Controls on flank erosion rates during valley widening: an application of cosmogenic nuclides in Andean and French valleys

Any keen observer has noticed that valleys show a large variability in their shapes: mountain rivers flow in valleys ranging from dramatic narrow gorges to valley floors that are several times river width. This geometry implies that the river has not only carved vertically but has also laterally eroded the valley flanks, depending on the action of lateral and vertical erosion on the flanks and valley floor respectively. Valley widening occurs when a river channel comes into contact with the valley edges and erode them, therefore, it depends on processes acting at the scale of a river bank but also at the scale of the river. However, valley widening processes and rate are still poorly documented while valley evolution has a key role in landscape evolution. Lateral erosion during catastrophic floods can drive valley flank erosion. Valleys are also wetlands with a voluminous aquifer and where ecosystems evolve as valleys widen or narrow over millennia. Moreover, the storage of carbon increases in wide valleys which, therefore, play a role in the carbon cycle and in the evolution of the global climate. Valley width strongly affects sediment residence times on continents, with large valleys potentially buffering sediment fluxes exported to the oceans. Finally, valleys widening is responsible for the formation of abrasion terraces and contribute to the development of pediments. The dynamics of valley widening is therefore a fundamental component of the evolution of relief and sediment flux. Given these challenges, it is becoming necessary to better understand valley widening rate and its controls. We have used and further developed the approach presented and tested in northern Chile by Zavala et al. (2021 doi:e2020GL0899).  For that, we focused on 13 river valleys (in the Arequipa Province, Peru, in the Atacama Desert in Chile and in the plateau of Valensole, France) where geomorphic, geological and climatological parameters offer a wide range of conditions to compare with erosion rates. In total, we analysed 79 samples of sands from valley flanks to measure the millennial erosion rates, by using in-situ produced Beryllium-10 (10-Be) and Aluminium-26 (26-Al). We used the detrital mean 10-Be concentration to calculate a mean local valley flank erosion rate, in a similar way that cosmogenic nuclide concentrations have been used to quantify catchment wide denudation rates. We also extracted factors that may control widening rate (valley width, slope of flanks and valley floors, incision and drainage area) for comparison with 10-Be and 26-Al concentrations. Our preliminary results in 31 samples show comparable 10-Be concentrations along a single stretch of valley, except for several outliers, for different valleys in the Andes and France, indicating some robustness in the sampling method. These results are promising and should provide new constraints on factors that control the widening rate of valleys.