Wu et al. presented an interesting study showing observations and characterization of the thunderstorm based on measurements from coherent Doppler wind lidar (CDWL), Doppler weather radar and other instruments. Application of CDWL to study the atmosphere during thunderstorm is a relatively novel topic in remote sensing. I recommend acceptance for publication in AMT after minor revision.

1. VAD scanning mode is used for the CDWL in the experiment. The observation path is titled at an elevation angle of 60 degrees rather than vertical, why?
2. How extensive is attenuation of lidar signals?
3. CDWL cannot penetrate thick clouds so the results here can only reveal (with high uncertainty) the lower part of the thundercloud (mostly <5 km). Please summarize the limitations of the CDWL measurements.
4. During thunderstorms, the strength of vertical wind shear is correlated with the strength of thunderstorms. Some regrets fail to discuss the impact from turbulence on the velocity spectrum.

There are still some grammatical problems need to be carefully checked.

1. Line 162-163: "The images of lightning and hail recorded from Hefei in Figure 2" is not a readable sentence.
2. Line 195-196: "at" needs to be deleted.
3. Line 199: replace “in” with “on board”.
4. Line 203: replace “of” with “including”.
5. Line 215-216: the sentence “the real cloud environment is different when higher clouds are detected, so measurement results of the DWR can also give a cloud environment changes over the USTC” can't be understood.
6. Line 240: “exists” should be “exist”
7. Line 333: “0 m s^-1”.
8. Line 345: “most of the IC lightning occurs height is higher” is not a readable sentence.

13 Line 358: “Combined with the lightning detected by multiple sensors, it was found that when there has additional graupel with a speed greater than 5 m/s in the thundercloud when a CG lightning within 10 km nearby.” is not a readable sentence.

The paper presents a novel method to study thundercloud structures using coherent Doppler lidar wind. A particular study case at Hefei (China) is used to illustrate the potential of the method. The method is novel and valuable for publication. In general, I agree with all the comments made by the previous referee. But I add a major revision that need to be addressed and is related with the structure of the paper. Basically, it is difficult to find what is the new methodology. A flow chart (Figure 8) is presented with details of the methodology are given in section 4.2.1, and they should be moved to a methodology section before. I would recommend joining with the section ‘Principle of CDWL detetion’ and make a more detailed and consistent methodology section, highlighting the novelties versus other developments/applications

Apart of that, I would like to add a few minor comments:

• Between lines 39-60, there is a long discussion with lack of references. This is in the introduction section, and I believe that the discussion is based on previous studies that must be cited.
• I miss details of the CDWL. Is the instrument commercial or home-made? In the last case, more details about its configuration are needed.
• Conclusion section poorly reflects the novelty of the work, what are the limitations found and what is the future work.