The morphodynamic model developed by DHI Water and Environment has been tested against several field and laboratory data sets. Modelled and measured sea bed elevations, currents, water level gradients and sediment transport rates were compared using three different data sets.
The sea bed elevation comparisons used the field data sets from the Tweed River sand bypassing project which included wave, tide and sea bed measurements. This data set was combined with sediment size, river flow, rainfall and water level measurements from government agencies and developed into morphodynamic model covering the Tweed entrance ebb tide delta and the surrounding coastline. The model results show limited agreement with the measured sea bed levels. To improve the model stability, useability and results, several improvements were implemented, including a new stable bed updating scheme, an accelerated sediment transport rate tabulation, improved boundary conditions and refined spatial resolution. Further, it is shown that dramatic changes in forcing neither significantly improve the results nor change them.
This lack of response to changes in the forcing is attributed to underestimations of gradients in water levels, currents and waves, ultimately leading to sediment transport gradient being underestimated. This is confirmed by detailed water level gradients obtained from Moreton Island rip current experiment, where the current model underestimated water level gradients by at least a factor two. However, it is also shown that these gradients can be well predicted when wave-current and current-current (cross-shore current mixing the longshore momentum) interactions are included in the model. Unfortunately, at this time, these two features require significant computational resources and are not practical with current computers and broad area models such as the Tweed River.
Finally, the sediment transport model was compared to field scale laboratory experiments from the Hannover wave flume. These measurements differ from the traditional wave tunnel tests in that the near-bed velocity exhibits acceleration-asymmetry. The sediment transport model compares well to the measurements if driven by the measured near-bed velocities, as does a new simple formulation by Nielsen and Callaghan . The robustness of predictions diminishes when the model is used to estimate the near-bed velocity. In particular, the model predictions show that different near-bed velocity formulations are required to closely reproduce the measurements. Again, this indicates that when the near-bed velocity formulation is held constant, as required in the morphodynamic model, incorrect sediment transport gradients will results.
This study also provides practical methods to; reduce the computation effort involved in calculating sediment transport rate tables; incorporate complex cross-shore boundaries; and stably incorporate sediment sinks and sources. The study also provides a thorough review of bed updating schemes and the key parameters in terms of numerical practicalities that make up the sediment transport model. Finally, the study generates a list of morphodynarnic model modifications and parameter changes that do not significantly improve morphological results.
The model review and testing indicated that including; wave-current interactions (wave modelling); and improved sediment transport formulation offer two possible model improvements that would provide morphodynamic predictions. The model review and testing indicated that including; wave-current interactions (wave modelling); and an improved sediment transport formulation offer two possible model improvements to achieve improved morphodynamic predictions.