As noted in lines 250-255, this is a complex system where the behavior of one interferometer impacts the downstream performance of the others.  As such, this detailed description is helpful and insightful (and necessary).   In fact, I would argue this is an important piece of work to have documented for posterity, particularly the mathematical descriptions of the specific Aeolus interferometer implementation and the detailed on-orbit operational performance.  The mathematical approach appears correct and robust. 

Doppler lidar is challenging because so many effects can appear as Doppler shifts.  Changes in laser frequency, thermal shifts, plate spacing – all can appear as Doppler shifts and have to be separated from the actual atmosphere-induced Doppler shift.  There is a comment early in the manuscript about thermal impacts on the telescope.  Are there any thermal impacts being noted on the interferometers?

In discussing the results shown in Figure 4 (e.g., manuscript lines 351-364), it would be helpful to know, from forward modeling, how good the fit has to be to maintain bias at <1 m/s, <5 m/s, <10 m/s, etc.   In my experience, the absolute wind determination is highly sensitive to the FPI fit.  Thus, even the small amount of variability as shown in Figure 5, for example, can have large impact on the wind retrieval.   This manuscript does a good job of explaining and quantifying the measured instrument response.  What is not answered is the question, “Is this good enough?”  In other words, perhaps add a succinct description of how well the errors have to be minimized to have less than X m/s impact on the final product.  This would help the reader understand how close the team is to best possible performance, or if this is best possible and is at the limit of noise.  There must have been an initial expectation of controlling instrument parameters to some limits, to maintain wind error less than some specified amount. How well did the initial expectation or modeling match the measured results?

Minor point:  Figure 4 is referenced in line 183, well before it appears in the document and before Figure 3.  Best practice is to have the figures in the order they are referenced.

Grammatical issue:  The authors use the phrase “in order to” nearly 40 times in this manuscript.  Use of “in order to” is poor grammar.  It’s one of the few useful things I retain from grammar school, but “in order to” should never be used.  Instead, just “to.”  For example, “In order to provide valuable input data…” is more appropriately simply “To provide valuable input data…”

This article analyzes the doppler frequency discrimination performance of the ALADIN. The detailed description is helpful and insightful. It makes a big contribution to the design and construction of spaceborne Doppler lidar receiver. The suggestions of this article are listed as bellow:

1.The optical frequency discrimination performance of Doppler lidar seriously affects the accuracy of wind speed measurement. This paper does not mention how the frequency discrimination performance affects the measurement accuracy with the system operation.

2.This paper mentioned that the frequency discrimination performance of Doppler lidar is mainly caused by the alignment stability and the laser quality. In my experience, the performance of the detector also affects the efficiency of photoelectric conversion, especially for ACCD.

3.Why don’t chose PMT detectors in this system, and apply FPI for Doppler frequency discrimination of Mie doppler signals?

4.How to normalize the long-term FPI transmission curves, in the line 380 and why the direct channel normalization forms of Fig. 5(a) and Fig. 5(b) are inconsistent?

5.The beam divergence angle of the incident FPI has a serious impact on the transmittance of the FPI in line 570. The larger divergence angle corresponds to the greater FWHM and the lower transmittance of the FPI. According to the optical path setting in Fig. 1, only one collimating mirror is used to collimate the received signal light. The signal light is reflected by the Fizeau interferometer and then enters the direct channel of the FPI. The reflected light of the direct channel enters the reflection channel of the FPI, and the optical path is too long. How to ensure the incident FPI with minimal divergence angle.

6.In line 135, 20 laser pulses for each measurement are used, while 190 pulses is used in 145 line, please explain it clearly.

7.Lines 155-190 describe the process of laser energy drift correction. For the reader's convenience, a simple system schematic diagram including optical and electrical devices is recommended.

8.Line 330, "in the top panel (a)" may refer to "in the top panel of Fig. 4 (a)", and line 365 "in the bottom panel (b)" may refer to "in the bottom panel of Fig. 4 (a)".