Scale-dependent dynamics of migrating bedforms and sediment transport in fluvial channels

Abstract

Dynamics of migrating bedforms and sediment transport are the fundamental building blocks in river morphodynamic processes. However, quantifying bedform characteristics and predicting sediment transport rate are nontrivial due to their large spatio-temporal variability. More importantly, recent advances in bedform quantification methods have shown that bedforms migrate and deform at different rates depending on their sizes. Yet, most existing bedform characterization methods in the literature are less effective in capturing such scale-dependent behaviors in superimposed multi-scale bedform fields. In the present thesis, the scale-dependent bedform characteristics and their resulting contribution to sediment transport are quantified using synchronized spatio-temporal bathymetry evolution and temporal sediment flux measurement data from large-scale open channel experiments. A new bedform tracking algorithm is proposed, which independently characterizes two distinct bedform scales using the Fourier decomposition: rapidly migrating small secondary bedforms and slowly translating large bedforms underneath. This new tracking method confirms the scale-dependent bedform behaviors and the results are used to validate an existing 2D spectra migration velocity model. The 2D spectra migration velocity model is then coupled with corresponding harmonic amplitudes to quantify the scale-dependent bedform contribution to sediment transport, namely a spectral bedform transport model. The new sediment transport model is essentially equivalent to the Simon's bedform transport equation in a spectral domain. The experiments reveal that large bedforms can shelter small bedforms downstream, resulting in reduced bed shear stress and migration velocity. In addition, the results show that not all bedform scales contribute to sediment transport. Instead, it is found that small bedforms are primarily responsible for sediment transport and induce migration of large bedforms. The spectra bedform transport model has great practical potential since it only requires easily measurable hydraulic parameters and bathymetric data (e.g. water depth, water surface slope, sediment grain size, and temporal bed elevation changes at fixed locations). The frameworks developed here will not only help advance our understanding on fundamental dynamics of water-driven earth surface processes, but also will be used to provide better guidelines for installing hydraulic structures and mitigating sediment management issues in rivers.