Abstract
Almost three decades of bottom pressure recorder (BPR) measurements at the Drake Passage, and 31 years of hourly tide gauge data from the Vernadsky Research Base on the Antarctic Peninsula, have been used to investigate the temporal and spatial variations in this region of the three main long-period tides Mf, Mm and Mt (in order of decreasing amplitude, with periods of a fortnight, a month and one-third of a month, respectively). The amplitudes of Mf and Mt, and the phase lags for all three constituents, vary over the nodal cycle (18.61 years) in essentially the same way as in the equilibrium tide, so confirming the validity of Doodson's “nodal factors” for these constituents. The amplitude of Mm is found to be essentially constant, and so inconsistent at the 3σ level from the ±13 % (or ∼±0.15 mbar) anticipated variation over the nodal cycle, which can probably be explained by energetic non-tidal variability in the records at monthly timescales and longer. The north–south differences in amplitude for all three constituents are consistent with those in a modern ocean tide model (FES2014), as are those in phase lag for Mf and Mt, while the phase difference for Mm is smaller than in the model. BPR measurements are shown to be considerably superior to coastal tide gauge data in such studies, due to the larger proportion of non-tidal variability in the latter. However, correction of the tide gauge records for non-tidal variability results in the uncertainties in nodal parameters being reduced by a factor of 2 (for Mf at least) to a magnitude comparable (approximately twice) to those obtained from the BPR data.