Comment on

“Water masses in the Atlantic Ocean: water mass ages and ventilation”

submitted by Mian Liu and Toste Tanhua to EGUsphere

Overview:

Mian Liu and Toste Tanhua submitted a manuscript about Atlantic water mass distribution and their (different types of) water mass ages, using transient tracers (the older anthropogenic CFC-12, the younger anthropogenic SF6, and the radioactive 39Ar) and the well-established TTD-method.

They introduce the paper with a short review of Atlantic water masses and their relevance in ocean circulation and for biogeochemical processes, followed by introducing the transient tracers and relevant methods. Moreover, they introduce the oxygen utilization and oxygen utilization rate and their importance in biogeochemical cycles.

Next, they introduce in more detail the application of the TTD method to derive water mass ages, they differentiate different types or definitions of ages and other technical aspects to compute them from the transient tracer data, which they obtained from the GLODAP data repository.

In the large results-section they show and discuss their computed water mass ages along two prominent WOCE/GO-SHIP sections and on maps in various density layers incorporating a large amount of GLODAP tracer data. They do that for several relevant Atlantic water masses, starting from the near surface layer to the deepest Atlantic water layers.

Additionally, they use their previously discussed water mass ages to derive the oxygen utilization rates for the most prominent Atlantic water masses. Finally, they compare the CFC and SF6 based water mass ages with ages derived from the radio nuclide 39Ar and discuss the deviations between the two.

Lastly, they summarize and discuss their findings and interpretation in a conclusion-section.

General comment:

The manuscript is very well written, easy to read and good to understand. The strength of the manuscript is the systematic application of the transient tracer-based TTD method on a Atlantic-wide large data set and the comprehensive discussion of the most relevant Atlantic water masses from surface to bottom. Moreover, the authors apply their findings by calculating and discussing oxygen utilization rates Atlantic-wide. Both might be a good data base for further investigations of biogeochemical cycles in the Atlantic Ocean.

I’d like to recommend publication after minor revision.

For minor revision I have listed a few specific comments (or better suggestions):

General comments:

Since I am a non-native speaker; I do not comment on wording, grammar, spelling, etc. I’d like to leave these issues to the editorial board.

In my pdf all figures are of minor, very blurred quality. I guess, that happened during crating the pdf for the reviewers from the submitted manuscript and submitted figures. So, this comment goes more to the editorial board than to the authors.

Specific comments:

Line 22-23: “eastern basin exhibiting younger ages compared to…” If this is meant in general, it should be rather opposite. Or is this specific for NEABW? Please, clarify.

Line 110-112: “The solubility (F)…” Find a better place in the methods’ section.

Line 113ff: You use the term “diffusive”, as widely common in this context. But wouldn’t it be more logic to use “disperse”. Many people automatically think on “molecular diffusion”, but here something different is meant, which might be better named “dispersion”.

Line 129-130: “Based on the TTD determination…” This sentence stays a little bit alone at this place. Skip or find better location.

Line 140: “the TTD is only a spike.” Moreover, in such a case the Δ becomes infinite small and the IG (Eq. in line 122) formally becomes a δ-function and can be computed the same way, so it is just a special case of the same formalism and not something completely different. Eventually worth to note here.

Line 141: “…underestimated by the tracer-age.” Eventually worth to refer to Sonnerup, 2001, who diels with different “tracer-ages” and the systematic deviation between tracer-ages from old and young or low and high concentrations.

Line 151-152: “The atmospheric concentration…” The sentence is important, bit at this place it stands a bit out of context. Please, find better location in section 3.

Line 152-175: The entire Δ/Γ-section could be shortened or focused. 1. The trouble is not enough tracers to compute Δ and Γ and there is almost no way out. Many authors previously discussed this issue and using Δ/Γ=1 is widely common and accepted. 2. The entire discussion here, if Δ/Γ= 0.6, 0.8, 1.0, 1.2 etc. does not lead to a final conclusion. Moreover, why not 0.1, 0.5, 1, 2, 5? 3. In principle and in theory the authors HAVE TWO tracers (or even three with the 39Ar) and COULD compute both Δ and Γ. But I guess, that might not help, since at least CFC-12 and SF6 do not provide enough independent information. So, keep it short. All tracer people know, why, and all others don’t care.

After line 175:

One would like to know, HOW the Γ is computed from the tracers. I guess a least square fit or something similar. But here is my most urgent question, if the authors do that by using CFC-12 and SF6 individually or at once. It is stated somewhere further below in the manuscript, but it should already be mentioned in the methods.

Which atmospheric data do the authors use? Pleas add a reference.

Line 177-178: In this study … vertical layers…” Could find a better place or, since independent from the following, mage an own paragraph of this sentence.

Line 188ff: As mentioned above, move this to methods-section?

Line 197-200: Move this paragraph to a better location, e.g. together with the sentence in lines 177-178.

