Quantifying the impact of low flow periods on bed topography and bedload transport rates

Abstract:

Despite the large body of scientific work focused on bedload transport, it remains difficult to predict accurately. Most empirical formulations cast the bedload flux rate as a function of shear stress acting on the bed in excess of a defined threshold value. This critical stress divides the hydrograph into two portions, an above threshold regime associated with bedload transport and a below threshold regime, where transport does not occur. The bulk of previous studies have concentrated on the portion of the hydrograph above threshold, leaving the below threshold region largely unexplored. A small number of previous studies suggest that increased time between transport events results in decreases in bedload flux rates. In many cases, these decreases in transport are attributed to compaction of the bed. Compaction potentially decreases mobility by increasing the friction on individual grains and the friction angle of particles that have settled into pockets. However, changes in bed topography during low flow have not been explicitly linked to changes in subsequent behavior during transport events.

In this study, we carry out flume experiments to examine the impact of sub-critical threshold flow duration on surface grain size, bed topography, and bedload flux rates. Preliminary experimental findings using a 5 m tilting flume with an 8 mm D₅₀ grain size mixture indicate that transport rate is very sensitive to low flow duration, consistent with previous studies. We generate high-resolution topographic maps of the bed surface using Structure from Motion photogrammetry. Analysis of bed topography indicates that reduced transport rates are associated with lower mean bed elevation and standard deviation of bed topography. Specifically, we document reductions in transport rates of ~30% associated with reductions in bed topography on the order of 50% of D₅₀ and changes in the standard deviation of bed topography on the order of 100% of D₅₀. Automated analysis of median surface grain size does not show a significant increase during low flow periods, suggesting armoring is not the cause of the observed changes in transport rate. These results have implications for river restoration efforts where controlled dam releases govern the recurrence interval of sediment transport events.