ALBERT

Description: The worlds 30 largest rivers represent half of the total runoff to the ocean and thus integrate the fluxes of Earth surface weathering and erosion over a large portion of global tectonic, geomorphic, and climatic zones. In-situ produced cosmogenic nuclides (10Be, 26Al) in detrital quartz sand can be used to constrain the mean millennial-scale denudation of these large basins. Yet cosmogenic nuclides have mostly been applied to small and intermediate size basins of significant relief. One reason is that in these settings, lowland sediment storage and burial are short compared to the half life of the nuclide (e.g. 1.4 Myr for 10Be). However, if sediment storage is long compared to the half-life, paired nuclides (e.g. 26Al/10Be), through their differential decay, allow to assess the duration of sediment transfer and burial ages from source to sink[1]. Here we present a new dataset of cosmogenic nuclides from 60 large rivers that integrate over ~30% of Earth’s terrestrial surface. 26Al/10Be ratios of around 6 to 7.5 for most rivers reveal burial durations shorter than the nuclides’ decay time scales, indicating high source-sink connectivity. In slowly-eroding basins such as the tectonically quiescent Australian Murray-Darling or the central African Okavango and Congo rivers, 26Al/10Be ratios of 〈6 indicate decay of nuclide concentrations. Such low nuclide ratios evolve during Myr-scale sediment burial during slow source to sink transfer. We converted denudation rates to sediment fluxes by estimating their actively eroding source areas. Extrapolating these millennial-scale sediment fluxes to global source areas provides an estimate of the global sediment flux. The comparison with estimates of modern sediment fluxes from river load gauging offers to deciphering the controls of sediment generation versus sediment transport across large basins.