Impacts of vegetation configuration on flow structure and resistance in a rectangular open channel

Abstract

Riparian vegetation exerts significant effects on the flow structure and the geomorphology of rivers. While it provides a rich wildlife habitat as well as a protection from bank erosion, vegetation can increase the risk of flood inundation by reducing the carrying capacity of flood. A proper balance between flood control and riparian eco-diversity is therefore indispensable for better riparian management. In the present study, the effect of vegetation configuration on turbulent flows in a rectangular open channel, especially on the resulting flow resistance, has been investigated both experimentally and numerically. The laboratory experiments are conducted for two types of vegetation configurations, where vegetation is established either continuously or in a discrete way along the centerline of the flume. A regular array of emergent circular cylinders of diameter 3 mm with the stem-to-stem centerline spacing of 3 cm is used as a laboratory model for vegetation. The applicability of a basic large-eddy simulation (LES) method together with a canopy model is critically examined against the experimental data. A great agreement between the present three-dimensional LES results and the experimental data is achieved in the continuous case. On the other hand, the LES is found to fail in reproducing the streamwise variation of the mean flow structures accurately in the discrete case. This failure is mainly due to the arrangement of a uniform distribution of the drag coefficient required in the canopy model. The LES is applied to flows with a wide variety of vegetation configurations to examine the relationship between the presence of vegetation and the resulting flow resistance. It is found that, in most cases, the flow resistance solely depends on the total volume of vegetation, irrespective of the configuration details in the streamwise direction. However, shifting vegetation toward the side wall achieves a considerable drag reduction. This is partly because the primary velocity is decreased with approaching the side wall and partly because the lateral mixing is suppressed by the presence of the side wall.