Increase of dissolved inorganic carbon and decrease in pH in near-surface waters in the Mediterranean Sea during the past two decades

Merlivat, Liliane; Boutin, Jacqueline; Antoine, David; Beaumont, Laurence; Golbol, Melek; Vellucci, Vincenzo

doi:https://doi.org/10.5194/bg-15-5653-2018

Articles | Volume 15, issue 18

https://doi.org/10.5194/bg-15-5653-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/bg-15-5653-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 18

Research article

|

21 Sep 2018

Research article |

| 21 Sep 2018

Increase of dissolved inorganic carbon and decrease in pH in near-surface waters in the Mediterranean Sea during the past two decades

Liliane Merlivat, Jacqueline Boutin, David Antoine, Laurence Beaumont, Melek Golbol, and Vincenzo Vellucci

Download

Final revised paper (published on 21 Sep 2018)
Preprint (discussion started on 06 Jul 2017)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review of Merlivat et al., “Increase of dissolved inorganic carbon and decrease of pH in near surface waters of the Mediterranean Sea during the past two decades”', Anonymous Referee #1, 13 Aug 2017
- AC3: 'Answer to the comment of the reviewer 1. The text of a revised manuscript is attached as a file.', Liliane Merlivat, 07 Dec 2017
  - AC4: 'Reformatted text', Liliane Merlivat, 11 Dec 2017
RC2: 'Review of "Increase of dissolved inorganic carbon and decrease of pH in near surface waters of the Mediterranean Sea during the past two decades" by Merlivat et al.', Anonymous Referee #2, 22 Sep 2017
- AC1: 'Answer to the comments of the reviewer 2. The text of a revised manuscript is attached as a pdf file.', Liliane Merlivat, 05 Dec 2017
- AC2: 'A correction to the answer to the comments of the Reviewer 2 just sent', Liliane Merlivat, 05 Dec 2017
  - AC5: 'Reformatted text', Liliane Merlivat, 11 Dec 2017

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (12 Dec 2017) by Jack Middelburg

AR by Liliane Merlivat on behalf of the Authors (08 Jan 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (09 Jan 2018) by Jack Middelburg

RR by Anonymous Referee #1 (13 Jan 2018)

RR by Anonymous Referee #2 (03 Apr 2018)

Suggestions for revision or reasons for rejection

MAJOR CONCERNS:

In my previous review of the submitted manuscript from Merlivat et al., I mentioned two general concerns that led me to recommend that in-depth revisions were needed. First, uncertainties appeared underestimated and poorly described, and second natural variablity was ignored as a possible explanation for some of the change seen between the two 3-year periods.

In the revised manuscript, the authors have tried to clarify their text in regards to my first concern. Yet the discussion of these results, appears unclear, imprecise, and does not offer a clear statistical demonstration that the differences between the two time periods are significant. More detailed concerns about the generally poor description of uncertainty analyses are provided below in the comments concerning lines 108-109, 241-246, 254-255, and 275-276.

As for my second major concern, the authors response is unsatisfactory. Although this concern is mentioned briefly in the Introduction, the authors have just cut and paste an entire sentence from my Review, word for word. The same concern is mentioned briefly in the Abstract, Discussion, and Conclusion. Much more text is devoted to the explanation of anthropogenic change rather natural decadal variability.

Overall, I am disappointed with the authors responses to both of my previous major concerns. Substantial improvements would still be needed to clarify these points to a satisfactory level before I could recommend that the manuscript would be publishable.

DETAILS:

lines 51-53: Change
"The quantitative estimation of anthropogenic CO2 storage in the
Mediterranean Sea based on experimental data is very inaccurate, of the order of a factor two [Huertas et al., 2009; Touratier and Goyet, 2009]."
to
"Estimates of anthropogenic storage in the in the Mediterranean Sea differ by about a factor of two [Huertas et al., 2009; Touratier and Goyet, 2009]."

lines 56-57: "[McKinley et al, 2011]" is repeated 5 times. This sentence is plagiarized from my review.

lines 58-59: Suggest changing "of time is a way to detect a possible trend in DIC." to "is useful to assess trends and variability of DIC."

line 62: change "very close" to "nearby" or "adjacent".

line 86: delete "depth"

line 96: Incorrect format in sentence for citation.

line 101: change "meters" to "m"

lines 108-109: Errors on fCO2 calculated from DIC and Alk, each having uncertainties of about 3 umol/kg, is about 5% based on results from Dickson (2010, Table 1.6). For a base value of fCO2=400 uatm, that would imply that the uncertainty in calculated fCO2 from Merlivat et al. is +/- 20 uatm. This is four times larger than the estimate quoted by the authors based on the paper from Millero et al. (2007). The authors need to mention this more recent study and the much higher uncertainty in calculated fCO2 that is implied from that.

line 138: change "this last effect should be negligible" to "the effect of nutrients may be neglected".

line 150: change "meters" to "m".

