show that if this hypothesis is correct, then the spatio-temporal distributions of the seismic events must have several specific features related to the effective permeability of the rock. As a rule the fluid injection-induced seismicity obeys such features. These features can be indications of the diffusive and even hydraulic nature of the seismicity triggering process.From this point of view we analyze the spatio-temporal distribution of the late aftershocks of the Antofagasta (northern Chile) 1995 earthquake. These aftershocks were concentrated in a plane, an approximately 3 km-thick spatial zone. This thin seismogenic layer is a part of the South American subduction zone. The time-distance distribution of the aftershocks along this layer indicates that they could be triggered by a diffusion-like process. Possibly, such a process is the relaxation of the pressure perturbation caused in the pore fluid by the main Antofagasta event. We estimated the permeability required to explain the spatio-temporal distribution of the aftershocks by such a triggering mechanism. The obtained value, 60 mD, is very large. However, it is realistic for a long-time existing and large-scale fault zone.

In this paper, we present a compilation of modern seismic and seismological methods applied to image the subduction process in North Chile, South America. We use data from active and passive seismic experiments that were acquired within the framework of the German Collaborative Research Center SFB267 ‘Deformation Processes in the Andes’. The investigation area is located between 20° and 25°S and extends from the trench down to 100 km depth. In the depth range between the sea bottom and 15 km, we process an offshore seismic reflection profile using a recently developed velocity-model-independent stacking procedure. We find that the upper part of the subducting oceanic lithosphere in this depth range is characterized by a horst-and-graben structure. This structure supports an approximately 3 km thick coupling zone between the plates. In the depth range between 15 and 45 km, we analyse the spatial distribution of aftershocks of the Antofagasta earthquake (1995). The aftershock hypocenters are concentrated in an approximately 3 km thick layer. Finally, in the depth range between 45 and 100 km, we apply Kirchhoff prestack depth migration to the onshore ANCORP profile. A double reflection zone is observed between 45 and 60 km depth, which may represent the upper and lower boundary of the subducted oceanic crust. Over the whole range down to more than 80–90 km depth, we obtain an image of the subducting slab. At that depth, the hypocenters of local earthquakes deviate significantly in the direction perpendicular to the slab face from the reflective parts of the slab. Consequently, our results yield a complete seismic image of the downgoing plate and the associated seismic coupling zone.

A temporary seismological network, which was located in the Antofagasta (northern Chile) region, recorded a large part of the aftershocks from a MW = 8.0 earthquake in 1995. Results from a tomographic inversion show that the aftershocks were concentrated in a plane, approximately 3 km thick spatial zone. Comparison of our results with earlier processed wide‐angle reflection data, shows clearly that this zone is positioned at the top of the oceanic crust. Therefore, it represents the shearing boundary between the subducting oceanic and the continental plate, and gives an in‐situ estimation of the thickness of the boundary. Outside of this zone, events were detected in the continental crust, as well as within the oceanic mantle.