Improving the predictive power when modelling physical effects of human interventions in estuarine systems

Abstract

To make substantiated decisions on human interventions in estuarine systems, it is nowadays common practice to assess the effects of proposed changes using state-of-the-art numerical models. Unfortunately, these models cannot be calibrated to represent the new dynamics in an optimal way, as there is no data available yet for the system after intervention. To overcome this problem, it is advisable to acquire model results for a wider range of parameters than those obtained from calibration. However, it is generally too costly and time-consuming to make the required sensitivity analyses with a state-of-the-art model. To overcome this complication, we propose to combine the results of idealised models, state-of-the-art numerical models and observations in an integrated way. Idealised models are semi-analytical, developed to assess the importance of various mechanisms and investigate the sensitivity to model parameters, parameterisations and geometry. Based on the proposed combination, a well-founded choice can be made for a few selected scenarios to be investigated using state-of-the-art numerical models. Thus, the combined use of these different types of models leads to a clear synergy.As an example, we investigate the influence of the location of the weir in the Ems Estuary on the water motion and sediment dynamics. The idealised model results show that an increase of estuarine length will deform the tidal characteristics (the M2 semi-diurnal tidal constituent and its first overtide) so that the lowest low water increases, the tidal range is decreased, and the sediment trapping in the freshwater region will decrease. These results are qualitatively reproduced by a state-of-the-art model, resulting in a new possibility to renaturalize the Ems Estuary.