Search

Large rivers are the main conveyors of continental material to the oceans through sediment and dissolved fluxes. The redistribution of elements is fundamental in Earth surface processes and central in various biogeochemical cycles. The nature of the exported continental material is a function of the processes operating in the river’s catchment. From external forcings such as climate, tectonics or anthropogenic activity, having a strong control on erosion and weathering, to transport dynamics and sediment storage in flood plain, chemical elements are segregated from source to sink. Evaluating the composition of the exported sediments is thus essential in our understanding of large scale processes. This raises the problem of integrating the sediments chemical composition both spatially in a river section and temporally during the hydrogram. Efforts have been made to precisely determine de total flux of transported material of major world rivers but the determination of associated chemical solid fluxes is still limited. Also it is now recognized that surface sediments composition does not reflect in most cases the average sediment composition (Galy et al. 2007; Bouchez et al. in press). Nevertheless, global chemical budgets still rely on simple averaging of the available data counting mainly on surface sediment samples. Hydrodynamic sorting of minerals exerts a first order control on the chemistry of sediments in the water column, segregating elements according to the flow dynamics and the water depth. This has to be accounted for in order to derive accurate chemical fluxes. This work is an attempt to integrate the chemical flux of the sediments transported by the Ganga. The river was sampled at various stages of the monsoon at the same location in Bangladesh between 2004 and 2010 using a point sampler to collect sediments throughout the water column. Sediments were analyzed for major elements and grain-size. The systematic use of Acoustic Doppler Current Profilers (ADCP) surveys during sampling allows us to document the hydrodynamic conditions at the sampling location. Building upon a Rousean model we are able to predict the grain size of the vertical water column and extrapolate it through the river section. Then, using the strong correlation between grain size parameters and major element composition (Al, Si, Fe) we infer the chemical flux associated with the river section. Extrapolating these fluxes through the annual hydrologic cycle we attempt to constrain the chemical flux of the sediments exported by the Ganga. The chemical composition of this exported material is finally compared to the composition of the Himalayan crust. The significant difference between these two compositions is principally interpreted as the result of the sequestration of coarse silicic material in the flood plain. J. Bouchez et al. in press, Hydrological Processes V. Galy et al., 2007, Nature 450 : 407-411