the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Apr 2025
Interactions between ocean alkalinity enhancement and phytoplankton in an Earth System Model
Abstract. Ocean alkalinity enhancement (OAE) as a CO2 removal strategy is well investigated in model studies, but risks for the ecosystem are presently not considered in models. Our study examines OAE-phytoplankton feedbacks in an Earth System Model by adding carbonate system dependencies to the phytoplankton growth term. OAE is performed between 2040 and 2100 in the exclusive economic zones of Europe, the US, and China, with alkalinity additions reaching 103.2 Tmol year−1 by the end of the century. Atmospheric pCO2 is reduced by 3–8 µatm. The excess ocean CO2 sink is mainly chemically driven, but can additionally be altered by biological feedbacks. Further, net primary production decreases by up to 15 % due to indirect effects of OAE. Our results do not confirm the direct positive effect of OAE on calcifying coccolithophores. Limiting alkalinity addition in locations with high aragonite saturation states is beneficial as it not only reduces the OAE impact on phytoplankton but also increases the reduction in atmospheric pCO2. Our study highlights the need to take ecosystem responses into account when evaluating the effectiveness of OAE.
- Preprint
(7085 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-1495', Lennart Bach, 12 May 2025
Seifert et al. study the response of plankton communities to OAE in an Earth System Model. The study is very interesting and I enjoyed reviewing. I have some minor/moderate comments that may help the authors to further improve their study.
Line 1: Have been investigated for what aspects? Efficiency?
Line 28: Consider double-checking this number. I think in IPCC 2013 the number is 0.25, which I think is based on global alkalinity discharge into the ocean via rivers as in Amiotte-Suchet et al 2003.
Line 37: I think there can still be some doubt on the efficiency of OAE, since models lack many of the relevant efficiency-modulating processes.
Is equation 1 showing GPP as a biomass concentration change per time or a division rate? If a division rate then it would be somewhat different to what is generally considered GPP.
Line 119: “both parameterizations…” It took me quite a while to understand what “both” refers to. Consider providing a clearer description (some degree of repetition seems justifiable).
Fig. 1: Why can the CO2 factor range between 0 and 3 but Fig. 1a suggests ~1.2 is the maximum?
Fig. 1: Is the alkalinity on the x-axis “equilibrated” with atmospheric CO2 or “unequilibrated”? I would assume that stronger effects occur in the unequilibrated case. I think this needs to be clarified in the caption.
Line 144: How can a t-test be done without any replicates? The statistical description probably needs more clarity.
The two paragraphs starting line 191 on OAE efficiency. The trends described here refer to where the air-sea CO2 influx occurred, whether in an EEZ of a deployment region or outside. I find this discussion somewhat artificial. Why does it matter where it occurs and how the efficiency is in a certain region? What matters is what happens over the entire affected area. The authors may consider deleting/condensing this section. These paragraphs also contain some discussion and not only results. Some readers may not like this (doesn’t bother me but some really don’t like interpretations in the results).
Table 2 (see above comment). Also, does the relatively high global efficiency of e.g. 0.72 include efficiency losses that would presumably come through OAE-induced reductions in the land-CO2 sink (or other Earth system feedbacks)? I would find this aspect more interesting than the text on air-sea CO2 flux.
Line 211: why is the land-sink weakened by an amount that overcompensates the phytoplankton effect? I would understand a mitigation of the land-sink but why is it relatively weaker than in the scenario without phytoplankton feedbacks? Adding an explanation would be interesting here.
Line 212: Does that mean the differences in the run with the phytoplankton feedback is a spin-up artifact?
Line 239: Does “globally” include land?
Fig. 5: “carbonate chemistry artifacts” sounds as if this is due to a glitch in the model. Was it an “effect” or really something weird in the model?
Line 168: “…, rejection the hypothesis of strengthened CO2 limitation due to OAE.” CO2 limitation can be seen in plankton community studies (usually a transient effect). So, I think the model cannot reject it. Rather you could say it plays no role for NPP on a regional/global level under stustained OAE?
Line 271: It would be nice to have reminder here that a CO2 factor leads to higher growth (calcification).
Line 284: Indirect effects would be MLD, light, temp, grazing etc.? Perhaps spell out.
Line 290: The indirect effect response could be explained a bit better. It cannot be fully indirect because the origin must somehow be induced through the CSE-modifications in the model, right?
Line 325: I am not sure if it really complicates things because it is already clear that constraining air-sea CO2 fluxes for individual deployments occurs at the basin scale and depends on large-scale (global) models. I think the discussion you are having here has already happened, and came to a conclusion.
Section 4.2: I think this is an interesting finding and needs further discussion. How does the increased NPP increase CO2 uptake? Nutrient concentrations are presumably the same between the runs, right? Hence, the shift in phytoplankton community composition somehow needs to export C more effectively to depth? Have you looked at preformed nutrient fields in the surface ocean? Are there indications of a more efficient BCP?
