Riverbed morphology and flow resistance: integrating UAV-based 3D mapping and flow modeling

Abstract

A comprehensive understanding of flow resistance is essential for understanding the hydrodynamical functioning of a river channel and predicting its evolution within the landscape. It governs the distribution of shear stress along the channel's boundaries, which in turn impacts the extent and distribution of erosion on the riverbed and banks. Flow resistance in steep mountain streams is primarily form drag and they exhibit high relative roughness. They tend to exhibit complex bed morphologies due to presence of large immobile boulders, aggregates of sediment particles, and specific bed or channel configurations. This roughness leads to the development of intricate three-dimensional flow patterns that may modify lift and drag forces acting on bed sediment. These alterations in forces can, in turn, impact water velocities and sediment fluxes within the stream. Despite extensive research, there is a lack of widely accepted theory addressing flow resistance in rough channels. Empirical evidence has established that conventional statistical description of bed morphology prove insufficient in fully describing riverbed roughness. This is why it is important to establish whether the morphology of riverbeds could exhibit a degree of inherent organization rather than appearing random. If this were to be true, riverbed morphology would exhibit a degree of consistency across different rivers and environments, enabling the formulation of generally applicable expressions for surface roughness that can be implemented in flow-resistance equations. Recent advances in uncrewed aerial vehicle (UAV) based structure from motion (SfM) allow for centimeter accurate measurement of the 3D morphology of river channels over hundreds of meters of river length. Obtaining accurate representation of the river channel's 3D topography serves as the starting point for examination of channel geometry and river bed roughness. Taking advantage of this, we have developed a workflow to characterize the geometry of natural river reaches, including shape metrics, estimates of hydraulic geometry, and grain size and location distributions based on automated image segmentation. We present our findings from a number of river channels in Taiwan. However, the most relevant way to characterize the geometry of a riverbed is through the interaction with the flow. The application of 3D flow modeling has proven to be of great importance in understanding flow patterns and sediment transport. We present plans to conduct 3D flow modelling over the measured river reaches in an effort to understand the most relevant aspect of the riverbed geometry to flow resistance.