18.6-year lunar nodal tides from altimeter data

Abstract

Harmonic analyses of 16-year long time series of collocated TOPEX, Poseidon and Jason-1 altimeter data were carried out in the Pacific and western Atlantic Oceans and their marginal seas. These time series are sufficiently long to adequately separate the 18.6-year nodal satellites Q 1 n , O 1 n , K 1 n , M 2 n and K 2 n from their parent constituents Q 1 , O 1 , K 1 , M 2 and K 2 . Editing criteria were used to eliminate results in areas where these satellites are weak (i.e., smaller than their formal error estimates), or where they are strongly affected by aliased low-frequency signals (e.g. in the Kuroshio, in the Gulfstream and in their extension regions). As expected from tidal theory, the phases of the altimetry-derived nodal satellites agree reasonably well with the phases of their parents. However, due to their relatively small amplitudes and the remaining influence from low-frequency aliased signals, the altimeter observed amplitude ratios between the nodal satellites and their parent constituents tend to exceed the values predicted by the theory. Examination of diurnal and semidiurnal nodal amplitudes in select coastal areas and marginal seas around the Pacific and the western Atlantic Ocean allowed the assignment of a nodal character to regions, which were each classified as nodal diurnal, nodal semidiurnal, or nodal mixed, based on the nodal amplitudes in each band. While the areas with predominant diurnal tides are all nodal diurnal, the small nodal ratio of 0.037 for M 2 n resulted in some regions with strong M 2 tides being classified as nodal diurnal or nodal mixed. The amplitude ratio between K 2 n and K 2 is 0.30, making the K 2 n amplitudes sometimes comparable to those of M 2 n . However, this effect was not sufficient to make all the areas with dominant M 2 to be dominant nodal semidiurnal. The observed amplitudes of the 18.6-year nodal constituent M n are relatively small, 1.5–3.5 cm. These values significantly exceed its theoretical amplitudes, which are less than 1 cm almost everywhere. The analysed signals at M n frequency are therefore of mostly non-tidal origin, part of the broad-band decadal ocean variability.