Modeling considerations for research on Ocean Alkalinity Enhancement (OAE)

Fennel, Katja; Long, Matthew C.; Algar, Christopher; Carter, Brendan; Keller, David; Laurent, Arnaud; Mattern, Jann Paul; Musgrave, Ruth; Oschlies, Andreas; Ostiguy, Josiane; Palter, Jamie; Whitt, Daniel B.

doi:https://doi.org/10.5194/sp-2023-10

Preprints

https://doi.org/10.5194/sp-2023-10

Preprints

28 Jun 2023

| 28 Jun 2023

Status: this preprint is currently under review for the journal SP.

Modeling considerations for research on Ocean Alkalinity Enhancement (OAE)

Katja Fennel, Matthew C. Long, Christopher Algar, Brendan Carter, David Keller, Arnaud Laurent, Jann Paul Mattern, Ruth Musgrave, Andreas Oschlies, Josiane Ostiguy, Jamie Palter, and Daniel B. Whitt

Abstract. The deliberate increase of ocean alkalinity (referred to as Ocean Alkalinity Enhancement or OAE) has been proposed as a method for removing CO₂ from the atmosphere. Before OAE can be implemented safely, efficiently, and at scale several research questions have to be addressed including: 1) which alkaline feedstocks are best suited and in what doses can they be added safely, 2) how can net carbon uptake be measured and verified, and 3) what are the potential ecosystem impacts. These research questions cannot be addressed by direct observation alone but will require skillful and fit-for-purpose models. This chapter provides an overview of the most relevant modeling tools, including turbulence-, regional- and global-scale biogeochemical models, and techniques including approaches for model validation, data assimilation, and uncertainty estimation. Typical biogeochemical model assumptions and their limitations are discussed in the context of OAE research, which leads to an identification of further development needs to make models more applicable to OAE research questions. A description of typical steps in model validation is followed by proposed minimum criteria for what constitutes a model that is fit for its intended purpose. After providing an overview of approaches for sound integration of models and observations via data assimilation, the application of Observing System Simulation Experiments (OSSEs) for observing system design is described within the context of OAE research. Criteria for model validation and intercomparison studies are presented. The article concludes with a summary of recommendations and potential pitfalls to be avoided.

Received: 23 Jun 2023 – Discussion started: 28 Jun 2023

Download & links

Katja Fennel et al.

Status: final response (author comments only)

RC1:
'Comment on sp-2023-10', Anonymous Referee #1, 14 Aug 2023
The manuscript provides a comprehensive review and discussion point about how the biogeochemical coupled hydrodynamic could; be used in an Ocean Alkalinization experiment s.
The manuscript lays out all the processes and questions any project about OAE should explore. It provides a Guide to Best Practices.
I found the manuscript well written, with only a few points of discussion that could be added, listed below.
A few dotted points related to the manuscript:
Any OAE experiment will require the modelling exercise to be complete as soon as the experiment is underway (near real-time) or, even better, in a forecast mode. There are many operational systems available at the global scale but few at the scale of OAE (regional to sub-regional). This is an essential point as the availability of forcing data initial conditions are crucial elements to any modelling system.

In Section 2.2.3. I would add the resuspension due to waves as a significant process in representing the sediment-water exchange. While it only applies to particles, resuspension can be crucial in enhancing carbonate particle dissolution. (Eyre, B. D., Cyronak, T., Drupp, P., De Carlo, E. H., Sachs, J., & Andersson, A. J. (2018). Coral reefs will transition to net dissolving before end of century. Science, 359(6378), 908–911. https://doi.org/10.1126/science.aao1118)

In section 2.3, I would add a point about the development of unstructured model mesh. Using unstructured mesh allows increasing resolution in a specific area while retaining the ability to have a lower resolution elsewhere to capture larger-scale processes. It can act as an alternative strategy to multiple model nests.

The discussion about the use of DA in OAE could be simplified. While there is room to develop news DA technic to assist OAE, I don’t see DA as a major player in OAE.

On the other hand, the author could discuss the use of DA in the optimisation phase of any OAE, for example, some of the DA machinery could be used to optimise the amount of Alkalinity being released to maximise the impact area.
Citation: https://doi.org/10.5194/sp-2023-10-RC1
CC1:
'Comment on sp-2023-10', Veronica Tamsitt, 22 Aug 2023

Thank you to the authors for a detailed and thorough review of the current state and challenges for modeling for OAE research.

