General Comments

This manuscript presents a hydrodynamic dataset for the Gambier Islands (French Polynesia) in the Pacific Ocean, detailing waves, currents, temperature, and water levels at multiple locations around an atoll for 9 months in 2019-2020. The entire atoll seems well-instrumented, with instruments suitable to capture a wide range of hydrodynamic phenomena.

This article furthers ESSD’s goals of clearly presenting and documenting a high quality earth science dataset. The authors’ dataset is unique, capturing the hydrodynamics of a remote coral atoll in with relatively wide spatial coverage and fine temporal resolution. Such a dataset is not trivial to produce and will provide a valuable addition to the scientific literature. The dataset is also useful in itself for understanding atoll hydrodynamics, but has already apparently been used as a basis for model validation (Bruyere et al 2023b). I have made many comments below but I believe they are mostly minor in scope and in the spirit of helping a strong dataset and manuscript shine brighter.

As coral atolls are at high risk to the effects of climate change, such a dataset provides a welcome snapshot of a vulnerable location that may be valuable for researchers in other fields (e.g., ecology, coastal engineering) or studying other atolls. I can think of several colleagues who may be interested in using this dataset for their research on coral reef hydrodynamics (including myself), although it seems strange to filter out low-frequency wave data from this dataset if it was measured (see detailed comments below). As far as I can tell, the dataset here is complete and makes a coherent collection, and is sufficiently presented by this manuscript.

Data quality

The data is very clearly available in well-formatted netcdf files from the website. I downloaded several of the files to check the metadata and accessibility of the data, and it was very easy to find what I was looking for. I will actually share this with my students as an example of good practice for sharing research data.

However, I think that more attention to error/uncertainty estimates and processing procedures should be given in the manuscript (see detailed comments below). The current quality control section (Section 4.5) is only two sentences long and would benefit from more detail. There was not much discussion of errors or data cleaning/processing, beyond some mentions of unlinked Python datasets. 

Presentation quality

The article is not too long and is clearly written, covering most of the major questions that I had when I started reading. The figures are all generally clear and well-formatted. Overall the presentation quality is good.

As per ESSD guidelines, “The authors should point to suitable software or services for simple visualization and analysis, keeping in mind that neither the reviewer nor the casual "reader" will install or pay for it.”  Perhaps then Section 6 (Data Availability) should here indicate again that the data is stored as NetCDF files and that they can be readily analyzed and visualized using a wide range of freely available code/software packages, or something along those lines.

Specific Comments

• L149-151: General question: how were your instruments mounted? E.g., were they placed on frames, laid directly on the seabed, bolted to a reef, weighted down? If placed on a metal frame, were your compasses calibrated and did you account for interference from the metal frame (e.g. for ADCP current direction)?
• L167: Can you provide more detail on the processing (“processed with Python routines”)? Is the code available online? If so, that would be good to mention here and link.
• L178/184: How were your pressure/wave data processed? Can you provide more details?
• L182: Why did you use a constant atmospheric pressure to offset your wave gauges instead of measurements from the nearby weather station? What are the limitations of this assumption?
• L184: You only examine waves with periods 3-25 s, but infragravity (25-250 s) and very low-frequency (>250 s) waves are often very important on coral reef flats. Why did you make this choice? Is the lower-frequency data still available? This data would be extremely valuable for e.g. predicting flooding on atoll islands and validating existing theories about low frequency wave hydrodynamics on coral reefs. If you still have the raw data at these frequencies, I would strongly encourage you to add it to your dataset as I can think of quite a few people (myself included) who would find it useful.
• Figure 5: Is there any temporal lag in the two signals? They look very close, so it’s a bit hard to tell, but I think this is useful information.
• L182: How long were your bursts? (e.g. 20 mins every hour)?
• L209: See my earlier comments re: compass calibration if you used a metal frame.
• L221: Again, is this code available?
• L226: Please provide more detail on the data cleaning process.
• Section 7 (Conclusions): Something that I missed was a sentence or two describing how similar/different this location is to other sites (and therefore how can it be applied/useful for scientists/engineers at different locations besides the Gambier Islands)?  Perhaps also circle back to the other French Polynesian sites mentioned in L43-53?
• Relatedly, a relatively limited pool of references was used, so I think the manuscript would benefit from a few additional references to other similar atoll measurements from other corners of the literature, such as:
  • Rogers, J. S., Monismith, S. G., Koweek, D. A., Torres, W. I., & Dunbar, R. B. (2016). Thermodynamics and hydrodynamics in an atoll reef system and their influence on coral cover. Limnology and oceanography, 61(6), 2191-2206.
  • Grimaldi, C. M., Lowe, R. J., Benthuysen, J. A., Cuttler, M. V. W., Green, R. H., & Gilmour, J. P. (2023). Hydrodynamic and atmospheric drivers create distinct thermal environments within a coral reef atoll. Coral Reefs, 1-14.

