It is important to  assess SGLI-observed cloud properties using surface data,such as SKYNET data. The authors discuss  the quality of the two most fundamental cloud properties—COD and  CER of both water and ice clouds From SGLI. In fact, it is not easy to do this kind of assessment. In particular, the author should pay attention to the following issues：

1. Objects （clouds） seen from satellites and the ground need to be substantially identical.

2. The algorithm that distinguishes between ice and water clouds requires precision and rigor to achieve good results.

3. As the authors say, the SGLI-observed CER exhibits poorer agreement than does the COD, with the SGLI values being generally higher than the sky radiometer values. And what's the reason? and how to improve? should be discussed in deatail.

4. some errors such as: ... "and (iii) the SGLI COD can be underestimated (resp. underestimated)  for optically thick (resp. thin) clouds. in line 386;" ..."for ice clouds and the tendency to underestimate (resp. overestimate) the COD in SGLI observations for optically  thick (resp. thin) clouds." in line 35; 

This paper validates the quality of SGLI cloud products through comparing with ground-based sky radiometer measurements for a few years. The authors have experiences for this sort of comparing work, thus their method seems adequate and contains enough content for publication. I, however, notice the following points that need to be clarified.

L174-176

‘we have underestimated values from SGLI for relatively high COD for both water and ice clouds, whereas most of the data samples show an overestimated COD from SGLI when they are less than ~20 and ~10 for water and ice clouds, respectively.’  Could the authors explain the mechanism to cause this tendency?

L189-190

‘However, note that CER depends on not only the reflectance function of the absorbing wavelength but also the COD (Zhang and Platnick, 2011).’ If so, how about showing CER comparison like Fig. 2 for thinner and thicker cloud cases separately?

L218-219

‘the larger fractions of water in the middle and lower cloud portions may make considering the water cloud phase reasonably valid for surface observations.’ Due to unreadability, could the authors rephrase the sentence?

L247  ‘Deep convective and nimbostratus clouds have COD values of greater than 23. ‘ But COD less than 20 seems to be included in Fig.4(c) (deep convection).