line 236: delete "It is interesting to note that"

Table 1: Over the final 2 columns it is marked "Temporal trend". Actually this is not a trend (change per unit time) but just the difference between the 2 periods.

lines 241-246: Confusing explanation of the uncertainties in the estimates of the difference dfCO213. just because there are 3 CARIOCA sensors and each is supposed to have an accuracy of +/-2 uatm does not mean that the accuracy of the estimate (presumably the time average) for each period is 2/(3)^0.5. The standard deviation of the measurements in each time period is much larger (7 to 28 uatm). The standard deviation of the difference is much larger still. The authors are not clear about what they are referring to exactly when they say that "the accuracy on the difference is estimated to be 1.6 uatm." They may be referring to the standard error of the mean, but they do not say so explicitly. In any case, judging from the numbers in Table 1, the last 2 columns suggest that the difference between the 2 time periods is not always even significant. The authors would need to provide a significance test to show that the means of the two time periods actually differ significantly.

lines 254-255: The authors do not provide enough detail on how they made their calculations for the changes in DIC and pH and the corresponding uncertainties. Are they using mean values of fCO213 and their for the two time periods? Are they using the standard error of the mean for the uncertainty? Without substantially more detail, I am left with the impression that they are underestimating the uncertainties. The uncertainties they do provide for the differences in DIC and pH are much smaller than the measurement uncertainties.

lines 275-276: change "the corresponding amount of anthropogenic carbon taken up from the atmosphere in order to maintain a chemical equilibrium at the sea surface would be equal"
to
"the corresponding change in surface DIC, assuming air-sea equiliibrium, would be would be"

It is unsure that these annual mean calculations are adequate since it has been shown previously that the air-sea flux of anthropogenic carbon varies seasonally. This doubt is further supported by the authors own statements in their lines 291-297.

line 332 : improper format for citation.

line 333: improper format for citation.

line 339: improper format for citation.

line 345: improper format for both citations.

line 355: The authors still use the symbol "pH_T", unlike what they say in their response to my previous comments.

line 356: improper format for citation.

line 360: improper format for citation.

line 364: improper format for citation.

Section 4.4: The discussion of changes in pH is inadequate. It does not consider that the change in pH depends not only on the change in

line 354-370: Section 4.4
This subsection appears particularly weak. It is descriptive but offers no real discussion. Why is the magnitude of the change of pH in the Mediterranean Sea more than the global ocean average. Is that difference significant? Why is it more than in the Iceland Sea but less than in the Irminger Sea. The authors have made no effort to discuss the causes of these differences. The last sentence only talks about the change in anthropogenic carbon, but that is not so clearly related to the change in pH since the alkalinity in the Mediterranean Sea is higher than at the other sites that are mentioned. Thus anthropogenic DIC increase there should be higher there although the pH change may be similar to that for the global ocean (Palmieri et al., 2015). But changes in pH are confusing because of the log scale. An absolute change in pH actually represents a relative change in H+.

Hide

ED: Reconsider after major revisions (06 Apr 2018) by Jack Middelburg

AR by Liliane Merlivat on behalf of the Authors (31 May 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (05 Jun 2018) by Jack Middelburg

RR by Anonymous Referee #2 (21 Aug 2018)

Suggestions for revision or reasons for rejection

In their second revision, the authors have addressed my two major concerns: (1) apparently underestimated uncertainties, and (2) the need to discuss natural variability as a possible explanation of part of the change seen between the two 3-year periods two decades apart. My recommendation would then be to publish this study after my remaining minor concerns, listed below, have been addresssed. Most but not all of these concerns relate to trying to improve the English.

ABSTRACT

Page 1, line 20: change "Two three-year-long time series" to "Two 3-year time series".

1, 21: change "10m" to "10 m"

1, 31-32: The last sentence of the abstract is not clear.

INTRODUCTION

2, 41-42: The three groups of citations in brackets with semicolons in between each group is strange. It does not follow BG conventions.

2, lines 47 and 49: Citations are given with incorrect formatting.

2, 55: change "after" to "only after"

3, 68: change "18 years" to "18-year"

4, 93: change "installed" to "attached"

4, 98-103: accuracy vs. precision may need revisiting

4, 113: The authors need to be more specific about what they mean by "Error on fCO2". I think they mean the propagated "uncertainty" (not the "error") based only on their previous uncertainty estimates for measurements of ALK and DIC, neglecting uncertainties in the equlibrium constants. This should be spelled out because it makes a big difference in the propagated uncertainty estimate. The authors should not use the word 'error', as detailed further below.

5, 132-137: Fig. 2 caption:
- The caption does not mention where the data were collected.
- "(d), (e), (f), seasonal variability" is unclear and not a sentence.

6, 142: change "extrema" to "extremes"

6, 147: change "thermodynamic" to "temperature"

6, 150: citation has incorrect format.

6, 153: "Biology accounts for" is vague.

