Tide Predictions in Shelf and Coastal Waters: Status and Prospects

Abstract

The challenges of supplying accurate tide corrections for satellite altimetry in shelf and near-coastal environments are reviewed. Relative to the deep ocean, tides in shallow water are generally larger and shorter wavelength. Nonlinearity further complicates tide prediction because of the multitude of additional compound tides and overtides that must be accounted for and because standard deep-ocean methods of inferring minor tides are inapplicable. Development of accurate shelf and coastal tide models with high spatial resolution requires assimilation of high-quality data; for most of the globe, the Topex/Poseidon and Jason time series constitutes the primary source. Data assimilation with nested high-resolution models tied to lower resolution global models appears feasible, but it places severe demands on the accuracy and resolution of bathymetry data. Some envisioned next-generation altimeter missions, capable of mapping the ocean topography at very high resolution, could relax these demands by providing direct tidal measurements at the requisite scales.

Appendix

The tide-gauge validation dataset used in Sect. 8.3 consists of tidal harmonic constants at 179 sites. This appendix briefly describes the sources of these data. Our present compilation should be considered preliminary owing to its inadequate spatial distribution (Fig. 8.5) and an incomplete catalog of constituents at many of the sites. For these reasons we are endeavoring to build on this initial dataset.

For reasons discussed above our dataset focuses on shelf tides as opposed to truly coastal tides. Most of our data are therefore tidal constants from bottom-pressure recorders sitting in shelf waters well away from coastlines. We also use data from small offshore islands.

Of the 179 stations, 50 were extracted from the Global Undersea Pressure (GLOUP) archives, housed at the Proudman Oceanographic Laboratory (POL), U.K. In addition to the pressure data, the GLOUP site stores outputs of tidal harmonic analyzes, which we have adopted for selected sites. We have selected only those GLOUP stations with at least 1 month of bottom-pressure data and have rejected a number of stations with evident errors. The most common errors appear to be time-tag problems. GLOUP analyses having poor agreement with the Cartwright-Zetler IAPSO (Cartwright and Zetler 1985) compilation of tidal constants were also rejected (or corrected if possible). An additional 32 bottom-pressure stations were extracted from the IAPSO compilations (Cartwright and Zetler 1985; Smithson 1992), although these sources usually list only the eight major constituents, but sometimes include M₄. Again only tidal constants based on at least 1 month of measurements are included.

Although the GLOUP database covers well the northwest European Shelf, we have included an additional 26 stations in the North Sea and offshore Belgium and The Netherlands, originally analyzed and compiled by Mariene Informatie Service, Rijswijk, The Netherlands, and distributed as “North Sea Tidal Data 1984–1993.” Some of these stations are bottom gauges, others are from North Sea oil platforms. Nine time series are from 1 to 5 years duration; all the remaining are at least 1 month and thus probably of comparable quality to the majority of our GLOUP stations.

Another large set of stations (41 sites) were extracted from the Australian National Tide Tables (Anonymous 2007). For the tests of Sect. 8.3 we tried to select only offshore gauges that appear to be open-ocean sites, although it is possible that a few selected stations too close to large barrier reefs have been inadvertently included. For Sect. 8.5 we also added a large number of stations on the main northern Australian coast, although we attempted to avoid small bays and estuaries or stations located in large rivers. The Australian tables include a number of minor tides often susceptible to nonlinear perturbations (e.g., µ ₂), which are useful to our Table 8.2 tests, and they also include compound tides M₄, MS₄, and 2MS₆.

Additional bottom-pressure sites from the western North Atlantic shelf were extracted from the compilation of Moody et al. (1984) and from one Woods Hole report (Irish and Signell 1992). One station from the latter (located in Wilkinson Basin) appears highly suspicious of a timing problem, but we have kept the site for now. Unfortunately, the Moody report tabulates only the five constituents M₂, N₂, S₂, K₁, and O₁. To obtain a more complete tidal spectrum at these important sites, a reanalysis of the original pressure time series would be highly desirable.

The remaining stations were taken from miscellaneous sources, e.g. (Teague et al. 1998, 2001).