The giant 2011 Tohoku-oki earthquake has been inferred to remobilise fine-grained, young surface sediment enriched in organic matter from the slope into the >7 km deep Japan Trench. Yet, this hypothesis and assessment of its significance for the carbon cycle has been hindered by limited data density and resolution in the hadal zone. Here we combine new high-resolution bathymetry data with sub-bottom profiler images and sediment cores taken during 2012–2016 in order to map for the first time the spatial extent of the earthquake-triggered event deposit along the hadal Japan Trench. We quantify a sediment volume of ~0.2 km3 deposited from spatially-widespread remobilisation of young surficial seafloor slope sediments triggered by the 2011 earthquake and its aftershock sequence. The mapped volume and organic carbon content in sediment cores encompassing the 2011 event reveals that this single tectonic event delivered >1 Tg of organic carbon to the hadal trench. This carbon supply is comparable to high carbon fluxes described for other Earth system processes, shedding new light on the impact of large earthquakes on long-term carbon cycling in the deep-sea.

Subaqueous slope failure mechanisms are still poorly understood partly because they are difficult to study due to the remote location of submarine landslides. Landslides in lakes are smaller in size and more readily accessible and therefore represent a good alternative to their marine counterparts. Lake Villarrica, located in South-Central Chile, experienced significant slope failure and serves here as an exemplary study area for subaqueous landslide initiation mechanisms in tectonically active settings. Coring and CPTU testing were undertaken with the MARUM free-fall CPTU deployed adjacent to the coring sites where all lithological units involved in the slope failure were sampled. Using geotechnical methods such as pseudo-static factor of safety analysis and cyclic triaxial testing, three types of soils (i.e., diatomaceous ooze, volcanic ash, and quick clay) were analyzed for their role in slope failure, and earthquake shaking was identified as the primary trigger mechanism. The investigated landslide consisted of two distinct phases. During the first phase, slope failure was initiated above a tephra layer. In the second phase, retrogression led to the shoreward extension of the slide scarp along a second failure plane located in a stratigraphically deeper, extremely sensitive lithology (i.e., quick clay). Results show that liquefaction of buried tephra layers was unlikely, but such layers might still have contributed to a reduction in shear strength along the contact area with the neighboring sediment. Furthermore, cyclic shaking-induced pore pressure in diatomaceous ooze may be similar to that in granular soils. We generally infer that failure mechanisms observed in this study are equally important for landslide initiation in submarine settings as diatomaceous ooze intercalated with volcanic ash may be abundantly present along active continental margins

Strong earthquakes at active ocean margins can remobilize vast amounts of surficial slope sediments and dynamically strengthen the margin sequences. Current process understanding is obtained from resulting event deposits and low‐resolution shear strength data, respectively. Here we directly target a site offshore Japan where both processes are expected to initiate, that is, at the uppermost part (15 cm) of a sedimentary slope sequence. Based on a novel application of short‐lived radionuclide data, we identified, dated, and quantified centimeter‐scale gaps related to surficial remobilization. Temporal correlation to the three largest regional earthquakes attest triggering by strong earthquakes ( M w >8). Also, extremely elevated shear strength values suggest a strong influence of seismic strengthening on shallow sediments. We show that despite enhanced slope stability by seismic strengthening, earthquake‐induced sediment transport can occur through surficial remobilization, which has large implications for the assessment of turbidite paleoseismology and carbon cycling at active margins.

Diatom microfossils have been detected in many natural marine sediment deposits around the globe and are held responsible for the disobedience to well-established geotechnical relationships between index-properties and shear strength. We revisit the static shear strength and present the first cyclic undrained shear strength experiments on diatom microfossil-clayey-silt mixtures to study the role of diatoms on submarine slope stability. It is attested that the angle of internal friction (U) increases with diatom content, however, we provide evidence for a significant overestimation of U in previous studies. Based on direct shear tests at varying normal stresses 600 kPa we find U 5 328 in contrast to 438 in pure diatom. Similarly, to static shear strength, cyclic shear strength increases with diatom content, however, in contrast to static shear strength the most drastic increase does not occur from 0% to 25% diatoms but from 75% to 100%. Interestingly, diatomaceous sediments tend to fail by liquefaction although well-established relations between index properties and liquefaction susceptibility predict the opposite. Liquefaction failure is observed solely in samples containing 50% diatoms whereas samples with lower diatom content fail by cyclic softening. We conclude diatom microfossils in marine sediments significantly contribute to an increased slope stability under static and cyclic loading conditions since diatoms lead to higher resistance independently of the loading mode. The strength increase is interpreted as a result of particle interlocking and surface roughness, which is very efficient given the highly variable habitus of diatom species.

Subaqueous landslides are common features at active and passive ocean margins, in fjords and lakes. They can develop on very gentle slope gradients (<2°) and the presence of sandy tephra layers seems to facilitate the development of translational failure. Despite numerous investigations, it remains elusive how different slope preconditioning factors act and interact over time and how different triggering mechanisms can lead to slope failure. In settings of low to moderate seismicity, stratigraphic sequences with sublacustrine masstransport deposits (MTDs) have successfully been used for constructing prehistorical earthquake catalogues.In high seismicity areas, it is inferred that not all strong earthquakes succeed in triggering landslides on the investigated slope segments, and MTD records do not fully represent their complete recurrence pattern.Here, we present the spatio-temporal distribution of MTDs in two large glacigenic Chilean lakes (Villarrica and Calafquén) based on a detailed seismic-stratigraphic analysis and several radiocarbon-dated piston cores (up to 14 m long). We find a strong influence of slope gradient on the occurrence and volume of landslide 2 events; i.e. most (small) landslides take place on slopes of 5-20°, whereas the few large (potentially tsunamigenic) landslides exclusively occur on slopes of <4°. Liquefaction of sandy tephra layers facilitates the development of thin (<0.5 m) in-situ deformations during earthquake shaking. When sandy tephra layers get progressively buried, liquefaction becomes unlikely, but repeated excess pore pressure transfer to overlying units facilitates the development of translational sliding. The occurrence of voluminous landslides seems to follow a "landslide cycle" which starts with the deposition of a tephra layer and the development of in-situ deformations directly on top. Once the slope sequence reaches a critical thickness, the end of the cycle is indicated by incipient scarp development, and subsequent major sliding event(s). The duration of the landslide cycle is defined by the rate of gradual sedimentation, but may be affected by sudden geological events (e.g., volcanic eruptions), expediting the end of the cycle. Despite the many methodological challenges inherent to the construction of a MTD stratigraphy, we propose that well-dated multiple MTD events can be used as positive evidence to strengthen and specify the regional paleoseismic record, concerning the largest events in a high-seismicity region. This method is most successful when targeting the base of relatively steep slopes (5-20°) with frequent, minor landsliding, and complementing this with seismic-stratigraphic analysis of fluid-escape features and correlation with distal turbidite records.