Review

Turbulent dissipation from AMAZOMIX off the Amazon shelf along internal tides paths 

Kouogang et al.

Summary

This manuscript describes measurements of currents, hydrography, and turbulence at numerous sites  near the Amazon outflow, where internal tides are also generated at the slope.  The dataset appears suited to study the questions posed by the authors.  Do internal tides affect the mixing?  Does the mixing affect nutrient fluxes?  

However, I have some concerns about the analysis listed below.  Since many of the details of the analysis are not presented, it is difficult for me to evaluate what has been done.   Some of the more prominent ones include: tidal decomposition of the currents, use/decomposition of shear, description/applicability of the finescale parameterizations used, and shaded color plots of SADCP velocity that include transits and times on station.

There are also presentation/stylistic problems that make the manuscript very difficult to read. Many of them are major, but also getting down to such points as the labelling of figures and sections.   Some examples are listed below.  It is well beyond the scope of a review to point all of these out.  Furthermore, the cumulative difficulty of numerous minor things (let alone the more major ones) turns into a major distraction to the detriment of the hard work done and the science. The senior authors must make considerably more effort in this regard.

Major comments

1. SI units should be written W kg^-1 not W.kg^-1 and 90 km not 90km and so on.

2. 50.4% on line 29.  Please state the error and adjust the significant figures accordingly.

3.  line 34 - A “guide for the mixing parameterization in future numerical simulation” is mentioned.  What is the guidance? 

4.  Section 2.2- some of the microstructure processing steps should be shown in either the main text, appendix, or supplementary material.  In particular, I would like to see some examples of the spectral fits.

5.  Semidiurnal currents are obtained by removing the mean current over a tidal period from the baroclinic velocities.  This will include other frequencies.  In fact Figure 4 has everything: 6 hrs, 12 hrs, …  It would be better to make a least squares fit of sines and cosines to the various tidal periods. With short time series of 1 day, the frequency resolution is 1 cycle/day.  So you could try fitting diurnal and semidiurnal, but recognizing that they are not formally distinct.  Also the inertial period here is at least 5 days and so cannot be determined in your dataset.  This is a key limitation that should be pointed out.  At the equator every frequency just about is in the internal wave band.   How or even whether the currents should be separated in frequency seems to be a major question given the limitations of the data and the desire to examine dissipation. 

6.  Shear.  Based on the calculations for barotropic u, baroclinic u, and tidal u. I do not see why shear is different for baroclinic and tidal u.  Tidal u = baroclinic u - lowpassed u.  It seems like it should be almost the same as the shear in baroclinic u.  I’m confused.  It would be helpful to see more of the processing steps and their justifications for these steps.  Perhaps there is a problem with the short time series or large tidal isopycnal displacements.  I suggest plotting some profiles of u and shear for barotropic, baroclinic, and tidal u to make sure these calculations are correct.  Also you could try doing these calculations on an isopycnal since it looks like the velocity structure is being heaved up and down by the internal waves.  That could lead to artifacts in the calculations on depth surfaces.  For example in Figure 4a1, on sigma = 26, current is negative.  At 120 m, current switches sign multiple times.  As far as I can tell, there are no plots of shear.  Also I do not understand why only one component of the current is shown.  Using both components is needed to understand propagation of internal tides either vertically or horizontally (alongshelf vs cross-shelf).  Section 3.2.2 uses a lot of words that could be instead expressed concisely in a map with current vectors.  Different colours could indicate different depths.

7. Dissipation is estimated using parameterizations based on shear in the tidal current.  Actual dissipation depends on the total shear or strain.    

8.  Some more explanation of the applicability of MacKinnon-Gregg parameterization vs those based on Gregg (1989) is needed.  The former is related to the limited bandwidth of internal waves (in shallow water for example), while the latter is for a typical open ocean environment away from generation sites. How does this apply to your study sites?

