Understanding how fluid flow drives sediment transport in a river is fundamental for predicting erosion and landscape evolution, managing river infrastructure, and protecting and promoting fish habitat. Bed load transport - the movement of sediment in frequent contact with the river bottom - remains notoriously unpredictable, despite almost a century of quantitative research. The problem is particularly acute in steep mountain streams where transport occurs close to the threshold of motion, and the resulting large-magnitude bed load pulses cannot be predicted using existing mathematical models. There are two unique aspects of mountain streams that we will examine here: (1) the presence of large, rarely-mobile boulders that generate large-scale fluid turbulence structures; and (2) highly energetic grain-grain collisions that cause coordinated motion of many particles. We will perform two parallel sets of laboratory experiments to examine the influence of these two factors on the formation and breakup of granular bed structures. Flow will be mapped using particle image velocimetry, while acoustic sensors will quantify bed load transport, pressure sensors will measure fluid stresses, and laser-induced fluorescence will be used to image the granular bed. These measurements will allow us to determine the relative contributions of fluid and granular stresses to bed load transport, and to develop a refined bed load transport equation that is better suited for prediction in mountain streams. We will test whether a recent framework developed in soft-matter physics, which unites the description of transport in a wide class of disordered systems, can be useful for understanding the statistical mechanics of sediment transport in rivers.
Floods erode streambanks, take out bridges and fill reservoirs with sediment. In steep mountain streams, large boulders form natural dams that can suddenly give way during a rain storm, sending catastrophic pulses of sediment downstream that endanger property and lives. These events occur infrequently, making it difficult to obtain the measurements required to understand the physical processes and to build predictive models. In this study we will use a set of scaled laboratory experiments, which allow us to effectively speed up time by scaling down size, to understand the physical processes driving the movement of boulders in mountain streams. We will measure the forces exerted on boulders by the moving water, and among boulders as they collide with each other, under carefully controlled conditions. We will then build a mathematical model that uses this physical understanding to predict when, and how much, boulders will move in floods of different magnitudes. This model will help Earth scientists studying the long-timescale evolution of eroding landscapes. Results will also be useful for engineers and managers charged with protecting and maintaining infrastructure, and restoring natural river function, in mountain settings. In addition, this grant will provide important training for future scientists, as the experiments will be led by a doctoral student and post-doctoral researcher, and also involve several undergraduate students.
