Abstract The Tuaheni Landslide Complex (TLC) is characterized by areas of compression upslope and extension downslope. It has been thought to consist of a stack of two genetically linked landslide units identified from seismic data. We used 3D seismic reflection, bathymetry data and International Ocean Discovery Program Core U1517C (Expedition 372) to understand the internal structures, deformation mechanisms and depositional processes of the TLC deposits. Units II and III of U1517C correspond to the two chaotic units in 3D seismic data. In the core, Unit II shows deformation, whereas Unit III appears more like an in situ sequence. Variance attribute analysis showed that Unit II is split into lobes around a coherent stratified central ridge and is bounded by scarps. By contrast, we found that Unit III is continuous beneath the central ridge and has an upslope geometry, which we interpreted as a channel–levee system. Both units show evidence of lateral spreading due to the presence of the Tuaheni Canyon removing support from the toe. Our results suggest that Units II and III are not genetically linked, are separated substantially in time and had different emplacement mechanisms, but they fail under similar circumstances.

Abstract Understanding how the characteristics of individual turbidites measured along a 1D (vertical) section may relate to reservoir-scale system geometry remains a significant upscaling problem, yet the ability to make this link is fundamentally important when evaluating turbidite reservoirs. Insights into the key relationships are perhaps best gained from well-exposed outcrops in which bed-to-bed correlations can be established. The Peïra Cava outlier of the Tertiary Annot Sandstone contains sheetform turbidites that were deposited in a confined and ponded basin. They are now exposed over a 10 by 6 km area. Bed-to-bed correlations have been established throughout a selected 420 m stratigraphic interval, allowing the 3D geometry of the system to be constrained. We test the significance of bed sand thickness, mud-cap thickness, sand percentage, grain-size, and the presence or absence of erosional structures and cross-stratification for their value as predictors of up-stream and down-stream bed geometry. The results are compared with models currently used to predict the spatial distribution of sediment properties, such as flow-efficiency concepts and the influence of topographic control.