A review of sinuous channel avulsion patterns in some major deep-sea fans and factors controlling them

This paper reviews side-scan sonar, multi- and single-beam echo-sounder, and high-resolution seismic reflection data, and discusses the anatomy and patterns of the most recent sinuous channel avulsions in the Amazon, Zaire, Indus and Bengal Fans, all located along passive continental margins and all fed by major river systems. Channel patterns formed by fore- (seaward) or back- (landward) stepping avulsions are common in the western Amazon Fan and in the Northern and Axial Zaire Fan. However, channels in the Indus and Bengal Fans, and in the Southern Zaire Fan, appear to form radial patterns, with most of the main avulsions having occurred in relatively restricted areas (avulsion nodes).High degrees of development of channel instabilities or avulsion-threshold conditions (e.g. channel sinuosity increase and lengthening, channel thalweg aggradation, differential channel fill and decrease in channel relief, slumping and channel plugging, bank cohesion), and increased peak volumes and speeds of turbidity currents, determined where avulsions occurred in the fan channels. During the Late Quaternary intervals considered here, several global sea-level falls and rises occurred and these, along with climatic changes, are likely to have influenced sediment-gravity flows and may have thus timed many, if not all, significant avulsions. However, available data do not allow documentation of this timing in many cases.The main differences between fan areas with (1) fore-and back-stepping and (2) radial channel avulsion patterns are that, in the former case, fan gradients are steeper with frequent slope breaks, and channel thalweg aggradation occurs along much of the depositional profile. In the latter case, downdip of avulsion nodes, fan gradients are either flat or possibly have no breaks, and channels are incised below the fan surface with no thalweg aggradation in the middle and lower fan regions. These characteristics may have caused differences in avulsion-triggering turbidity flow volumes and avulsion threshold conditions that, in turn, may have resulted in different channel avulsion patterns.