Defining a single set of calibration parameters and prestorm bathymetry in the modeling of volumetric changes on the southern Baltic Sea dune coast

Abstract

The studies described herein aimed to estimate the accuracy of determination of the volumetric changes on the dune coast of the southern Baltic Sea through the application of the XBeach numerical model, which is crucial for coastal engineering. In the first phase of the study, the profile (1D) mode of the model was adapted to 19 cross-shore profiles located along the Dziwnów Spit.The model was calibrated with a storm event in 2009 that caused significant changes to dunes and beaches. Cross-shore profiles were measured approximately one and a half months before and after the storm. An evaluation of model performance was made based on the Brier skill score (BSS), the visual match of the profile shape (VMS), the absolute volumetric change error (m3/m) and the relative volumetric change error (%). In this study, parameters related to the asymmetry transport (facua) and the dune erosion algorithm (wetslp) were taken into account.The best results for model calibration on all 19 cross-shore profiles were obtained with facua values ranging from 0.16 to 0.40 and wetslp values from 0.35 to 0.60. The calibration of individual profiles yielded good results, with an average absolute error of approximately 4 m3/m and an average relative error of ca. 20%. The poorest results were collected for the profiles situated near coastal engineering structures, where the average absolute error was 10 m3/m and the relative error was 60%. The possibility of accepting one set of parameter values for all the profiles at once was also investigated. These studies revealed that the application of one set of facua and wetslp values for all profiles simultaneously resulted in a relative volumetric change error of ca. 25% on average, with the maximum of about 40%.Due to the difficulty of collecting data just before and after the storm event, complex studies using all available bathymetric data were performed. Using a joint dataset composed of prestorm topography recorded before that storm and bathymetry from different years, a simulation of the 2009 storm event was carried out. The studies revealed that the prestorm bathymetry and the randomness of the selection of calibration parameters have similar effects on the accuracy of volumetric changes.Moreover, the impact of the nearshore bathymetry (to a depth of 2 m) on modeling the volumetric changes in the terrestrial part of the shore is evident. A change in the sea bottom inclination and a successive change in the nearshore sediment volume can increase the difference between modeled and actual volumetric changes.