Review of “CANOPS-GRB v1.0: a new Earth 1 system model for simulating the evolution of ocean-atmosphere chemistry over geologic timescales”

This paper describes a new version of the CANOPS Earth system model, CANOPS-GRB (Global Redox Budget). The main improvements of this model over previous versions include dynamic atmospheric O₂, sedimentary reservoirs that evolve self-consistently with the atmosphere-ocean system, and a simplified global redox budget. Collectively, these improvements make the model suitable for investigating long timescale geochemical evolution, but with the benefits of resolving ocean transport and sediment diagenesis. Model outputs are validated against the observed modern Earth steady state, and example time-evolution calculations are also presented to illustrate feedbacks on multiple timescales.

I will preface this review by saying evaluating the validity of model assumptions in the abstract is challenging; the appropriateness of the chosen functions, parameterizations, and simplifications are strongly context dependent. However, I have no doubt that this manuscript and open-source model it describes will be incredibly valuable to the biogeochemistry community. The authors ought to be commended for such a thorough description of model features – this will be invaluable for trying to understand future applications of the model. I don’t have any major criticisms – the primary limitations of the model are largely already identified and discussed in the manuscript. However, I do have several minor suggestions for improvement:

The Introduction motivates the CANOPS-GRB model through a discussion of terrestrial exoplanets and exoplanet biosignatures such as oxygen and methane. Moreover, the abstract mentions exoplanet biogeochemistry as one potential application of CANOPS-GRB. It would be helpful if the discussion also had a paragraph or two on both the potential and limitations of the current model for terrestrial exoplanets. For example, by adopting the Claire et al. photochemical parameterization and by omitting a climate model, the ability to investigate O2-CH4 biosignatures is limited to Earth-twins around a sun-like stars (see also line 351). Obviously, these parameterizations could be modified to accommodate different stellar types etc., but I think it is helpful to make the current limitations of the model clear to the exoplanet community.

Line 229-230: I appreciate that the omission of Fe²⁺ is discussed later in the paper, but it might be worth mentioning this here (e.g. see discussion of iron species below) since upon first reading I immediately wondered why the iron budget was being ignored.

Line 291-292: Neglecting the inorganic carbon cycle is an interesting choice that merits more discussion. To be clear, I understand the reasoning behind this simplification – carbon cycle feedbacks are uncertain and introduce more unconstrained complexity to an already complex model. However, changes in inorganic carbon cycling have been proposed as drivers of atmospheric oxygenation (e.g. Williams et al. 2019; Nature Communications), and the 13C record provides a useful sanity check on any proposed oxygenation story. Neglecting the inorganic carbon cycle also means climate feedbacks are absent, which prevents the model from exploring the coevolution of life and the environment in many deep time contexts. More discussion of these limitations would be helpful in preventing misapplications or misunderstandings of the model.

Line 329: 14.3 Tmol/yr total organic carbon burial feels a bit on the high side. For example, see the compilation of estimates in Table 1 of Kipp et al. (2020; Global Biogeochemical Cycles). Similarly, organic weathering in CANOPS-GRB is about double most literature estimates. The discrepancy is noted on line 1368, but I am left wondering why such a high organic carbon flux is required. Given isotope mass balance, a high organic burial flux would also imply quite a high total carbon burial flux, which potentially conflicts with other literature estimates.

Equation 2: This is another place where it might be worth mentioning that climate feedbacks on silicate weathering are an important omission that could have consequences for global redox.

Line 458: Although this is stated later in the manuscript, this discussion of sulfur oxidation might be a good place to mention that the model in its current form is not suitable for modeling the Archean atmosphere-ocean system.

Section 2.4.2: I appreciate that the vertical and horizontal transport parameterizations developed here are carefully validated later in the paper. However, one unanswered question I had is how well do the authors expect these parameterizations to work when they are applied to radically different surface climates in deep time (e.g. hothouse Earth, snowball Earth)? Some discussion of this might be helpful for readers considering applying the model to different times in Earth’s history (or indeed different exoplanet climates, continental configurations etc.)

Equation 108 and 109: It’s possible I’m misunderstanding a sign convention here, but it’s not clear to me why H escape represents a loss of oxygen mass from the atmosphere system. The signs in equations (13) and (14) seem more intuitive. A word of explanation would be helpful.

