The manuscript is of good scientific quality and is clearly and well written, except for some minor issues, which include, but are not limited to:

• All symbols and acronyms should be defined.
• A few sentences are missing a verb..
• We need to provide the necessary details, e.g., the polynomial (in h) correction in eq. (7).
• It appears that extinction (and backscatter) efficiency is calculated using a temperature-dependent index of refraction m≡m(T), but the dielectric factor K, which also depends on m, is kept as a constant, i.e., no temperature dependency. Within 0-10 °C, |K|² (used in calculating Z) has a ~2% variation, considerably smaller than the variation of m in the same temperature range. Is this the reason that a constant |K|² is assumed? 
• Blue on black (Figure 5) is difficult to see. Consider using a lighter color than the blue used.

Please see the attached PDF for more comprehensive and detailed comments and suggestions.

TRANSLATE with x

English

 

TRANSLATE with

COPY THE URL BELOW

Back

EMBED THE SNIPPET BELOW IN YOUR SITE

 

Enable collaborative features and customize widget:

 

TRANSLATE with x

English

 

TRANSLATE with

COPY THE URL BELOW

Back

EMBED THE SNIPPET BELOW IN YOUR SITE

 

Enable collaborative features and customize widget:

 

TRANSLATE with x

English

 

TRANSLATE with

COPY THE URL BELOW

Back

EMBED THE SNIPPET BELOW IN YOUR SITE

 

Enable collaborative features and customize widget:

 

TRANSLATE with x

English

TRANSLATE with

COPY THE URL BELOW

Back

EMBED THE SNIPPET BELOW IN YOUR SITE

Enable collaborative features and customize widget:

The manuscript by Dong et al. examines rain microphysical properties derived from vertically pointing Ka band radar data on the Tibetan Plateau.  The authors use an optimal estimation (OE) technique constrained by surface disdrometer data to derive the properties of the rainfall DSD in vertical columns above the radar during a single summer season.  They show interesting statistical results of the rainfall properties and discuss possible physical processes that explain the results they find.  The paper is well done.  I do find that more information regarding the OE algorithm should be provided so that an interested reader could replicate their algorithm and fully understand their method.  The details that I think should be provided are given in my specific comments following this paragraph.  Subject to the minor revisions suggested, I find the paper to be a significant contribution to the scientific literature on this topic and worth of publication.  

Specific Comments:

 Line 100:  Should note that the velocity given is the Doppler Velocity since it is weighted by the backscatter cross section.

Line 104:  Does the Haynes et al. algorithm account for the flattening of raindrops as they fall?  This would certainly influence the details of the retrievals for heavier rain. 

Line 133:  It would be useful to see examples of the covarinace matrices or frequency distribution plots of the terms in Sa. 

Line 133: Is the covariance matrix constant or is there a state dependent Sa?  Are the terms considered to be correlated?

Line 134: Additional information should be provided regarding the observational error covariance matrix.  Are the errors considered correlated or uncorrelated.  What is the source of the uncertainty? 

Line 137:  There needs to be more information regarding how the terms of Jacobian, Kx, are calculated and please provide typical examples of the values of Kx.  These details are useful because they show the degree to which the quantities to be retrieved are sensitive to the observations used to constrain them.

Figure 3: The authors should comment on the cause and influence of the systematic biases between the forward calculation and the observations that show up in this plot.  the explanation given regarding the height difference and wind may be reasonable but how those issues would result in the biases shown should be demonstrated. 

The fact that the calculated values do not lie on the 1:1 line  suggest that there may be forward model errors that are not accounted for - i.e. the assumption of spherical raindrops versus reality.  Accounting for forward model errors is key in these types of OE inversions.  Very often the forward model errors are larger (usually significantly larger) than the uncertainties in the observations.  

Figure 4 and 5: Should explain the meaning of the error bars in figures 4 and 5.  How are they derived?

Line 221: The contention made by the authors that N0 is higher and Dm smaller at the melting layer is not obvious in the data shown.  If the authors think this is important, they should devise a way of showing it more clearly.

Line 242: I wonder if it would make sense to separate this into warm rain events versus cold?  It seems likely that the Tsukuba results were associated with rain just below the melting layer when drops derived from large aggregate snow would begin to break up.  That is not what is implied by Figure 7 of this paper.

Figure 8: Explain the meaning of the box plot.  i.e. Median, Interquartile, 90'th, and 10'th percentiles?