Figure 6: Could you add “mean age” and “mode age” into the figure? (Holds for all such figures.)

Figure 7: Limit from 0 to 1000 dbar?

Figure 9: Limits from 0 to 3000 dbar?

Line 290: “over THE ridge”?

Line 359-366: The discussion why to favor the mode age instead the mean age does not convince me. In my understanding, the mode age and the mean age are almost one order of magnitude apart, hence, the OUR would also be almost 1 order of magnitude apart. Thus, an argument is needed, why the one and not the other. As I understand the authors, the mean age is subject to mixing with adjacent water masses. True. But the mode-age is, too. So, this argument does not count. I have also no clue, what is the right argument for favoring the mode-age. I guess it is related to either the “right” order of magnitude OR it has something to do with the integrative effect of the mean-age (the mean age sums all the damn history) and the more advective character of the mode-age.

Line 428-429: If the authors decide to put the technical aspect of how to compute Γ into the method-section, this could also go there. As I understand right, it is just the equation from line 122 with an additional decay factor e^-λ/t ? Might be worth to note clearly.

Line 442: “(Fig. 18)” does not exist.

Line 455-465: What I would really enjoy would be a Figure 18 as a property-property plot as CFC/SF6 mean-age versus 39Ar mean-age and CFC/SF6 mode-age versus 39Ar mode-age and a short related discussion of such a figure. I’m sure one could learn something from it.

Line 467: I would put this sentence to the end of this paragraph.

Line 468: Add 39Ar to the bracket.

Line 470: Replace “investigated” by “discussed”?

Line 472: Replace “different parts” by “different fractions” or “different pathways”?

Line 498-505: Here I was puzzled. I can’t put this into the context of the manuscript. I can’t see climate change in the manuscript, since no repeat sections or stations are compared. Numerical simulation methods were not applied here. Please, either skip or clarify in the manuscript’s context.

Review of “Water masses in the Atlantic Ocean: water mass ages and ventilation” by Mian Liu and Toste Tanhua.

General comments

The study is of the GLODAP data product, in particular transient tracers and dissolved oxygen data are used to estimate water mass ages and ventilation. The transient tracers of CFC-12, SF6 and 39Ar are examined with the assumption of an Inverse Gaussian function in order to find the mean- and mode-age of different water masses. The resulting calculations of age and ventilation of different water masses is outlined in great detail. Further inferences regarding the oxygen utilization rate are outlined with implications for biogeochemistry.

I think that this study could be worthy of publication in EGU Ocean Sciences, once my below comments have been addressed. As further outlined below, the Data and methods section was confusing and needs a strong revision. The figures were not of a high production quality (often too small, low resolution) but could be improved with some effort. Finally, the authors should take more care in stating the limitations of the results and when the conclusions are more speculative. I have tried to note specific examples in the below comments.

On a positive note, I will say that much of the paper was clear and easy to follow (e.g. Introduction and Results sections). While the paper reads a little like a report at times, with results described in detail one after another, I think that this is probably unavoidable due to the nature of the study. Overall, I think it would be beneficial to the community to have the work published, as work on water mass age and ventilation is important to the community.

Major comments

1. Data and methods section: I am not an expert in age tracer methods, so I found that this section was very confusing to follow. I suggest to strongly revise this section, keeping in mind to explain the overall methods and reasoning to the reader at each step. For example, the equations (Lines 116-122) were introduced abruptly with little explanation. I was left with basic questions e.g. is width defined as \Delta or \Delta^2? What are the units on the mixing ratio? What is 2-IG? These are just some examples. Some information later in the manuscript (e.g. Lines 352-366) would have been helpful earlier in Section 2.
2. Figure quality was low: figures were quite low resolution (at least in the version I had). They were often too small (e.g. Figure 6) and required zooming in to see the details. I also suggest that different colors could be used for the different parameters in Figure 14, in order to more easily distinguish between the parameters in the panels. Some captions could also do with more clarity, and I have tried to note these points in more detail in the comments below. The actual data on the figures was interesting, and I think it would be a relatively straightforward fix to improve the figure quality across the entire paper.
3. Being more careful outlining limitations of results and speculation: Some of the content in Section 2.1 and 2.3 outlined some of the main limitations of the results, such as assuming an Inverse Gaussian profile and assuming a mixing ratio of 1. I understand that these assumptions are required in order to get useful results out of the data, but I think that the results and conclusions section could help with re-iterating these limitations. There were also a few times where the wording was a bit strong on speculative statements, for example in the conclusion “The results show that with global warming, the age of the Atlantic water mass shows a certain degree of change.” But my take on the paper was that it was showing the water age and ventilation, rather than changes with climate? So this seems a little speculative and I have tried to note these parts in the comments below.

