The beach barrier coast of Holland protects a large portion of The Netherlands lying below mean sea level. Without proper maintenance of this coastal system, coastal towns such as Egmond aan Zee (Egmond at Sea) would long ago have turned into Egmond in Zee (Egmond under Sea). Beach barrier coasts show a wide variety of morphodynamic processes, including tide-, wave- and wind-driven transport of sediment. The combined effect of these processes yields trends as well as (quasi-) rhythmic fluctuations in the position of the coastline and other coastal phenomema. It is important to assess trends in natural dynamics to identify any structural sediment deficit. Since the fluctuations are often of large magnitude, they may easily obscure the overall trends in coastal behaviour.

Fluctuations may be of standing or propagating nature and may have periodicities on time scales ranging from the very small (connected to the variations in wave conditions) to the very large (connect to overall sediment availability). Effective and efficient management of coastal zones benefits from a sound understanding of trends and fluctuations and the ability to predict them.

Because of the variety of spatial and temporal scales involved in coastal evolution, NCK chooses to structure its research on beach barrier coasts on the basis of three scales. Starting point is the concept of a scale cascade, which assumes that it is possible to distinguish a spatially and temporally bounded domain existing within the boundary conditions established by larger scales and governed by smaller-scale intrinsic dynamics which may be either ignored or captured by aggregation assumptions. The theme seeks to formulate dynamic models for such domains, using both theoretical and observational information. This concept of 'appropriate modelling' aims to make models both as simple as possible and yet as complex as necessary.

The theme identifies three evolutionary scales, while noting that it is yet uncertain whether these scales may indeed be distinguished as bounded scale domains. These are:

• Beach state and seasonal scale (time scales ranging from the duration of wave events to seasons);
• Management and intervention scale (time scales ranging from years to decades, in the order of the life time of hard and soft interventions in the coastal system);
• Historic and recent Holocene scale (time scales ranging from a century to thousands of years).

On the beach state and seasonal scale, research activities are based on the use of process-based modelling approaches, supported by high-resolution in-situ and remote sensing observations. On the historic and recent Holocene scale, where process-based models become inaccurate, behaviour-oriented, empirical models are used, supported by historic and geological data. On the management and intervention scale, research activities aim at furthering the prediction horizon of morphological simulations through combined use of process-based models, empirical models and data-driven approaches. Long-term data sets such as the JARKUS data set of annually-surveyed coastal profiles are an important asset in this respect.