Review of “Advantages of using multiple Doppler radars with different wavelengths for three-dimensional wind retrieval.” By Tsai et al.

This paper provides an overview of multi-Doppler analyses of a bow-echo that passed through South Korea on 2 August 2020. This convective system was sampled by a network of 11 radars of varying wavelengths. The authors conduct experiments where variational wind retrievals are made using only specific wavelength radars for given field experiments. While the authors conduct an exercise that would have potentially useful implications for how wind retrievals are calculated, there are numerous problems that prevent me from recommending this paper for publication.  

Fatal flaws:

For one, the results are not placed in the context of the scanning strategies of the radars. Were the radars on a synchronous scan strategy? In addition, The S-band radars are all placed relatively close together, while the C and X band radars are further out, making a more optimal baseline for multiple Doppler retrievals. Could this also be a factor as well?

In the analysis of the updraft cores I found it hard to determine the number of updraft cores simply by eye. Have the authors considered counting these using thresholding techniques (see Varble et al (2014?)). Finally, The MB and RMSD values in the quantitative analysis in Table 4 do not clearly favor the SCX regime, and do not seem to demonstrate any quantitative improvement of using SCX over just S.  Do the authors have statistics for earlier and later stages of this storm, or other cases, that would provide a larger amount of data for analysis?

Major comments:

Lines 87: Cha and Bell (2023) also added the IBM method to SAMURAI. Please mention their work in your literature review.

The authors should also mention the 3DVAR work done by Shapiro and Potvin that are now in PyDDA (Jackson et al. 2020). These works should also be mentioned in the literature review.

Minor/technical comments:

Line 100-103: Run on sentence.

Line 104: “lower” should be “coarser”

174: “frozing” should be “freezing”

Figure 2: The station measurements are difficult to read on the figure. I would suggest removing some and making the font size bigger, or removing all of them.

Line 216: Extra “.”

Line 361: “An”

Line 545: “leading edge”

Figure 4: “reflectivity”

References:

Varble, A., E. J. Zipser, A. M. Fridlind, P. Zhu, A. S. Ackerman, J.-P. Chaboureau, S. Collis, J. Fan, A. Hill, and B. Shipway (2014), Evaluation of cloud-resolving and limited area model intercomparison simulations using TWP-ICE observations: 1. Deep convective updraft properties, J. Geophys. Res. Atmos., 119, 13,891–13,918, doi:10.1002/2013JD021371.

Cha, T., and M. M. Bell, 2023: Three-Dimensional Variational Multi-Doppler Wind Retrieval over Complex Terrain. J. Atmos. Oceanic Technol., 40, 1381–1405, https://doi.org/10.1175/JTECH-D-23-0019.1.

Jackson, R., Collis, S., Lang, T., Potvin, C. and Munson, T. (2020) ‘PyDDA: A Pythonic Direct Data Assimilation Framework for Wind Retrievals’, Journal of Open Research Software, 8(1), p. 20. Available at: https://doi.org/10.5334/jors.264.

The manuscript Advantages of using multiple Doppler radars with different wavelengths for three dimensional wind retrieval by Tsai et al. presents a valuable effort to evaluate the use of different radar bands configuration in the WISSDOM multi-Doppler configurations for observing a severe weather event  around Seoul. However, several aspects of the study require clarification to enhance its scientific rigor and broader applicability. In particular, the intent of the analysis, and the methods used to evaluate the results. The following major and minor comments aim to help the authors improve the clarity, accuracy, and overall impact of their work.

Major comments:

