Abstract. The Amazon shelf break is a key region for internal tide (IT) generation. It also shows a large seasonal variation in circulation and associated stratification. This study, based on a high-resolution model (1/36∘) explicitly forced by tide, aims to better characterize how the ITs vary between two contrasted seasons. During the season from March to July (MAMJJ) the currents and mesoscale eddies are weak while the pycnocline is shallower and stronger. From August to December (ASOND) mean currents and mesoscale eddies are strong, and the pycnocline is deeper and weaker than in MAMJJ. For both seasons, semi-diurnal M2 ITs are generated on the shelf break mainly between the 100 and 1000 m isobath in the model. South of 2∘ N, the conversion from barotropic to baroclinic tide is more efficient in MAMJJ than in ASOND. Local dissipation of the coherent M2 at the generation sites is higher in MAMJJ (30 %) than in ASOND (22 %), because higher modes are favourably generated (modes 2 and 3), making the internal wave packet more dispersive. The remaining fraction (70 %–80 %) propagates away from the generation sites and mainly dissipates locally every ∼ 100 km, which corresponds to the mode 1 reflection beams. About 13 %, 30 %, and 40 % of the M2 coherent IT dissipates at the first, second, and third beams. M2 coherent baroclinic flux propagates more northward during MAMJJ while it seems to be blocked at 6∘ N during ASOND. There is no intensified dissipation of the coherent M2 that could explain the disappearance of the coherent flux. In fact, the flux at this location becomes more incoherent because of strong interaction with the currents. This has been shown in the paper using 25 h mean snapshots of the baroclinic flux that shows branching and stronger eastward deviation of the IT when interacting with mesoscale eddies and stratification during ASOND. Finally, we evaluated sea surface height (SSH) frequency and wavenumber spectra for subtidal (f<1/28 h−1), tidal (1/28 h−1 < f<1/11 h−1), and supertidal (f>1/11 h−1) frequencies. Tidal frequencies explain most of the SSH variability for wavelengths between 250 and 70 km. Below 70 km, the SSH is mainly incoherent and supertidal. The length scale at which the SSH becomes dominated by unbalanced (non-geostrophic) IT was estimated to be around 250 km. Our results highlight the complexity of correctly predicting IT SSH in order to better observe mesoscale and sub-mesoscale eddies from existing and upcoming altimetric missions, notably the Surface Water Ocean Topography (SWOT) mission.

Abstract. The forthcoming SWOT altimetric missions aim to resolve the mesoscale with an unprecedented spatial resolution and swath. However, high-frequency processes, such as tides, are undersampled in time and aliased onto lower frequencies, so they need to be corrected properly. Unlike barotropic tides, internal tides (ITs) are not completely stationary and have significant temporal variability due to their interactions with the ocean circulation and the stratification variability. Stratification changes impact both the generation and the propagation of ITs. The present study proposes a methodology to quantify the impacts of background stratification using a clustering method for the classification of a broad range of stratification and idealized modeling of ITs in the frequency domain. The methodology is successfully tested in the western equatorial Atlantic and in the Bay of Biscay. For the western equatorial Atlantic, a single pycnocline is observed and only the two first vertical modes of ITs have significant amplitudes. With no variation in the stratification intensity, the variation in the depth of this single pycnocline linearly impacts the elevation amplitude, energy fluxes and surface wavelength of the two modes. In the Bay of Biscay, there is a permanent deep pycnocline and secondary seasonal pycnoclines near the surface. No proxy have been found to describe the changes in ITs, so a seasonal climatology is explored. The seasonality of the stratification strongly affects the elevation amplitudes as well as the energy fluxes of modes 1, 2 and 3. The distribution of the modes vary with the background stratification, changing the horizontal scales of the ITs.

Abstract. The Amazon shelf break is a key region for internal tides (IT) generation. The region also shows a large seasonal variation of circulation and associated stratification. The objective of this study is to document how these variations will impact IT generation and propagation properties. A high-resolution regional model (1/36° horizontal resolution), explicitly resolving IT is analyzed to investigate their interactions with the background circulation and stratification, over two seasons: first MAMJJ (March to July), with weaker mesoscale currents, shallower and stronger pycnocline, and second ASOND (August to December) with stronger mesoscale currents, deeper and weaker pycnocline. IT are generated on the shelf break between the 100 and 1800 m isobaths, with a maximum on average at about 10 km offshore. South of 2° N, the conversion from barotropic to baroclinic tide is more efficient in MAMJJ than in ASOND. At the eight main IT generations sites, the local dissipation is higher in MAMJJ (30 %) than in ASOND (22 %). The remaining fraction propagates away from the generation sites and mainly dissipates locally every 90–120 km. The remote dissipation increases slightly during ASOND and the coherent M2 fluxes seem blocked between 4°–6° N west of 47° W. Further analysis of 25 hours mean snapshots of the baroclinic flux shows deviation and branching of the IT when interacting with strong mesoscale and stratification. We evaluated sea surface height (SSH) frequency and wavenumber spectra for subtidal (f < 1/28h−1), tidal (1/28h−1 < f < 1/11h−1) and super tidal (f > 1/11h−1) frequencies. Tidal frequencies explain most of the SSH variability for wavelengths between 300 km and 70 km. Below 70 km, the SSH is mainly incoherent and supertidal. The length scale at which the SSH becomes dominated by unbalanced IT was estimated to be around 250 km. Our results highlight the complexity of correctly predicting IT SSH in order to better observe mesoscale and submesoscale from existing and upcoming altmetrics missions, notably the Surface Water Ocean Topography (SWOT) mission.