I have a nuanced comment/question about the statement in paragraph lines 728-735 and line 927 that 'assimilation of carbonate system parameters is not appropriate when models are applied for MRV.' I agree that care has to be taken with this, and that when air-sea CO2 flux is quantified by calculating the difference between two simulations then it is not appropriate to assimilate biogeochemical observations into only one of the pair as they are no longer comparable. However, I think there could be cases in which assimilation of BGC observations is possible. The signal-to-noise ratio is such that we expect that changes to carbonate system variables due to OAE will not detectable in observations except right at/near the alkalinity addition site. In a regional model designed to quantify air-sea CO2 flux on a broader scale (not for a near scale model close to the alkalinity addition site), then couldn't the same carbonate system observations be assimilated for both experiments, if the OAE signal is not detectable in these observations? For example, if a 1 year long state estimate using an adjoint method could be run, starting with running the model forward for 1 year, then running the adjoint to calculate the cost function (observation-model misfit) and adjust the controls (e.g. initial conditions and atmospheric forcing). Then using the adjusted controls, two free-running 1-year long forward experiments could be run in parallel, one with the OAE intervention, and the other counterfactual without the intervention. The experiments would have identical initial conditions and forcing, the only difference would be the OAE addition, and the intervention happens all within one assimilation window. It seems that assimilation of BGC observations might be possible in this case.

Citation: https://doi.org/10.5194/sp-2023-10-CC1
- AC1:
  'Reply on CC1', Katja Fennel, 22 Aug 2023
  
  Thanks so much for you comment!
  Agree that the case as described would be reasonable. Since the simulations with and without OAE intervention are two free-running simulations, they are not subject to updates due to DA. We don’t see a contradiction to our statement.
  
  Citation: https://doi.org/10.5194/sp-2023-10-AC1
  - CC2: 'Reply on AC1', Veronica Tamsitt, 23 Aug 2023
    
    Hi Katja,
    Thanks for your response and clarification, it sounds like we are in agreement.
    Veronica
    
    Citation: https://doi.org/10.5194/sp-2023-10-CC2
  - RC2: 'Reply on AC1', Anonymous Referee #2, 01 Sep 2023
    
    I reviewed the original version of this manuscript. I found the paper to be easy to read and given that the paper is a perspective, there was little technical material to evaluate. I have a few suggestions that the authors could consider if they lightly revise the manuscript.
    (1) On line 66, the authors speak to observations that "contain sufficient meaningful information". This seems like a vague and undefined statement. Could you offer some specific examples that help tell the reader what you consider to be "meaningful"?
    (2) Beginning on Line 122: Here, the authors make a seemingly authoritative statement that the direct impacts of OAE on the carbonate system is greatest in the nearfield early in the OAE experiment. Sure, this seems intuitive, but perhaps you could cite literature that supports this statement. Otherwise you could alter the text to reflected that this is an assumption.
    (3) Paragraph on line 418: It might be worth mentioning here that many models in estuaries represent these processes, so I don't think the gap is as big as stated, at least for regional models
    (4) Line 669-670: It may be worth pointing out here that one might specifically design an observational program that can fully validate the wide-range of impacts of the particular OAE being applied. Perhaps that is implied, but given that different OAE approaches may have different impacts (metals, injections of particulate material), one can tailor their measurements to track the specific impacts of the specific OAE. If CO2 drawdown was all that you cared about, chemical measures of the carbonate system would suffice, but if ecosystem effects were important, you might measure everything the authors describe previously.
    
    Citation: https://doi.org/10.5194/sp-2023-10-RC2
RC3:
'Comment on sp-2023-10', Steve Rackley, 17 Sep 2023
I feel honoured to be asked to review the work of this distinguished group, and I hope that some at least of the following comments and suggestions—coming as they do from a non-oceanographer—will be helpful.
Line 51: I would argue that MRV is primarily a deployment rather than a research challenge, and that “skillful and fit-for-purpose models” will be essential in meeting this challenge, rather than merely “valuable”.

Line 66: It would be helpful to expand on what is “sufficient”, or how one would go about establishing that.

Line 77: Is the conflation of “scenarios” and “counterfactuals” standard nomenclature in this field? I am used to a different definition of scenarios, as not being limited to counterfactual cases. Where do simple sensitivities—essential in uncertainty analysis—fit into these four general types?

Line 84-85: “especially when considering” seems superfluous, as successful model implementation will be even more of a challenge for eventual OAE deployment. Also, modelling of OAE field trials involving small alkalinity additions may still involve large spatial and temporal scales, depending on local circulation and given the long gas exchange timescale.

Line 115: Suggest to replace “mCDR” with “CDR”, since terrestrial methods will also affect Earth system feedbacks.

