The atmospheric bridge communicated the <i>δ</i><sup>13</sup>C decline during the last deglaciation to the global upper ocean

Shao, Jun; Stott, Lowell D.; Menviel, Laurie; Ridgwell, Andy; Ödalen, Malin; Mohtadi, Mayhar

doi:https://doi.org/10.5194/cp-17-1507-2021

Articles | Volume 17, issue 4

https://doi.org/10.5194/cp-17-1507-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/cp-17-1507-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 17, issue 4

Research article

|

19 Jul 2021

Research article |

| 19 Jul 2021

The atmospheric bridge communicated the δ¹³C decline during the last deglaciation to the global upper ocean

Jun Shao, Lowell D. Stott, Laurie Menviel, Andy Ridgwell, Malin Ödalen, and Mayhar Mohtadi

Download

Final revised paper (published on 19 Jul 2021)
Supplement to the final revised paper
Preprint (discussion started on 28 Jul 2020)
Supplement to the preprint

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'Review of CPD manuscript 10.5194/cp-2020-95 by Shao et al.', Anonymous Referee #1, 14 Sep 2020
- AC1: 'response to the referee comments by Anonymous Referee #1', Jun Shao, 09 Feb 2021
RC2: 'Review of MS by Jun Shao et al.', Fortunat Joos, 02 Oct 2020
- AC2: 'response to the referee comments by Fortunat Joos', Jun Shao, 09 Feb 2021
EC1: 'editor assessement of the reviews', Hubertus Fischer, 02 Oct 2020
- AC3: 'Reply to EC', Jun Shao, 09 Feb 2021

Peer-review completion

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

ED: Reconsider after major revisions (09 Feb 2021) by Hubertus Fischer

AR by Jun Shao on behalf of the Authors (23 Mar 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (24 Mar 2021) by Hubertus Fischer

RR by Fortunat Joos (21 Apr 2021)

Suggestions for revision or reasons for rejection

I thank the authors for revising the manuscript. This manuscript has now developed into a nice paper that illustrates forcefully and convincingly the important imprint of air-sea gas exchange on upper ocean d13C.

It is nice to see that the mass balance of 13C is now treated correctly.

I have a few remaining comments that the authors may wish to consider.

Main manuscript (line numbering refers to track changed MS)

1) l. 14 and other places: suggest to write: “a ~35ppm rise in atmospheric CO2” instead of “..pCO2” as ppm units are for a mixing ratio and not a partial pressure.

2) Line 487: “Since there is no 13C fractionation during CaCO3 formation in the LOVECLIM model, the last term on the RHS can be assumed to be zero (see Supplement).”

This holds for steady. The isotopic signature will change during the experiment and this change is carried downward by the CaCO3 flux (irrespective of the assumed fractionation). As described in the SI, the contribution of CaCO3 dissolution to DIC is small in the upper 1000 m and therefore the last term can also be neglected in transient experiments and the upper ocean. The authors may provide a more complete explanation.

Supplementary Information:
1) SI: line 57: “From this we confirm, as expected, that DIC = DIC(pref) + DIC(Csoft), and d13C(DIC) = d13C(pref) + δ13C(Csoft).” I guess a qualifier is needed: “in the upper 1000 m” as CaCO3 dissolution may contribute significantly in the deep.

2) SI, line 70-90: In my opinion, it would be helpful to show here explicitly the equations how the different components are computed.

3) Please include #7 and provide the equation describing how d13C_soft is computed from the explicitly simulated DI13C_soft and DIC_soft and DIC tracers.

4) SI, line 93: “As is widely appreciated, AOU overestimates the consumption of oxygen through respiration as a result of incomplete equilibrium occurring between the ocean surface and overlying atmosphere.” This explanation is only partly correct. It should also be mentioned in addition that the solubility of O2 is nonlinear. For example, mixing two O2 saturated water bodies with different temperatures will lead to an O2 concentration that deviates from the saturated concentration.

Hide

RR by Anonymous Referee #1 (25 Apr 2021)

Suggestions for revision or reasons for rejection

Rereview of CPD manuscript 10.5194/cp-2020-95 by Shao et al.

With great interest I have reread the revised manuscript of Shao et al. and their response to my comments on the first submitted draft of the manuscript. The authors have responded efficiently and satisfactorily to many of my earlier points of criticism. The use of the LOVECLIM simulation LH1-SHW-SO is much better motivated, the logic behind a comparison of LOVECLIM and cGENIE output data is much clearer, and the paper benefits from the additional error analysis (although those should be part of the discussion rather than just a bunch of figures in the supplement). I am further glad to see that new simulations were performed with cGENIE and that the paper now focusses on the initial deglacial d13CO2 decline, which makes it clearer and more streamlined.

