The Dynamics of Upland River Confluences

Abstract

River confluences are common features of the landscape as a result of the dendritic nature of river catchments. Much of the research to date has investigated flow at lowland confluences, whereas the most dynamic river confluences are likely to be found in upland areas. Therefore, in order to improve the understanding of the flow, sediment transport and morphological processes at river confluences, it is necessary to investigate the applicability of the findings from lowland confluences to their upland counterparts.

This thesis uses field and numerical modelling experiments to increase the understanding of the dynamics of these upland river confluences. Flow and particle tracking data have been obtained from two different field sites: the more stable upland river confluences of Moor House National Nature Reserve and the steeper confluences of the Borgne d’Arolla braid plain. These data have been used to test two new numerical modelling developments. The first is the application of an explicit free surface model within a Computational Fluid Dynamics framework to enable the simulation of flow in natural rivers with steep water surface slopes. The second is the development of a new, reduced-complexity Discrete Particle Model that uses the three-dimensional flow data to simulate individual particle trajectories and the Generalised Likelihood Uncertainty Estimation method to estimate the uncertainty in these trajectories.

There are three main findings in this thesis. Firstly, at upland confluences changes in the flow direction over the less well-defined banks have the greatest impact on individual particle trajectories and thus the morphological evolution of these confluences. Secondly, as water surface slopes increase, topographic features have less impact on the flow field than undulations in the water surface. Finally, turbulent flow structures and helical flow cells are found to be less important for the dynamics of these upland river confluences.