How to improve numerical modelling of gravel replenishment with non-uniform sediment? Proof of concept with a comprehensive experimental dataset

Abstract

Gravel replenishment is a common solution to restore river channel experiencing sediment deficit. However, its effects on channel morphology and surface grain size sorting are rarely considered in restoration projects. In this study, we investigated, both experimentally and numerically, the channel responses and surface grain size sorting resulting from gravel replenishment in a straight channel with alternate and mid-channel bars. Flume experiments showed a perturbation of the initial surface grain size sorting due to the full mobility of all sediment classes under high flows of a flood and the local morphological forcing induced by the replenishment. Using the Telemac2D-Gaia code, we developed a 2D numerical morphodynamic model of the experimental flume with non-uniform sediment. A calibrated model with a classic sediment mixture module successfully reproduced the erosion dynamics of the replenishment and the channel morphological changes observed during the experiments. By testing various grain size mixtures for the stockpile and the initial bed, the simulation results suggested that coarser stockpiles provided more persistent morphological forcing against flow erosion, which continuously promoting channel instability and triggered sediment transport. The finer sediment fraction within the stockpile was transported farthest downstream during the flood, highlighting its important role in fining the downstream bed and the need of thoughtful design in restoration projects. Our study demonstrated the current model’s ability to reproduce morphological processes in the context of non-uniform sediment and it is recommended as an operational tool for optimising gravel replenishment strategies.