Comments on revised m/s from Reviewer 1. Pls respond to these when submitting your final m/s.

The authors have thoroughly responded to the comments from all reviewers. As outlined below, there are a number of issues for which further clarification is required, but these are relatively minor.

Major comments from initial review:
1. Identification of deficiencies in SNODAS: I agree that detailed analysis of this issue is beyond the scope of the paper, and is obviously hampered by no access to the source code. But you need to be careful, therefore, with the wording in the paper. If you can’t diagnose model deficiencies then you need to modify the justification provided on line 397: “…a much more spatially extensive dataset, such as the CLPX-2 LiDAR, is required for determining the underlying causes of model error.”

2. Extend analysis to other uses with SnoTel: OK. I agree with the response that use of SnoTel data is extremely limited due to the pointwise nature of the measurements. I was simply wondering if a long time series of SnoTel versus SNODAS would have any value.

3. Analyst intervention in the SNODAS assimilation: it’s still not clear whether these are objective data QC routines (i.e. to remove erroneous observations) or more subjective analyst-driven decisions. Can this be clarified?

4. Clearly state objectives in Introduction: The introduction reads well, but I still think this section would be strengthened by a more detailed/itemized statement of the objectives that occurs before the final sentence of the section.

5. Hourglass sampling: OK. But next time give them magnaprobes and tell the samplers to make more measurements!

6. Snowmelt: OK

7. Snowmelt/sublimation: OK

8. Combine Figures 5/6: OK

9. Potential bias in LiDAR measurements: “…the 5-10 cm underestimation in the LiDAR measurements is well within the vendor-recommended uncertainty and does not ultimately affect the final comparison to SNODAS in a strong manner.” I agree that the impact of this bias is minor. But were this random instrument/retrieval, one would expect a randomly distributed error around zero, not a consistent bias. The bias is certainly consistent with a systematic over-probe in the in situ dataset…
On line 422, I suggest the negative bias value be provided in the text.

10. Regression results: I don’t feel super strongly about this, but in the absence of a summary table with actual statistical results, I’m left with the subjective statement that “…upscaled snow depth changes were overwhelmingly found to best predict the discrepancy.” Even if only one variable provided significant results, I still think a summary should be provided.

11. Figure 8/9: I don’t see different symbols for each region in the scatterplot…

Editorial Comments
Line 60: “…this study purports to examine…” ‘Purports’ is not the right word choice here.
Line 66: …”over a vast extent.” I’m not sure I would describe this study are as ‘vast’.
Line 175: change to “…date are used…”

Congratulations! It was a lengthy process but I think you have a much stronger paper as a result. Best regards, Ross B