Dear Butzin et al.,

Please find attached my review.

Yours sincerely,

Anne Morée

This manuscript describes results from a new implementation of carbon isotopes in an ocean model. The paper is well written, nicely illustrated and the conclusions are backed up with the evidence provided. I only have a few minor comments and leave it to the discretion of the authors how much they want to change the manuscript.

Line 197: The reconstruction by Kwon et al. (2022, https://doi.org/10.1038/s43247-022-00388-8) includes the surface and could be used to compare with the model results there.

Line 218: The decomposition by Broecker and Maier-Reimer (1992) is problematic as it ignores the effect of differences in preformed d13C and PO4. Interior ocean d13C and PO4 include preformed components. For d13C the preformed component is impacted by air-sea gas exchange, whereas for PO4 it isn't. Thus d13C_BIO is not equal to d13C_rem (which doesn't include a preformed component). d13C_BIO doesn't include the correct biological preformed component of d13C either since it was calculated using PO4. In other words, this decomposition, although widely used, is pretty much useless to understand d13C.

Line 223: Since d13C_BIO (as calculated following Broecker and Maier-Reimer, 1992) includes effects of air-sea gas exchange (from preformed d13C), d13C_AS includes effects of biology.

Line 247: Why not run model with Suess effect included for a proper comparison to observations? Suess effect can be expected to decrease d13CPOC by about 2 permil (e.g. Fig. 8 Schmittner et al., 2013). What is the global difference (model-obs)?