Modelling intertidal area evolution in the Western Scheldt : Hindcast (1964-2010) and forecast (2020-2100) under sea level rise scenarios
Author(s) |
M. van der Wegen
|
M.G. Aguilera Chaves
Publication type | Report Deltares
The Western Scheldt covers about 92 km² of intertidal areas, including sandy and muddy shoals and large marsh regions like the Land van Saeftinghe. These intertidal zones hold significant ecological value and contribute to flood protection through wave attenuation. Over time, the intertidal area evolves due to tidal and wave forces, as well as human interventions such as dredging and disposal activities. From 1965 to 2010, there was a 2% decrease in intertidal area and a 30% increase in mean intertidal area height.
This study assesses the performance of the Delft3D model in predicting the evolution of intertidal flats in the Western Scheldt and explores the potential impacts of sea level rise (SLR) and dredge-disposal strategies. The analysis focuses on morphodynamic hindcast results from 1965 to 2010 and forecast results from 2020 to 2100. During the hindcast period, the model’s deviations from observations were about 10% of the observed changes, indicating challenges in accurately reproducing variations. The model initially underestimated the increase in intertidal area and later overestimated the decrease, with performance generally worse when subregions were analyzed.
The 80-year model forecast shows a declining trend in intertidal area under various SLR scenarios, with significant losses under extreme SLR. Strategies disposing more sediments in the eastern part and deeper parts of the main channel lead to more intertidal area loss compared to disposing sediments near intertidal areas throughout the estuary. The study suggests that process-based modelling has potential in predicting long-term trends of intertidal flat evolution, especially when considering larger-scale interventions and long-term changes in forcing. Future work may include 3D dynamics, mud, and more advanced schematizations to improve model performance.