9. Naming of the sites and transects.  Maybe these are what are used on the cruise but they should be rearranged in a logical order to help the reader.  It is completely confusing as it is. Consider another system, such as A B C D … for generation sites. If there’s a transect near site A, it’s transect A.  If there are 3 stations on transect A, they are sites A1 A2 A3 and counting higher offshore. Or labelled by isobath, A500, A1000, A2000… There may not be a generation site for each transect.  For example lines 265-267 read: “The highest baroclinic tidal current velocities  were observed (between 25-48 cm.s-1) at sites Aa and Ab along T1-T2. Whereas lower tidal velocities (< 25 cm.s-1) are found  in site F along T4 (e.g., at S20 and S21) compared to OUT-ITs stations (e.g., at S24).”  So I have to use the map.  This could read instead as: “The highest baroclinic tidal current velocities  were observed (between 25-48 cm.s-1) at sites D and E along transects D and E. Whereas lower tidal velocities (< 25 cm.s-1) are found  in site A along transect A (e.g., at A2 and A3) compared to OUT-ITs stations (e.g., at site B1).” 

10.  Mixing is invoked as the explanation for increased nutrients.  How about coastal upwelling?  Where is the euphotic zone?  

11.  Were adjustments to the shear parameterization based on the Gregg (2003), where it was shown that internal wave interactions may lead to little mixing at the equator?

Minor comments

1.  Line 107- While the bin size of the LADCP may be 8 m, its resolution will be about 50 m.  Examine where the vertical wavenumber spectra fall off

2. I would recommend a native english speaker help with some of the grammar, if possible.  It’s ok as is but could be much better.  It will be less distracting for the reader.  Then the reader will pay more attention to the science.

3.  In sections 2.1 and 2.2, there is some jumping around from instrument to instrument.  Please collect into sections for each instrument.

4.  MacKinnon and Gregg (2003) provided validation in their paper.  There needs to be some explanation of the advantages/conditions for this parameterization over those following from Gregg 1989.  Gregg (2003) describe how latitude affects these mixing parameterizations.  I saw no mention of this.

5.  It is unclear how mixing layer depth is chosen.  There is a statement about choosing a minimum but the profiles show a lot of variance.  

6.  It is stated the mixing layer depth is always less than the mixed layer depth.  This seems like it cannot be true by definition- you need mixing to make a mixed layer.  Maybe the mixed layer depth criterion is too small.  How about 0.1 kg m^-3?

7. Figure caption does not correspond to what is plotted.  Shear is missing?  Figure labelling: how about (a), (b), …  instead 4.a.1?  There are also faint grey lines around the figures.  All the dots in the shear vs N figures are hard to see.  Consider binning the results.  

8.  There are lots of seemingly correct facts and figures in the text but I am unclear what is the point of them.  Take section 3.2.2, for example.  Currents are going this way and that at various depths and locations.  True.  I would suggest a format for each paragraph as follows. “In this paragraph, we examine the flow patterns over the slope to show something [topic sentence].  Here are the relevant facts and skip ones not immediately related to the topic sentence.  In summary, we have shown flow is in this direction here which is important for something.  Something is described in the next paragraph.”

Comments by line number

19 - twelve hours is the semidiurnal period but not the M₂ period = 12.42 hrs

129 - fft_length, etc.  Perhaps you could just give these a symbol if they come up more than once, e.g., L or n.  Otherwise skip the variable names in a manuscript.

136 - what is the high pass? and the low pass?

150 - mixing layer depth

170 - I don’t understand.  Please rephrase

172  - what is H?

193 - references? Also strong internal tides propagating upward and impinging on the thermocline make ISW.  No need for bottom reflections.

200- You have cross-shore measurements of N.  Are those not enough to make a horizontally varying N?

