| 21 Jan 2026
Modulation of Internal Tides Properties off the Vitória–Trindade Ridge during Contrasted Seasons from Altimetry and a Regional Ocean Model
Abstract. The incoherent fraction of internal tides, generated through interactions with mesoscale eddies and other transient oceanic features, remains poorly understood and challenging to predict. This knowledge gap limits our ability to accurately represent energy transfers and mixing processes in the ocean induced by these waves. The Vitória–Trindade Ridge, located off the Brazilian shelf, provides a particularly relevant natural laboratory to investigate these processes, as it constitutes a hotspot for internal tides generation embedded in a region characterized by strong mesoscale activity and vigorous eddy fluxes. To assess how seasonal stratification and mesoscale variability modulate internal tide properties, we combined two complementary approaches: internal tide signals were extracted from a 27-year satellite altimetry record and compared with a high-resolution (1/36°) regional simulation performed with the ocean model NEMO v4.0.2. This joint analysis allowed for a consistent characterization of the generation, propagation, and dissipation of internal tides under two contrasted oceanic regimes. Austral winter (defined here from May to October) is marked by a deep pycnocline, whereas austral summer (defined here from November to April) is associated with a shallower, sharper seasonal pycnocline. Both the model and observations depict six intense, in-phase reflection beams propagating southward from the ridge. The first two beams of reflection are associated with a wavelength of 100 km approximately corresponding to the mode 1 of propagation, while the beams further from the ridge are separated by about 50 km only, which likely corresponds to the second mode of propagation. Quantification from the model show that internal tide generation rates are 5 to 15 % higher in summer than in winter. Dissipation occurs predominantly in the vicinity of the ridge (45 %) but also extends offshore (40 %), reaching beyond 2 to 3 mode-1 wavelengths. In the open ocean, dissipation rates are up to 40 % higher during winter than in summer. In the model, these seasonal differences result in stronger baroclinic fluxes that propagate farther south during summer, whereas winter fluxes are more rapidly dissipated. Altimetric observations further confirm pronounced seasonal variations in both wavelength and amplitude, in particular for mode-2 internal tides. In addition, a representative case of interaction between internal tides and a mesoscale eddy is documented under summer conditions, showing deviation and diffraction of the baroclinic flux as it encounters the eddy. This study demonstrates that mesoscale variability and seasonal stratification act jointly to modulate the coherence and energy pathways of internal tides. These findings are essential for improving predictions of the incoherent tide and for guiding the interpretation of recent high-resolution altimetric observations.
This paper looks at internal tides in the Vitoria-Trindade region using both satellite altimetry and a high-resolution ocean model. The approach is straightforward and the results seem reasonable to me. I recommend some minor changes before publication.
I was especially intrigued by the propagation study that examined how a large mesoscale eddy impacted internal tide fluxes -- Figures 12 and 13. The authors emphasized the change in propagation direction (clear in the cartoon of Fig 13, less clear to me in the actual data of Fig 12). But even more striking to me was the nearly complete disappearance in panel (b) [Feb 24] of the energy fluxes south of the eddy. This seems more significant than the small perturbations to direction. But it also raises a number of questions. Most of this paper focuses on M2 alone. Is Fig 12 showing just M2, or is it a general "semidiurnal" flux? If just M2, then I'd like more details about how M2 could be isolated from all the other constituents in their model. If all tides are shown, is the lack of energy in panel (b) just the spring-neap cycle, and thus of minor significance? Perhaps the authors could fill in more details behind these diagrams. (They could also improve the aesthetics of Figure 12, since some fonts are too tiny to see. Maybe some tiny text is not needed?)
I also had some concerns with Fig 5 that compared model and altimetry. The altimetry has mostly smaller amplitudes. I agree that this could stem from the altimetry being an average over 20+ years. But the altimetry is also somewhat noisy and one naturally wonders about error bars. It would be good to see error estimates here, which should be easily derivable from their altimeter analysis.
Some other minor points (with line numbers):
The instructions to OS say the Abstract should be "short, clear, concise." It is written clearly, but it does seem somewhat long.
32 - main -> main mechanism? missing word?
38 - the reference to Carter et al. does not seem appropriate, since that work was focused on the Hawaiian Ridge, not the global energetics.
43 - "to a day" - It can be many days, as shown by some waves crossing ocean basins.
88 - The altimeter data extend past 2020, even if not all of it is used here.
90 - "without aliasing" is not correct. The tidal signals are still aliased, no matter how long the time series.
97 - "contribution of baroclinic tides" is part of the noise?? Perhaps the authors mean "incoherent" baroclininc tides.
99 - Arbic reference not needed for the frequency of M2, surely.
101 - I think Step 2 is actually a part of Step 3 and can be removed. It is just a part of any filtering step. There are really only two important steps here.
113 - I did not understand what is "in adequacy"
Fig 5 panels for barotropic tide. The phase plot is not informative as it just shows big ±180° jumps. Changing the y-axis to some other range would show the data better.
136 "boundaries" might be better than "frontiers"
140 - Was FES2014 model used for forcing only on the open boundaries?
144 - I think it would be useful to say how barotropic and baroclinic signals were separated. This can be sometimes a source of confusion, so it is good to state what was done.
Fig 3, right panels: Are these profiles a MEAN over the whole region? And over what time range?
Fig 3, left panels: Is this the "barotropic" SSH for the model, or the "full" SSH? I suspect it is the full external+internal tide, which would explain the phase jitters.
Also, what is contour interval for phase lines?
163 - why "global"?
Fig 4. Box 1 is mostly red, boxes 3-6 mostly blue. Why are all boxes negative in the bar chart? It seems Box 1 should be positive.
Fig 5 and corresponding text, for wavelength determined from altimetry. The peak in the wavenumber spectrum is used. Was this corrected for the orientation of the T/P track? The track crosses the main beams at an angle. (This same issue arises in the Conclusions, too -- line 336.)
Fig 10 - aside from a small change in wavelength, summer versus winter, more pronounced is the differences in the energy level. Summer peak is much reduced in magnitude.
Fig 11 - seems rather cramped and hard to see. Would a log scale for dissipation be better?
474 - The reference is wrong. This cites it as a OS Discussion paper. The actual publication was in OS in 2022.