The study applies the HydroBlocks model to simulate soil conditions in Finland, and the technical implementation appears thorough. However, I have significant concerns regarding the scientific novelty, methodological clarity, and overall focus of the paper. In its current form, the manuscript does not meet the standards for publication in this journal.

Below are my specific comments:

1. Lack of Clear Innovation and Advantage. The study appears to be a regional validation of the HydroBlocks model over Finland. However, the manuscript does not clearly articulate what specific advantages HydroBlocks offers compared to existing reanalysis products or other high-resolution models. Without a clear comparative analysis against established datasets (e.g., ERA5, satellite-derived SWE, in situ observations), it is difficult to assess the added value of this work.

1. Misdirected Introduction Focus. The introduction extensively discusses frost quakes, yet the manuscript does not directly address this phenomenon in the results or discussion. If the goal is to model soil conditions relevant to frost quakes, the connection should be explicitly demonstrated. Otherwise, the lengthy background on frost quakes seems disconnected from the actual content.

1. Insufficient Model Performance Evaluation. While the authors note that HydroBlocks performs well in some respects, several KGE values are negative or low. For hydrological modeling, KGE > 0.5 is often considered acceptable for streamflow, but for other variables (e.g., soil moisture, temperature), benchmarks are less clear. The manuscript should include a systematic comparison with independent observational or reanalysis products (e.g., satellite SWE, soil moisture from SMAP or ERA5) to objectively demonstrate model superiority.

1. Unclear Purpose of Spatial Comparisons Across Years. The comparison of spatial patterns for specific days in different years (e.g., Fig. X) does not convincingly illustrate trends or model skill. Given interannual variability in meteorological conditions, such snapshots may not be representative. A more statistically robust analysis of temporal trends or climatological comparisons is needed.

1. Title and Focus Mismatch. The title promises an analysis of “temporal and spatial trends,” but the manuscript lacks a dedicated trend analysis. The results are largely descriptive and do not systematically quantify or discuss long-term changes in soil conditions. This disconnect should be addressed.

1. Redundant Figures. Figures 10, 11, and 12 contain repeated subplots of meteorological and SWE time series, which adds little value and distracts from the key messages. These could be consolidated or moved to supplementary material.

1. Weak Scientific Narrative. The manuscript reads more like a technical report than a cohesive scientific paper. The focus is diffuse, shifting between frost quakes, model validation, and regional climatology without a clear central question. I recommend reframing the study to emphasize either (a) advancements in land surface modeling at high resolution, or (b) a focused investigation of frost quake drivers using HydroBlocks. The current version would benefit from substantial restructuring and refocusing before reconsideration.

The paper “Modelling of temporal and spatial trends in soil conditions in Finland using HydroBlocks model” uses outputs from a land surface model to estimate thermal stress in frozen soil to analyze the co-occurrence with frost quakes in Finland. The analysis includes comparing model outputs of snow water equivalence, soil water content, soil temperature and soil ice content between the model and observations for several sites in Finland. The paper is well written and has a logical organization that makes it very comprehensible. The subject matter is interesting and provides a unique application for land surface model outputs. The paper can be approved by addressing inconsistency in terminology, revising several plots, and strengthening the conclusions. Specific comments for revisions are given below.

Revisions:

Line 48: There is a lot of text devoted to the description of the thermal stress and even including an equation for thermal stress. This made me think that the analysis would be analyzing thermal stress in the soils. However, this paper is really about validating a land surface model for the potential use of estimating thermal stress. It could be helpful to cut back some of the text describing the thermal stress and focus more on how land surface models estimate the key inputs needed for thermal stress.

Line 67 – Here you first introduce soil moisture content (SMC) and soil water content (SWC). It is not clear what the difference is here and is likely a model specific designation and sometimes the two are used interchangeably in broader hydrology. The readability of the paper would improve if you define specifically what you mean by SMC and SWC and how the two are fundamentally different. Also, from later analysis it is clear that what you really want is soil ice content, so it may make sense to start with that and describe that first.

Line 72: The word paring “multiple different” is redundant and should be changed to either “multiple” or “different”.

Line 78: “soil water equivalent” should be “snow water equivalent”.

Line 105: It is unclear what the sentence that starts with “The boundaries of the subdomains” is trying to communicate. It would be helpful if this was revised.

Line 276: Figure 7 adds very little information that cannot be assessed from Table 3. In particular, the scale of the x-axis is such so that it extremely difficult to see differences between the models and the observations. One way to improve this is to show a short time period for the gauge that does the best (Vaala) and the gauge that does the worst (Konnevesi). This will give the reader more insight as to why the model performs well at one location and not so well at another.

Line 322: To help avoid confusion, make the column headers Table 4 consistent with how they are described in the paper and the caption (i.e. SWC-5cm, not W5cm). Same for Table 5 and 6.

Line 326: As mentioned above, the description of SMC, SWC, and SIC is very confusing. This seems to be more of nuance from the specific modeling framework. It would be clearer if in the methods section you say that you use SMC and SWC from the model to get SIC. Then just treat SIC as another model output. Given that most of the analysis is focused on SWC, then you can remove SMC from the paper entirely. This could help clean up the terminology and make the analysis clearer.

Line 392: Conclusion 5 is not really a conclusion. It just uses two outputs from the model to create an output that the model doesn’t current output.

Line 394: Conclusion 6 is too broad for what is analyzed in this paper as there was no analysis of thermal stress or predictability and risk of frost quakes. This paper did validate the inputs need to calculate the thermal stress and demonstrates that there could be potential for calculating thermal stress and frost quakes, but it does not actually validating the model against observed thermal stress and predicting frost quakes. Rewording this to focus on what was shown in the results would be beneficial.