This study investigates snow growth characteristics using triple-frequency radar observations from multiple field campaigns. Based on early seminal concept connecting snow microphysics to triple-frequency signatures, this study demonstrates how these early findings can be used to compare the climatology of snow microphysics and benefit future triple-frequency satellite missions. In particular, the measurements at Southern China are unique, and I congratulate the authors for generating an important dataset.The manuscript is well written, organized, and clear. I have nothing substantive to add, aside from some minor suggestions listed below.

L66: The Introduction is a good summary of current understanding of triple frequency technology and its applications. I suggest adding some discussion on how such observation may benefit model development.

L198: Why a single A-Z parameterization is used?

L279: A more detailed physical interpretation of the variations of DWR-Z dependence, as well as how cloud-top temperature affects the radar threshold for non-Rayleigh scattering, is needed.

Fig. 6: The comparison is important. It demonstrates that the velocity-based FR estimate is a good approximate for heavy riming. Does this mean that we do not need W-band radar considering the attenuation effect, if the velocity is accurately measured?

L320: Ideally, much more data should be used for a climatological analysis. So, the impact of relatively short observation period for current results should be discussed.

General Comment

This study investigated triple frequency radar data from five locations and discussed ice microphysics for snow riming in those locations. The results seem reasonable and to be interpret reasonably. Figures are clear, and the text is well organized. Some of the datasets (e.g. AWARE, TRIPEx-pol, METRICs) might not be enough to explain climatology or general characteristics of snow at the site, but this study got foot in door of analyzing climatological/statistical characteristics of snow riming using the multi-frequency radar datasets. The manuscript should be published after addressing minor comments below.

Specific comments

1. Please add an advantage of using triple frequency rather measurements than dual frequency measurements or Dual frequency + Doppler measurements. The discussions of snow riming and aggregation are well explained by the triple frequency data analysis, but I felt that the discussions in the manuscript can also be explained using the dual frequency result only (DWR Ka,W and Zka + FR). Please more highlight the advantage of adding the third frequency (i.e. X or C) and what could be clarified by the triple frequency data analysis and not be clarified by the dual frequency (+FR) analysis only.
2. Could you explain more why non-Rayleigh scattering signatures appear at cases when large radar reflectivity at a colder cloud top? I am interesting in discussion of microphysics of this characteristics accounting for the environments at the site.
3. Lines 285-286: I cannot see this signature for the METRICs and BAECC datasets.
4. I would prefer stating “warmer/colder” than a specific temperature rather than using “above/below” a specific temperature. Because I assumed that the colder temperature represents higher altitude, I was sometime confused with this about the altitudes.
5. I am interested in the following discussions, and it would be great to add the discussions in the manuscript.

• How/Why the warmer cloud top temperature correlates with high DWR?
• Why high latitude sites have warmer cloud top and heterogeneous nucleation? Considering that high latitude sites could have lower aerosol concentrations, heterogeneous nucleation could be less. Please give some comments about this.