Dear Authors,

Thank you for your responses to your referees and for your revised manuscript draft.

I have now reviewed these materials, and while I judge that your answers address the majority of the points raised, I believe the discussion around the performance of the model’s silicon cycle could be more thorough. This aspect of the model is the focus of this study after all and, as the referees remark, the model has some clear deficiencies.

One way to address this could be to contextualise the model’s performance against comparable global-scale models, such as those produced as part of the CMIP process. Many of these models now include a silicon cycle, and the CMIP repository include a number of relevant diagnostics, including silicic acid and diatom biomass / production distributions. Models are, inevitably, always deficient, but a straightforward way to illustrate the utility of one is to show how it improves our ability to represent a real world system relative to the current state-of-the-art (or, at least, compares with this).

I would appreciate it if you could expand on this portion of your response to your referees. As both of the referees judged your manuscript to require major revisions, I will be returning it to them subsequently, but I believe more completely addressing this issue of performance (i.e. relative to other silicon cycle models) would strengthen the manuscript draft ahead of this.

Please get in contact with me should you have any questions or require any assistance about any of the above.

Thank you again for your response and revisions.

With best regards,

Andrew Yool.

Dear Authors,

As I previously indicated, further to receiving and reviewing your revised manuscript, I returned it for an additional round of referee review. While one of the original referees was available, the second was unable to help at this time and a new referee was approached.

Both referees have now returned their reports on your revised manuscript. The returning referee was satisfied with the revisions made to your manuscript in response to their original comments. However, the new referee identified several remaining significant issues, including aspects of the model description, the justification given for the temperature-dependence of opal dissolution, and some discrepancies in the modelled silicon budget.

I would appreciate if you could respond fully to these remaining issues. Should any of these be unclear, please contact me and I will seek clarification from the referee. Also, I appreciate that there may be continuing difficulties imposed by the pandemic, so please do not hesitate to contact me should you encounter any difficulty or delay in completing these revisions.

Thank you once again for your response and revisions.

With best regards,

Andrew Yool.

Dear Authors,

Thank you once again for your manuscript revisions and responses to referees.

Given the significant criticism levelled by one referee, and the specific and detailed points raised in their review, I returned the manuscript to them for an evaluation of your revisions.

In reply, the referee is broadly pleased with your revised manuscript, and finds that it addresses much of their criticism. Nonetheless, they still identify a significant lack of clarity in the description, and rationale, of the dissolution parameterisation used. As this aspect of your model is a key determinant of its operation and skill, having clarity in details of its formulation is highly important.

To this end, I am returning the manuscript to you to address these remaining points from your referee. As ever, should you have any questions about the referee’s remarks, please feel free to get in contact with me and I will endeavour to have them clarified.

Finally, please accept my apologies for this extended review process of your manuscript.

With best regards,

Andrew Yool.

Dear Authors,

Thank you for your revised manuscript and response to your referee.

I have examined both and find that most of the issues raised by the referee have been sufficiently addressed. However, I have listed below a few points where I believe that the manuscript could be made clearer:

- Regarding your statement “We can not plot the vertical distribution of sinking opal because we do not model this quantity explicitly”, it should be possible, even with implicit variables, to plot the vertical distribution of sinking opal (and/or its dissolution). It may be that you have not stored this diagnostic output, but that is different. As the description of this process is clearly complicated, if you could plot a profile of the opal flux or its dissolution, that would help readers.
- Regarding Table 6, the seafloor Si fluxes at the 3 depth-bands don't sum to the total. Since there's no overlap in the bands, this is a little surprising. What's the explanation?
- Also regarding Table 6, the KMBM3 model appears to have completely flipped the dominance of phytoplankton from LP (96.2% to 5.5%) to diazotrophs (3.8% to 81.8%), but there is only slightly more N2-fixation (+25%). Is this the correct interpretation? While your text already acknowledges that your diazotrophs can uptake nitrate rather than fix N2, this isn't as clear as it could be when the phytoplankton types are introduced (ln. 75-78).
- Ln. 82: your amendment here is still ambiguous. Could you rephrase as "lost due to burial at the seafloor"? The referee rightly questions whether all material reaching the seafloor is “lost”, or whether it is just a fraction of this.
- Ln. 118: Are you just describing Leibig's law here? That is, realised growth is a function of potential growth multiplied by the strongest limitation factor. If so, it may be worth simply noting this.
- Regarding eq. 8 and ln. 129, the amendments are OK, but would these read more straightforwardly if the numbers were expressed in mmol Si / m3 rather than mol Si / m3? Eq. 8 would then become something like ... k_si = 0.8 + 7.2 * ( [Si] / ( 30 + [Si] ) )
- Regarding your statement “As with CaCO3, opal produced by diatom losses to the bacterial loop are not considered”, perhaps it may help to add a qualification like the following: "In linking opal production to diatom mortality, our model essentially focuses on the importance of its export pathway in establishing vertical gradients. Opal that is not exported is effectively assumed to be recycled to dissolved silicate within the surface layer."

As an aside, I also noticed that your revised manuscript also contains mark-up that should really only appear in the tracked-changes version.

If any of the above points are unclear, please get in contact with me. Hopefully these should only be minor amendments.

With best regards,

Andrew Yool

Dear Authors,

Thank you for your revised manuscript and responses to the issues I raised.

I have now reviewed these and I am happy – with one caveat – that they satisfy these issues. My remaining caveat is that the new Figure 2 is not directly referenced in the text. I would appreciate it if you could add a sentence (or two) at an appropriate point in the main body to introduce and explain what it shows. It would also be useful to explain in the caption the units used in the panels as they are not immediately obvious.

Consequently, I am happy to accept the manuscript subject to this technical correction.

Going forwards, after this correction the manuscript will be passed over to our production team. I am not formally involved at this stage, but if you need any assistance, please feel free to get back in contact.

Finally, thanks once again for selecting GMD for publishing your work.

With best regards,

Andrew Yool