Exhumed lateral margins and increasing flow confinement of a submarine landslide complex

Submarine landslides, including the basal shear surfaces along which they fail, and their subsequent infill, are commonly observed in modern seabed and seismic reflection data sets; their resultant relief impacts sediment routing and storage patterns on continental margins. Here, three stacked submarine landslides are documented from the Permian Ecca Group, Laingsburg depocentre, Karoo Basin, South Africa, including two superimposed lateral margins. The stratigraphic framework includes measured sections and correlated surfaces along a 3 km long, 150 m high outcrop. Two stacked 2·0 to 4·5 km wide and 90 m and 60 m deep erosion surfaces are recognized, with lateral gradients of 8° and 4°, respectively. The aim of this study was to understand the evolution of a submarine landslide complex, including: evolution of basal shear surfaces/zones; variation of infill confinement; and location of the submarine landslides in the context of basin‐scale sedimentation and degradation rates. Three stages of formation are identified: (i) failure of submarine landslide 1, with deposition of unconfined remobilized deposits; (ii) failure of submarine landslide 2, forming basal shear surface/zone 1, with infill of remobilized deposits and weakly confined turbidites; and (iii) failure of submarine landslide 3, forming basal shear surface/zone 2, with infill of remobilized deposits and confined turbidites, transitioning stratigraphically to unconfined deposits. The expression of basal shear varies laterally, from metres thick zones in silt‐rich strata to sharp stepped surfaces in sand‐rich strata. Faulting and rotation of overlying bedding suggest that the shear surfaces/zones were dynamic. Stacking of landslides resulted from multi‐phase slope failure, increasing down‐dip topography and confinement of infilling deposits. The failure slope was probably a low supply tilted basin margin evidenced by megaclast entrainment from underlying basin‐floor successions and the lack of channel systems. This study develops a generic model of landslide infill, as a function of sedimentation and degradation rates, which can be applied globally.