The study quantifies 1) the efficiency of a continuous global application ocean alkalinity enhancement using several metrics, 2) the subsequent response of the global carbon fluxes, and 3) the long-term potential of OAE to mitigate ocean acidification. The authors used an Earth System Model to simulate three atmospheric temperature stabilization scenarios to year 2500. For each scenario, five realizations with and five realizations without OAE were produced in order to quantify the direct response of the system to OAE, and to quantify the impact of climate feedbacks.

Results indicate that OAE efficiency estimated with the metric that accounts for climate feedbacks is much lower than the maximum potential efficiency only based on carbonate chemistry. The authors rightly discuss in which circumstances one metric should be favoured than the other. The study also shows that in centennial timescales, under continuous global deployment of TA, it is the reduction in atmospheric CO2 the main driver of acidification mitigation, while the impact of OAE itself is less prominent. This is a relevant message given that ocean acidification mitigation is typically presented as a beneficial side-effect of OAE deployment.

The manuscript is mostly clearly written and the messages are relevant and worth publishing. Some comments and suggestions below:

1. The authors acknowledge in their caveats section that the CO2 emissions related to producing the alkaline material need to be taken into account in future studies. I would suggest bringing this caveat earlier in the manuscript (maybe in the setting of the simulations). Should the reader interpret that the production of alkaline material in this idealized framework does not produce further emissions?

2. Please clarify in which way a continuous global alkalinity addition “approximates” discrete OAE pulses. Please elaborate on the broad shared characteristics mentioned in L470-471. With the experiments presented, is it possible to state that the climate and carbon cycle responses will be similar with continuous global vs discrete OAE pulses?

Other comments:

L25: Ongoing mitigation efforts such as?

L25: I agree that emissions reductions remain inadequate but not despite ongoing mitigation efforts, as these two are different activities. Please consider rephrasing.

L105-106: Please clarify the difference between the observed CO2 emissions for the 1861-2005 period and the observed CO2 emissions from 2005 to 2020.

L108: is “current” the right word here? Or maybe “the warming level for the corresponding period”?

L114-116: the CO2 emissions followed for the CDRMIP protocol do not account for any CO2 produced during the production of TA. If this is correct, please state it here.

L155, Equation 2. the fluxes from which simulation were removed from which other simulation? For the same ensemble? Or from the mean? Please include this clarification. Also, are the differences estimated on a year by year basis?

L220: But considering the ensemble ranges, the peak is only truly different in the 3 deg OAE scenario compared to the REF.

L245: “opposite sign” was confusing on a first read, as the global warming pattern is not shown. Please rephrase, if possible.

L272 and 273: revise the plurals, change “signals” to “signal” and “oceans” to “ocean”

L292 and Figure 7: In Figure 7 the decline in net ocean capture efficiency seems linear and is hard to note the “before and after the peak in atm CO2”. Does it make sense to indicate with a vertical line the timing of the peak atm CO2 for each scenario? Also because this pre-peak and after peak measure is used in the following paragraphs.

L301-303: stronger stratification would be expected in a warming scenario, and therefore shallower mixed layer depths. How is it that the mixed layer depth increases 10 times by 2500?

L309: Please include a noun after “This”, maybe “efflux”? Same comment for L316.

L317: both processes? Would not calcification be considered one process?

L339: but also the CO2-driven was extended? It appears so, as the gross efficiency is shown. Please state so.

L369: “arising”? Or you could also just remove arise altogether.

L411: an end of sentence period is missing after the references.

Figure 1. If the reference simulations stabilize air temperature without OAE, what do they do?

Figure 3. The word “anomaly” is missing from the caption in (a).

Figure 6. if possible, consider including the frame of reference in the caption (e.g., positive means loss from the atmosphere) in the caption.

Grosselindemann et al. employ a fully coupled Earth System Model to simulate OAE scenarios from 2026 to 2500 under different warming scenarios (1.5–3 °C), including a baseline simulation, OAE perturbation experiments, and additional decoupling experiments. Following the OAE protocol defined by CDRMIP, the study systematically evaluates carbon sink flux changes under OAE, particularly focusing on the responses of the ocean, and especially the atmosphere and land carbon reservoirs.

The manuscript highlights several key findings:

1. OAE efficiency depends strongly on the evaluation metric. Whether ocean and land feedbacks induced by reductions in atmospheric CO2 are considered leads to substantial differences in estimated OAE efficiency.

2. Under higher warming scenarios, the OAE-induced reduction in atmospheric CO2 is larger.

3. OAE leads to a linear reduction in global mean temperature compared with the control.

A major strength of the manuscript is the systematic clarification of four different efficiency metrics for OAE. The study clearly distinguishes these definitions and quantitatively demonstrates their differences. Figures 1 and 7 illustrate these distinctions particularly well. In addition, the OAE 2 °C reference simulation (Ref∗) experiment effectively separates the ocean feedback / anomalous outgassing driven by atmospheric feedback, which is crucial for quantitatively understanding how atmospheric responses after OAE can reduce overall OAE efficiency through the ocean and land feedbacks.

Overall, the manuscript is clearly written and presents a complete set of results, providing useful quantitative insights that will serve as a valuable reference for future studies on OAE. I recommend publication.

Comments:

1. Line 355: Regions where efficiency locally exceeds unity are often also regions with high gas transfer velocity (e.g., Zhou et al., 2023). This may suggest that the enhanced local gross uptake results from the combined effects of ocean circulation and locally elevated gas transfer rates.

Minor comments:

Figures 3, 4, 5, 7, 8, 9: It is recommended to add gridlines and minor ticks on the time axis to improve readability.

Figure 9c: The vertical grid line at year 2500 is too similar in color to the background. Please adjust the color to make it more visible.