However, on some aspects, as outlined below, the authors made, in my view, an insufficient effort to improve the impact and clarity of the study. I also noticed one other weakness, namely in the results section that includes a number of discussion elements and does not give a full account of the model results and observations that are relevant for the study.

Earlier I have raised concerns regarding the premise of the paper, the distinction between preformed and regenerated ocean carbon in an early deglacial transient simulation with LOVECLIM, when this model version does not simulate those species explicitly. This has been echoed by the second reviewer, who even recommended to rerun the simulations with an explicit tracer scheme for preformed and regenerated carbon. I accept the authors notion of going forward with the paper setup as is, but I recommend to specify in line 175-177 the restrictions of the AOU approach. This is important and acknowledges that the authors approach for carbon partitioning of the LOVECLIM data is imperfect. I suggest to clearly outline the reasons why AOU (as difference between in-situ and calculated oxygen) can in some instances not be a faithful representation of regenerated carbon, and may overestimate it, as discussed in the mentioned literature.

I am somewhat disappointed by the authors response to my suggestion to better carve out the relative contributions of top-down and bottom-up processes on oceanic d13C, because of the fact that this “may very much depend on the models used”. The authors would have two models at hand to discuss this. I cannot follow statements in the paper such as “Subsequently, air-sea exchange dominates the δ13C decline in the global upper ocean. (line 253)” when the cause of the d13C decline is outgassing in the Southern Ocean (LOVECLIM model). Should it say in the global upper ocean outside the Southern Ocean? I am slightly confused, and was hoping for more clarification on the regions where outgassing occurs and where the water column is overprinted. In other words, the trigger for an atmospheric d13C bridge must be outgassing in parts of the ocean, where upper ocean d13C must see the regenerated d13C and DIC from below. The water masses are likely overprinted by the atmosphere (essentially by their own signal), but it is not clear by how much. I find explanations around this issue confusing, e.g. in line 316-319: why would the atmospheric d13C_CO2 signal be compensated by the upwelling d13C_CO2 signal from the deep, when both should have a similar negative signature? Are we talking in the EEP about a gas exchange of water masses from the sub-surface (with a regenerated signal) that equilibrates with an already decreased atmospheric d13C_CO2 (acquired somewhere else)? Or has that sub-surface signal itself acquired a negative preformed signal through atmospheric CO2 overprints outside the EEP? In general, I believe the paper would benefit from a more careful and in-depth explanation of the processes at play in the different ocean region, with model data support in the form of informative figures/maps. The authors mention the study of Martinez-Boti et al., (2015) which show an pCO2 oversaturation with respect to the atmospheric of 80 ppm at the beginning of the early deglaciation. This is evidence for the fact that the water mass has not fully equilibrated with the atmosphere. How can this be reconciled with the authors notion that isotopically it has acquired its signal from the atmosphere?

I further recommend that the results section is not mixed with discussion elements. At its current stage, the entire Results section below line 236 (to 260) is actually discussion. The only observation on carbon partitioning that is described and that would qualify for the results section is that from the North Pacific, which is incomplete. Even more worrisome is that none of the results of the cGENIE benchmark test has been described properly in the results section. This needs to be revised. Model papers have the luxury of presenting a wealth of data. The reader will appreciate the distillation of the main observations of the model output data.

Minor comments regarding language and clarity of the text.
Line 39-40: I’d suggest to acknowledge the existing large body of literature on this. Despite extensive research efforts over the last decades, the chain of events leading to the atmospheric changes recorded in ice cores remains incompletely understood.

Line 57: influencing

Line 62: move “until recently (Lynch-Stieglitz et al., 2019)” to the end, otherwise this sentence makes no sense

Line 88: I am confused by the word “theoretical”. I believe that the model calculates oxygen saturation directly, as function of ocean temperature and salinity etc. Also, “at every grid point in the model” can be dropped.

Line 92: it is better to write “have used the separation of DIC species into regenerated and preformed as starting point/basis to study […]” because Khatiwala et al. for instance use a much more sophisticated framework also considering disequilibrium effects.

Line 98-101: this sentence needs to be revised. E.g.: Here, we use the Earth system model of intermediate complexity cGENIE with a comprehensive diagnostic tracer framework (including for the first time a respired organic matter δ13C tracer) in order to fully evaluate the AOU-based off-line approach made based on LOVECLIM model data.

Line 117: Define abbreviation VPDB

Line 140: first mention of NADW, acronym needs to be defined

Line 151: AOU has been defined earlier.

Line 167: This may seem trivial, but in my view it should be mentioned how d13C_pref and DIC_pref was calculated.

Line 179-181: This goes against the statement in the introduction whereby the cGENIE model is employed to analyse a simulation with numerical tracer scheme.

