Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
Braided rivers form some of the most fascinating fluvial patterns found on Earth. They are identifiable by their unique morphology of complex networks of intertwined channels that spread across wide floodplains. Detailed knowledge of their dynamics is needed to define proper river management strategies that can address both human needs (e.g. protection against floods, bank migration, etc.) and natural needs (e.g. the preservation of fauna and flora, river restoration, etc.).
Recently, the study of braided rivers has undergone significant progress. Developments in the areas of laboratory experiments, monitoring techniques and field surveys, in addition to new paradigms in the field of geosciences and mathematical modelling have greatly improved our understanding of braided rivers. However, many questions remain unanswered. Is it possible to predict the long-term evolution of a braided river under steady flow conditions? More fundamentally, where do the braided pattern emerge from? Does it grow out of an intrinsic flow instability? And, if this is the case, which one? The present work aims to fill two specific gaps in the current state of knowledge: the dynamics of braided river networks and the development of a morphodynamic model that uses a non-equilibrium bedload formula that can predict bedforms that ultimately produce braiding.
This thesis studied the dynamics of the braided networks experimentally. Two laboratory-scale experiments were performed from which we extracted and investigated the braided network's temporal evolution. A set of variables describing the network was determined -namely the number of nodes, the number of links and the network's total link length. These variables were shown to relate to the flow conditions. Moreover, the evolution of the braided network was described by identifying similar network configurations as modes. The modes' evolution was well captured by their probability. Using a Markov process, we were ultimately able to reproduce the probability of occurrence of those modes.
A morphodynamic model based on the shallow-water equations and a stochastic-based bedload transport formula was developed. Applying linear stability theory, we were able to obtain marginal stability curves that predicted the development of bedforms. Two types of bedforms were identified: two-dimensional bedforms (antidunes and dunes) and three-dimensional bedforms (bars). The results agreed well with the literature. The present work was the first morphodynamical model to predict the development of both dunes and bars within the same framework using shallow-water equations. Moreover, we were able to show, albeit qualitatively, the influence of particle diffusion-present in the bedload transport equation-in the development of bedforms.