General comments:

This is a straightforward paper that identifies fake cloudiness in the CloudSat radar data record due to mirror echoes and multiple scattering tails. The author identifies that the effect is relatively small in the CloudSat record and is located at very high altitudes where it can be easily identified due to the specific PRF of CloudSat. EarthCARE however will have a much higher PRF in order to enable reasonable Doppler performance. The author shows that as a result of the increased PRF a non-negligible impact of these second trip echoes will both create fake cloudiness and occasionally artificially increase the single-scattered reflectivity value in real clouds. The methods are well described and appropriate for the study. The relevant papers have been cited where appropriate. The study is timely ahead of the EarthCARE launch and has immediate specific implications for the CloudSat data products. I only have a few minor comments below.

Specific Comments:

Page 1, line 22: phase 0 -> pre-phase A

Page 5, Line 6: The two sentences beginning with ‘The folding’ are not clearly written. I Suggest: ‘The folding mechanism is further explained in Panel b for the reflectivity profile at -32 latitude (corresponding to the red line in Panel a). The Level-1 CPR product reports the satellite height is 715 km and the PRF is 4.37 kHz which corresponds to an unambiguous range of 34.3 km. By distributing an integer number of these 34.3 km long unambiguous ranges from the satellite height downwards, the folding window’

Page 6, Line 5: It isn’t labeled in the legend which percentage goes with each latitude band.

Page, Line 11: It’s also pretty straightforward to use CALIPSO to identify these clouds.

Page 16, Line 21: could you elaborate on how you think the mirror echoes add any information. I think the fact that you can predict the magnitude and location of the mirror echoes with some accuracy suggests that there isn’t much additional information 

The Autor produced a structured and excellent written paper addressing an essential problem for space-borne cloud radars. The ‘fake clouds’ due to second trip echoes can appear in the CloudSat data. The Autor demonstrated the reasons for the generation of such echoes and how their detection is performed. This is followed by a statistical analysis of CloudSat data to identify the ‘fake cloud’ appearance in data over the ocean. While the effect in high latitudes is relatively tiny for tropical regions, it can be significant.

This study is in a second step also applied to the future satellite mission EarthCare, which has a higher PRF than CloudSat. Therefore, the probability of second trip echoes and fake clouds in the data set is higher than CloudSat. This work is also based on statistics and forward modeling possible second trip echoes into the CloudSat data, assuming EarthCare settings.

All the results show that for future satellites which are planned to run with a higher PRF, the detection and filtering of such echoes have to be considered. Because for higher PRFs the possible window of such echoes will also move closer to the ground and therefore more into the area of interest.

I only have two minor questions and suggestions:

Could the Autor explain why only CloudSat profiles over the ocean considered for the statistical analysis? 

And is it also possible to do such detection and analysis for profiles overland?

I do not have any further questions and would suggest the paper for publication.

This is a well written manuscript describing observation, simulation, and prediction of second trip echoes in W-band satellite radar data. It was very useful and insightful to use of CloudSat data showing how second trip echoes appear in non-physical heights in CloudSat observations, yet, these second trip echoes would appear at realistic height ranges when using EarthCare radar operating parameters. The cloud community needs to understand the impact of second trip echoes and this manuscript will help with that education. Also, I am appreciative of the references pointing to prior work.

Two minor comments to help my understanding:

As someone who would like to repeat these simulations so that I understand the math better, can the constants used to generate Figs. 1 and 2b be added to the figure caption or in the text? If I can repeat Fig. 2b, then I could repeat the EarthCare simulations. 

Also, I may not understand how profiles with MS scattering are being counted (page 10 line 12). Should a frequency of 10^(-3.7) be one on 5000 profiles (not one in 50 profiles)? Maybe 10^(-3.7) is relative to all profiles in 2008.

I look forward to seeing this manuscript published.