Review of Subhas, Lehmann and Rickaby

Enhancing ocean alkalinity is one of the potential carbon dioxide removal pathways needed to meet the Paris maximum global warming target of +1.5 and 2.0 ^oC.  To develop a verifiable carbon removal technique based on ocean alkalinity enhancement deployment, we must adopt a multifaceted research approach comprising laboratory studies, mesocosm, field experiments and natural analogues studies. This paper focuses on the latter and complements other contributions in this dedicated State of the Planet volume on Ocean Alkalinity enhancement deployments.

The paper is topical, timely and well-written, but it is not entirely clear what the target audience is. The text is for the major part quite generic and understandable for a wide audience, but sometimes too concise for a non-expert to grasp the message or too detailed given other parts of the text. As an example of the latter, the multivariate analysis is quite generic, and it is unclear why it should be discussed in two pages A4 given that OAE specific discussions might need some more elaboration (see below).

Another major point of attention is that the manuscript needs some more articulation of the pros and cons of natural analogs vs. field experimentation/mesocosm studies. Field experiments and mesocosm studies do offer almost similar complexity while being more conclusive (manipulation experiments). However, these studies do not provide information on long-term feedbacks acting over larger space scale, neglect the adaptation of communities and organisms to enhanced alkalinity, and might be biased because of missing events (resuspension-deposition following storms, etc).  The iron fertilization and ocean acidification literature could serve as inspiration.

Lines 21-25: These lines could be read as if alkalinity is primarily linked to carbon dioxide cycling and carbonate mineral dissolution and precipitation, and that other processes such as anaerobic mineralization are only locally of importance. There are multiple processes impacting alkalinity and for the presumed wide audience I would have expected at least something about the role of primary production/organic matter mineralization balance (after all we correct measured TA before we can use it to infer carbonate mineral dynamics) and that aerobic and anaerobic mineralization have different effects on alkalinity.

Line 26:  The statement 60 Tmol alk y-1 needs a reference.

Line 27: I guess it should be Middelburg et al. 2020 not Middleburg et al. 2022.

Figure 2 and lines 92-109 (text on mineral dissolution and precipitation). The authors propose that mineral dissolution is occurring at smaller spatial scales than precipitation. Both can be studied at the micron-to-millimeter scales to regional/global scale, depending on the approach (experimental in the laboratory or computational chemistry for the small scale and budgeting for the ecosystem to global scale).  The presumed difference in spatial scale is unclear and needs further documentation. The alkalinity increase in ocean interior is primarily due to global scale calcite/aragonite dissolution.

River plumes are discussed as a potential natural analog. It would be instructive if the authors would communicate to the non-specialist audience that rivers often have similar DIC and alk concentration. Many freshwater chemists consider bicarbonate=DIC=alk. Rivers deliver alkalinity to the ocean, but in a 1 to 1 ratio with DIC.

Lines 110-115: There is an extensive literature on how dredging and dumping of mud can modify ecosystem functioning.

Line 140: Air-sea CO2 exchange.  One factor that might be mentioned here is the sensitivity of the seawater CO2 to alkalinity addition. This can be explicitly calculated, and it might be useful to track in natural analogs.

Section 2.2. natural analogs.

When reading this list, I wondered why the Mediterranean or Red Sea with their natural high alkalinity are not mentioned. The combination Red Sea, Mediterranean, Black, Baltic and Atlantic ocean might perhaps be useful to tear apart the salinity-alkalinity co-variance.

Some other thoughts: alkaline seeps are less common than CO2-rich seeps, but they do occur. Solar evaporation ponds with or without carbonate in the background might offer an excellent natural analog because at an evaporation stage of 2-2.5 times seawater one would expect a factor 2-3 difference in alkalinity between systems with or without aragonite (see seminal Laskar paper in L&O of early 80-ies).

Lines 221-242 The section on geological targets to study OAE is largely without references (or is there a reference limit to this publication outlet?).

Lines 256-285 The conservative vs non-conservative estuarine mixing text needs some revision. Why is equation 1 presented explicitly? Is it really needed? The lines 280-285 could better articulate that (only) net removal or additions can be traced by non-conservative mixing.

For the non-specialist, it might be needed to explain here that DIC and alkalinity are conservative upon changes in mixing, T, P changes, while the species contribution to it, are not. (It is mentioned later, but not elaborated enough for a marine biologist, I guess).

Line 296: the abbreviation MRV appears without an introduction. Moreover, the Rau et al. paper could be useful to cite here.

Line 305: Ho et al. is not in reference list

Line 335: it would be useful to explain that satellites can track whitenings, cocco blooms, salinity etc, but alkalinity only if proxied by salinity.

Line 366: perhaps reformulate or add:… may not be informative to take daily samples without taking tides into account.

Line 389: Alkalinity often co-varies with salinity, and this is probably the most challenging confounding factor to account for. Co-variance with temperature, nutrients etc will be far less. It would be useful to inform the reader about this well-known co-variance. (Salinity is often used as proxy for alkalinity in global mapping studies).

Section 3.6 is rather long for its relevance for this paper. Yet machine learning approaches that better deal with non-linearities are not discussed.

Section 3.7. Many regional and global models resolving alkalinity dynamics lack sediment biogeochemical processes, process description for precipitation and dissolution and are limited to calcite. A few resolve aragonite as well, but high Mg calcite is not, which may be suboptimal when studying high alkalinity seas.