Line 351: Table A5 suggests that PIC:POC decreases in the Chinese EEZ.
Line 350-352: I think the discussion non coccolithophore proliferation needs a slight adjustment: Instead of suggesting coccolithophores do not benefit I would rather say that they benefit, but relatively less than other groups in the model. You mention this above but I think it needs to be repeated here that a proliferation is physiologically possible but not ecologically realized. I think this also needs to be mentioned in the abstract as it is an important aspect of your study (and a very nice finding!).
Lien 355: It is worth noting that the study by Lehmann and Bach also found that the correlation between PIC:POC and carbonate chemistry only holds when looking from “a very high altitude” (i.e. averaging over very large areas) but that regionally other factors override the CSYS response. I think this is at least conceptually a plausible explanation for differences, since under a limited perturbation the CSYS effect may be lost in noise (i.e. overridden by other factors that play a stronger role).
Line 363: You reduce by half and get 50% of CDR sounds not like a surprising outcome. I think you need to say what you would expect (in a one-dimensional assumption) to make clear why this is unexpected.
Section 4.4.: Did you have secondary precipitation in the model? I don’t think so but the omega <10 discussion implies there is.
Line 370: I don’t think that the development for ocean acidification means that using the factor for OAE is a weakness. In the Bach et al study the mechanistic model was derived from treatments that are essentially “OAE”, since alkalinity was manipulated in many different ways. They just didn’t call it OAE back then, but this exact dataset was repurposed in the OAE context because it explored OAE-relevant conditions (Bach et al., 2019).
The conclusion section could be improved. It currently just seems to open up an new discussion point (see next comment). Perhaps consider flashing out the key findings and novelties and then provide recommendations, what type of data exactly is needed to improve this model.
Line 392: In the context of closer collaboration: Can community studies really inform the development of these types of models? All you usually get from these studies is the ups and downs of some plankton groups, and the explanations WHY this happens are usually not particularly robust. It would be helpful to state what exactly the modellers need. My understanding is that models are driven by rates, and these come from physiological (or simplified ecological (grazing)) studies. Can you let the reader know what the big points for improvement are?
Fig. A4: The figure implies that the model offset to GLODAP got worse upon implementation of the CSYS feedbacks. Does this need to be mentioned in the discussion or limitations section somewhere? Is that a concerning outcome, or just within the noise of general model competence?
Citation: https://doi.org/10.5194/egusphere-2025-1495-RC1 - AC1: 'Reply on RC1', Miriam Seifert, 18 Jun 2025
-
RC2: 'Comment on egusphere-2025-1495', Wentai Zhang, 27 May 2025
General summary:
The authors investigate the potential impacts of Ocean Alkalinity Enhancement (OAE) on the Earth system by incorporating carbonate chemistry dependencies into the phytoplankton growth term within an Earth system model (ESM). They perform multiple ESM simulations to assess the environmental impacts of OAE, with particular emphasis on ecosystem responses. This approach highlights the importance of accounting for biological feedback when evaluating geoengineering strategies such as OAE.
This is a timely and important contribution to the fields of climate change and geoengineering. I recommend publication of this work after the authors address the concerns outlined below.
Comments:
Line 20: Please add a reference to support the statement “Efforts…”
Line 40: I suggest deleting the phrase “, although it was identified as a major risk” to improve clarity and conciseness.
Line 45: Consider changing the word “minimal” to “little”.
Line 62: The sentence beginning with “In a modeling study, …” is unclear. Please revise to clarify it.
Line 70: Please specify which Earth system model was used in this study.
Line 210: I was unable to locate the 16% in Table 1. Please clarify where this number comes from.
Line 221: Instead of ranges or general terms, please provide the exact values here.
Line 239: The sentence “NPP anomalies globally…” is unclear. Please rephrase.
Line 247: The origin of the values “3% vs 97%, 40% vs 58%” is unclear. Please indicate their source or explain how they were derived.
Line 301: This sentence should be rewritten for clarify. Also, please specify where the associated values can be found in the text or tables.
Figure A4B and A4C: These figures show the difference between modeled alkalinity and observational data. Please revise caption accordingly.
Section 2.3: It would be helpful to include a table summarizing the key details of all simulations conducted in this study.
Figure 2,3, and 4: I recommend revising the captions to first provide an overview that describes the figure as a whole, followed by brief description for each subfigure individually. This will help readers better understand both the overall context and the specific content shown in each panel.
Please improve the abstract and conclusion section.
Citation: https://doi.org/10.5194/egusphere-2025-1495-RC2 - AC2: 'Reply on RC2', Miriam Seifert, 18 Jun 2025
Model code and software
REcoM code Miriam Seifert and Judith Hauck https://zenodo.org/records/7457987
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 219 | 89 | 18 | 326 | 9 | 20 |
- HTML: 219
- PDF: 89
- XML: 18
- Total: 326
- BibTeX: 9
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1