General comments:

Based on one year of simultaneous wind measurements acquired from 17 Doppler wind lidars across China, this manuscript by Wu et al. conducted a comprehensive comparison study against Aeolus wind products. Overall, this topic fits well the AMT. The instruments and data are reliable, and the analysis methods are scientifically sound. In my opinion, the manuscript is well organized except for some typos and grammar errors. The comparison results are of great importance to better understand the performance of Aeolus wind products in China, even though the measurements are obtained from coherent Doppler lidar (with 1550 nm wavelength) over China. However, before the manuscript can be recommended for acceptance for publication, I have several suggestions and comments here that need to be addressed.

Specific comments:

1. Section 2.2.2: Line 128 says “The measurement heights selected for comparison are 50 m, 100m.” Nevertheless, I only see the comparison results at 50 m in Figure 2. I am curious why not showing the results at 100 m AGL? 2 only compares the wind speed without considering the wind direction. I suggest the authors compare the u-component wind. Also, more details about the site Haiyang and the specification of the mast that is mounted cup anemometers and wind vanes at Haiyang needs to be added.
2. My biggest concern is on section 3.2. In this section, the authors proposed an important result that the vertical velocity could impact the HLOS wind velocity retrieval from Aeolus. But there are neither references nor experimental demonstrations here, only theoretical derivation. I suggest authors add a result about the deviation of Aeolus and CDL HLOS wind varies with vertical velocity.
3. The number of Mie-cloudy comparison pairs and 387 Rayleigh-clear comparison pairs were 52 and 387, respectively. In my opinion, the sample size is too small. Therefore, the significance test must be performed in Fig.10-12.
4. L144: “For Aeolus, only observations with the corresponding “validity flag” of TRUE are considered.” Please clarify what is the “validity flag” of Aeolus. Similarly, why the estimated errors threshold of Mie-cloudy and Rayleigh-clear wind velocities were set to 4 and 8 m/s? There are neither references nor experimental demonstrations here.
5. 3 needs to be modified. In the case studies, the HLOS wind and vertical correction HLOS wind from CDL are both used to compare with Aeolus HLOS wind. It needs to express the data matching process in a clearer way.
6. What conclusions should the reader make from Fig. 6. It is just an observation case by Aeolus and CDL. I suggest that the author remove this picture, or draw the wind profiles that needs to be compared.
7. If my understanding is correct, Figs. 7 and 8 should be the case comparison. What do the error bars on the red and blue curves mean?

Technical corrections:

L18: “atmospheric boundary layer” and “planetary boundary layer” appear several times in this MS, if both have the same meaning, just keeping one expressing is more appropriate.

L27-28: It is not appropriate to say “better than” when describing MAD and bias.

L41-42: the second and third“from” are redundant and can be dropped. Reference supports are needed for the statement “wind profiles from global radiosonde network and aircraft”. The authors may refer to Guo et al. 2021(https://doi.org/10.5194/acp-2021-257); and Zhang et al., 2020 (https://doi.org/10.1029/2020JD032803), respectively.

L61-62: what is missing is the predicate in the sentence begin with “An example of early validation”

L83: “During 14 January and 14” -> “During the period from 14 January to 14 ”

L85: delete “)”

L174 and 177: Typos in Eqs. 2 and 3: “cot” should be revised to “cos”

Figure 6a,b: “DEM altitude”s -> “DEM”

Table 7a and 7b: error in the citation for “RS over China (Guo et al. 2021),” which should be revised to “Liu et al. 2021”. Also, The ACPD by Guo et al. 2020 in the reference list has been finally published as ACP in 2021, and is suggested to be corrected as well.

L332: grammar errors in “by used with” and needs to be corrected.

L341: “By” -> “Using”

L364: the article “the” is missing before “planetary boundary layer”

L364-365: The two sentence can be rephrased as “In the planetary boundary layer, the vertical velocity from convection and turbulence could influence the comparison, due to the impact of vertical velocity on the HLOS wind velocity retrieval from Aeolus.”   

First of all I am very happy with this paper and the presented results. Especially the time evolution in Aeolus wind quality seen for the different baselines is impressive and valuable for other users.

Some detailed comments:

line 30:

you mention that you used data for different baselines. However, later in the paper for example on line 195 you only refer to a date for which Aeolus data was used. This makes it hard and in some cases impossible for users at a later time to know which data was used.

For eaxmple the date of 10-May-2020 will be part of the 2nd reprocessing campaign, and therefore the same day will be available for 2 different baselines (the NRT data stream which baseline was baseline 2B10 and the reprocessing baseline 2B11).

Therefore please state clearly which baseline and date you used if you refer to the Aeolus data for a specific date or period..

line 52:

* typo: the emotion  => the motion

line 68:

* typo: in the worldwide => in the world

line 79:

* typo: section 5 summaries => section 5 summarizes

page 5: table 1: what is DBS (measurement mode)?

line 163:

HLOS is compared to HLOS, and in section 3.2 you try to also take vertical velocity as observed by the CDL in to account. It is not yet clear to me how you apply this vertical velocity. In the example given in figure 5 is seems you apply the vertical velocity as measured at the exact time of the collocated Aeolus measurement. This I think is not correct. From figure 5 it can clearly be seen that there is a lot of variation in the vertical wind. This is probably due to convection present in the boundary layer giving sometimes updrafts and sometimes downdrafts. Aeolus accumulates over 15 to 85 km along track (in just of few seconds), so if the typical dimension of this convection is smaller than this averaging length, than the average vertical wind observed by Aeolus will be much closer to zero than what is observed by the CDL, and the net effect will be a broadening of the observed Doppler shifted spectral line only. So before applying a correction for vertical wind, first the typical size of the convective cells should be determined. Only then it can be decided if applying a correction is useful or not.

figure 8(a) shows a case where Aeolus data from 21-Jan-2020 is used. You should mention als here that the adaptive bias correction based on ECMWF data and M1 telescope temperatures that was added with baseline 10 was not yet in place for this date (even though you do explain this in a later section). This explains the noticable bias for the Rayleigh channel winds. For this date the NRT data stream was baseline 2B07.

line 255:

you mention that you apply a data selection using "the original standard deviation" but it is not clear to me what this is or how it is defined. Please give more details.

line 279: here you mention the different results for ascending and descending tracks of Aeolus data. This orbit phase depending bias has been largely solved by the adaptive bias correction based on ECMWF data and M1 telescope temperatures that was added with baseline 10. So If you only use data after 20-Apr-2020 (baseline 2B09 or newer) this should be much improved.

line 286:

you write:

     "The baseline 10 dataset consists of baseline 09 data from April 2020 to October 2020 and the FM-B low bias reprocessed dataset of 2019"

this should be:

     "The bias corrected dataset consists of baseline 09 data from 1 to 20 April 2020 and baseline 10 data from 20 April 2020 to 8 October 2020 and the FM-B low bias reprocessed dataset of 28 June 2019 to 31 December 2019."

summary tables 7a and 7b do not really belong to this study. It raises much more questions, i.e. are alle these studies just looking at the boundary layer as well? (I guess not). How much data did they use? What extra quality controls did they implement, etc. My feeling is that comparing CalVal results from many teams should be published in another report or paper and not here.