Dear authors,

I had sent your manuscript for re-review to the original referees. One of the referees is happy with the revision, but the other one still has some objections. Thus, please consider the comments below, provide an appropriate reply, and update your manuscript accordingly.

Thank you very much!

Referee #2 comments:

I would like to thank the authors for their reply to my comments, although some of the comments were not fully addressed.

On comment 1, I agree that this method has potential to improve the current techniques for lidar data analysis. However, this has not been shown. It should be quite straightforward to use the LES data set to show that the LiSBOA retrieved winds and turbulence intensity are closer to the true wind field than other methods.

The response to comment 2 does not respond to the question. If the outer scale non-turbulent eddies (see e.g. Fig. 1 in O’Connor et al., 2010) cannot be filtered out with this method, it should be stated in the manuscript.

On comment 5, the authors clarify the use of equal spacing for azimuth and elevation angles in the optimisation in this case. However, my request to include both angles in the optimisation process is not implemented. Furthermore, I note that in Part II the authors use a scan schedule where they first set elevation angle spacing to [5, 6, 7, 8, 10, 12, 15] and then optimise the azimuth angle. In Part II also the sampling time per profile is decided prior to applying the LiSBOA optimisation. So it seems that only a subset of the scan parameters are provided through LiSBOA, though optimal scan design is one of the main arguments for using LiSBOA.

On comment 6, the reply to comment 2 seems more relevant, i.e. zero instrument error appears to be assumed. This is a major drawback, as lidar measurement will always have some instrumental error. However, by filtering out small-scale variability LiSBOA possibly allows using relatively noisy measurements, increasing the amount of useful data. Therefore, the effect of instrumental error must be investigated in more detail. Please use the LES data to determine maximum acceptable instrumental uncertainty in radial wind measurement. It should be relatively simple to add random noise to the lidar simulator (Eq. 17 in the revised manuscript) to find limits for acceptable noise. This will also determine minimum CNR threshold, which is essential for determination of sampling time per profile.

On comments 7 and 8: The code is still not available.
L682-683: “Code availability. The LiSBOA algorithm is implemented in a publicly available code which can be downloaded at the following URL: https://www.utdallas.edu/windflux/.” This seems to be a group home page. Underneath it there is a page ”Software/Datasets”: https://www.utdallas.edu/windflux/software-datasets/ but even that does not contain the codes used here. Please use a permanent repository for the code or include it as supplement so that it is permanently stored. Group home page is ok for keeping the latest version available, but it is not a permanent storage.