Study of the wind drag coefficient during the storm Xaver in the German Bight using data assimilation

Wind stress is the most important driving force for storm surge in coastal waters. Currently, the parameterization of wind stress is still not well developed and is the most uncertain parameters for storm surge models. In this paper, a storm surge model for the German Bight is developed using the model Delft3D. The data assimilation method, adjoint-free 4Dvar, which is relatively easier to be implemented than the adjoint method, is used to assimilate measured water level data into Delft3D. The storm surge prediction is more accurate when we adjusted the drag coefficient using the adjoint free data assimilation method, than in a companion forward model. The drag coefficient is consistent with results from previous studies. The wind drag coefficient is also calculated using results of a wave model. It has been found that the relation between the drag coefficient and wind speed is approximately linear in the deep water. But in the shallow water, wind drag coefficient shows larger variability due to the wave shoaling effect. The linear relation adjusted by data assimilation agrees well with that derived from a wave-related formula. It is also found that wave-related parameterization scheme is necessary for shallow waters.