Pore geometry, avalanching, and subsurface flow: A sand infiltration model

The deposition of sand into gravel riverbeds has been well-documented, along with its negative impacts on developing salmon eggs and riverbank extraction for water supplies. Dam releases may be used on regulated rivers to flush the bed of fine sediment, but it is not generally known how deep the sand deposit extends or how much sand is there. One-dimensional (plane-bed) experiments consistently show that the depth of infiltration is a function of the sand and gravel grain size distributions and that the saturation sand fraction is near 8-10%. However, precise empirical relationships developed in individual studies do poorly at predicting the results of other experiments. Furthermore, no infiltration model includes the effect of flow conditions in the water column, although flow conditions clearly impact the deposit characteristics. We propose a mechanistic model for the infiltration of fine sediment and compare its predictions to the results of two recent infiltration experiments. This model is based on geometric arguments about pore and particle shape and five mechanisms: particle settling, particle capture, subsurface avalanching, average subsurface flow, and subsurface pressure fluctuations. The model successfully predicts for both experiments the fraction of sand deposited and the shape of that deposit as a function of depth.