Review of “Carbonate pump feedbacks on alkalinity and the carbon cycle in the 21^st century and beyond” by Planchat et al.

Using CMIP6 models, the authors explored how the biogenic CaCO3 export at a 100 m depth responds to projected increases in atmospheric CO2 and how the different responses impact ocean surface alkalinity, the saturation state with respect to calcite, and oceanic CO2 uptake during the 21^st century. Motivated by the CMIP6 model spread in the projected CaCO3 export, the authors further explored the oceanic responses to imposed CaCO3 export changes for the extended time-period of 2100-2300 using an offline ocean biogeochemistry model. The authors nicely showed that the carbonate pump is one of the least constrained processes for the oceanic carbon pump and its projected uncertainty is very large. By assessing how the projected uncertainty can propagate into uncertainties in simulated oceanic carbon cycle on decadal to multi-centennial timescales, this study elucidates potential feedbacks from the carbonate pump on future carbon cycles. It is especially interesting to see that CaCO3 dissolution could respond abruptly to ocean acidification and that the sudden shift in CaCO3 dissolution could impact the regional patterns of oceanic CO2 uptake within the next century. The relatively minor effects of changing carbonate pumps on the oceanic CO2 uptake, compared to changing ocean physical dynamics, seem also novel and insightful. Overall, I have minor points.

1. The long-term response of CaCO3 dissolution to ocean acidification and its feedback on the Southern Ocean CO2 uptake seem very interesting. A question is how realistic this could be. Although a detailed model description is already presented in a previous publication by the same author, it might be worth highlighting here how the CaCO3 dissolution is parameterized in the PISCES model and how this parameterization compares with some observational constraints and/or models (e.g., Subhas et al. (2017); Liang et al. (2023) ). Related with this, the authors might say something about how CaCO3 dissolution is parameterized among the CMIP models in Introduction, as the effects of CaCO3 dissolution would become increasingly more important for the oceanic carbon cycle on longer timescales than the effects of CaCO3 export.
2. Please revise figures. In many figures (e.g., Figure 3), X- and Y-axes labels and units seem incomplete.

Line #288-289: From Fig. 4b, I don’t see that the indirect acidification rises from the deep to the surface. Instead, I see that the depth of CaCO3 dissolution (therefore basification) rises from the deep ocean towards the surface.   

Line #299-300: Increasing subsurface POC remineralization itself would cause subsurface ocean acidification, opposing the effects of reduced CO2 uptake.

Data availability: Perhaps, the authors can make the NEMO-PISCES sensitivity experiments available to the public?

References:

Liang, H., Lunstrum, A. M., Dong, S., Berelson, W. M., & John, S. G. (2023). Constraining CaCO₃ export and dissolution with an ocean alkalinity inverse model. Global Bigeochem. Cycles, 37, e2022GB007535. https://doi.org/10.1029/2022GB007535

Subhas, A. V., Adkins, J. F., Rollins, N. E., Naviaux, J., Erez, J., & Berelson, W. M. (2017). Catalysis and chemical mechanisms of calcite dissolution in seawater. Proceedings of the National Academy of Sciences, 114(31), 8175-8180. https://doi.org/doi:10.1073/pnas.1703604114

The manuscript “Carbonate pump feedbacks on alkalinity and the carbon cycle in the 21st century and beyond” by Planchat et al analyzes ocean carbonate cycling in CMIP6 ESMs and performs an additional idealized analysis with NEMO-PISCES simulations to explore the underlying mechanisms.  The manuscript is an important addition quantifying trends and uncertainties from the CMIP6 generation ESMs on ocean carbonate cycling changes out to 2100 where the impact on carbon uptake is shown to be small, as well as post 2100 in which internal ocean changes to carbonate saturation states lead to fundamental shifts in carbonate cycling and impacts on surface alkalinity.  I recommend publication with minor technical revision as outlined below.

Technical points:

20 – I think “basis” should be “base”

35 – “towards” should be “to” and “, which explains why it” should be “in what”

43-44 – I am not sure if this statement is appropriate as a blanket statement or only in the context of the steady state/preindustrial ocean.  I also do not think the current Zeebe reference about Boron is relevant to the argument… perhaps the authors intended this one instead? Zeebe, R. E., & Wolf-Gladrow, D. (2001). CO2 in seawater: equilibrium, kinetics, isotopes (No. 65). Gulf Professional Publishing.

132 and Equation 1 – The text says that alpha is multiplied by the PIC production, but inspection of the equation for alpha looks like it is zero at preindustrial CO2 of 285 ppm which would mean zero PIC production… should “multiplied by” be “added to” – in which case alpha has units of PIC production, or perhaps the equation goes to 1 at preindustrial?  Also, the authors should specify how they derived the value of 0.15.

145 – add “, respectively” after “sDIC”

146-147 – The authors should specify what tolerance for “drift” was used to exclude these models but include others with less drift.

198 – The “effect” should be more specific – do the authors mean chemistry “acidification”, or fluxes “carbon uptake” or both?

Figure 2 – yellow line “crab_low” should be “carb_low”

249 – “that” should be “those”

250 – I think “affected” should be “driven”

Figure 3 – Why no symbols for NEMO-PISCES simulations on panel A?  Unclear why a few symbols in A and C have a black box around them, “7” in panel C and “yr” in axis labels are not showing up, and Y-axis legend for panel D also seems to have formatting issues in my pdf version.

278 – remove second “is”

281 – This sentence does not make sense to me and should be reworked “Although a calcite saturation state threshold can robustly be pointed out, the shift itself should be reversible regarding this environmental control parameter, and should remain dependent to it.”

282-284 – “If subsurface Ωcalc is back at higher values than the threshold, then the vertical PIC dissolution should shift back at depth, probably without hysteresis.” I should be made clear that this is resented as a hypothesis untested in he present study and thus speculation.  As such, the next sentence shouldn’t begin with “Therefore” as if the previous assertion had been proven.  Also, it is not clear whether the Chen study was evidence for or against a tipping point.  This should be clarified.

288 – add comma after “uptake”

293 – remove comma after “uptake”

316-317 “As there is a robust negative correlation between CaCO3 export anomalies and salinity-normalized surface alkalinity in the

CMIP6 ensemble, salinity-normalized surface alkalinity observations could be used to identify historical trends in PIC export.”  Are the authors sure about that?  Carter et al (https://agupubs.onlinelibrary.wiley.com/doi/10.1002/2015GB005308) argue that the earliest historical signals only emerge in 2030.