In this first part of a two-part paper we present results from a high-resolution, data-assimilative, fully nonlinear barotropic tidal model of the global oceans that excludes the Arctic. The model assimilates, in waters deeper than 1000 m, altimetric tides derived from the analysis of 2 years of TOPEX altimetric data. It also assimilates tide gage data from coastal tide gages. The model domain includes that covered by the altimeter and extends to the Antarctic. In the first part we present tidal results for the primary semidiurnal (M2, S2, N2, and K2) and diurnal (K1, O1, P1, and Q1) constituents. The second part deals with applications (Kantha, et al., this issue). The model results are compared with observational data from pelagic gages. These comparisons show that overall in the open ocean, in deep waters away from the margins of the primary basins, the model performance is comparable to other tidal models derived using TOPEX data, except for M2 and S2. However, one advantage of this model compared with those based solely on the analysis of altimetric data is that the altimetry-derived tides are subjected to dynamical constraints by the model. This results in reduction of subtidal variability often folded into tidal signals by TOPEX data analysis. It is also shown that in shallow waters along the margins, especially in east Asian marginal seas, the model differs substantially from these other models.

In addition to sea surface heights, we also present dynamically consistent barotropic currents in the form of tidal ellipses for M2 and K1 constituents. Since the goal of this research is a global tidal model uniformly valid in both deep and shallow waters, we finally present tidal elevations and currents on two well-known shallow water areas, the northwest European shelf and the northeast American shelf, and a semienclosed western Pacific marginal sea, the Bering Sea. The model results are compared with independent observations and, where possible, with results from other numerical models. The results highlight the importance of bottom topography; on the northwest European shelf, where accurate bottom topography was included in the model, the results are much better than on the northeast American shelf, where the Digital Bathymetric Data Base 5 (DBDB5) topographic database has significant errors. Results from the Bering Sea, where the topography is more accurate than the DBDB5, compare well with known tides in the region. ¿ American Geophysical Union 1995