Jack Middelburg, June 26, 2023

Review of “Natural Analogs to Ocean Alkalinity Enhancement” by A. V. Subhas, N. Lehmann, and R. E. M. Rickaby

The authors provide an thorough overview of natural OAE analogs and how researchers could use these to study OAE effectiveness at ocean CO2 uptake, the impacts of OAE on ecosystems, mixing and spreading of alkalinity additions, and the interaction of these and other processes. They identify some examples of natural OAE analogs and discuss the benefits and challenges of using such systems to answer OAE questions. There are a variety of scales at which natural analogs can be found, from small scale river plumes or glacial fjords to large scale (and long timescale) processes such as changing carbonate compensation depth in the ocean. They provide an idealized example of a river with high alkalinity mixing with ocean water and explain how we can identify when more processes that affect alkalinity are at work than just mixing. The chapter further discusses how to determine if a location is a good candidate of a natural OAE study and proposes a framework for observing/measuring ocean variables and analyzing complex data to isolate the effects of changing alkalinity in a natural analog. The authors also touch on modeling efforts and highlight the utility of using models along with observations of natural systems.

This chapter is highly informative, well written, and I learned a lot reading it. It provides a high level summary of natural OAE processes and how we might use these to inform OAE research questions. Overall, I enjoyed reading the chapter, but I think there are few modifications the authors could make to make it even more enjoyable and understandable to a broad audience of researchers.

The biggest issue to me is that there a several places in the chapter where the authors touch on something that seems like it could be interesting and relevant but then don’t provide enough background information for the reader to fully appreciate the concept. I realize that the chapter provides a high-level overview but sometimes mentioning something and then not providing enough detail can be a bit frustrating for the reader. Here are a few examples:

• Line ~175: What are whiting events and what do we know about them? Is the suspended CaCO3 coming from chemical precipitation or is the suspended CaCO3 delivered somehow? How long do they last and how frequently do they occur? Whiting events seem very intriguing as a OAE analog, but they need more of a background introduction. Perhaps a lot is unknown but that needs to stated explicitly.
• Line ~180: What happens with seafloor weathering of basalts? Is there an example location of this and how does it change the water chemistry?
• Line ~200: Figure 3 needs a more in-depth explanation to be useful to the reader. The 3-point list on lines 198 to 204 is helpful but perhaps reference the figure in this list. E.g., explain the difference between the tan and blue lines in each panel.
• Line ~209: It would be helpful to actually describe what happened during the PETM and how this affected dissolved alkalinity in the ocean. Just another sentence or two would be enough.
• Line ~235: What happened to biomineralizers during the Permo-Triassic ? Please describe more in detail.
• Line ~450: the Gomez study needs to be more thoroughly described.. Please use a few sentences to say the purpose of their study and how the Mississippi river is bringing more alkalinity, and what they found.

All of the above points are interesting but need just a bit more explanation to be intriguing to the reader.

Minor comments:

Line 20: the authors say “at scale” several times in this section but it’s not clear what this means. Do you mean at the scales meaningful for OAE deployments? Natural analogs vary in temporal and spatial scales so it’s just not clear..

Line 28/29: define alk. it’s obvious but should be defined before using it… or just write out alkalinity to be consistent with most of the rest of the chapter.. TA is also used quite a bit later in the chapter. Just be consistent with these abbreviations..

Line 58/59: run on sentence. Perhaps add a semicolon after “limited”

Figure 2: the x-axis needs a label (spatial coverage?)

Line 81: Remove the word “always”.

Line 87: Add “of OAE” after “natural analogs” just to remind the reader what you’re talking about.

Line 93: replace “should be” with “are”

Line 203: add “natural” before “environment”

Line 134: replace “are” with “is”

Line 140: what is meant by “surface expression”? the surface area?

Line 146: biological or chemical CaCO3 precipitation? or both?

Lines 227 to 231: This sentence is really long and hard to follow. Please break it up and simplify or expand to make the subject matter understandable. So many subjects (e.g., snowball earth or carbon perturbations during the Mesozoic) are brought up that need more explanation to be meaningful.

Line 242: Could you describe a bit more about the CCD deepening after the LGM with the regrowth of forests? Just needs a bit more to be intriguing to the reader who might not have a deep background in paleo research.

Line 245: remove the word “here”

Line 256: concentrations of what? Alkalinity?

There are two “Figure 4”s

The first Figure 4: Please define the red and blue squares in the figure caption. I realize that this is described in the text but it should also be in the caption.

Line 265-275: Define “f_river” and “f_ocean”… at first I thought it meant fluxes but it makes more sense as fractions… so please describe better.

The second Figure 4: I found this flowchart a bit confusing.. I’m just wondering what the appropriate answers need to be to get to “Suitable Natural Analog”…  So if the answer is “no” (or “none”) for some of these questions, then it’s not a suitable analog? To me, it just seems like a list of relevant questions to ask before selecting a OAE analog…

Line 369: by “unmixed” do you mean differentiated?

Line 429: I have no idea what homoscedasticity means.. perhaps explain in simpler language?

Line 438: Reword using only the part in parentheses: “Given this higher flexibility, one of the drawbacks of GAMs is that non-linear features are potentially less intuitive and more complicated to interpret.”

Line 474: Is there a citation you could add for a reference that demonstrates that models cannot capture the adaptive response of phytoplankton to long term high-TA exposure?

A final question for the authors: A variety of laboratory studies have shown that pelagic calcifiers, such as coccolithophores, calcify less as the ocean becomes more acidic. If natural coccolithophore populations start calcifying less due to ongoing ocean acidification, is this considered a natural OAE analog (since they start leaving more alkalinity in the surface waters since they calcify less)?