Multiple Characteristics of Precipitation Inferred from Wind Profiler Radar Doppler Spectra

Round 1

Reviewer 1 Report

Please see the attachment for the comments.

Comments for author File: Comments.pdf

Author Response

Please find our reply in the file attached.

Author Response File: Author Response.docx

Reviewer 2 Report

This manuscript discusses a new methodology for estimating of precipitation characteristics like type and rate from radar wind profiler data. The result has been compared against data collected from ground and claimed that the new method is simpler and accurate than the exist method which can also classify the precipitation type. Based on the presented methodology, the novelty of this research is related to the classification of precipitation to rain, snow and mix of both. The manuscript and analysis may need a minor revision based on the following comment.

MRR and RWP use different frequencies of 24.23 and 1247 MHz respectively, which are sensitive to different droplet size in the atmosphere. How the analysis from these two different data sources are comparable?

The data used in this study is just for a 48h period in 2017. Although the authors explained the reason of using this period in lines 99 to 101, I believe that different case study is needed to claim that the new methods works better.  

In lines 158 and 15 the authors interduces some thresholds for detecting rain, snow and mix of both. Although it referred to a paper, how much these values are reliable?

Author Response

Please find our reply in the file attached.

Author Response File: Author Response.docx

Reviewer 3 Report

This manuscript has sufficient merits to be considered for publication, revisions are needed in these aspects:

-          The introduction of instrumentation is overly detailed and can be simplified.

-          Need to clearly separate the instrumentation processing already done by manufacturers (and existing method) and the new processing methods done by the authors. For example, any new method applied for spectrum filtering?

-          Method 2 vs Method 1: It is not clear if the Method 2 has better results than Method 1, or not, for the wind estimation and Doppler processing.

-          Table 2, MRR2, frequency value is in GHz unit.

-          Figure 4 needs more explanation, why (a) and (c) are totally identical?

-          Need more discussions on A73 and R95, what physical basis for each method are? Why choose them and how they compare fundamentally?

-          Understand the verification of classification is only at lowest altitude based on the disdrometer (Table 6). The MRR seems did not provide much data and support in the validation. Ideally, this should be compared with polarimetric radar classifications in the same region.  

Author Response

Please find our reply in the file attached.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Thanks for the quick revisions from the authors, 

the improved revision seems appropriate to publish, there are still concerns on the validation of the classification, but the improvement of functionality from the existing approach is acknowledged. 

Author Response

Thank you very much for the new review of the corrected version of the manuscript. To take into account your comment about the limitations of the validation of the classification, we have modified the following sentences in the Conclusions “Despite some limitations on the comparison procedure, qualitative and quantitative results based on contingency table verification scores indicated an overall good performance of the estimated precipitation for snow and rain types showing promise for further application unlike mixed types that were not correctly diagnosed”. We also introduced this idea in the following sentence of the Abstract: “Results indicate (…) a good skill to determine precipitation type when comparing the lowest estimate with disdrometer at ground data for snow and rain, but not for mixed precipitation cases that were not correctly identified”.