Technical Corrections

• L27: Cite Bruyère et al (2023a) here just to make it clear that the biophysical model is in that paper and not in the current manuscript.
• L27: Missing word à “connectivity [OF] oyster larvae”
• L42: Missing something here re: flushing vs stagnation- what’s most important for oyster larvae?
• L52: Add some words to clarify? “Similar data collected on [OTHER FRENCH POLYNESIAN] atolls…”
• L53: Maybe add something here about human influences (or lack thereof) on the atoll? E.g. have the channels between islands been dredged or land reclaimed for runways etc? This sort of information could also be appropriate elsewhere in Ch. 2.
• L83: Missing word – “Mean and maximum [DEPTHS] are…”
• Figure 2: Indicate location of weather station (and ERA reanalysis point) on map with a dot and label?
• L100: Indicate what the vertical datum is? (e.g. measurements relative to mean sea level or some other local datum?)
• L100: I know it’s in Figure 3 implicitly (and later in L124-125), but since this is the first time it comes up, I think you should explicitly state here in the text which dates your data spans during Leg 1 and 2.
• L112: Clarify “…probably related to wave events.” Is it worth mentioning the difference here between local wind waves and remotely-generated swell?
• L118 (and Figure 4): What is the (cross-shore) position on the reef? E.g., were you measuring on the shallow reef flats or deeper on the fore reef or in the lagoon etc?
• Figure 4: Could you add a sentence or two to the text of the manuscript about how Andrefouet & Bionaz (2021) did the seabed classification? E.g. was it just based on depth or also on ecological parameters?
• Table 1: Great, very comprehensive!
• L224: RIP P01. I am impressed that you only lost one instrument during such an extensive campaign, well done!
• L227: What is the range beyond which something was considered “out of range”?
• L238: Capitalize “python” à “Python”; can you cite the T-Tide package?
• Figure 6: Could you indicate the side of the atoll next to each station in the legend? E.g., “O03 (SE); O04 (N)”, etc. I think this would help the reader with interpretation, especially with connecting the long-period swell vs local sea in panel (C) with your text in Section 5.2.
• Highlight that this data could also be useful for flood hazard estimation or validation of early warning systems (see Winter et al…)
• L324: the delta T described here is with respect to time for a single sensor across the full period (i.e. max dT/dt), rather than the max temperature difference between the surface and bottom during the period (i.e. max dT/dz), correct? Some clarification here would be helpful.
• L347: Given that you describe thermally stratified conditions in L314 and Figure 9, how appropriate is it to show depth-averaged flow here? Is there any shear in the velocity? Perhaps it is good to mention in ~L211-216 whether only the depth-averaged flow is included in the ADCP dataset or if you have provided all the bins.

Once these comments have been addressed, I think that this manuscript will make a welcome addition to ESSD and the coral reef hydrodynamics community.

Sincerely,

Stuart Pearson

Summary

The authors document the design, deployment, and curation of an array of oceanographic motioning instruments and the various data streams that they generated. This was done specifically within and around the lagoons of Gambier Islands, French Polynesia for the purpose of better quantifying the processes that are likely having an (adverse) impact on oyster spat. Though as the authors point out, such a wide array of high temporal resolution data can be used for many different investigations. The authors quickly move through the background to the project, the deployment/retrieval, QC, and a description of each type of data. Providing insights into the importance of the findings as they go. The conclusion is rather short, but that is fine. As the authors also point out that more in-depth studies utilising these data have already been (or are in the process of being) published.

The effort needed to manage a project of this scope is commendable and the dataset (as a series of NetCDF files) is well packaged and openly accessible online. The data are easy to download, extract, and work with. Though I think it would have been better to package the data into one single NetCDF file. Or at least just one NetCDF file per instrument type. For example, it puts unnecessary encumbrance on the user to have to separately download, load, and combine files for L01_3m and L01_21m when the only difference between them is depth, which is already one of the attributes in the NetCDF file. By creating a spray of files like this it reduces the R of the data (i.e. Reusability; FAIR). That being said, the data are already published and I don’t think it’s necessary that they be recombined, even though it would be of some benefit to the community, and to the posterity of the data themselves.