Finally, somewhere in the paper (perhaps in the introduction) it would be helpful to provide a brief history of the CANOPS model and its previous applications. This would help give the reader a better appreciation of the recent GRB improvements, as well as how the improved model might be applied.

This carefully prepared paper describes an Earth System model of intermediate complexity that captures Earth’s C, N, P, O, and S cycles. The model is well-described and two relevant applications are presented to demonstrate the capabilities of the model. The model is particularly useful for simulations on long (geological) time scales. The model code is made available allowing the work to be reproduced (but see additional comment below).  

I read this paper with pleasure and have relatively few comments. Most of the detailed comments below are suggestions to rephrase the text to improve the clarity. The manuscript is quite long but because it is well-written, I don’t think that is a problem.

Detailed comments

1. Lines 25 and 1650. The authors write that “The model source code is available on GitHub and represents a unique community tool” and later “The CANOPS-GRB code is provided freely but with the requirement that prospective users contact the corresponding author with their research plans….”. I don’t think the latter requirement is appropriate. Other scientists should be free to use the model without the author’s consent or knowledge. That’s open science.
2. Line 151. “preferred” instead of “required” since this will depend on the application.
3. Line 173. Suggested change: “of the Earth system”
4. Line 258: “to describe the overall design of the biogeochemical cycles”
5. Figure 2. caption needs editing. Line 267: “anaerobic” instead of “anoxygenic”
6. Line 278 and 283: “are transformed each other” needs rephrasing.
7. Line 312: change to “the water column”
8. Line 318. “2x60 sediment segments.” Can you explain why this is two times 60?
9. Table 2: Better to use the term “Fe bound P” since a lot of the P is not sorbed and instead is more strongly bound.
10. Table 2: typo in “simulated” and “denitrification”
11. Line 344. AOM is more commonly referred to as “anaerobic oxidation of methane”
12. Line 345. “as a CH4 degassing flux”
13. Line 399-401. There is more work that is relevant in this context. You could cite Algeo and Ingall (2007) since they clearly showed the role of hydrogen sulfide in P retention and/or Papadomanolaki et al. (2022; Science Advances) who show a potential role for acidification and warming.
14. Line 544: change to “papers”
15. Line 797. Here, methanogenesis is termed “organic matter production from carbon dioxide” whereas this is a degradation process. Needs to be reformulated.
16. Line 802: remove “most”
17. Line 803: You write: “oxidants for organic matter decomposition change gradually, depending on the amount of each oxidant”. Gradually with time? Or depth? Any change will also depend on the supply of reductants.
18. Line 804: “on previous studies of early…”
19. Line 817: change to “in the geological past”
20. Line 821: “oxidation to” not “oxidation by”
21. Equation (32) remove the dot.
22. Line 831: change to “using a bimolecular..”
23. Line 834: by definition, the suboxic layer does not contain hydrogen sulfide so it would be better to reformulate.
24. Line 844 Sentence needs reformulation for clarity
25. Line 858. Change to “are listed in”
26. Figure 6: the line corresponding to “This study (oxic)” is not visible in the figure.
27. Line 925. Here the authors write “such as” but it appears that these are the only terminal electron acceptors considered.
28. Line 960 See earlier comment on “Fe-sorbed P”
29. Line 1011-. Better to write “sediments overlain by oxic bottom waters”. This will only hold for sediments with a relatively low input of organic matter. Otherwise the Fe-bound P will be lost, as observed in many modern coastal sediments.
30. Line 1021-1025. See earlier comment.
31. Line 1270 change to “until reaching steady state”
32. Line 1291 change to “LD and HD regions”
33. Line 1380-1381. It’s not clear what is meant by “(Gray dots represent the unknown dissolved O2 concentrations)”.
34. Figure 13b and lines 1404 and further. The authors focus on the benthic flux of P here but this gives a very large spread – as is commonly the case with benthic fluxes. No attempt was made to compile P burial fluxes from the literature. This needs some justification in the text, i.e. were there insufficient data to compare to or was there another reason?
35. Lines 1450. Here you mention the role of the coastal ocean for N cycling. This is also relevant for the P cycle, since most P burial takes place in the coastal ocean. This could be mentioned somewhere in the text.
36. Line 1508. “….leads to a lower”
37. Line 1511. “leads”