HR: 1340h
AN: B23B-1270    [Abstracts]
TI: Assessing the Impact of Floodplains in Sedimentary Transport in the Amazon Basin Using the Chemistry of Riverine Suspended Load and Bedload
AU: * Bouchez, J
EM: bouchez@ipgp.jussieu.fr
AF: IPGP - Geochimie-Cosmochimie, 4, place Jussieu, Paris, 75005, France
AU: Gaillardet, J
EM: gaillard@ipgp.jussieu.fr
AF: IPGP - Geochimie-Cosmochimie, 4, place Jussieu, Paris, 75005, France
AB: Large rivers integrate over time and space the geologic events affecting their catchment. But, as dynamic systems, they also have their own characteristic time-scale. Large rivers such as the Amazon are thus likely to record important geologic events such as glacial-interglacial alternance or tectonic uplift. It has been demonstrated by several independent methods that most of the Earth's largest rivers are not eroding at steady-state, exporting amounts of material different from what has been inferred from chemical, geomorphological and tectonic studies. Floodplains can act as major sediments export filters, and could be one of the systems making rivers not to erode continents at steady-state. Nevertheless, the exact impact on riverine sedimentary storage and transfer, as well on weathering, are still a matter of debate. To assess the impact of floodplains, we propose a simple model for the Amazon basin main rivers (Amazon, Madeira and Solimões), based on the inversion of data on total and elemental riverine fluxes. This inversion method has been widely used (e.g. in Gaillardet et al., 1999) to solve geochemical issues. From the assumption of an equilibrium between erosion in the Andes, floodplain deposition/erosion, and export by the river, a set of equations is obtained and inverted, the most largely unknown parameters being deposition/erosion in floodplains and the bedload export flux. The inversion method will allow to better constrain these fluxes. A priori values are needed for all the parameters of the equations, as well as for their uncertainties. Particulate and elemental fluxes are estimated through integration of particulate concentration and composition, and water velocity and composition, over the water column of the rivers near their mouth. The chemical composition of the floodplain river sediments is constrained by the litterature (e.g. by Martinelli et al., 1993) and by correlations obtained on analyzed suspended sediments and bedload sands of the river. The composition of the drained continental crust is taken from litterature models or composite rocks, following the tributary. Large uncertainties are applied to the most unknown parameters. The results of the inversion process show that :

1. the floodplain is accumulating material;
2. and/or the bedload flux export of the Amazon is very high.