An advanced signal processing technique for deriving grain size information of bedload transport from impact plate vibration measurements

AbstractA reliable characterization of bedload transport is required to gauge the engineering and theoretical issues related to the dynamics of sediments transport in rivers. However, while significant advances have been made in the development of monitoring techniques, robust quantitative predictive relationships have proven difficult to derive. In this article, we develop a dedicated signal processing technique aimed at improving the usage of impact plate measurements for material transport characterization. Our set‐up consists of a piezoelectric hydrophone mounted on the bottom side of a stainless steel plate, thus acting as a ‘sediment vibration sensor’. While the classical analysis with such systems is usually limited to rather simple procedures, such as impact counting, a large amount of useful information is contained in the actual waveform of the impact signal, which conveys the force and the contact time that the bedload imposes on the plate. An advanced signal processing technique called ‘first arrival atomic decomposition’ is used to improve the characterization of bedload transport by analysing the amplitude and frequency attributes of each single impact. This new processing approach proves to be well suited for bedload transport monitoring using plate systems and allows us to establish a relationship between the median grain size (D50) and the impact signal properties. This link is first observed and validated with controlled flume experiments and then applied to continuous impact records in a small gravel‐bed river during a flood event. The estimated D50 offers a novel possibility to observe the time‐varying grain size distribution of bedload transport. Copyright © 2014 John Wiley & Sons, Ltd.