Scientific interpretation of Fig.5 (spectral analysis, no-tide vs tide)

[stephanieleroux]

(general question:) Not clear to me how far you intend to go into scientific interpretations/comments of the diagnostics figures here in the github repo?

Example with Fig.5 on the spectral analysis page :

• i'd be curious to know what is the interpretation for the fact that the seasonal difference in the spectra seems to vanish in the simulation with tide compared to the one without tide in the Gulf Stream region. The observations seems to show a seasonal difference in the fine scales though.

• i would also be curious to know why the change in the fine scales of the spectra between no-tide and tide is much obvious in the Azores region than in the GS region. Is the Azores region a internal wave hotspot ?

So, do those questions fit in the github issue section here? Or should i just ask you guys in private?

(PS: In any case: really cool to see this eNATL60 repository and all the tools and diags explained and opened. I like the concept very much! )

[lesommer]

Glad you like the concept of this repos !

From my perspective, questions on the results of the comparisons with obs. should indeed be formulated as GitHub issues. The issues can then host discussions that may eventually lead to modifications of the README.md files (+other files if needed).

The idea is to use the functionality of GitHub to gradually build a shared understanding of the strengths and weaknesses of this simulations. But we want do do so through a process that is open and visible to all.

Thanks for your question which helps clarifying how this repos should be used !

[lesommer]

Regarding the seasonality of SSH wavenumber spectra in the model and in observations, I think the plots can be rationalized if we keep in mind that there are three distinct geophysical signals that vary seasonally at fine scales (two of them being explicitly represented in the model) :

• the intensity of balanced submesoscale turbulence is expected to vary seasonally because of the changes in surface stratification through mixed layer instability, it is expected to increase SSH variance in JFM vs JAS;
• the surface signature of internal tides is also expected to vary seasonally because of the changes in surface stratification, but the expected signal is opposite with stronger variance in JAS vs JFM;
• surface waves are also generally larger in JFM vs JAS because the winds are usually stronger in winter, this signal is not represented in the model but it induces a seasonality in the noise level of altimeter signal.

Re your questions :

• the weaker seasonality in GS region in the runs with tides is simply due to the fact that the seasonality of internal tides surface signature partly covers the signal associated with submesoscale balanced turbulence; the reason why altimeter signals exhibits a stronger seasonality is from my perspective due to the seasonality of the noise level;
• a major difference between GS vs AZO regions is actually the intensity of mesoscale turbulence; in GS the mesoscale signal is so large that internal tides are too weak to reverse the seasonality of SSH variance; the situation is different in AZO where mesoscale turbulence is weak enough for the seasonality of internal tides to take over and drive the seasonality of SSH variance at fine scale. Coming back to your question, the difference between tides vs no tides is weaker in the GS region because mesoscale turbulence is stronger there, so that tides don’t change much the signals.

We may think about including these explanations in the README.md if you think it is appropriate.

Thanks again for your questions !

[stephanieleroux]

ok, thanks for the explanations/comments!
No strong opinion on whether it is appropriate or not to add it in the readme. If it is clear for all the future Readers that they can browse the issue page to find scientific discussions/interpretation, then it might be enough, i guess?