207- there has to be some minimal description of amazon36 here even if it is referenced

210- sensitivity not sensibility test, although the latter could also be used

263- 3-5 not 03-05

267-272 - Shear obviously varies with N because if shear were bigger than it would have lower Ri and be unstable.  It would be more instructive to consider Ri or reduced shear = S² - 4N²

You could use some mean N or interpolated N and see what it looks like.  If the text talks a lot about shear, it would be helpful to plot shear.  In fact the baroclinic vs tidal shear is discussed butthere are no plots.

Fig 4- caption and figure all on same page please.  Add tick marks to the x axis to show when your CTD/VMP casts took place.  Also this figure seems to include transits and time on station.  In Fig 4b1 there are some strange looking changes in the current.  Perhaps it would be better to do your analysis station by station.  Or are there artifacts of the SADCP processing going from on station to transit?

301- what sort of of mean? time, depth, etc

312-314- More explanation is needed here.

Fig 5- the tidal rays look to have shallower slope than the topography.  It would be helpful to plot ray slope vs topographic slope to verify.  The dissipation seems unrelated to the ray paths as far as I can tell.  Perhaps it is not 2D.  There appear to be 2 sources at least, which can constructively interfere.  Rays may propagate at an angle to the coast.

384 - “for the first time”  How sure are you of this?  It would be conservative to add: to the best of our knowledge

Review of “Turbulent dissipation from AMAZOMIX off the Amazon shelf along internal tides paths”, by Kouogang et al. Submitted to EGUsphere.

By Kouogang et al.

The paper reports oceanographic observations in the tropical Atlantic ocean within the framework of project AMAZOMIX, which aims to quantify mixing, identify the associated processes, and investigate their impact on nutrient fluxes off the Amazon shelf. The observations were collected during a field campaign in August-October 2021, on board the RV ANTEA with a dedicated satellite coverage (reported elsewhere). The region is well known for large amplitude internal tides and Internal Solitary Waves (ISWs), that were first measured and reported in Brandt et al., 2002, with wave amplitudes exceeding 100 m. Those large amplitude ISWs are clearly seen in Synthetic Aperture Radar (SAR) satellite images, as illustrated in the author’s Figure 1. Clearly, the cruise measurements were carefully planned in advance based on the knowledge provided by the SAR image archive. The most relevant aspect reported in the paper is the relative increase of mixing found about 225 km away from two distinct generation sites of internal tidal waves at the shelf break. The site where the relative mixing increases coincides with the surfacing of Internal Tidal (IT) rays emanating from different sites at the shelf break. But it also coincides with the appearence of large amplitude ISWs seen in the satellite record. While the constructive interference of IT rays has been suggested, and indeed is consistent with ocean colour images, in the central region of the Bay of Biscay (Western Europe), this aspect has not been documented elsewhere until now. It is therefore remarkable the association made by the authors between enhanced mixing and IT interference. However, a more comprehensive comment should be made concerning the presence of ISWs near Station S14, and its possible relation with the measured increased mixing.

The authors calculate the turbulent kinetic energy (TKE) dissipation rates and vertical diffusivities making use of microstructure and hydrography data collected at a series of stations and transects in a vast study region, being able to compare these values within internal tide pathways, and away from those paths. The observations also reveal a great deal of mixing at the shelf break, as expected. Baroclinic shear currents were calculated from ADCP data collected between stations and transects. It was then possible to estimate the ratio contribution to mixing between IT and Baroclinic Currents (BC). Interestingly, ITs dominate the ratio for IT/BC near the generation sites of ITs at the shelf break (65%/35%) and also at Station S14, 225 km away from the shelf break, admittedly where the IT beams coming from different sites may interfere (60%/40%). Although this fact is suggestive of the IT rays interference being capable of impacting mixing well away from the shelf break, the authors do not attempt comparison with the impact ISWs alone would have, on mixing.

The manuscript is reasonably written (with some minor pitfalls indicated below), and the data analysis certainly merits publication at EGUsphere. The hypotheses tested in the paper are well explained and presented, representing a milestone in the oceanography of the study region. The paper also includes an extensive appendix with auxiliary results. This reviewer recomends publication after some minor corrections (please see list below).

