I commend the authors on submitting such a nicely organized and written manuscript that presents impactful scientific results.   The manuscript was economical, easy to read, conveyed results in a logical order, and included well crafted figures that were easy to digest.   I quite enjoyed consuming this manuscript.

On a purely scientific level, this manuscript presents important findings that quantify CloudSat spaceborne radar deficiencies in the lowest atmospheric layers above ground level when compared to high resolution airborne radar observations that effectively sample the boundary layer and CloudSat blind zone.  Cold air outbreak (CAO) conditions are particularly highlighted in this study since they produce shallow convective cloud features that might not be sampled well by CloudSat observations.  CloudSat observations and products are instrumental model evaluation datasets, but contain known deficiencies.  This study should instigate further research that will enable CloudSat products to be improved by accounting for blind zone features that are systematically undersampled  (e.g., cloud and precipitation occurrence and blind zone evolution).

I recommend that the manuscript be published after some very minor issues are addressed.  I would be happy to review the manuscript again if needed, but also defer to the editor’s judgement on whether the following minor comments warrant another review. 

1.  Line 93:   Are the SIC thresholds chosen arbitrarily, or based on previous studies?   Relatedly, how are these percentages calculated (I assume daily AMSR2 SIC products)?

2.  Lines 414-416:  While I think it is totally acceptable to apply the Maahn et al. (2014) Z-S relation to the MiRAC observations for back-of-the-envelope calcuations, it is likely that the rosette habit assumption used to derive Z-S does not translate very well to the microphysical composition of oceanic snow-producing clouds generated under CAO conditions, especially when comparing snow event categories differentiated by snowrate intensity.   Acknowledging this methodological shortcoming is advised, but its overall effect does not detract from the larger message conveyed in the manuscript.

3.  Line 432:  ClaudSat → CloudSat

4.  Line 470:  Kulie et al. (2016) and Kulie and Milani (2018) partition CloudSat-observed snow events by “shallow” and “deep” categories, with special emphasis on high latitude regions prone to CAO’s.  They highlight the light nature of shallow snow in CAO regions with appropriate (but unresolved) blind zone related caveats.  This study clearly indicates that CloudSat estimated snowfall occurrence and rate/amount are significantly impacted by blind zone limitations that hamper efforts to quantify snowfall with the best available spaceborne instruments.  

5.  General comment:   It might be worth mentioning that a combined CloudSat/CALIPSO product exists that will more successfully identify low-level cloud structures compared to the CloudSat 2B-Geoprof product.   

6.  General comment:   Mateling et al. (2023; JGR) was just published.   It focuses on CAO snowfall production in the North Atlantic Ocean using CloudSat products - another highly relevant manuscript that would benefit from the information gained from the current study.

Review on the manuscript:

Assessing Arctic low-level clouds and precipitation from above – a radar perspective

by I. Schirmacher et al.

This paper presents a study on the evaluation of CloudSat measurements in the very low altitude levels based on collocated airborne radar measurements in arctic clouds.

The method consists in applying a forward-simulation methodology to high resolution airborne radar measurements (MIRAC instrument) to convert these observations in a “CloudSat mode” and then assess the CloudSat measurements.

The authors used first 4 collocated flights with CloudSat track to evaluate and validate their methodology. Then, they used airborne radar observations from 4 airborne campaigns in the Arctic region to evaluate CloudSat limitations.

Surface conditions (open water or ice) and meteorological conditions (cold air outbreak index and circulation weather type) are taking into account in the analysis of the results.

It is a very well-known issue that CloudSat measurements are limited near the ground, which implies very large uncertainties when deriving snowfall amount at ground level on a global scale. This study propose a nice methodology to assess this issue.

The results of this study are very important and should be take into account for the future studies using CloudSat products (reflectivity, snowfall rate or amount) near the surface.

The paper is very well organized, well written and easy to read.

In particular, the methodology is clearly and concisely described in the section 2.3, even if the figure 4 in the further section 4.1 helps to visualize the effect of the four steps.

I recommend this paper for publication, after answering the following minor comments and corrections:

• Line 142-143: Please rephrase the end of the sentence
• Line 432 : replace “ClaudSat” by “CloudSat”
• Figure 3: You could add the histogram with all data (separated by CWT)?
• Figure 7 and lines 288-289: I suppose the 2 dB difference you mention between Z_sim and Z_C comes from a linear fitting ? I suggest you add it on the figure.
• Line 410: what is the assumption of crystal habit for Z-S law in the 2C-Snow-Profile product? Please add a few details on this product to facilitate de further discussion on the differences.
• Figure 12: I suggest adding the profile of S_M along that of A_M,norm.