Modelling Suspended Sand Transport Under Breaking Waves

The modelling of coastal morphodynamics has often been hindered by the lack of robustness/accuracy of constituent formulae, especially sediment transport formulae in the breaking and swash zones. Consequently, modellers are often forced to rely on crude calibration efforts and practical models consisting of empirical tuning-constants, to obtain favourable model results. Such methods are often unavoidable however due to theoretical limitations of existing models. The aim of this thesis is therefore to improve accuracy and applicability of suspended sand transport models for breaking wave conditions, for implementation into morphodynamic modelling studies.
Several existing suspended sand transport models (6 reference concentration C₀ + 5 concentration profile C[z]) models were evaluated quantitatively and qualitatively against one another, and against state-of-the-art high-resolution datasets which were collected under large-scale breaking wave conditions. Numerous limitations were observed in existing models, with the most common of these being their inability to accurately replicate suspended transport patterns in multiple cross-shore regions. This was due to various issues, such as not adequately accounting for the effects of breaking-induced turbulent kinetic energy on resulting sand transport. This resulted in large discrepancies between computed and measured transport particularly in the highly turbulent breaking zone. Such poor performance in computing C₀ and C[z] had residual effects on the resulting suspended flux (uC[z]) and current-related transport rate (qsc) computations also, which are essential to the accurate modelling of morphodynamics, particularly in the medium- to long-term.
A novel set of suspended sand transport (C₀ + C[z]) models (“L19”) were developed for breaking wave conditions and evaluated against the aforementioned existing models and datasets. The L19 formulae showed significantly greater performance than all existing models, indicating excellent agreement with measured data in all tested cross-shore regions. These improvements led to considerably better estimations of uC[z] and qsc, which have promising implications for future morphodynamic modelling.