HR: 0830h
AN: V51C-0311    [PDF]
TI: U-Series in Large River Basins : Contraints on Processes and Timescales of Physical and Chemical Erosion
AU: * Bourdon, B
EM: bourdon@ipgp.jussieu.fr
AF: Laboratoire de Geochimie et Cosmochimie, IPGP-CNRS, 4 Place Jussieu, Paris, 75252 France
AU: Dosseto, A
EM: dosseto@ipgp.jussieu.fr
AF: Laboratoire de Geochimie et Cosmochimie, IPGP-CNRS, 4 Place Jussieu, Paris, 75252 France
AU: Gaillardet, J
EM: gaillard@ipgp.jussieu.fr
AF: Laboratoire de Geochimie et Cosmochimie, IPGP-CNRS, 4 Place Jussieu, Paris, 75252 France
AU: Vigier, N
EM: nvigier@crpg.cnrs-nancy.fr
AF: Laboratoire de Geochimie et Cosmochimie, IPGP-CNRS, 4 Place Jussieu, Paris, 75252 France
AU: Vigier, N
EM: nvigier@crpg.cnrs-nancy.fr
AF: CRPG-CNRS, BP20, Vandoeuvre-les-Nancy, 54501 France
AU: Allegre, C J
EM: allegre@ipgp.jussieu.fr
AF: Laboratoire de Geochimie et Cosmochimie, IPGP-CNRS, 4 Place Jussieu, Paris, 75252 France
AB: It has become increasingly clear that erosion is a complex function of climate, relief and lithology. All these parameters impact on the rates of chemical and physical erosion; yet, it is largely unresolved how fast the system respond to these forcing functions at the scale of a watershed. To address these questions, we have initiated over the past few years, the study of U-series nuclides ($^{238}$U-$^{234}$U-$^{230}$Th-$^{226}$Ra) in about 50 rivers of variable sizes (Mackenzie, Amazon, Narmada and Tapti, India). We have also measured complementary trace, major elements and radiogenic isotope data. In order to be able to constrain mass balances properly, we have analyzed both dissolved and suspended loads and, in some cases, bed sediments. The suspended load and bedload are in most cases depleted in the more mobile nuclides ($^{238}$U-$^{234}$U-$^{226}$Ra) that are thought to be leached during chemical weathering relative to the more immobile $^{230}$Th. By comparison, the dissolved load is enriched in the mobile nuclides. ($^{238}$U/$^{230}$Th) and ($^{226}$Ra/$^{230}$Th) in the dissolved load correlate positively. One must note that there is no correlation between ($^{238}$U/$^{230}$Th) and ($^{226}$Ra/$^{230}$Th) in the suspended load, which indicates that the timescale of particle weathering and transport must be greater than the half-life of $^{226}$Ra (1600 a). If one makes the hypothesis that the bedrock was initially in secular equilibrium, the relative fraction of U-series nuclides transported in rivers as a result of chemical (dissolved load) or physical erosion (suspended and bed loads) can be estimated. An important result of these calculations is that erosion does not always operate at steady-state as inferred by geomorphologists. By comparing the present and past rate of denudation derived from U-series, we can infer constraints about the evolution of erosion in the watershed. In Andean rivers, the denudation rate would tend to destroy currently existing soils and the residence time of particles in soils is short. Simple models for the release of nuclides using first-order release rates provide rough estimates of the timescale of weathering for the suspended load. This timescale is short (a few ka to 10-20 ka) for mountain rivers (Andes) and recently glaciated areas (Mackenzie) and commensurate with the half life of $^{230}$Th (or greater) for shield areas (lowlands of the Amazon basin) or rivers flowing on the Deccan trap (Narmada, Tapti). Bed sand transport times are in general greater than for suspended load (tens of ka).
DE: 1045 Low-temperature geochemistry
DE: 1806 Chemistry of fresh water
DE: 1832 Groundwater transport
DE: 1886 Weathering (1625)
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting