REDUCING UNCERTAINTY IN PREDICTION OF DUNE EROSION DURING EXTREME CONDITIONS

Coastal dunes protect low lying coastal areas against the sea. Extreme waves and water levels during severe storms may cause breaching of the dunes. Consequently, serious damage due to flooding and direct wave attack could occur, resulting in loss of life and property. Proper coastal management implies that reinforcement measures will be taken if the actual safety level does not meet the agreed standard. In order to cope with small probabilities of failure, which are relevant for the Dutch dune coast, a proper safety assessment method is required. Various aspects, which are currently considered as relevant for dune erosion, are not included in the present safety assessment method. This study concerns (1) an approach to reduce the uncertainty in dune erosion prediction as well as (2) a probabilistic sensitivity analysis of various variables that are included in the current Dutch safety assessment method. The aim of the latter part is to get more insight in the influence of the stochastic characteristics of the various variables which are taken into account in the current method. The calculation values which are used for the actual safety assessment, in a semi-deterministic way, are based on a full probabilistic investigation. This full probabilistic investigation has been used as a reference for the present sensitivity analysis, in which all stochastic characteristics have been varied. Both the deterministic DUROS+ model, which is used in the current safety assessment method, as well as the process based DUROSTA model have been applied. Main conclusion is that for both DUROS+ and DUROSTA the stochastic characteristics for the water level and the grain size are the most important for the prediction of dune erosion.