Line 176: Might be useful to mention Lagrangian methods as also being relevant to modelling the physical dynamics of particles.

Lines 246: “significant” is superfluous here, as even small-scale field trials are subject to the slow air-sea gas exchange. See note 4.

Line 248: I think it is incorrect to state that transportation distance is not relevant in the case of alkalinity “added to seawater that is oversaturated in CO₂”. Such waters can also be transported large distances before equilibration (same gas exchange timescale applies!), and the seawater conditions at the point of equilibration (not at the point of alkalinity addition) are what determine OAE efficiency.

Line 295: “wastewater treatment plants” could also be mentioned in lines 290-293 in relation to reactive mineral addition (Planetary's approach).

Line 313: “precludes” is perhaps too strong. Regional models may be sufficient if they fully contain the area of ocean circulation prior to subduction.

Lines 364-367: I do not think that local variations in relation to carbonate chemistry equilibrium coefficients impact eventual OAE efficiency. The relevant equilibrium coefficients are those at the point of air-sea gas equilibration or, more strictly, at the point of eventual subduction of the DIC enhanced waters, in the case that there are significant changes in seawater conditions between equilibration and subduction.

Line 383: Challenging indeed, but perhaps not required for deployment modelling since, as suggested in line 542, there are likely reduced-complexity approaches that will be sufficient.

Line 463: I guess this should be “and decreasing pH”?

Line 496: Does this really stand out as “critical” among the many areas needing improvement?

Lines 535-536: The concern regarding prescribed atmospheric pCO₂ is really only relevant in the case of conceptual global studies. For actual OAE deployments, hindcast modelling with the actual pCO₂ history will be required for the quantification of removals.

Lines 575-577: Re. “ultra-high-resolution modeling tools” if might be of interest to note (as a “personal communication”?) that Planetary is developing a Lagrangian particle tracking approach to determine the “alkalinity release field” generated by the advection, sinking, and dissolution of reactive mineral particles with a generic size distribution, which will then be imposed on a regional scale Eulerian model.

Lines 580-581: “The mean state …”. See points 4 and 11 above.

Lines 650-652: “different points …” seems a bit too indeterminate, given that some points in space and time are far more relevant than others! (points 4 and 11 again).

Line 910: While the pCO₂-pH pair results in higher parameter uncertainties, I don’t think it is correct to characterize this as a “high” uncertainty in the overall context. (See e.g. CarbonPlan’s Verification Framework, which characterizes carbonate system uncertainty as Low (1 to 5% impact) https://carbonplan.org/research/cdr-verification/ocean-alkalinity-enhancement-mineral Component 2; Mineral dissolution).

Line 929: I think ensemble-based methodologies, in general, provide a useful framework for uncertainty analysis, whether incorporating DA or not.

Typographical corrections
Line 67: An array “is”, rather than “are”. Alternatively, “Many methods … are …”

Line 104: Replace “.” by “,” after “2.2”.

Line 314: Suggest to replace “alone and” with “alone, which”.

Line 414: Replace “104” with “10⁴”.

Figure 4: Bottom right text, “desirable”.

Line 614: “pCO₂” has been used multiple times already and does not need to be defined here.

Line 713: Replace “properties. But” with “properties, but”

Line 906: Replace “an” with “and”.

Line 932: Replace “to made” with “to be made”.
Citation: https://doi.org/10.5194/sp-2023-10-RC3

Katja Fennel et al.

Viewed

Total article views: 684 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
425	240	19	684	10	9

HTML: 425
PDF: 240
XML: 19
Total: 684
BibTeX: 10
EndNote: 9

Views and downloads (calculated since 28 Jun 2023)

Month	HTML	PDF	XML	Total
Jun 2023	16	5	3	24
Jul 2023	185	118	0	303
Aug 2023	132	60	11	203
Sep 2023	92	57	5	154

Cumulative views and downloads (calculated since 28 Jun 2023)

Month	HTML	PDF	XML	Total
Jun 2023	16	5	3	24
Jul 2023	185	118	0	303
Aug 2023	132	60	11	203
Sep 2023	92	57	5	154

Viewed (geographical distribution)

Total article views: 671 (including HTML, PDF, and XML) Thereof 671 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 29 Sep 2023

Short summary

This chapter describes biogeochemical models and modeling techniques for applications related to Ocean Alkalinity Enhancement (OAE) research. Many of the most pressing OAE-related research questions cannot be addressed by observation along and will require a combination of skillful models and observations. We present illustrative examples with references to further information, describe limitations, caveats, and future research needs, and provide practical recommendations.


Total:	0
HTML:	0
PDF:	0
XML:	0