Review

Internal tides vertical structure and steric sea surface height signature south of New Caledonia revealed by glider observations

Arne Bendinger et al.

Summary

This manuscript describes glider observations of internal tides and compares them with numerical models.  Generally the comparison is good and discrepancies are explained via mesoscale eddy refraction. There are two major points that need to be addressed on the CTD and near inertial waves, but otherwise the manuscript is generally clear and to the point.  A nice feature is that observations from a glider are compared to the model which is selectively sampled as the glider did the ocean. 

Major comments

1. The glider CTD inlet is located on top and so upward profiles experience clean flow.  Downward profiles experience distorted flow.  Some justification/statistics are needed to show downward profiles are not contaminated (i.e., both raw and binned data, T-S plots of upward vs downward profiles, rms difference vs p, evidence of hysteresis, and so on).  

For future consideration, even if the downward profiles are ok, their advantage is limited.  There is no gain in resolution at the turning points; there is a factor of 2 increase in resolution at the mid point; and everywhere else it’s in between.  So at best the noise is reduced by a factor of √2. Then there is the energy use for marginal gain in my opinion, which could be used instead to extend the sampling with clean, upward profiles.

2.  (a) From Fig 5 it looks like you should also fit at the inertial period ~ 29 hrs, even though this will overlap with the diurnal signal though for the 3-day window because the frequency resolution is 1/(3 days).  This raises two points. (b)   In line 316, this is linked to submesoscale dynamics, which appears to ignore the inertial signal, which is also expected to be larger also in the upper ocean.  Since the near-inertial displacements are similar in size to the semidiurnal displacements, the near-inertial signal will have much more kinetic energy and may then be much more energetic internal wave signal than the semi diurnal signal. (c) Any comparison with the model should be a comparison of semi diurnal and not individual constituents, ie M2, S2, which it seems to me is done most/all of the time.  

Minor comment

3.  It would be worth mentioning how the internal tides near New Caledonia show up in previous global models and observations by Zhao, Niwa and Hibiya, Ray and Cartwright, etc. and how your models and observations compare.  

4. For interference patterns there is a paper by Rainville et al about Hawaii.   

Comments by line

44 - The smearing of temporal variability into spatial variability by slowly moving gliders has been examined by Rudnick & Cole (2011).  You can extract some wavelengths based on your dominant diurnal and semidiurnal periods.  https://doi.org/10.5194/egusphere-2024-247

48 - Could lead with the advantages of using gliders for studying internal waves and then follow with limitations (line 37).

58 - This paragraph could benefit from the addition of some of the original references related to Hawaii, theory, altimetry, interference, and coherence.

127 - Some more explanation is needed of how forcing by tidal potential and at the boundaries works. Usually only one type of forcing is used because the 2 methods may provide conflicting information, which would produce spurious waves at the boundaries and perhaps elsewhere. Or perhaps only at certain times when the mismatches are more apparent.

148 - Absolute geostrophic currents referenced to glider depth mean velocity can be calculated too.  

168 - Also smearing of the semidiurnal signal goes into scales < 30 km (Rudnick & Cole, 2011) and so averaging should be done over larger scales.

265 - This isn’t a big deal because it’s the time differences (ie, ~ 6 hrs) between profiles that is the main thing.

275 - It would be helpful to see the full depth structure of modes 1-2.

Fig 7- no advantage to using grey shading

Fig 11 - there are some spurious signals in the model or interpolating it onto the glider path especially near 15/8, 8/10 and 16/10.  

Fig 11 - explain why the upper 100 m is missing.  

310 - semi diurnal displacement variance. See comment 2 about total internal wave signal.  

405- Do these variance include the upper 100 m which are missing in Fig 11?  If not, then this is where the residual/possibly near internal wave variance is largest 

Review

Internal tides vertical structure and steric sea surface height signature south of New Caledonia revealed by glider observations

Summary

This paper describes an effort to investigate internal tide structure using glider observations. I found this paper to be interesting and not hard to follow.

I have no major comments, only several minor comments, and they are listed below

Line 8-9 – eddy-internal tide interactions are certainly one source of discrepancy. I don’t discount this mechanism, but there are other potential sources of error as well (such as topography and stratification). What about those?

Line 14 – “glider observations’ predominating coherent nature”. Why would this be the case? I assume there isn’t something intrinsic about the glider observations that would make them have a coherent nature. Is the observation record too short to observe much incoherence, was this a time period of low variability or some other reason?

Lines 44-46 – problems separating high and low frequency signals. Can these be separated enough to get insight into the coherent/incoherent question?

Line 121 - the numerical model used, has it been compared against other models and/or observations to show it is sufficiently accurate/realistic to aid in this study?

Line 127 – tidal forcing along boundary. It appears you’re forcing with FES along the boundary. This is barotropic forcing only. Mazloff et al (2020-JGR) talk about the importance of remote forcing for regional modeling of internal waves. Can you comment on the results of this paper and either it’s relevance or lack of relevance to this work?

Figure 7 – I find this gray scale plot hard to interpret. The other plots in the paper are color, was there a reason why gray scale was used? I may be missing something that this color bar choice was intended to help convey.

Line 255 – “have never been used the SSH signature” doesn’t read right. Did you intend to convey “have never been used to derive the SSH signature”?

Line 344 – There are other sources of potential disagreement as well, such as forcing, topography and stratification errors and unconstrained variability. Have you explored or considered those?

I had some additional thoughts, but those were enumerated by reviewer 1, so I'll omit including those.