The paper couples soil physical properties with dissolved organic carbon parameters to establish a mechanism-based understanding of soil C dynamics and changes to DOC during freeze-thaw cycles in early spring and late autumn arctic tundra soils. Significant findings include increased soil respiration and soil microaggregation following 7 freeze-thaw cycles. The paper provides a much needed and valuable contribution to the freeze-thaw literature. The authors expand the pre-established concept of changeability of the physical arrangement of the soil to dissolved organic carbon response (also generally studied in separation in freeze-thaw experiments). The paper is well written and high quality, the experimental design is elegant and well-supported, the results are impactful, and the discussion is insightful, further emphasizing the value of the authors' contribution. The paper is well suited for publication in the Cryosphere. There are some points of clarification that should be addressed prior to publication, outlined below. 

1. It is clear from context clues throughout the paper that the authors sampled active layer soil. However, this needs to be more clear. Please provide the depths of sampling (or, if I missed them somewhere, please make them more visible/clear). Overall, the introduction and discussion both do a nice job of setting up and discussing the use of active layer soil, but it just needs to be explicitly stated.  
2. Why were organic soils sampled, rather than mineral? Apologies if I missed this somewhere. It seems like it would have made more sense to do PSD and aggregate stability analysis on mineral soils. Or both organic and mineral. Could you add a sentence defending the focus on organic soils rather than mineral? I believe some of the papers you are citing focused on mineral deformation by freeze-thaw cycles, but please correct me if I'm wrong.
3. Personally, I would like to see a diagram of each of the three cores collected via SIPRE corer with depths of organic and mineral soils as well as active layer and permafrost shown. I would also be interested in knowing how much of the organic soil you subsampled for the experiment. Did you include peat layers? Or just the more decomposed muck?

4. I agree that 12 hours is enough to freeze and thaw 120 g of soil, but please cite another incubation for the method or provide test data where you found that soil was able to completely freeze and thaw in that time. It will make your method more citable/reproducible.

5. I am unclear on the significance of 7 freeze-thaw cycles in the context of your paper. Ma et al., 2021 found unpredictable freeze-thaw response up to 7 freeze-thaw cycles, after which freeze-thaw resulted in increased pore connectivity. Placing your experiment on this cusp, at 7 freeze-thaw cycles, is a great contribution, but I think it needs to be a little bit more clear why you chose 7 freeze-thaw cycles. Was it because you were targeting that unpredictable pore network response that ends at around 7 freeze-thaw cycles? (line 94)

6. I would like a little more information about the soil processing pre-incubation. No sieving occurred? Existing aggregate structure was maintained from coring to subsampling to incubation? (lines 80-87)
7. Can you expand on what you mean by "adjusting soil bulk densities" (line 98) How was this done? Do you just mean you packed the sample containers with a measured mass of soil to achieve a known bulk density? 
8. Please be clear in the discussion when you are comparing findings from a paper that looked at mineral soil with your organic soil-based experiment. It's okay to have both, but I think it's important to be really clear if there are contrasting material types. I thought Ma et al., 2019; 2021 and Liu et al., 2021 generally looked at mineral soil but I could be wrong.
9. Lines 223-224: I understand that speculation, but I would point out that in Rooney et al., 2022, changes in pore structure were observed at both low and high water contents. So that speculation could be an oversimplification of how the geometry and architecture of the pore network or even the direction of the freezing front influence pore deformation during freeze-thaw. 
10. Lines 223-224: Do you expect that much evapotranspiration to have taken place? It wasn't clear to me in your methods that water was continously evaporating throughout the experiment. Could you specify if the samples were generally kept sealed from the atmosphere in some way to prevent evaporation?
11. Figures 1-2: I really like both of these figures. 
12. Figure 3: I think the figure caption could have a little bit more information in it. Specifically could you explain the legend? For a couple minutes I thought the colors had something to do with temperature, especially with the placement of the color bar in the figure.
13. Figure 4: I hope someone else can provide feedback to you on this. I am unfamiliar with structural equation model analysis. 
14. I was disappointed to not see PSD plots (y axis = pore volume, x axis = pore diameter). Maybe the plots could be included in supplemental? Personally, I would like to see a PSD graph for each individual sample to get an idea of overall sample heterogeneity, rather than just the standard error for each size class. Showing sample heterogeneity seems especially important in organic soils. 
15. Check for typos throughout, although I didn't see many.
16. This is a very exciting and cool paper! Looking forward to the published version.

This is a very interesting study looking at the impact increased freeze thaw cycles (FTCs) may have on carbon dynamics in surface tundra soils, via impacts to organic matter quality and soil structure. The study is well designed, and the results are very interesting with significant importance. 

My recommendation for this manuscript is to resubmit following major revisions. While the results are of high interest to The Cryosphere, the manuscript is currently lacking in information to be considered for immediate publication. It it's current form, it is unclear exactly how the authors conducted the experiment, and it would not be possible to repeat this study given the information provided. The authors have also not fully described, presented, and explained their results. Both the methods and results section of this manuscript can be greatly improved with more information. 

One issue I have currently with the interpretation of the results is the presentation and conclusions drawn from the structural equation model (Figure 4). I suggest the authors reconsider how this is presented, provide a detailed a prior model that includes the hypothesis and justification behind each pathway, and the soundness of each pathway.

I have provided more detailed comments throughout the manuscript in the attached PDF. 

Looking forward to seeing this manuscript again,

Liam Heffernan.