SIMULATION, DETECTION AND PREDICTION OF SEA FLOOR DYNAMICS P. Menting, L. Dorst, R. Lindenbergh and H. Wüst Peter Menting, Fugro-Inpark B.V., Leidschendam, The Netherlands. E-mail: peter.menting@fugro-inpark.nl Leendert Dorst, Netherlands Hydrographic Service, Den Haag, The Netherlands and University of Twente, Enschede, The Netherlands. E-mail: ll.dorst@mindef.nl, Roderik Lindenbergh, Delft University of Technology, Delft, The Netherlands. E-mail: r.c.lindenbergh@tudelft.nl, Hans Wüst, Directorate-General of Public Works and Water Management-AVV, Rotterdam, The Netherlands. E-mail: j.c.wust@avv.rws.minvenw.nl Abstract. The Dutch part of the North Sea is mapped by two authorities. Both authorities predict the behavior of the sea floor by analyzing time series of bathymetric data. The Hydrographic Service determines parameters for sand wave dynamics from an area, based on deformation analysis. The North Sea Directorate uses a multivariate state space modelling approach to estimate, update and extrapolate trends per grid point. Here, both approaches are described and combined in order to obtain a prediction method that incorporates a sand wave propagation model. The new method is tested on two data sets of different characteristics. 1. Introduction The Dutch part of the North Sea is being mapped by two authorities, the Hydrographic Service of the Royal Netherlands Navy and the North Sea Directorate, part of the Directorate-General of Public Works and Water Management. Monitoring the depth in the Southern North Sea is essential, because it is heavily used by shipping, but shallow. To ensure that main ports, such as Rotterdam, remain safely accessible, reliable depth information is needed. However, the depth is not constant in time, due to e.g. the development of sand waves. These are regular wave patterns, having amplitudes of several meters. These sand waves tend to grow and migrate, which may have consequences for the navigable channel depth. As it is expensive to survey large parts of the sea on a regular basis, insight in the dynamic behaviour of the sea floor is required. Both authorities have developed their own method to monitor the sea floor by using time series of bathymetric echo sounding data. In this article, parts of both methods will be described and combined to a new method that benefits from both. The Hydrographic Service is responsible for the production and upkeep of nautical publications, such as nautical charts. In order to make a more efficient planning of the bathymetric surveys used for the production of these charts, insight in the dynamics of the sea floor is needed. Therefore a method is developed, [1], for analyzing time series, based on geodetic deformation analysis. The core of this method is a testing procedure to determine if the sea floor is static or contains some kind of dynamics. Two kinds of dynamics are considered: outlying surveys and linear trends. This procedure can be applied at several scales, for single grid-points or for a