Description: At convergent margins, the structure of the subducting oceanic plate is one of the key factors controlling the morphology of the upper plate. We use high-resolution seafloor mapping and multichannel seismic reflection data along the accretionary Sumatra trench system to investigate the morphotectonic response of the upper plate to the subduction of lower plate fabric. Upper plate segmentation is reflected in varying modes of mass transfer. The deformation front in the southern Enggano segment is characterized by neotectonic formation of a broad and shallow fold-and-thrust belt consistent with the resumption of frontal sediment accretion in the wake of oceanic relief subduction. Conversely, surface erosion increasingly shapes the morphology of the lower slope and accretionary prism towards the north where significant oceanic relief is subducted. Subduction of the Investigator Fracture Zone and the fossil Wharton spreading centre in the Siberut segment exemplifies this. Such features also correlate with an irregularly trending deformation front suggesting active frontal erosion of the upper plate. Lower plate fabric extensively modulates upper plate morphology and the large-scale morphotectonic segmentation of the Sumatra trench system is linked to the subduction of reactivated fracture zones and aseismic ridges of the Wharton Basin. In general, increasing intensity of mass-wasting processes, from south to north, correlates with the extent of oversteepening of the lower slope (lower slope angle of 3.8 degrees in the south compared with 7.6 degrees in the north), probably in response to alternating phases of frontal accretion and sediment underthrusting. Accretionary mechanics thus pose a second-order factor in shaping upper plate morphology near the trench.

Description: Fluids are suspected to play a major role in the nucleation and rupture propagation of earthquakes. In Chile, seismological data were previously interpreted to indicate that fluids captured in the fault zone are released periodically during large underthrust earthquakes, leading to post-seismic fluid flow. In central Chile, heat flow derived from the presence of a bottom simulating reflector (BSR) show a smooth trend across the margin. BSR-derived data are in excellent agreement with thermal subduction zone models. Over the young accretionary prism, both BSR-derived and measured surface heat flow support a common trend. Landwards of the backstop, however, measured heat flow triples over a distance of 20–30 km, producing a profound discrepancy to the BSR-derived data. We suggest that this disparity is related to transient flow of warm fluids through the gas hydrate stability zone possibly caused by fluids released after large underthrust earthquakes. Such flow events may inherently affect the distribution of solid gas hydrates between the seafloor and the BSR.

Description: We examine micro-earthquake records from a dense temporary array of ocean bottom seismometers (OBS) and hydrophones that has been installed from September to November 2005 at the trench outer rise offshore Nicaragua. Approximately 1.5 locatable earthquakes per day within the array of 110 × 120 km show the high seismic activity in this region. Seismicity is restricted to the upper ∼15 km of the mantle and hence where temperatures reach 350–400 °C, which is smaller than values observed for large mantle intraplate events (650 °C). Determination of moment tensor solutions suggest a change of the stress region from tensional in the upper layers of the oceanic plate to compressional beneath. The neutral plane between both regimes is located at ∼6–9 km beneath Moho and thus very shallow. Fluids, which are thought to travel through the tensional fault system into the upper mantle, may not be able to penetrate any deeper. The earthquake catalogue, which seems to be complete for magnitudes above Mw = 1.6–1.8, suggests a strong change of the lithospheric rheology when approaching the trench. And b-factors, that is the ratio between small and large earthquakes increase significantly in the closest 20 km to the trench axis, implying that the crust and upper mantle is massively weakened and hence ruptures more frequently but under less release of stress. We explain this with a partly serpentinized upper mantle.

Description: The 1992 Nicaragua earthquake was a ‘tsunami earthquake’, which generated tsunamis disproportionately large for its surface wave magnitude Ms = 7.2. Seismological studies and tsunami simulation indicated that the event was a slow earthquake, which occurred on the plate boundary between the subducting Cocos plate and the overriding Caribbean plate. We present a finite element model that enables us to estimate for the first time the temperature and inferred frictional conditions in the rupture area of a tsunami earthquake. Direct and indirect observations are used to constrain all model parameters, and surface heat-flux measurements provide independent information to verify the model results. Furthermore, we used a genetic algorithm to perform a sensitivity analysis of all model parameters and to define the spatial range of thermally defined updip limit of the seismogenic zone. The earthquake nucleated in the seismogenic zone at temperatures of ∼150 °C and propagated updip towards the trench axis. The centroid or centre of mass of moment release was located in a region characterized by temperatures of ∼50 °C. Thus, the rupture propagated through a region where plate motion is normally accommodated by aseismic creep. Our observations support a model in which tsunami earthquakes nucleate in the seismogenic zone near its updip limit. However, in such an environment coupled asperities are perhaps too small to cause large earthquakes. Seamounts, however, are abundant on the incoming Cocos plate. Therefore, in addition to temperature-dependent metamorphic induration of sediments, increased normal stress by seamount subduction may contribute to accumulate stress sufficiently large to release enough energy near the updip limit of the seismogenic zone to promote dynamic slip along a normally aseismic décollement all way to the ocean.

