Résumé
Bedload monitoring based on seismological observations has recently emerged as a viable non-invasive measurement technique. However, its applications have only been shown for sediment transport in steep mountain rivers. We evaluated for the first time the potential of seismology for bedload transport monitoring in a rural gravel bed stream (Koulbich river in Luxembourg), characterized by low gradient (around 0.6%), small flowrate (<2.3 m (super 3) /s) and D50 transported (around 5 mm on average). The dominant anthropogenic noise in the area of investigation greatly complexified the seismic analysis but was successfully eliminated in order to extract the river's imprint on the ambient seismic field. Additional in-situ hydroacoustic measurements of bedload motion (co-located impact-plate device) and 3D hydro-morphodynamic modeling were performed to help interpreting this river's seismic signature likely dominated by water turbulences. From previous laboratory flume experiments, a calibration model was obtained between the amplitude/spectral attributes of impact measurements and the grain size of transported material, leading to an unprecedented estimation of bedload median grain size (D50) at high temporal resolution (minutes ranges) using impact-plate system. For the studied natural flood event, both simulated and estimated D50 from impact-plate measurements exhibit near-simultaneous increases, a similar temporal evolution and the same order of magnitude. The joint analysis of seismic data with hydroacoustic records and sediment transport simulation showed that the seismic records in close proximity of the stream contain evidence of bedload transport, especially in form of hysteresis behavior. These results underline the potential of non-invasive seismic measurements to estimate the triggering of bedload transport for a broader range of river systems and grain sizes than previously investigated. This study also confirms the potential of impact-plate device as robust quantitative bedload monitoring technique.