General comments

Nominally (see below), this manuscript describes a dataset of soil incubations focusing on carbon mineralization and temperature sensitivity (Q10) calculation. This is interesting and important for reasons well laid out in the introduction, as such incubations have been a major source of information about this process and informed models and understanding at many scales; an analysis-ready dataset of incubations is valuable. The authors’ dataset is publicly posted, has almost 22,000 rows, and seems clearly laid out (although see #2 below).

That said, there are several significant problems here. First, the ms is oddly structured. It essentially has three parts: (i) a description of the dataset; (ii) data summaries and comparison with ancillary data (in particular, incubation temperatures compared with the mean annual temperature of sampling location); and, very unexpectedly, (iii) an extended summary of earth system model approaches to decomposition and simple modeling exercise involving the dataset. From https://www.earth-system-science-data.net/about/aims_and_scope.html, the scope of ESSD is “Articles in the data section may pertain to the planning, instrumentation, and execution of experiments or collection of data. Any interpretation of data is outside the scope of regular articles.” Based on this, I think that (iii) above is clearly out of scope; it’s extremely odd to find this ESM algorithm analysis in an ESSD ms, and it should be removed. Even (ii) strikes me as marginal in terms of scope—it’s analysis, not data description!

Second, there’s no mention of SIDb (https://soilbgc-datashare.github.io/sidb/). The SIDb paper (Schädel et al. 2020) is cited but it’s bizarre not to note and discuss *at length* this pre-existing and seemingly very similar effort. How much overlap is there between the authors’ work and SIDb? Why not contribute these data to SIDb, rather than duplicate work and confuse researchers?

Finally, as already noted I have concerns about the structure of the data and how it doesn’t support easy reproducibility in terms of finding the source studies.

In summary, while I appreciate the large amount of work here, and believe this dataset will be valuable, the current ms should be rejected or subject to fundamental revisions.

Specific comments

1. Line 35: “expressed as”
2. The dataset structure as posted on Figshare is a little odd. The study information is combined with the observational data, i.e. it’s all in a single CSV file, so many rows are duplicated; having separate “data” and “studies” files might be clearer and cleaner. In addition, there’s no DOI, URL, or volume/issue information…to find a paper, are users supposed to search the title? Having a machine-searchable link or DOI seems crucial
3. 492-498: duplicated reference

This study compiles a global dataset of 22,000 observations from 191 incubation experiments on soil organic carbon (SOC) mineralization to assess how temperature controls soil CO₂ release, a key climate–carbon feedback. The dataset, largely biased toward surface soils, short incubations, and mid-latitude regions, reveals major gaps in deep soils, extreme ecosystems, and Africa.

The author suggest that Earth System Models generally misrepresent SOC temperature sensitivity, especially under warming extremes, though multi-term exponential functions perform best. Using a two-pool carbon model, the authors find that both intrinsic factors (SOC quality, microbial traits) and extrinsic constraints (oxygen, mineral protection, moisture) contribute equally to global SOC responses, but their relative importance varies by ecosystem (e.g., croplands vs. wetlands). The work highlights the urgent need for more representative experiments and improved model formulations to reduce uncertainty in carbon–climate feedback projections.

Although I find the article interesting and the dataset very useful, I have a number of comments that could help improve the readability of the manuscript.

Firstly, the mathematical approach used to distinguish between internal and external effects is rather poorly explained in section 5.4, which makes section 5.5 more difficult to understand. I suggest that the authors provide more detail in section 5.4 and perhaps simplify the vocabulary used.

It is noted in several places that you tested the temperature response functions of 69 ESMs. This is inaccurate, as most of the models whose temperature response functions you tested are not ESMs. This approximation should therefore be corrected throughout the manuscript.

Table 2 also needs to be simplified because some of the information is incorrect. For example, for Jules, the centre is described as the UK, which is a country, whereas for the other models, the authors give a research group instead.

CENRUTY-> CENTURY

It is also unclear in this table what the difference is between ‘land carbon’ and ‘land surface models’.

Figure 4 also needs improvement as it lacks clarity, particularly as M1 and M2 have not been defined.

How eq. 3 affect eq. 2 in the model developed by the authors?

Section 2 L. 78 point 2) More details are needed here, for instance do you accept when the same samples were incubated at 2 different temperatures? Do you use equal time or equal C (Hamdi et al., 2012)?

Hamdi, S., Moyano, F., Sall, S., Bernoux, M., Chevallier, T., 2012. Synthesis analysis of the temperature sensitivity of soil respiration from laboratory studies in relation to incubation methods and soil conditions. Soil Biol Biochem 58, 115–126. https://doi.org/10.1016/j.soilbio.2012.11.012 

L323: “There were no significant differences of the relative importance…” how this was tested?