General comments

This manuscript describes an analysis of CMIP6 outputs, specifically presenting a computation of the soil, vegetation, and land carbon sensitivity to CO2 (beta) and climate (gamma) changes. This follows earlier, similar analyses of previous CMIP generations, and is a useful diagnostic of model behavior and to help understand earth system response to anthropogenic CO2 emissions. The ms is well written and interesting, focused, and generally clear; I applaud the inclusion of a link to the analytical code.

In addition to some minor issues, my overall concern is that the analysis is quite limited, misses some very interesting possibilities, and doesn’t always display its results well. Specifically, a comparison of CMIP5 and CMIP6 values—ideally quantitatively in a figure, but at least treated in the discussion—would add a lot of value. In addition, the bar graphs are not particularly illuminating, and consider better visualization options (e.g. #9 below).

In summary, this is an interesting and valuable contribution, but needs moderate revisions for concision and clarity; to improve how it conveys its results; and, ideally, to expand its scope a bit.

Specific comments

1. Line 11: Jones et al. 2018 http://dx.doi.org/10.1029/2018GL079350 might be a good citation here
2. 37: “has been”
3. 52: This is a long time ago! If any models have been added since then, would it be possible to include them on revision? That said, I’m not trying to make a huge amount of new work for the authors
4. 141-142: the results section has a certain amount of restating things that have already been defined/said in the introduction and methods; consider trimming. This is one example
5. 239: do you mean “explicit” here?
6. 256: missing word? “that beta and gamma linearity is a valid assumption”?
7. 264: thanks for the code transparency. Adding a README to this repo would be useful, and I suggest permanently archiving it (i.e., generating a DOI) using Zenodo
8. Figure 2: move to SI? Not sure how useful this is; maps are very small
9. Consider whether Figure 4 could be re-thought for clarity and impact. For example, what about plotting deltaCs (x) versus beta+gamma (y) with a 1:1 line, coloring points by 2xCO2 or 4x? That might be a better way to visualize for readers
10. You don’t need to say “Bar chart”, “Maps”, etc. in the figure captions. Readers can see what type of plot it is

Review for by Soil carbon-concentration and carbon-climate feedbacks in CMIP6 Earth system models by Varney et al., submitted to Biogeosciences (EGUsphere)

Please see the attached supplement.