Experimental investigation on submarine channels formation

Turbidity currents are recognized as being a primary mechanism through which sediment is distributed in deep-water marine environments. A large number of studies have been devoted to the understanding of these sedimentary systems, using various approaches: (i) interpretation of large-scale geometries and sedimentary structures in preserved outcrops or using acoustical imaging of the subsurface, (ii) in-situ measurements of modern turbidity current, (iii) theoretical and numerical modelling of the flow and associated sediment transport, and (iv) laboratory experiments. To date, most of the physical experiments focus on the depositional stage of turbidity currents, by releasing in a flume a mixture of sediment and water on a fixed or mobile inclined bed. Some recent studies investigate the behaviour of gravity currents in fixed, pre-formed, straight or sinuous channels. However, no or very few studies report the mechanisms of erosion and incision of a turbidity current over a mobile bed. The present study investigates the formation of micro-scale submarine channels in a small tank (2x0.5x0.5 m). Brine is injected in the fresh water tank at the top of an inclined plane covered with low density (1080 kg.m-3) plastic sediment (d50=0.1 mm). Regular acquisition of the topography using a fringe projection technique allows quantifying precisely the channel incision dynamics and equilibrium morphology. Gravity current height and velocity profiles in the boundary layer are measured using an acoustic Doppler velocimeter. Conditions for submarine channel formation are investigated through systematic exploration of the slope / input discharge rate phase space, and reported in terms of the non-dimensional Shields number. Observations reveal that incision may be triggered by longitudinal instabilities in the dense plume that focus erosion. The width of the gravity current decreases and velocity increases through time as the channel propagates down-slope and deepens. Eroded sediment is transported by traction on the channel bed and is deposited into a prograding lobe that can be re-incised by the channel. When a longitudinal equilibrium profile is reached, the dense current is totally contained in the incised channel, and no sediment is transported. This study reveals the potential of micro-scale laminar experiments to investigate the dynamics of large natural systems such as submarine canyons. By using low density sediment, the shear stress required to put grains in motion is greatly reduced, thus reducing the characteristic length scale of the morphological system.