The accumulation and decay of near-bed suspended sand concentration due to waves and wave groups

Abstract

High resolution acoustic measurements were made of suspended sand and bedform dimensions caused by prototype-scale waves, both regular and in groups, over a mobile sand bed, in a very large wave channel. The changes in wave height at the beginning of the regular waves and within wave groups provides an opportunity to examine the time lag in the response of the sediment. For regular waves suspended sand concentrations lagged the forcing waves with the lag increasing with distance from the seabed. Typically, near-bed (1–2 cm) concentrations reached an equilibrium one to two wave-periods after the waves themselves had reached their steady height while at elevations of 10–15 cm the lag was longer. This lag was interpreted as due to the continual injection of turbulence into the water column from vortex processes associated with the oscillatory wave boundary layer over bedforms. A similar pattern was seen for wave groups, with the sand concentration near the bed lagging by the waves by one to two wave-periods and increasing with distance from the bed. Despite the controlled nature of these prototype-scale suspension experiments, with detailed measurements of bedforms and attempts to achieve ‘equilibrium’ bedforms, considerable variability (±30%) in the suspended sand concentration occurred between ‘similar’ forcing conditions, both at a wave-to-wave level and on the scale of groups and longer. The results suggest that considerable variability (a factor of two or more) should be expected in the suspension due to turbulence produced from wave boundary layers in natural environments, where bedforms are frequently continually evolving as the waves change their height, period and direction. A simple wave-average suspended-load model is used to describe the major temporal features of the suspension and to quantify the lag of the suspended sediment in relation to the waves and wave groups. Quantification of the lag is essential for assessing the transport of sand at infra-gravity frequencies. A decay rate of 0.06 s −1, applied to antecedent waves, decreased the mean average error (MAE) by a factor of 3 when predicting the suspended load of the repeating wave groups. When tested against five further data sets (including random waves and wave records from the SANDYDUCK field experiment) including the decay rate of 0.06 s −1, resulted in a decrease in the MAE of a factor of 1.5–2 (compared to the same model with no lag). The entrainment (pickup) constant in the same model was variable and no consistent pattern was found, although there were suggestions of a link to the location of the measured profile relative to the bedforms.