This paper used co- and cross-polarized backscatter data from Sentinel-1 SAR C-band images to estimate snow depth variations over northern Finland from 2019 to 2022. From the report of this paper, snow depth estimated from Sentinel-1 images tended to underestimate compared to measurements from automatic weather stations. Additionally, snow depth increased with higher canopy density. Their findings provide technical and theoretical references for estimating snow depth from C-band SAR images to some extent. However, several issues still need to be resolved before publication.

Comments and suggestions:

1. Line 57, it is suggested to provide brief introduction about the limitations in estimating snow depth using C-band SAR.
2. Line 58, the abbreviation S1 has been provided here; henceforth, please use the abbreviation when referring to Sentinel-1.
3. Line 91-92, please provide descriptions of the terrain characteristics, meteorological conditions, vegetation cover, and other relevant information for these three weather stations.
4. Line 124, for the empirical methodology developed by Lievens et al. (2019), in comparison to the original version, what specific improvements did this study make?
5. Line 166, this paper states “Overall, we observed clear underestimations in the shallow snow depth regions”. However, Figure 2 clearly demonstrates a more pronounced underestimation in regions characterized by deep snow depth. Please provide possible reasons.
6. Please delineate the boundaries of the water bodies in Figure 1.
7. There may have been rapid changes in snow depth during the period from April 3 to 7, 2002. It is not appropriate to compare the measured snow depth data from this period with the estimated snow depth from S1 on April 6.
8. L193, it is confused to state “thicker snow depth values over dense vegetation and water bodies areas, where the canopy density is equal to 0%”. The canopy density is equal to 0% means there is no vegetation.
9. Line 197, change “Figure 5b” to “Figure 4b”.
10. L213-214, it is suggested to analyze the reasons why the correlation of the year 2019-2020 and 2021-2022 is lower than that of 2020-2021?
11. Line219-220, to achieve more accurate estimation of snow depth, what improvements should be considered for the estimation method?

This manuscript estimated snow depth over the northern Finland area from Sentinel-1 observation utilizing an existing retrieval algorithm, and compared the estimation with in-situ measurements from ground stations. Overall, the subject this study deals with is relevant to this journal, and the manuscript was easy to read and concise. However, the main issue was difficulty in capturing the main message or contributions that the authors are trying to deliver with this study. 

General comments

1. I thought the main contribution might be an algorithm development. However, the authors utilized an adapted version of Lievens et al. (2019), but it was not clearly stated which components of the original algorithm were adapted (modified/changed). How did the authors improve the algorithm? How did the modification affect the retrieval performance in reference to the original algorithm?

2. On the other hand, the comparison method was not rigorous enough to make this work a validation study. Weather station observation would represent the snow depth a few meters around the station, while the Sentinel-1-derived snow depth has much coarser resolution (it even went through a 990 m by 990 m kernel moving mean filter). How could those data be comparable? What was the authors' strategy to adequately address this issue? 

3. In addition, while there are many hypotheses for snow depth underestimation, the supporting evidence was rarely provided. The authors mention that temperature and precipitation are important factors, but they were not taken into account in the analysis. The retrieval algorithm also doesn’t consider temperature. Could this be one of the reasons for underestimation?

4. Lastly, the review of previous studies seems to be incomplete. For example, I have the following questions. Is this the first Sentinel-1 snow depth retrieval study that covers northern Finland? Is Lievens et al. (2019) the only C-band SAR snow depth retrieval algorithm? Why was this algorithm chosen for this study? How good or bad are Sentinel-1 snow depth retrievals over different areas or regions based on different algorithms? Are there any previous studies/reports available?

Specific comments

L194: There are two periods ("..") next to "40".

Figure 1: Please consider changing the color of letters (currently in black) over the dark blue background. It is difficult to read.