Relations between bed recharge and magnitude of mountain streams erosions

• The effect of bed recharge on mountain streams erosions was investigated.
• A long flume experiment was run with a bimodal sediment mixture.
• Successive bed armor construction and destruction was responsible for erosion events.
• No correlation was observed between transport rate intensity and eroded volumes.
• A conceptual model was proposed linking eroded volumes and bed recharge.

Because flood risks in mountain streams depend not only on flow intensity, but also on the volume of transported sediments, this paper presents the results of a long flume experiment aiming to investigate the relations between bed recharge and magnitude of erosion events. This 46-h experiment conducted with a natural bimodal mixture produced quasi-periodical fluctuations of transport rates and slope, despite constant feeding conditions (for water and sediments). These fluctuations were associated with successive bed armouring and erosion as well as bedload sheet production and migration.Short and long erosion events were recorded, covering a wide range of transport rates and sediment volumes. No correlation was observed between transport rates and transported volumes. Analysis showed that transported volumes were related to the degree of bed recharge only. Superposed periods were measured in the time evolution of the slope signal. A Fourier transform analysis showed that these periods reflect the influence of the different parts of the flume, which do not erode and aggrade at the same rate. All this complex behaviour was observed despite constant feeding rates and was strongly controlled by grain sorting. By extrapolating the mechanisms observed locally to the entire flume length, it was possible to propose a simple conceptual model linking bed deposition and magnitude of erosion events.Although this flume experiment still needs to be reproduced to account for all variability affecting natural flows, these results could have strong implications in risk mitigation strategies where not only exceptional flow events should be considered because (1) the volume of transported sediments depends essentially on the bed recharge and (2) the frequency of moderate flows able to transport sediments can play an important role over the long term by limiting in-channel accumulations between two major floods. Recommendations are made for risk mitigation strategies.