ALBERT

Description: The main goal of MSM21/4 was the study of gas hydrate system off Svalbard. We addressed this through a comprehensive scientific programme comprising dives with the manned submersible JAGO, seismic and heat flow measurements, sediment coring, water column biogeochemistry and bathymetric mapping. At the interception of the Knipovich Ridge and the continental margin of Svalbard we collected seismic data and four heat flow measurements. These measurements revealed that the extent of hydrates is significantly larger than previously thought and that the gas hydrate system is influenced by heat from the oceanic spreading centre, which may promote thermogenic methane production and thus explain the large extent of hydrates. At the landward termination of the hydrate stability zone we investigated the mechanisms that lead to degassing by taking sediment cores, sampling of carbonates during dives, and measuring the methane turn-over rates in the water column. It turned out that the observed gas seepage must have been ongoing for a long time and that decadal scale warming is an unlikely explanation for the observed seeps. Instead seasonal variations in water temperatures seem to control episodic hydrate formation and dissociation explaining the location of the observed seeps. The water column above the gas flares is rich in methane and methanotrophic microorganisms turning over most of the methane that escapes from the sea floor. We also surveyed large, until then uncharted parts of the margin in the northern part of the gas hydrate province. Here, we discovered an almost 40 km wide submarine landslide complex. This slide is unusual in the sense that it is not located at the mouth of a cross shelf trough such as other submarine landslides on the glaciated continental margins around the North Atlantic. Thus, the most widely accepted explanation for the origin of such slides, i.e. overpressure development due to deposition of glacial sediments on top of water rich contourites, is not applicable. Instead we find gas-hydrate-related bottom simulating reflectors underneath the headwalls of this slide complex, possibly indicating that subsurface fluid migration plays a major role in its genesis.

Description: Formosa Ridge is one of many topographic ridges created by canyon incision into the eastern South China Sea margin. The northwestern termination of the ridge is caused by beheading of the ridge due to a westward shift of the canyon that originally formed to the eastern flank of Formosa Ridge. Below Formosa Ridge a bottom simulating reflector (BSR) exists. Its depth below sea floor coincides with the theoretical base of the gas hydrate stability zone and the reflection has reverse polarity suggesting that it is caused by free gas below gas hydrate accumulations. The BSR is ubiquitous but shows significant variations in depth below sea floor ranging from 150 ms TWT (or approximately 180 m) underneath the incised canyon in the north to up to 500 ms (or approximately 460 m) underneath the crest of Formosa Ridge. Predominantly this depth variation is the result of topography on subsurface temperature, but comparison with the average BSR depth underneath the surrounding canyons suggests that recent canyon incision in the north has perturbed the thermal state of the sediments. Formosa Ridge consists of a northern half that is dominated by refilled older canyons and a southern half that consists mainly of contourite deposits. However, judging by the reflection seismic data this difference in origin seems to have little effect on the distribution of gas hydrate.

Description: The Chilean subduction zone is among the seismically most active plate boundaries in the world and coastal ranges suffer from a magnitude 8 or larger megathrust earthquake about every ten years. The Constitución-Concepción or Maule segment in central Chile between about 35.5°S and 37°S was considered to be a mature seismic gap, rupturing last in 1835 and being seismically quiet without any magnitude 4.5 or larger earthquakes reported in global catalogues. It is located to the north of the nucleation area of the 1960 magnitude 9.5 Valdivia earthquake and to the south of the 1928 magnitude 8 earthquake near Talca. On 27 February 2010 this segment ruptured in a Mw 8.8 earthquake, nucleating near 36°S and affecting a 500 km-long segment of the margin between 34°S and 38.°S. Most of the aftershocks occurred offshore. Therefore, a network of 30 ocean-bottom seismometers (OBS) was deployed in the northern part of the rupture area for a three month period, recording local offshore aftershocks between 20 September 2010 and 25 December 2010. In addition, data of a network consisting of 33 land stations of the GeoForschungsZentrum Potsdam were included into the network, providing an ideal coverage of both the rupture plane and areas affected by post-seismic slip as deduced from geodetic data. Two years prior to the Maule event the Collaborative Research Center SFB 574 "Volatiles and Fluids in Subduction Zones" operated an amphibious seismic network in the same area. Both data sets gave a great opportunity to compare seismicity and stress distributions before and after a megathrust event and to study the evolution of a subduction zone within the seismic cycle of a megathrust event. In this study the aftershocks of the Mw 8.8 Maule earthquake are analysed in order to gain information about the rupture zone, stress distributions, and faulting in the forearc after a megathrust event. As most of the temporary and permanent seismic networks are located on land, automatic picking routines have been developed with land station data and there are few studies with automatically determined phase arrivals from OBSs in the literature. The analysis of aftershocks in this study is performed in an automated approach to show that an automated determination of phase arrivals and polarisation, focal mechanisms and magnitudes can be accomplished with OBS data as well. Aftershock seismicity analysis in the northern part of the survey area reveals a well resolved seismically active splay fault in the accretionary prism of the Chilean forearc. Splay faults, large thrust faults emerging from the plate boundary to the sea floor in subduction zones, are considered to enhance tsunami generation by transferring slip from the very shallow dip of the megathrust onto steeper faults, thus increasing vertical displacement of the sea floor. These structures are predominantly found offshore, and therefore, hard to detect in seismicity studies as most seismometer stations are located onshore. Application of critical taper theory analysis suggests that in the northernmost part of the rupture zone, co-seismic slip likely propagated along the splay fault and not the subduction thrust fault while in the southern part it propagated along the subduction thrust fault and not the splay fault. The most profound features of a comparison of aftershocks to data collected in 2008 before the Maule event are: (1) a sharp reduction in intraslab seismic activity after the Maule earthquake, (2) an increase in seismic activity at the slab interface above 50 km depth, where large parts of the rupture zone were mainly aseismic prior to the Maule earthquake. Further, the aftershock seismicity shows a broader depth distribution above 50 km depth, shifting the updip limit of the seismogenic zone about 30 km closer to the trench, and (3) an active seismic cluster in the 2008 data while in 2010 there is a seismic gap in about 40 to 50 km depth along the plate boundary probably related to a relic mantle body.

Description: On 27 February 2010 the Mw 8.8 Maule earthquake in Central Chile ruptured a well known seismic gap, which last broke in 1835. Shortly after the mainshock Chilean agencies (UC Santiago, UC Concepción) and the international seismological community (USA (IRIS), France (IPGP), UK (University of Liverpool), Germany (GFZ)) installed a total of 142 portable seismic landstations stations along the whole rupture zone in order to capture the aftershock activity. Additionally a network of 30 ocean bottom seismometers was deployed (IFM Geomar) in the northern portion of the rupture area for a three month period between 20 September 2010 and 25 December 2010. We present first results from local earthquake tomography based on arrival times from automatic detection algorithms and picking engines which are calibrated with manually picked events. The Vp and Vp/Vs velocity models will better illuminate the structure of the forearc including the downgoing slab, the sedimentary basins and the volcanic arc down to depths of 75 km. The 2D and 3D velocity models cover the northern part of the rupture Maule 2010 rupture zone between 33 ° and 36 °S. This region is characterized by pronounced crustal aftershock activity which started after a strong aftershock doublet (Mw=6.9) on 11 March 2010 near the city of Pichilemu (~34.5 °S) in the overriding plate. This pronounced cluster of crustal seismicity close to the city of Pichilemu will be the focus area with smaller node spacing of the velocity model.