Description: Trench-outer rise earthquakes occur by reactivation or creation of normal faults caused as the oceanic lithosphere approaches a subduction zone and bends into the deep-sea trench. These faults may cut deep enough into the mantle to allow sea water to penetrate into the lithosphere, causing serpentinization. The amount of water carried into the mantle is linked to the maximum depth that the tensional faults cut into the lithosphere, which in turn is directly linked to the maximum focal depths of outer rise normal faulting earthquakes. We analysed teleseismic P and S waves of seven earthquakes from the trench-outer rise offshore of Central America using teleseismic waveform inversion of broad-band data. For the computation of Green's functions for waveform inversion, probabilistic earthquake locations were calculated. To study the rupture process, earthquake centroid depths and focal mechanisms for a sequence of subevents were calculated. Both, hypocentral depths from the relocation process and the estimated centroid depths from the waveform inversion show that all events occur at shallow depths (〈30 km). Furthermore, the locations of the subevents relative to each other suggest that fault planes for Mw∼ 6 are in the order of 50 km in length and only 5–10 km in width. Rupture generally propagates downdip and the focal mechanisms change for most events from normal faulting to strike-slip or oblique thrusting with time. The depth at which this mechanism change is observed may represent the depth of the nodal plane between tensional and compressional regions in the incoming plate.

Description: Several trench-outer rise settings in subduction zones worldwide are characterized by a high degree of alteration, fracturing and hydration. These processes are induced by bending-related faulting in the upper part of the oceanic plate prior to its subduction. Mapping of P- and S-wave velocity structures in this complex tectonic setting provides crucial information for understanding the evolution of the incoming oceanic lithosphere, and serves as a baseline for comparison with seismic measurements elsewhere. Active source seismic investigations at the outer rise off Southern Central Chile (∼43°S) were carried out in order to study the seismic structure of the oceanic Nazca Plate. Seismic wide-angle data were used to derive 2-D velocity models of two seismic profiles located seaward of the trench axis on 14.5 Ma old crust; P01a approximately parallel to the direction of spreading and P03 approximately parallel to the spreading ridge and trench axes. We determined P- and S-velocity models using 2-D traveltime tomography. We found that the Poisson's ratio in the upper crust (layer 2) ranges between ∼0.33 at the top of the crust to ∼0.28 at the layer 2/3 interface, while in the lowermost crust and uppermost mantle it reaches values of ∼0.26 and ∼0.29, respectively. These features can be explained by an oceanic crust significantly weathered, altered and fractured. Relative high Poisson's ratios in the uppermost mantle may be likely related to partially hydrated mantle and hence serpentinization. Thus, the seismic structure of the oceanic lithosphere at the Southern Central Chile outer rise exhibits notable differences from the classic ophiolite seismic model (‘normal’ oceanic crust). These differences are primarily attributed to fracturing and hydration of the entire ocean crust, which are direct consequences of strong bending-related faulting at the outer rise. On the other hand, the comparison of the uppermost mantle P-wave velocities at the crossing point between the perpendicular profiles (∼90 km oceanward from the trench axis) reveals a low degree of Pn anisotropy (〈2 per cent).

Description: Crustal- and upper-mantle structures of the subduction zone in south central Chile, between 42 degrees S and 46 degrees S, are determined from seismic wide-angle reflection and refraction data, using the seismic ray tracing method to calculate minimum parameter models. Three profiles along differently aged segments of the subducting Nazca Plate were analysed in order to study subduction zone structure dependencies related to the age, that is, thermal state, of the incoming plate. The age of the oceanic crust at the trench ranges from 3 Ma on the southernmost profile, immediately north of the Chile triple junction, to 6.5 Ma old about 100 km to the north, and to 14.5 Ma old another 200 km further north, off the Island of Chiloe. Remarkable similarities appear in the structures of both the incoming as well as the overriding plate. The oceanic Nazca Plate is around 5 km thick, with a slightly increasing thickness northward, reflecting temperature changes at the time of crustal generation. The trench basin is about 2 km thick except in the south where the Chile Ridge is close to the deformation front and only a small, 800-m-thick trench infill could develop. In south central Chile, typically three quarters (1.5 km) of the trench sediments subduct below the decollement in the subduction channel. To the north and south of the study area, only about one quarter to one third of the sediments subducts, the rest is accreted above. Similarities in the overriding plate are the width of the active accretionary prism, 35-50 km, and a strong lateral crustal velocity gradient zone about 75-80 km landward from the deformation front, where landward upper-crustal velocities of over 5.0-5.4 km s〈SU-1〈/SU decrease seaward to around 4.5 km s〈SU-1〈/SU within about 10 km, which possibly represents a palaeo-backstop. This zone is also accompanied by strong intraplate seismicity. Differences in the subduction zone structures exist in the outer rise region, where the northern profile exhibits a clear bulge of uplifted oceanic lithosphere prior to subduction whereas the younger structures have a less developed outer rise. This plate bending is accompanied by strongly reduced rock velocities on the northern profile due to fracturing and possible hydration of the crust and upper mantle. The southern profiles do not exhibit such a strong alteration of the lithosphere, although this effect may be counteracted by plate cooling effects, which are reflected in increasing rock velocities away from the spreading centre. Overall there appears little influence of incoming plate age on the subduction zone structure which may explain why the M-w = 9.5 great Chile earthquake from 1960 ruptured through all these differing age segments. The rupture area, however, appears to coincide with a relatively thick subduction channel.