SEDIMENT MOTION AND DISPERSION IN RIVERS: INSIGHTS FROM LIDAR DIFFERENCING, PARTICLE TRACKING, AND NUMERICAL MODELING

Sediment transport theory has tended to focus on predicting bulk fluxes of sediment rather than the trajectories of particular grains or packages of material. Yet sometimes the fate and transport of particular grains matters. For example, sediment grains may carry information such as isotopic content, luminescence, composition, and degree of weathering. In rare cases, the sediment may constitute or contain dangerous contaminants. These examples highlight the need to understand not just how much sediment is moving, but where it could have come from and where it might end up. In this paper, we review strategies for measuring and modeling the displacement of parcels of sediment in rivers. One method is to compare lidar images from before and after a flood event. This method is illustrated with an example from the Rio Puerco, New Mexico, USA, where differencing of pre- and post-flood lidar reflects the fate of sediment entering a channel reach after being eroded from a recently de-vegetated section upstream. Comparing the depositional patterns with a simple one-dimensional model of sediment transport leads to the inference that the bulk of the eroded sediment was deposited on floodplains within a few kilometers of the source. A drawback of this method, however, is that it cannot directly reveal grain trajectories or their distributions. Obtaining such data requires particle-tracking technology such as passive-radio tagging. We illustrate the capability of this method with an example from a gravel-bed stream in the Colorado Rockies, USA. Here, a six-year tracer experiment reveals median annual grain travel distances ranging from several centimeters to nearly one hundred meters, depending on the magnitude and duration of the seasonal snowmelt hydrograph. The experiment shows a strong component of dispersion, with many particles remaining immobile while a few race well ahead of their peers. Strong dispersion in part reflects large cross-channel gradients in transport energy. Data such as these demand a theoretical framework that honors the probabilistic nature of sediment entrainment and dis-entrainment. Examples of such a framework, building on recent work in stochastic transport theory, are used to illustrate the potential for probabilistic mapping of transport trajectories using high-resolution topographic data.