The mansucript attempts to project global changes in carbonate production using a habitat suitability model coupled with a climate model. The methods regarding how carbonate production is calculated, and the exact methods forcing changes in net carbonate production under climate change need to be made explicit here in the methods. Currently, the reader must go through a number of other papers to determine why the results of this mansucript have played out how they have and what the authors have actually done. I also found the writing to not be very direct. Clarifying points that are trying to be made, and explaining clearly why, would make this mansucript more accessible to a general reader. This mansucript needs to be revised before its suitably could be assessed, but likely would provide the reader a useful alternative to other existing models of changes in carbonate production under climate change.  

I give specific comments below that hopefully will assist the authors:

Line 39 onwards: What does the plus/minus indicate? Standard error, range, standard deviation? And is this the variability globally spatially or variability in any one location depending on the model outcomes?

Line 53: The role of carbonate ions in seawater has largely been disproven, see Comeau et al. (Comeau et al. 2018) and the various opinion/discussion papers, e.g., Cyronak et al (2015), Jokiel (2013).  

Line 69: And some used pH instead of saturation state, which is likely more appropriate for most calcifying taxa (including corals).

Line 84: Please define GCM if it has not already been done so. But, does it need an acronym?

Line 86: Some intro to this model is required for the reader.

Line 120: There needs to be some more details here and further on regarding how this model is working for this to be a stand alone paper. Citing the previous paper and not including any details here means the reader must go through the Bouttes et al 2024 paper with a  fine tooth comb to understand some very important aspects of this paper (probably the two most important parts of the methods) 1) how carbonate production is calculated and what controls it within the model, and 2) how temperature and ocean acidification impact carbonate production in this model. If space is an issue, remove the previous text that describes components of the model that are not as important for understanding the results here please.

Paragraph around line 140: Please in clear language explain to the reader why this part of the methods is important? How does ESC matter in the context of coral reef carbonate production over absolute changes in temperature and pH? There is more details on this than on how carbonate production is estimated.

Line 173: Its difficult to determine the speed and extent to which corals will gain increased tolerance to higher temperatures. However, this method is just as good as those previously used.

Paragraph around line 200: Other than using the NOAA guidelines for bleaching, and possible reef habitability, how do these projections of coral presence/absence actually function? A grid either has coral carbonate production at full value or 0 if the grid is habitable? Again, not enough detail on what are the most important aspects of the model outputs for this manuscript here.

Line 330: Is global mean net erosion that same as complete cessation? If some locations still have positive net carbonate production, then perhaps this statement is misleading, as to me this means all locations stop producing carbonate.

Line 355: No model can project changes in in situ temperature on specific reefs at enough resolution anyway, so perhaps this does not matter.

References used here:

Comeau, S., C. E. Cornwall, T. M. DeCarlo, E. Krieger, and M. T. McCulloch. 2018. Similar controls on calcification under ocean acidification across unrelated coral reef taxa. Global Change Biology 24: 4857-4868.

Cyronak, T., K. G. Schulz, and P. L. Jokiel. 2015. The Omega myth: what really drives lower calcification rates in an acidifying ocean. ICES Journal of Marine Science 73: 558-562.

Jokiel, P. L. 2013. Coral reef calcification: carbonate, bicarbonate and proton flux under conditions of increasing ocean acidification. Proceedings of the Royal Society B: Biological Sciences 280: 1764.

This is a high-quality study that adds to the literature on projecting coral reef futures with climate models over the remainder of this century. Although the results are not highly novel, being quite similar to previous efforts, the different approach here makes it a valuable contribution to the literature. I have three main areas that could be improved:

Why this model? In the Introduction, the authors make the argument that previous works were limited by using only climate model output, which could be an issue because that approach does not allow interaction between corals and climate and is constrained to the temporal and spatial resolution of the provided output. That makes sense, however it doesn’t seem like the present work really improves on the previous works. The iLOVECLIM model seems quite coarse (3° ocean grids, larger than the climate model output used in at least some previous works) and the authors consider the effects of carbonate processes negligible on climate on centennial timescales. I think it is fine to present the results from this single model and explain its features and any new insights, but I am not seeing the argument that this is clearly an improvement over previous efforts in this field.

It would help readers if there were more explanation of the carbonate model embedded in iLOVECLIM. The results do seem highly similar to the Cornwall works, especially in the sense that the future state of reef carbonate production depends primarily on emissions scenario and heat-induced changes in coral communities and cover. The authors here note that their method of calculating carbonate production differs from the Cornwall approach of synthesizing laboratory studies. However, we never really get a clear explanation of how the carbonate module in the present study works. How was it parameterized and validated? Perhaps more emphasis could be placed on the similarity with Cornwall et al works from different approaches, but at present it is difficult to judge just how different the methods are.

It should be stated that the NOAA approach to estimating bleaching is indeed an estimate and it carries substantial uncertainty. It would be difficult to do, but ideally the current error of the NOAA method for estimating bleaching could be included in uncertainty assessment run into the future. But since the current uncertainty may not be well known, the authors should at least clearly describe this assumption. Along these lines, it is not clear how degree heating weeks are calculated in the present works: Is one value calculated for an entire 3x3° area? That does seem really quite coarse relative to spatial scales of marine heatwaves. At least, this should be discussed to a greater extent.