Minor comments:

Introduction:

Line 46 – 47: “They can dissipate where the energy beam reflects at the bottom, at the surface ora t the thermocline levels…”; The authors start describing bottom reflection of IT beams etc. without first explaining how those beams form and propagate in the vertical. A reader may like to read first how these beams are formed and why they propagate following rays that are determined by the well known formula that gives the angle to the horizontal. There should be a more comprehensive explanation about tidal beams before this point in the text. It may be worth refering to the work of Theo Gerkema (Gerkema, 2001 and Gerkema and Zimmerman, 2008).

Line 53: The reference Muacho et al, 2014 should be refered earlier, together with M’Hamdi et al., 2024, in preparation)

Line 62: Please include reference Brandt et al., 2002 in JPO about ISW in situ observations.

Line 69: “…no direct measurements of dissipation rates have been conducted.”. Do you mean before the presente paper?

Methods:

Line 170: “…time series with LADCP profiles glued below ~500m…”. later, in line 209 (page 8), the word “stitched” is used with a similar meaning? Could the authors use jargon more carefully, and consistently, please.

Line 174: “… with overbar the average…”. A word seems to be missing between “overbar” and “the average”.

Line 183: “… for M2 semi-diurnal baroclinic energy)…”. The parenthesis after the word “energy” appears to be unnecessary!

Line 187: “Previous study of Huang et al. (2019) shown…”. Please revise gramar.

Line 188: “… between 15-123 m in Ocean Atlantic”. Do you mean “Atlantic Ocean”?

Line 209: “ and stitched…”. Please be consistent with jargon language.

Results:

Line 218: “… . They are thicker along the IN-ITs…”. If you are referring to the “step-like” structures from the previous line, consider using another word for “thicker”, e.g. “larger”.

In Figure 3 the choice of colours for dissipation rates (epsilon) is somewhat difficult to grasp. The colours “red” and “magenta” are sometimes confusing, particularly in the vertical profiles in Figs. 3b to 3e. Could the “red” colour be changed to a “light Orange”, or similar?

Line 236: “Now, ….”; Unnecessary use of word “Now”.

Line 246: “….are found almost anywhere at S14…”; is there a better word than “anywhere” to be used in this sentence. Perhaps the words “at any depth” would be more appropriate?

Line 247: “…shelf stations in the ITs regions, S3 and S5, ….”  ; the vertical profile of station S3 does not appear in Fig. 3. (it appears only in Fig. A1.c). Please refer to where it appears.

Line 271: “… dissipation rates (epsilon) already presented in section 3.1.2, IS….”;  do you mean “ARE also reported in Fig. 3”?

In Figure 4b, the baroclinic tidal currents at S11 are less clear. Do you have any reason to suggest for this fact?

Line 287: “Now, …”; Unnecessary use of word “Now”.

Line 331: “IT ray paths were observed reflecting at the surface….”; The words “were observed” are not the best choice here, as you are talking about a model calculation; This reviewer suggests the use of the words “are expected to reflect” or “were predicted to reflect”.

Line 335: “flow becomes unstable beyond ….”; use the word “below” or beneath” instead of beyond?

Line 336: “Large epsilon are encountered where IT rays presumably interfere between them or….”

This reviewer does not see “IT ray path interference, since the different ray tracing computations (in red and blue colours) are a matter of different stratification choices, and hence you either have a “blue” ray or a “red” ray.

Line 338-339: “Some large epsilon are observed where IT rays radiated at the surface (e.g. at S12, S14 and S20)…..”; Large epsilon values are a common feature to all turbulence profiles of dissipation rates near the surface. It looks more like the effect of wind mixing, rather than surfacing and reflection of IT rays.

Line 353: “The next steps of our study IS…”; either use singular (“step”) or use correct tense (“ARE”).

Discussion and Conclusion.

