This thesis is concerned with the small-scale processes influencing sediment transport in the nearshore. These processes can be summarised by the 'sediment triad', a concept describing the complex interactions between hydrodynamic forcing, bed morphology and suspended sediment. The main original contribution to the literature is considered to be a new timeevolving procedure for the prediction of bedforms and suspended sediment concentration profiles (including the hitherto unknown contribution of current mega-ripples to sediment entrainment), and the formulations contained therein . The procedure, in the form of a selfcontained computer algorithm, is aimed at end-users requiring knowledge of sediment
transport in application to practical problems, such as coastal planners or engineers.
Much of the existing knowledge regarding near-shore processes has been obtained in controlled laboratory conditions, and bridging the apparent gap between this knowledge and observations in the field was a main motivation of the present research. In particular, consideration was given to processes thought to be important in the field, namely the time-lag
between wave/current forcing and bedform evolution, and the relationship between ripplegenerated roughness and sediment suspensions. The long-term, high resolution, mostly acoustic dataset of the LEACOAST2 (LC2) project forms the basis of the present analysis. The data were collected along the shoreline of Sea Palling, Norfolk, in water depth of
approximately 8m dominated by strong tidal currents and exposure to both local and swell waves, with median grain sizes of 0.39mm. The data comprise the results of Acoustic Doppler Velocimeters (ADV), an Acoustic Backscatter System (ABS), and a three-dimensional Acoustic Ripple Profiler (3D-ARP). The mean of the highest 1110th of wave heights and the
peak spectral wave period were found to best represent the influence of waves on sediment transport. In generating the current velocity profile, both the impacts of ripple roughness and waves were incorporated. A new method for the conversion of three-dimensional bed morphology data (including bimodal beds) into representative bedform heights and lengths, based on turning points, is presented. An existing method for predicting the suspended grain size is shown to compare well with ABS data, and the associated results are used to invert backscattered voltages to suspended sediment concentration.
Using the results of this data analysis, the following conclusions have been reached. Wave/current dominance at the bed is an important control on the bed roughness and associated sediment suspensions, and the boundary between the two regimes can be quantified using a non-dimensional stress parameter. Current ripple dimensions are dependent on grain size only, until a critical flow to grain size ratio is reached, when larger current mega-ripples form. The existence of these current mega-ripples is shown to be an important control on suspending sediment away from the bed. Wave ripples are dependent mainly on the near-bed orbital diameter, until ripple break-off occurs, which is itself controlled mainly by the wave period. Wave and current ripples can exist in the same locality, but most probably represent growth-limited forms of each individual component. The incorporation of time-dependency in ripple prediction allows almost all of the above observations to be included in the proposed predictive algorithm. By allowing bedforms to evolve in response to
changing forcing, both wave and current dominant bedforms can be predicted, and the rate of the response is shown to be related to the non-dimensional mobility of the grains. Regarding suspended sediment concentration profiles, the reference concentration can be predicted with reasonable accuracy using a wave-current grain roughness bed shear stress, and is related to velocity to the power six. In contrast, the rate of vertical sediment transfer (sediment diffusivity) is shown to be heavily dependent on the total ripple-related roughness, indicating the importance of convective entrainment processes. A semi-empirical model parameterised by the ripple-dependent sediment diffusivity is derived, and shown to have reasonable skill at predicting the related concentration profiles.
The final product, namely the time-dependent practical algorithm (termed nearShore_susSed), is internally validated using the LC2 data, and shown to have significant skill at simulating the observed bedforms, associated bed roughness and corresponding concentration profiles. Through testing the algorithm using wave-alone and current-alone inputs, randomly generated inputs, and sensitivity to uncertainty in inputs, the robustness of the simulation is confirmed . Favourable comparison with datasets collected as part of the SANDPIT project provides partial independent validation. Though there are a number of limitations that should not be disregarded, the algorithm is considered to fulfil the overall aim of the thesis, namely a simplified, but state-of-the-art, procedure for time-dependent bedforms and concentration profiles in the nearshore.