Line 185: in the supplement.

Lane 207: it should read Rae et al., (2020). Does that mean that the model was tuned to Pacific LGM data only? If yes, this should be explicitly stated.

Line 275: Why should this signal be seen first in the STGSP? Some explanation is needed here. Can you clarify what nature the d13C anomaly has, of regenerated or of preformed nature?

Line 303: Reference to figure is needed. Or in fact a description in the results section.

Line 317: where the δ13CO2 signal […] is / Line 319: atmospheric overprint.

Hide

ED: Reconsider after major revisions (26 Apr 2021) by Hubertus Fischer

AR by Jun Shao on behalf of the Authors (06 Jun 2021) Author's response Author's tracked changes Manuscript

ED: Publish subject to minor revisions (review by editor) (08 Jun 2021) by Hubertus Fischer

Dear authors

I have carefully read your replies and the re-revised version of the manuscript and am happy to say that except for some relatively minor points listed below, the paper is ready for acceptance. Please, quickly implement the changes listed below in the main text and the supplement and send the final paper back to CP including a quick list of the changes done.

Main Text:
line 27: I am not sure what you mean by "upper-deep" waters and in which region these upper-deep waters would be located. Be more specific
line 78: "at 566 m depth"
line 111: "After describing..."
line 139: "(Berger, 1978), changes in the freshwater surface balance as well as..."
lines 163-165: There is something wrong with this sentence, please fix it
line 188: "As for the preformed..."
line 195: "implemented as DICpref..."
line 197: Do you mean "(and more conveniently input)"?
line 205: There is something wrong with this sentence, please fix it
paragraph line 205: You say you used idealized glacial boundary conditions, but then you apply interglacial CO2 of 278 ppm. This is confusing. See also comment on the supplement below.
line 236: "and since there is no"
line 247: "always reaches..."
line 248:" that this is not always the case"
line 284: Here you say that that the Southern Ocean would be heavier by 0.1-0.2permille everywhere, but that is not true for the surface. Please be more specific
line 321: "inherent in the AOU-based..."
line 325: "and because O2 solubility..."
libe 333: "throughout the uppermost...
line 375: "reflect an enhanced..."
line 405: "evidence indicates that..."
paragraph line 405: This needs some clarification. I read it that although the EEP thermocline waters are supersaturated with respect to atmospheric CO2, their d13C is controlled by the top-down signal. This refers also to the question by referee #2 related to the work by Martinez-Boti, showing a supersaturation of waters with respect to CO2 in the atmosphere. The argument is that isotopic equilibration doesn't require net gas exchange and you should say so in this paragraph to explain this "conundrum".
lines 424-425: Here you mention an early decline in d13C between 18.3 and 17 ky and that LOVECLIM would be able to simulate this. However, reading the text I wasn't sure I could pinpoint always to which process at what time you refer to. Please be more specific with the timing and clarify the text.
line 442: "of estimating respired..."

Supplement:
line 4: Here you say that the model has only a EMBM but in the next point you say it has a dynamical sea ice model. I assume, you also have a prescribed (average) wind field? If yes, say so to clear up the contradiction
line 91: "for example for the Bern3D..."
line 105: "=O2(AOU)/138" ???
line 113: "reduction) the global..."
line 140: ""from regenerated organic..."
line 150: "parcels of water"
line 155: "high southern latitudes"
line 161: "30micromol kg-1 which dominates"
line 165: "(AOU-d13C) - deviates"
line 168: "the water column..."
paragraph line 211: Here you explain that you do not really do a deglaciation run, but that you are only interested in the temporal evolution in a perturbed transient run. In the main text this does not become so clear, and I wonder whether you could mention that more prominently in the main text.
line 228: "across the uppermost"
line 243: "throughout the uppermost"
line 246: "point is that this is..."

Hide

AR by Jun Shao on behalf of the Authors (10 Jun 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (10 Jun 2021) by Hubertus Fischer

AR by Jun Shao on behalf of the Authors (11 Jun 2021)

Short summary

Planktic and shallow benthic foraminiferal stable carbon isotope (δ¹³C) data show a rapid decline during the last deglaciation. This widespread signal was linked to respired carbon released from the deep ocean and its transport through the upper-ocean circulation. Using numerical simulations in which a stronger flux of respired carbon upwells and outcrops in the Southern Ocean, we find that the depleted δ¹³C signal is transmitted to the rest of the upper ocean through air–sea gas exchange.

The atmospheric bridge communicated the δ13C decline during the last deglaciation to the global upper ocean

Download

Interactive discussion

Peer-review completion

The atmospheric bridge communicated the δ¹³C decline during the last deglaciation to the global upper ocean