Field investigations of the threshold of grain motion by ocean waves and currents

Instrumented tripods were deployed on three occasions at continental shelf sites: twice in the United States (December, 1978, and March, 1979) and once in Australia (December, 1979). A total of 37 days of data were collected. Data included measurements of current speed and direction 100 cm off the seabed; mean bottom pressure and pressure fluctuations; water turbidity with nephelometer or transmissometer; and photographs of the seabed. Bottom sediment samples were also collected prior to each deployment. These data were analyzed to estimate the vector-averaged velocity ( U 100), tides, gravity waves, near-bottom sediment concentration, bed configuration, and bottom sediment texture. During the sampling period, sediment resuspension occurred frequently as a result of oscillatory currents due to surface gravity waves. The field data have been used to evaluate several existing relationships for predicting the threshold of grain motion under oscillatory flow conditions. The first method is an evaluation of the Shields (1936) entrainment function for unidirectional flows in which the boundary shear stress is computed using the wave friction factor of Jonsson (1966). The second is an equation presented by Komar and Miller (1973, 1975) that is based on the laboratory investigations of Bagnold (1946) and Manohar (1955). The third method computes the boundary shear stress as a nonlinear combination of stresses due to waves and currents as proposed by Grant and Madsen (1979) and further developed in this paper. The results show that the Shields diagram, in which the Shields entrainment function for unidirectional currents is plotted with respect to a grain Reynolds number, adequately predicts the threshold of grain motion on the continental shelf. Although the results of all three computational methods were within the scatter of results of the laboratory studies upon which the predictive relationships were based, the wave-current model of Grant and Madsen is preferred because it includes the influence of currents in the computation of boundary shear stress.