Specific comments

• Abstract: “The distribution of oceanic water masses and their properties, such as ventilation constitute fundamental parameters, for instance, the thermohaline circulation patterns and biogeochemical processes in the marine systems.” This sentence does not make sense and needs revision.
• Line 19-22: Can remove the brackets around “mean” and “mode” in this sentence.
• Paragraph from Line 45-55: There has been more work recently using ocean models to discuss the age tracer, please consider citing some of these (e.g. Li, Q., England, M. H., Hogg, A. M., Rintoul, S. R., & Morrison, A. K. (2023). Abyssal ocean overturning slowdown and warming driven by Antarctic meltwater. Nature, 615, (7954), 841-847).
• End of Introduction: Consider adding a paragraph with a brief outline of the rest of the paper to orientate the reader.
• Data and methods section: See Major Comment 1 (above). Unfortunately, by the end of Section 2 I still did not really understand why mean- and mode-ages are used for different applications. Please strongly revise Section 2, as it is by far the weakest section.
• When G(t) first introduced: Explain physically G(t) and the assumptions going in to G(t). My understanding is that you are assuming the shape of the profile as G(t) and then further assuming the mixing ratio, and either the width of the mixing or \Lambda? Then you have the profile such as in Figure 1b. But it is still not clear to me at this stage what measurement is required for Figure 1b. How does CFC-12=300ppt play into G(t)?
• Line 144-147: “The mean-age is obtained by the average value of all the different aliquots in one water sample by considering the transport time of different pathways (Fig. 1, b). This concept of water mass age is often useful for biogeochemical studies. In other cases, such as when discussing the transport times of water masses from the formation area, the mode-age concept is useful.” This seems directly in contrast with Lines 471-473: “The mean-age, which shows the average value of all the different parts in one water sample, is used to show the static distribution of water masses. The mode age, which shows the age of the dominant water mass in the sample, is used to trace the biogeochemical phenomena.”
• Figure 1: How does the description of (a) in the caption match with the figure in panel (a)? Text on the figure is very small also.
• Figure 2: Very small text and the vertical axis line was not present on panel (a).
• Figure 3: Low resolution and tricky to make out the results as the panels are small. What are the green lines? Please also state the units of age either on the figures or in the caption. Also, it would help to use the same mixing increments as in Figure 2(b) i.e. 0.6, 1.0 and 1.4. This would connect Figures 2 and 3 better.
• Line 156: “ration” -> “ratio”
• End of Section 2: The justification of mixing ratio of 1 is not particularly strong, as it does seem like a value that could change a lot regionally depending on physical processes in the ocean. I agree that the authors are limited here so a mixing ratio of 1 is probably the right choice, but it would help to highlight that this is a limitation of the study.
• Table 1 caption: Consider simplifying the text to something like: “Summary of the selected hydrographic cruises to the Atlantic Ocean chosen for the oceanographic sections in this study”.
• Figure 4: What are the green contour lines on the figures?
• Section 3.1: This sub-section starts with a long paragraph. Consider splitting it into two paragraphs for ease of reading.
• Figure 6: plots are small. Please make them bigger.
• Line 230: “… section through in the distribution …” Please remove “in”
• Line 235-236: “In principle, the AAIW is supposed to get higher ages towards the north.” Consider explaining why this is expected.
• Line 239-240: “The reason for the above result is the intervention from surrounding water masses.” I find this sentence strongly worded for this more speculative result. Other reasons for the result could be that the IG function is not a good assumption, or that mixing ratio really does change across the ocean.
• Line 240-241: “The maximum distance of AAIW to the north can reach 30 °N, but between 241 20 °N and 30 °N, this water mass mixes with the ENACW and upper NADW…” Consider saying “we speculate that this water mass mixes…” or “this water mass likely mixes…”
• Paragraph of Lines 253-262: Again, I find much of this discussion on mixing as speculative. Please consider mentioning this or weakening the statements a little.
• Figures 8, 10, 14: Panels are small, please make them bigger.
• Figure 14: AOU top panel. What does it mean for this value to go negative? And how does this influence the OUR result?
• Figure 15: Caption does not appear to line up with shown panels.
• Figure 16, 17: “pentagrams” symbols would more commonly be called “stars”.
• Figure 16,17 caption: “The solid isolines show the 50% fractions of water masses and the dashed lines show the 20% fractions.” Is this consistent with the lines shown in Figure 15?
• Line 442: Refers to Figure 18. Is this figure missing?
• Line 467-468: “As the continuous work…” perhaps replace with “In the present work…” ?
• Conclusions: “Three different definitions of water mass ages are investigated. The tracer-age assumes the ocean as a totally advective situation without diffusion and underestimates the actual age in the realistic ocean.” Tracer-age is briefly discussed in the manuscript, but no figures or results are shown of it, so I would disagree that it has been investigated here. I suggest to just be careful of the wording here.
• Final two paragraphs of the conclusions: Talk about effects of climate change, but (as far as I can see) this is not really touched on in the actual results of this paper. So perhaps it is a bit speculative? It would also help to remind the reader of the limitations of the study in the conclusions.