6, 160: change "CO2 air-sea flux" to "air-sea CO2 flux"

7, 164 and 167: Do the authors mean "internal waves" rather than "intertial waves"?

7, 169: change "leading" to "corresponding"

7, 172: change "over" to "during"

7, 199: add "s" to "temperature"

8, 204: delete "very"

8, 209 and 211: The authors need to inform readers what the uncertainties in salinity are supposed to represent (std. deviation, standard error of the mean, or something else).

8, 224: add a comma after "winter"

8, 226: change "is clearly highlighted for the whole range of temperature." to "is evident across the range of temperatures."

9, Fig. 3 caption:
- yellow dots are mentioned in the figure caption for panel (a) but none are visible in the actual figure.

10, 249-258: The authors have confused me entirely here. In English, paragraphs should be indicated either by an indentation or a blank line as a separation before the start of a new paragraph. The authors have done neither. Nonetheless, I was usually able to guess when there is a new paragraph in their manuscript, i..e., when the previous line does not extend all the way to the right margin. However, this sloppy formatting makes it impossibile to tell what is going on in lines 24ç-258 and perhaps beyond. There appear to be two 1-sentence pseudo-paragraphs in the beginning, followed by a 3-sentence paragraph. But maybe all of this is intended to be part of the Table 3 title? In any case, with the current structure, one cannot tell where the Table title ends and where the ensuing text in the manuscript begins. All very confusing!

10, line 259: What "mean" is being discussed here?

11, line 273: If the two methods give results that are not significantly different, i don't think that one can justify using one vs. the other because "the former method produces a more conservative estimate." Furthermore, just because the value is lower does not mean it is "more conservative".

11, 281: It is incorrect to use the word "error" here. The 'error' cannot be known because one cannot know the true value. The 'error' also has a sign; it cannot be reported as +/-x (unlike the uncertainty). Thus "error on" should be changed to "standard uncertainty of". Later in the same line, "uncertainty" should be changed to "combined uncertainty in".

11, 282-284: It should also be stated that uncertainties in the equlibrium constatnts are neglected in this propagation of uncertainties.

11, 285-287:
- units for DIC are given strangely (remove the spaces before "mol"
- pH has no units. Remove "unit" in lines 286-287.

What is meant is "the propagated uncertainty accounting only for standard uncertainties of the measurements and ignoring uncertainties in the constants".

12, 298: Problems with formatting of Table title and foonotes

DISCUSSION

13, 344: The authors say that "The difference between these two values is significant." Such statements should come with the name of the specific statistical test used and the resulting p value that allows the author to make such a statement.

13, 349: add "the" before "mixed layer depth".

13, 353: add hyphen between 'DIC' and 'rich'.

14, 354: 'This' what? Never use 'This' by itself at the beginning of a sentence. In this case, the authors could say 'This entrainment'.

14, 366: change 'of' to 'for'

15, 388: 'in excess of 17+/-10%' is unclear. Do you mean it could be more than 27%?

15, 396: delete 'would'

15, 400: add a hypen between 'Long' and 'term'

15, 419: delete 'have also to be taken into account'. That phrase is unnecessary and messes up the sentence.

15, 420: delte 'rather'

16, 424: delete 'absorption'

16, 428-429: change 'Mediterranean anthropogenic acidification' to 'anthropogenic acidification of the Mediterranean Sea'

16, 430: change add 'of the Mediterranean Sea' after 'pH'

16, 431: add 'that of' after 'from'

16, 436: 'considerable short-time' is an oxymoron. The meaning is unclear.

16, 442: change 'as a source' to 'a substantial amount'

16, 443-444:
- change 'towards' to 'to'
- delete '('
- rather than saying ", close to 10% ([Schneider et al., 2010] or 25% [Palmiéri et al., 2015]", the authors should provide their estimate and follow that with something like "which lies between estimates of 10% by Schneider et al. [2010] and 25% by Palmieri et al. [2015].

16, 451: the '2' shoud not be subscripted in CO2SYS.

REFERENCES

The references are hard to read. Please add a line space between them or provide them each with hanging indentation at the beginning.

Hide

ED: Publish subject to minor revisions (review by editor) (22 Aug 2018) by Jack Middelburg

AR by Liliane Merlivat on behalf of the Authors (03 Sep 2018) Author's response Manuscript

ED: Publish subject to technical corrections (04 Sep 2018) by Jack Middelburg

AR by Liliane Merlivat on behalf of the Authors (05 Sep 2018) Manuscript

Short summary

The fugacity of carbon dioxide in seawater (fCO₂) was measured hourly in the surface waters of the NW Mediterranean Sea during two 3-year sequences separated by 18 years. A decrease of pH of 0.0022 yr⁻¹ was computed. About 85 % of the accumulation of dissolved inorganic carbon (DIC) comes from chemical equilibration with increasing atmospheric CO₂; the remaining 15 % accumulation is consistent with estimates of transfer of Atlantic waters through the Gibraltar Strait.