This manuscript presents the detailed Aeolus lidar surface return (LSR) calculation algorithm and provide some meaningful results/conclusions. The outcomes of this study will surely support the researchers in investigating the radiative transfer, reflectivity and thus guard the further spaceborne lidars missions. The manuscript is of high quality and scientific interest. The benefits of this study would be great for the enriching of Aeolus data products. Hence, I recommend the acceptance of this manuscript after the necessary revisions.

The general comments are listed below:

1. The methodology in this manuscript should be well-constructed, it should work fine in this study. My main concern is the framework of this manuscript, regarding the sequence of the section 3.1 and section 3.2. Normally, one should firstly verify the method proposed in this paper before the statistical analysis. I mean, probably authors may think about describing section 3.2 before section 3.1.
2. For the sections of “Discussion” and “Conclusion”, I propose the authors to reorganize these two parts. Some information in Discussion part should be moved to the “Conclusion” part. In current manuscript, the conclusion somehow looks like outlook.

The technical corrections:

1. The symbols of variables in the article should be consistent. For example, qflag(sometimes quality flag)/LER_G(sometimes LERG)/LSR(sometimes lsr)/ODray(sometimes RayOD). Please make them consistent.
2. Line 317: why 1-sigma but not 3-sigma is selected? Is 1-sigma too strict for the calculation?
3. Figure 3: red frame: qflag>0, but the QF in the attenuated case with number 1 is 0. Please check it.
4. Figure 4: I cannot see the histogram distributions with magenta color. In the text, of course you explain only a few numbers of strong return remain, therefore please find another way to show the distributions of water reflectivity between -35 and 35.
5. Actually, the authors should reproduce all the figures in this manuscript. They are hard to be recognized, especially for the legends/values.
6. Figure 10: Is it necessary to use log scale for LSR?
7. Figure 13: Why only “LER snow TROPOMI” and “LER CLEAR GOME-2” are compared? The authors should also explain why LER with other features are not compared.
8. Figure 14: which y-label is designed for TROPOMI? Is the y-label of "GOME-2 LSR" shared for TROPOMI and GOME-2? Please make it clear.
9. Line 680: How the values of snow cover (SNW_1) are calculated? Is it a proportion?
10. Figure 17: please recheck this figure. The titles (month information) of the graph in the first row have been obscured and are not displayed entirely.
11. Line 738: Discussion should be listed as section 4.
12. Line 809: Conclusions should be listed as section 5.

This manuscript is a relevant piece of work paving the way for an effective design and usage of LSR in future lidar missions. It contributes to the expansion of the database for the research field of surface reflection properties from only passive to also active instruments. The results on the excellent sensitivity of Aeolus LSR's to white surfaces (snow and ice) are very promising for future studies. The manuscript addresses relevant scientific questions and presents a novel concept and data resulting in substantial and reasonable conclusions. The objectives of the paper are well described at the end of chapter 1. The method is described to a sufficient extent, so that reproduction by fellow scientists seems feasible. The structure of the overall presentation is reasonable, but the language might need some revision in certain paragraphs. Own new contributions are indicated and proper credit to related work is provided including a good number of references of high quality. The amount and quality of the supplementary material seems appropriate.

I recommend the acceptance of this manuscript after the necessary revisions. Specific minor comments and unfortunately a large number of technical corrections are suggested below.