Comments on manuscript gmd-2022-281,

“SedTrace 1.0: a Julia-based framework for generating and running reactive-transport models of marine sediment diagenesis specializing in trace elements and isotopes”,

by Jianghui Du

The author provided a Julia based code for generating diagenetic reactive transport codes. The goal of the model development is well described and the code realizes the flexibility to enable various usages targeting at different sediment environments with different tracers of interest. The model’s validity has been well examined and example simulations demonstrate the capacity of the model well. Accordingly, I think the paper is well-suited for publication in GMD. Followings are my comments on specifics that I hope the author may find of some use to improve the manuscript.

Potential internal inconsistency.  pH calculation in Sect. 4.2 does not seem to take into account detailed aqueous speciation (e.g., no accounts of Fe + CO3, Fe + HCO3, Fe +HS in Eqs. 27, 28, etc.). However, those are included in aqueous speciation calculation in Sect. 4.3. As long as concs. of included aqueous species with EIs or other tracked aqueous species are insignificant compared to total concs. of EIs/other tracked species, pH calculation as well as mass balance must not be significantly affected, but we can think of a situation where some ion pairs matter relative to EIs/other tracked species, e.g., ferruginous oceans in the Precambrian or deep depths of sediments where porewater chemistry is in equilibrium with some soluble solid phases. This may be my misunderstanding but potential problem here is model’s internal inconsistency between aqueous speciation and pH modeling, and potential violation of charge balance in porewater.  If this is the case, it should be clarified.

Numerical diffusion. Especially related to Sect. 4.4. It is well known that numerical diffusion affects proxy record and thus different models have attempted to deal with it together with diagenetic reactions (e.g., Kanzaki et al., 2021; Munhoven, 2021). It probably does not affect steady state age-profiles, but the author should make it clear how the model deals with numerical diffusion and its effects on proxy reading or comparison with observations.

L419-425. The author should show a plot of Logistic and Heaviside functions as a function of omega so that it is easy for the reader to compare the two functions and evaluate the approximation here as I have not seen this approximation elsewhere. Also, Eq. 24 does not look the same as that in L428. Please make sure they are consistent.  Also, it would be better to show how this approximation affect the results where omega calculation can affect sediment profiles (e.g., Fig. 4).

Technical comments:

Code. I have tried to install and use the code to test example simulations in this paper (from examples directories) but ended up not being able to (failed with windows Cygwin and virtual Ubuntu in windows). This could be only because of me not used so much to Julia but it might be better to indicate under what environments the code has been tested so far and is supposed to work. This does not have to be in the text but anywhere else like Supplement or docs in code repository.

L106. Why does it have to be in mathematica notebook, which is not open to everyone? It would be better to give some final form of derivation in the main text or even supplement (or any form available to the reader) rather than saying that the reader can check them if they can use mathematica.

L106. maybe --> may be

L831. Sensitivity --> sensitive

L1097. It would be easier to read results if the author also provides delta values (1000 x ln alpha) in per mil.

Please refer to the review in the attached PDF