Lines 468–470: The outcomes of continuous OAE deployment and pulse-based OAE deployment are not necessarily equivalent. The studies of Wang et al. (2023) and Burt et al. (2024) do not seem sufficient to support the claim that individual pulses are equivalent to continuous OAE deployment. In addition, the impacts of global-scale deployment and regional deployment of OAE may differ substantially.

Lines 470–471: It is recommended either to add a supporting reference (e.g., Tyka et al., 2025, which suggests that the atmospheric response may be relatively insensitive to OAE deployment scale) or to remove this statement.

References:

Tyka, Michael D. "Efficiency metrics for ocean alkalinity enhancements under responsive and prescribed atmospheric p CO2 conditions." Biogeosciences 22.1 (2025): 341-353.

Zhou, Xiaohui, et al. "A sea state dependent gas transfer velocity for CO2 unifying theory, model, and field data." Earth and Space Science 10.11 (2023): e2023EA003237.

Summary:

This study examines OAE’s long-term efficiency and potential to remove CO2 from the atmosphere by simulating OAE in the GFDL ESM2M model from 2026 until the year 2500 under three different warming scenarios. The simulations show that atmospheric CO2 drops by 73-130ppm until 2500 with a larger reduction contribution occurring earlier in the simulation and under higher warming scenarios, while SAT tends to decrease nearly linearly over time and independently of the background warming level. This study also goes into detail about different metrics for OAE efficiency and how the chosen metric can influence the narrative and examines the ocean acidification mitigation potential of OAE, which can be traced back primarily to the reduction in atmospheric CO2 especially over centennial time-scales. 

General comments: 

The paper is overall of high quality, well structured with high quality figures, addresses relevant scientific questions related to the mCDR method OAE with a state of the art fully-coupled Earth system model and fits the scope of BG. However, the manuscript could be improved by primarily clarifying / adding details in the Methods and Results sections as listed below.

Specific comments:

106-112:

A more detailed description of the AERA would be beneficial here. It is without being familiar with the cited paper unclear how the emission pathways were chosen by the algorithm. Why / how does the algorithm e.g., lead to significantly different time windows for reaching the target temperature? Is a maximum positive / negative CO2 emission or target year specified? 

L206:

The results section provides lots of details via numbers, which partly compromises readability and dilutes the key findings. Maybe offloading details mentioned in the text to a new table could be a solution to maintain details and allow for quicker and easier comparison between simulations, while improving text readability. 

L318-319:

In the model description the authors mention the possibility of sedimentary calcite shell deposition (L94). In combination with the statement here, it is unclear how much of the added alkalinity via OAE is actually removed from the water column via the ecosystem and enters the sediment. 

On that note, the authors mention in the Methods section that the model represents alkalinity well (L99), but don’t show a Figure of the background ocean alkalinity inventory and no comparison of how much the inventory changes as a response to the yearly alkalinity addition over time. 

L339-341:

This detail should be mentioned in the Methods section and be included in Figure1. Currently, Figure1 is inconsistent with Figures 8-10. 

Technical corrections:

L30-31:

Please add a citation to the statement or rewrite. 

L34:

“alkaline materials” -> not just materials, but also alkaline solutions, which would actually represent the model OAE implementation much better due to not modelled mineral dissolution processes in the surface ocean, but instantaneous dissolution.

L36:

‘bicarbonate and carbonate ions … stable … for 10000 to 100000 years’: This is a bit misleading formulation. First of all bicarbonate and carbonate ions are not conservative tracers, i.e. they are not even stable wrt mixing. More importantly, this quote ignores potential non-stability (or reduced durability) wrt to interactions with CaCO3 sediments and also wrt to modified terrestrial weathering. While this is often not considered in OAE studies (since they usually focus on centennial timescales) and also not by the paper under consideration, it has been studied by Köhler et al. 2020 (https://doi.org/10.3389/fclim.2020.575744) with a box model that can simulate such long time scales. I suggest to rephrase and refer to Köhler’s work here.

L43:

“CO2 sequestered” -> additional ocean carbon uptake

L52:

“decrease” -> reduce

L56/64:

‘most modelling studies’ (56) vs. ‘only a few … modelling studies’ (64), however, in both cases you cite six studies, each - I don’t see your rational for ‘most’ vs. ‘few’ here

L68:

Unclear what is meant with “single” Earth system models. Maybe “other” or “previous”?

L88:

Please change to ‘It simulates marine and terrestrial anthropogenic carbon uptake …’, to be explicit here. (I assume you refer to both marine and terrestrial uptake and storage.)

L126:

Can you please clarify how/weather non-CO2 GHG forcing was applied in the Ref* simulations.

L175:

‘to reduce atmospheric pCO2 and therefore global warming’, I find this formulation a bit awkward, since in a transient climate warming is related to cumulative emissions (i.e. the TCRE), not the atmospheric pCO2

L190-4:

I am not really convinced by the name of the term delpH-Hysteresis 

L212 / caption Fig. 3

‘air temperature anomalies’

Caption Fig. 4

alike: indicate that you show anomalies hier 

L255:

‘when global warming is stabilized’: sounds as if the warming (carbon-climate feedbacks) is the central point here, but I guess that the decrease in marine CO2 uptake (rate) is primarely carbon-concentration driven, i.e. related to having negative emissions in Ref

Fig. 5:

Interesting that in this model the land is a net source of carbon until about 2050 in OAE and Ref. Please add a comment on why this behavior occurs (c-climate vs. c-concentration feedback?) and point to a reference, if this was discussed in a previous paper. 

L272:

“signal(s)” typo

L417:

“experimental setups” -> such as carbon / temperature overshoot scenarios. 

L436:

Please add a comment if this difference is (in)-significant. 

Figure C2:

Please add temperature labels above the three columns similar as done in Figures 4-5.