The deposition and erosion of cohesive sediments determined by a multi-class model

  • La recherche

  • Recherche

14/12/2004

O. El Ganaoui, E. Schaaff, P. Boyer, M. Amielh, F. Anselmet, C. Grenz

Estuarine, Coastal and Shelf Science 60 (2004) 457–475

Type de document > *Article de revue
Mots clés publication scientifique > dépôt , érosion , sédiment
Unité de recherche > IRSN/DEI/SECRE/LME
Auteurs > BOYER Patrick

Sediment properties were studied experimentally using a laboratory flume and sediments sampled from both fresh and saltwater systems, i.e. the Rhone River and the adjacent continental shelf (Golfe du Lion, Mediterranean Sea). An experimental protocol is proposed which includes the development of an analytical method based on a multi-class model in order to determine the erosion and deposition fluxes of these sediments. This approach highlights the presence of a "fluff layer" characterized by a very low critical erosion shear stress and a second layer with a significantly larger critical erosion shear stress. If a global model is used, i.e. only one class of particles is considered, this distinction between the two layers cannot be made as sediment parameters are identified for the whole eroded particles including the "fluff layer". Critical shear stresses of erosion were 0.05, 0.04 and 0.025Nm-2, respectively, for the first class of particles eroded at Rhone 1, Rhone 2 and SOFI stations. The second class of particles showed critical shear stress 10-20 times higher, around 0.20Nm-2. The use of the multi-class model allowed estimation of the erosion rates in the order of 2.5×10-6, 6.2×10-5 and 2×10-6kgm-2s-1 for Rhone 1, Rhone 2 and SOFI. Settling velocities were in the order of 0.012-0.045mms-1 for the first class and between 0.05 and 0.09mms-1 for the second class. For the river experiments, it is shown that the "fluff layer", corresponding to the first class of particles being eroded, is mainly representative of recent deposits of suspended particles probably trapped in the overlying water column during sediment sampling. This highlights the difficulties in sampling sediments and transporting these from field to lab and more generally into flumes, and the limits of applying a global model to flume experiments when studying sediment properties.