Line 398: “…., both large (up to 60 m length) isopycnal displacements…”; perhaps the wording “…, both large (up to 60 m) AMPLITUDE isopycnal displacements…” is more apropriate?

Line 426: “… It showed bottom FICTION”; I think you mean “bottom friction”?

In Figure 7 the legend for NBC is given, but it is not drawn in Figure 7.

Line 446: “This relative diminution….”.; can you use another word for “diminution” that is more common in literature? Perhaps the word “decrease” suits what the authors want to say?

Line 471: “The really new aspect raised in this study…”; Please avoid writing “The really new”. Something is either new or “not new”.

Line 483: “ITs disintegrate into a package of ISWs events…”; Please use the word “packet” for ISWs because this is what became acknowledged in the literatute for groups of internal wave trains.

Line 510: “…, ITs act as an important supplier of nutrient into….”; please use plural for “nutrients”.

This paper reports on microstructure measurements off the Amazon shelf. This is an interesting area with a combination of various dynamical processes, internal tides, low frequency circulation and amazonian water lenses. Consistently with this dynamics contrasted dissipation rates are observed with the highest values at generation sites, and along internal tide pathways and the lowest values in no-tidal areas. The relative contribution to dissipation of the mean baroclinic current (North Brazil current) compared to that of internal tide was estimated : the contribution of IT shear was found larger than BC shear near generation sites, equal along IT pathways. In addition turbulent diffusive nutrient fluxes were computed : large values were obtained.

I think there is interesting material for publication but that part of the analysis must be revisited to provide convincing results and some irrelevances to be removed (details in the following). Some of the figures are overloaded and not easily readable, having many figures in the appendix does not facilitate a fluent reading. Part of the sections would need a careful polishing.

The main result to highlight is the spatial contrast of dissipation rates and  give insights on the origin of these variations.

I list in the following my main general comments:

-More information on the background state should be given: bathymetry, sea surface salinity (Amazone plume), as well as the mean current and for instance information on generation sites for internal tides (reorganization of Figure 1 which is not easy to read, subplot (b) does not seem necessary)

-Baroclinic currents and energy : it is unclear why a parameterization of e is referred to as baroclinic total energy.  The MG  parameterization does not provide any relevant information (the correlation with epsilon is not clear and it is used as a proxy to evaluate the contribution of tidal shear and low frequency shear which I find questionable)

-Ray tracing calculation is applied but horizontal density gradients are neglected : this assumption is surprising owing to the major influence of the mean baroclinic flow. Also in Figure 5 it would be more clear regarding the IT ray paths to consider the full water column

-Figure 3 : it is difficult to have a view on the evolution along the transects, why not show e profiles along the transect with density superimposed, some large values at the end of the e profiles would need to be checked (S12 and S14 in Fig. 3. and 3.e as well as S2 figA.1.c)

-the relationship between step-like structures and strong internal tides is not convincing as it is presented

-Figure 4 : Emphasis is made on shear instability, this is quite convincing at S10 but not at S14, this should be interesting to comment on

-Competitive processes to generate mixing: the hypothesis of shear-driven dissipation is followed with the aim to discriminate between the low frequency shear contribution and the IT shear one. I find this subsection difficult to follow, and I don’t understand why the MG parameterization is introduced to this aim.

-Figure 5 : the ray tracing approach should be revisited with taking into account the low frequency current, mean along shore current displays some spatial variations that may modify the ray structure ;

-Nutrients fluxes : profiles of Kz and nutrients fluxes are displayed. I don't see the point  in giving values of nutrients fluxes without showing the concentration profiles and introduce the motivations and the issues in the studied area.

-Discussion and conclusion : needs to be re-written and with convincing results for most part of it, one example : l442 « shear instabilities stronger >10-4s-2 », IT shear : high tidal modes are referred to but not shown etc

As a conclusion, I think this manuscript is no ready for publication and needs significant work to produce convincing results. I think the most efficient way to proceed is a new submission.