ALBERT

Description / Table of Contents: Dutch waters lack an accurate and easily accessible 3D description of the lowest astronomical tide (LAT) surface, i.e., a so-called “separation model”. This tidal datum, defined as “the lowest tide level that can be predicted to occur under average meteorological conditions and under any combination of astronomical conditions” (International Hydrographic Organization 2011a, Technical Resolution 3/1919), is, in those regions where tides have an appreciable effect on the water level, adopted as chart datum (CD) by the International Hydrographic Organization (IHO). Having an accurate separation model of LAT provides numerous benefits to a variety of commercial and non-commercial users, including faster, cheaper, and more accurate hydrographic survey data for nautical charts; more precise navigation, even over areas with submarine hazardous objects; merging of depth data with height data for coastal zone management; and an accurate planning of depth maintenance in port approach areas. The current practice to obtain a coastal-waters-inclusive continuous (CWIC) separation model of LAT is to express it as the sum of LAT values derived from a global/regional ocean tide model and an altimeter-derived mean sea level (MSL) model, complemented by LAT values derived from water level observations and GNSS ellipsoidal heights at tide gauges. The latter are the sole information source for deriving the separation model of LAT in coastal waters. This is due to the lack of information about the MSL in these waters as a consequence of the degrading accuracy of radar altimeter data in the vicinity of land (e.g., Andersen & Knudsen 2000, Deng et al. 2002). To obtain a continuous surface in coastal waters out of these point data, strong interpolation is required. This interpolation is not trivial as the tidal behavior at a particular tide gauge location, and hence the LAT value, is not necessarily representative for nearby locations. Furthermore, significant gaps may exist between the tide gauge locations. Alternatively, information about the MSL can be derived from GNSS surveys (e.g., Pineau-Guillou & Dorst 2011). These are, however, expensive and therefore this approach is not preferable. To derive an accurate and continuous separation model of LAT in Dutch waters including the coastal waters, Wadden Sea, and Eastern and Western Scheldt estuaries, avoiding interpolation, we developed an alternative approach. In this approach, the ellipsoidal heights of LAT are computed as the sum of the quasi-geoid heights and the heights of LAT relative to this quasi-geoid. The latter can be derived from modeled water levels of a vertically referenced hydrodynamic model, which comprise the tide and the time-averaged meteorological and steric contributions. In this study, a new quasi-geoid model is estimated that covers the whole Netherlands Continental Shelf and the Dutch mainland, because the European Gravimetric Geoid 2008 (EGG08) lacks accuracy for two reasons. First, EGG08 does not include data acquired by the Gravity field and steady-state Ocean Circulation Explorer (GOCE) satellite; these data have become publicly available in May 2010. Second, in the estimation of EGG08, satellite radar altimeter data have been corrected for using a global ocean tide model. In this study, we use a shallow water hydrodynamic model which provides apart from tides also the surge and steric contributions to the dynamic topography corrections. The fact that all contributions are obtained from the same model has the additional advantage that any non-linear interaction between the three contributions (e.g., between tide and surge (Prandle & Wolf 1978)) is accounted for. The fact that we have to use a hydrodynamic model both in deriving the separation between the quasi-geoid and LAT and in estimating the quasi-geoid itself requires a small, but important, change of our approach to derive a separation model of LAT. Instead of a one-way link between the quasi-geoid and the hydrodynamic model, a two-way link exists. This two-way link results in a chicken-and-egg problem. On the one hand, we aim to use a hydrodynamic model to derive a proper quasi-geoid, while on the other hand this quasi-geoid needs to be the model’s reference surface. The development of a methodology that solves this problem is our main research objective. In this thesis, we first present the overall procedure by means of a flowchart. Three steps can be distinguished: (i) the data preprocessing; (ii) the estimation of the quasigeoid and the realization of a hydrodynamic model that provides water levels relative to this quasi-geoid; and (iii) the realization of the LAT and mean dynamic topography (MDT) surfaces once the final quasi-geoid and the vertically referenced hydrodynamic model are available. The latter surface can be used to produce the ellipsoidal heights of the MSL (as the sum of quasi-geoid and MDT), which is often used as a vertical reference surface in the offshore industry. The model used in this study is the Dutch Continental Shelf Model version 5 (DCSM), which is a 2D storm surge model. In the remainder of the thesis, we elaborate on four key elements of this procedure.