Schweiger, C.; Koldrack, N.; Kaehler, C., and Schuettrumpf, H., 2020. Influence of nearshore bathymetry changes on the numerical modelling of dune erosion. Journal of Coastal Research, 36(3), 545–558. Coconut Creek (Florida), ISSN 0749-0208.

Coastal dunes are an essential protection measure at the German Baltic Sea coastline. As the breaching of dunes during a storm is very complex and detailed data is still rare regarding the evolution, width and the depth of the breach and how the hinterland will fill through a breach and empty after drawdown, the aim of the collaborative project PADO is to investigate how a real dune breaches during a storm and how the hinterland is flooded. Within the scope of a physical model test, a large-scale research dune was built at the beach of Rostock-Warnemuende in Germany. During the modelling, the nearshore bathymetry was measured monthly and a high variability was observed. The objective of this study was to investigate the influence of a varying initial nearshore bathymetry on the results of coastal erosion modelling. Therefore, four XBeach models which differed only in nearshore bathymetry were set up and forced with a storm surge event from October 2017. The results were analyzed with regard to the computed post-storm bed level. Using the nearshore bathymetry of July and November 2017, large amounts of the research dune are eroded. The eroded sediment minor decreases when the nearshore bathymetry of September 2017 is applied. On the contrary, the sole use of freely accessible bathymetry data leads to significant decrease in the computed dune erosion. This leads to the conclusion, that the initial nearshore bathymetry can have a significant influence on the computed post-storm bed level. Especially when only freely accessible bathymetry data can be used, deviations between post-storm measurements and computed dune erosion can occur. This might be due to out-of-date data or a too coarse resolution. The results also show that pre-storm bathymetric measurement does not necessarily have to have a high resolution in order to achieve reliable results.