Ocean tide models representing all major diurnal and semidiurnal tidal constituents with a spatial resolution of 0.75¿¿0.75¿ have been estimated using the first 1.5 years of ERS 1 and TOPEX/POSEIDON altimetry. The ocean tide model was derived from the combined use of ERS 1 and TOPEX/POSEIDON data by using a modified orthotide formulation that simultaneously solves for all diurnal and semidiurnal constituents as well as the annual signal. An additional adjustment of the solar semidiurnal harmonic of the gravitational potential was applied in order to account for radiational forcing, particularly in the S2 constituent. TOPEX/POSEIDON provides excellent ocean tide estimates in the open ocean. However, especially in coastal regions, the track spacing of TOPEX/POSEIDON (315 km at the equator) is too coarse to determine large parts of the ocean tide signal. For these regions the inclusion of data from the ERS 1 35-day repeat mission provides a valuable supplement, as the ERS 1 satellite has a track spacing which is around 3.6 times better than that of the TOPEX/POSEIDON satellite. The combined ERS 1 and TOPEX/POSEIDON ocean tide solution exhibits distinct sectoral geographical pattern of highs and lows when compared with the Cartwright and Ray (1990, 1991) ocean tide model.

This indicates the presence of small but fundamental orbit errors present in the Cartwright and Ray ocean tide solution. Compared with a new set of 104 tide gauge readings compiled by Le Provost, the RMS differences of the combined ERS 1 and TOPEX solution are 2.51, 1.67, 1.58, and 1.13 cm for the M2, S2, K1, and O1 constituents, respectively. The increased spatial resolution of the combined ERS 1 and TOPEX model as compared to a TOPEX-alone model is seen to reduce RMS differences from 37 to 22 cm for the M2 constituent, when compared to a selection of 90 pelagic and coastal tide gauges in the northwest European shelf region. ¿ American Geophysical Union 1995