When the fluid shear stress in the boundary layer exceeds a threshold value that depends on the details of the sediment and the bed, sediment grains are mobilized. The immersed weight can be supported by either fluid turbulence (suspended load) or inter-grannular collisions (bed load) or a mix (saltation). In the oscillatory flow conditions of the nearshore, the new direction and rate of transport is usually the small difference between large onshore and offshore components of the wave cycle. As a consequence, net transport under waves is a great deal harder to model than under unidirectional flow (where at least the net direction is known).
The response time of sediment can be rapid (often a fraction of a second), and space scales short (centimeters above the bed). Thus, sampling presented a difficulty that was solved for single point measurements with optical backscatter measurements in the early 1980's [ Downing, et al., 1981] and, for the multiple locations in a vertical transect, with acoustic backscatter techniques [ Young, et al., 1982; Hay, 1983].
In many ways, the results were not intuitive. For example, Jaffe et al. [1984] showed that the wave driven sediment flux could be larger than and the opposite sign from the flux product of the mean concentration and the mean velocity. Clearly the fact that these are waves is important. Huntley and Hanes [1987] further decomposed the wave-driven transport by frequency, using cross-spectral analysis. Intriguingly, they found that transport under the incident band was shoreward directed while the phase-locked infragravity motions drove offshore transport (through a mechanism previously proposed by Shi and Larsen [1983]). In addition, Hanes [1988] and many others have found that suspended sediment concentration is very intermittent and apparently not a simple response to a slowly modulating wave height.
Recent work has continued the discoveries of the complexity of small scale sediment transport. While Beach and Sternberg [1992] found reasonable agreement with fairly traditional boundary layer and sediment transport models under very strong wave forcing conditions, results from others have suggested greater complication. For example, using acoustic techniques, Hay and Sheng [1992] observed that the expected logarithmic decreases in suspended load with height above the bed were only observed for heights greater than 10 cm. At lower levels, a power law seemed more appropriate. Moreover, using several acoustic sensors spaced several meters apart in the cross-shore, Hay and Bowen [in press] found the dominant sediment response to be event-like with short-scale clouds of high concentration extending well above the boundary layer and advecting back and forth with the wave orbital motion. Conley and Inman [1992] suggest that cyclic ventilation through the bottom boundary can systematically influence boundary layer dynamics, perhaps inducing a new transport. Clearly the dynamics of suspended load transport are complex under wave and current forcing.
Little progress has been made in understanding bedload or saltation on natural beaches, primarily due to the unavailability of useful sensors. Instead, several alternate approaches are rapidly developing interest. One method is the monitoring of small scale bedform behavior, usually with acoustic techniques, on the assumption that bedform translation is primarily caused by bedload or saltation [ Hay and Bowen, 1993]. An intriguing, independent approach has been the use of cellular automaton modeling techniques [ Anderson, 1990]. These very simple, rule-based models operate under the assumption that the basic behavior of a system (for example, the generation and propagation of ripples) results not from details of boundary layer dynamics and sediment transport, but from a few elementary aspects of the interaction that can easily be incorporated into a few, simple rules.