The authors are perhaps a bit heavy on their use of figures. And the contrast between colour palettes and aesthetic styles between figures is sometimes a bit jarring. I did however enjoy most of the figures and found some of the visualisations to be quite interesting. I assume that some of the figures were made with different software applications, which will prevent the authors from maintaining a consistent aesthetic throughout. So I would recommend trying to have at least a consistent font face, and/or develop a border for the figures in post-processing that looks the same in order to ensure a more contiguous visual experience for the reader. And please replace rainbow colour palettes whenever possible. The ‘viridis’ colour palettes being one easy choice.

Other than that, my only recurring criticism is the quality of the writing. The text would benefit from being passed through a language correction software like deepL in order to bring it closer to a native English level. I was however able to progress through the manuscript and understand the authors meaning for every sentence on the first read (which I find is often not the case for physical oceanography papers). So I am not suggesting that any structural re-writing is necessary.

Overall I think the paper effectively communicates what data were collected, why they were collected, what they look like, and where to find them. Whereas I do have some minor points below, these are just comments on the grammar/syntax before I stopped editing the language too closely. I think the authors should re-consider how they visualise their data, but I don’t think this is absolutely necessary for this manuscript to be published. Therefore there are only minor revisions to make.

All the best,

-Robert Schlegel

Title

- No comments.

Abstract

- 27: “which aims at tackling the connectivity oyster larvae” Not quite sure what the authors want to say. Perhaps: “which aims to understand the connectivity of oyster larvae”

- ln 29: “data set” Here and throughout, I think that in a data paper the preferred spelling is now ‘dataset’. Even though this isn’t historically the proper English spelling, the language is changing via the influence of technical computer terms.

- 30: “into the” -> “via the”

Introduction

- 34: “as in” -> “as of”

- 35: “Islands, are” -> “Islands are”

- 37: delete “however rather”

- 37 : “term of” -> “terms of”

Study Site

- 55: I assume that the preferred editorial style for lon/lat coordinates is in degree decimals, not minutes seconds. As the authors/editor prefer.

- 57: “counts” -> “counted”

- 68: “10-meter” -> “10 m” Here and throughout, the authors connect the unit to the number (e.g. 100m), but the general standard is to put a space between the number and the standard unit (e.g. 100 m).

- 100: “at 91 altitude meters” -> “at an altitude of 91 m”

- 102: “from ERA5” -> “from the ERA5

- 103: “, data” -> “. Data”

Sampling Strategy

- 139: “Leg 1” -> “Leg1” Double check throughout the manuscript.

Instruments etc.

- 170-171: “Specific processing steps required for some datasets are presented here.”

- 180: “a data logger”

Hydrodynamic overviews

- 265: “excepted during few events” -> “except during a few events”

- 334: Insert line break between paragraphs

Data availability

- Link to data works well. Data can be downloaded and are in a standard NetCDF format. Though I wasn't sure what the ‘CORRELATION’ values were meant to show in the ADCP files.

- It would be preferable if the code used for the analyses / data QC were also made publicly available (e.g. GitHub etc.).

Conclusion

- No comments

Table 1

It’s very impressive to see how much work this is when summarised into a table. I recommend putting the rows of moored instruments only for Leg1 at the top of the table.

Table 2

No comments.

Figure 1

Very nice. I like the layout of the progressively zoomed in map panels. Nice attention to detail how there is more and more colour as the panels zoom in. I recommend removing it and just having the one land mass shown.

Figure 2

I don’t think the inset is necessary. It’s not much more zoomed in than the main image.

Figure 3

Rather use a non-rainbow colour palette to show wind direction. Such as one of the viridis colour palettes. “Wind directions follow meteorological convention” Meaning that the direction shown here indicates that is where the wind is coming from, or towards where it is going?

Figure 4

“. .” -> “.”

Figure 5

No comments

Figure 6+7

No comments.

Figure 8

Add names to y-axes, and panel labels (i.e. (A) (B) )

Figure 9

That is an interesting way to show the difference between the surface and the bottom temperature.

Figure 10

Interesting.

Figure 11

“(direction where the current is going)” This would be helpful to state in other locations in the text as well. As I noted in a comment above. For the Figure 10 caption as well.