CADA00.10 is a 10-constituent data assimilation model for Southern Ocean tides. The model uses TOPEX/Poseidon satellite altimetry and benthic and coastal tide gauge data to improve the tidal fields from the "Prior" model, CATS00.10. CADA00.10 is an application of the Oregon State University (OSU) Tidal Inversion Software (OTIS) [see Egbert et al., 1994; Egbert, 1997; http://www.coas.oregonstate.edu/research/po/research/tide/region.html ].

The grid is 1/4 x 1/12 degree (lon x lat), and covers the entire Antarctic, extending from 86^o S (to include the under-ice-shelf ocean cavities in the Ross and Weddell Seas) to 58^o S. Bathymetry (water column thickness under ice shelves) is based on ETOPO-5, updated in the Weddell and Ross Seas and under the Amery and George VI Ice Shelves ( more details on the bathymetry grid). The model northern boundary has been moved from 50^oS (in earlier CATS models) to 58^oS, since this provides a completely open, deep-water northern boundary which improves the stability of the solution. The spatial resolution near the Antarctic coast (~70^oS) is ~10 km. The 10 constituents are chosen to be the same as for the model providing the open ocean boundary conditions, TPXO5.1 (http://www.coas.oregonstate.edu/research/po/research/tide/region.html ). These are 4 diurnal (O₁, K₁, P₁, Q₁), four semidiurnal (M₂, S₂, K₂, N₂) and two fortnightly (long-period) (Mm, Mf).

Click on figure for higher resolution image (~100 Kb)

The quality of the model fit to tide height data (see Figure above), excluding T/P altimetry, is shown in the Table below. The comparison is broken into two categories: all stations excluding the Ross Ice Shelf (RIS), and the RIS by itself. The reason for keeping these categories separate is that the RIS stations are mainly gravity meter records from the 1970’s [Williams and Robinson, 1980] and we are not completely confident of their accuracy. The improvement from the forward model (CATS00.10) to CADA00.10 is not at all surprising: the data are included in the assimilation.  However, the model also significantly improves predictions for other data points including 3-D differential GPS measurements on ice shelves, and comparisons with differential synthetic aperture (SAR) interferometry [Rignot et al., 2000]. This leads us to conclude that CADA00.10 is a better model than CATS00.10 for tidal height fields around Antarctica. Preliminary studies also suggest that tidal currents are also improved by height data assimilation, although simple changes to the forward model, such as the use of linear rather than quadratic friction, can improve current predictions to a similar degree.

The CADA series of models will be extended through assimilation of current meter records.

At this time the model is available in Matlab™ and Fortran format. Further details of the model, including download instructions, can be found on our Polar Tides Model page.

We also solicit interest in comparing the model solution to new data sets, including traditional current meter, bottom pressure, and coastal tide height data, and also less traditional measurements such as ice shelf motion from differential 3-D GPS or SAR interferometry, and ice-mounted buoy drift velocities.