Prediction of surface flow hydrology and sediment retention upslope of a vetiver buffer strip

Abstract

Vegetated buffer strips are widely used to reduce fluxes of eroding soil and associated chemicals, from hillslopes into waterways. Sediment retention by buffers is time-dependent, with its effectiveness changing with the deposition process. Our research focuses on settling of sediment upslope of stiff grass buffers at three slopes, under subcritical flow conditions. A new model is developed which couples the hydraulics, sediment deposition and subsequent adjustment to topography in order to predict water and sediment profiles upslope of a buffer with time. Experiments to test the model were carried out in the Griffith University Tilting-Flume Simulated Rainfall facility using subcritical flows at 1, 3 and 5% slopes. Water and sediment profiles were measured at different times as Vertisol sediment was introduced upslope of a vetiver grass strip. A region of increased flow depth (backwater) was produced upslope of the strip which increased in depth and decreased in length with increasing slope. Backwater height could be predicted from flow rates and thus could be used as an input for the model in the absence of experimental data. As slope increased, sediment was deposited closer to the grass strip, moving into the grass strip itself at 5% slope. The grass strip was less effective in reducing sediment in the outflow as slope increased and differences between slopes were significant. Model prediction of water and sediment profiles compared reasonably well with measured data, giving low root mean square errors and high coefficients of model efficiency. Masses of deposited sediment were generally simulated within 20% of measured values. However, simulated particle size distributions of deposited sediment were less accurate.