1. There is some ambiguity regarding the intent of the study: is this a case study analysis or a broader investigation into network design? If this is intended as a single-case study, could the authors clarify why conclusions about radar network configuration are generalized? For example, the statement in L550–551 seems to imply a broader applicability, but the findings are drawn from a single squall-line event occurring in a specific region. Would it be more accurate to frame these conclusions in the context of this particular setup?
2. Section 2.4  could benefit from clarification. The content in lines 282–291 might be presented more clearly,  to enhance conciseness, and improve readability. Additionally, lines 292–296 repeat information already discussed in the previous paragraph.
3.  Understanding how the data are gridded and whether filters are applied is essential for assessing quality. For instance, are filters in WISDOMM applied that might mitigate aliasing errors? Additionally, how are the gridding parameters (e.g., horizontal/vertical resolution, interpolation scheme) chosen, and how might these influence the retrieved wind field?
4. “The bow shape of the squall line is not particularly evident across the different scenarios, and the assertion that SCX provides the most “accurate” representation of the wind” (L 353-355) field appears speculative, especially since it's based on visual inspection. Could the authors clarify the basis for determining "accuracy" in this context? 
5. The phrase “highlighting the positive impact of adding C and X-band radar obs to S-band radars” (L433) suggests a generalized improvement, but this is essentially a typical gap-filling outcome.
6. Maybe it would be good to clarify that while this set-up is beneficial in this specific case, similar improvements may not occur for other cases? (e.g., to acknowledge that this may vary case-by-case, rather than presenting it as a general characteristic?)
7. The statement about C-band radars providing more near-surface data needs clarification (L397). This behavior is not general and is highly dependent on radar configuration and terrain. Could you be more specific about the setup used in this study (e.g., scanning strategies, beam elevation angles, and terrain impact)? How does WISDOMM deal with terrain?
8. A reference/discussion to lower-boundary limitations due to topography and data availability, and how those may impact wind retrievals  would help contextualize the limitations of the analysis.
9. Given the limitations of using the sounding as “ground truth” (only valid for that grid point), might it be more robust to compare the dual-Doppler retrievals with high-resolution model output? Could a sensitivity analysis be conducted, testing the effects of vertical coverage and radar configuration on retrieval quality? This could provide a more systematic understanding of the strengths and limitations of the setup, especially in the absence of in-situ storm-scale observations.
10. L545-547 statement implies a core assumption, but it’s actually fundamental to the validity of the analysis. Could the authors be more explicit about how the radar sampling strategy, scanning configuration, and network geometry impact the results? For example, how much blockage is present per radar? Are there areas that are not well sampled at low levels by the S-band radar, but are captured by the X-band system?
11.  Understanding how the data are gridded and whether filters are applied is essential for assessing quality. For instance, are filters in WISDOMM applied that might mitigate aliasing errors? Additionally, how are the gridding parameters (e.g., horizontal/vertical resolution, interpolation scheme) chosen, and how might these influence the retrieved wind field?
    
    
    
    

Minor edits:

L58-59: Please, clarify "measure radar reflectivity of the documentation of precipitation structures”. 

L80-81: I suggest rewriting this line as it sounds like the variational method is uniquely a type of multi-doppler technique.

L84: Cha and Bell 2023 developed SAMURAI over complex terrain. This work should be mentioned here.

L113: Although the facts in the=is statement are correct, it reads as “smaller precipitation particle” detection and “gap filling” are directly correlated, when they are not. Please, rewrite this sentence. =

L141: Please clarify this sentence, I don’t understand the relationship between having more radars available and increasing the availability of thermodynamic fields being related.

L143: Please, change to read “storm dynamics and phenomena”.

L144: Please, clarify “their advantages”.

L151: There are two “area” words in the same sentence. 

L166: Please, change to read “spatial” instead of “horizontal”.

L155: Change to read: automatic weather stations (AWS)

L168: Remove the small “s” after AWS.

L174: Please, change to read “freezing”.

L193: Please, change to read “may be affected”.

L196: Please, change to read “convective” instead of “convection”.

L202: Please, change to read “The evolution of”

L207: Please, change to read “”stratiform precipitation was located behind the convective region.”

L208: I believe this would benefit from a reference (after squall line).

L213: Maybe “stratiform precipitation areas”? Instead of “formations”?

L218: Please, change to read “moved easterly”.

L231: Please, change to read: Liou and Chang (2019)

L263-264: Consider re-writing it. It is confusing as it is.

Table 2: Under “Data Implementation- Background”, change “linier interpolation” to “linear interpolation”

L277: Might be useful to cite those studies.

L285: Change to read:” [...] domain, the mean bias (MB) [...]”

L286: Please, clarify “[...] associated with the rate of rise of the sensors [...]”. I assume it is related to the vertical velocity of the sounding sensor, but I don’t understand the relationship between that and the sensor gridspacing (horizontally?) and the mean interpolation to 250m.

L287-288. The sentence does not have much sense on its own. No verb is found.

L300: Move the RWP explanation to L292, when it is referred to. It would be easier for the reader to understand.

L314: I suggest referring the gust front position to the main storm (e.g., 50 km away on the leading edge of the main squall line, at X=125km). 

L340: Please, change to read “reflectivity”, not “relativity”. Same in L403 (Figure 7 caption).

L348: There seems to be an inconsistency regarding the convergence area: if the scenario with only S and C-band data does not exhibit this feature, what, specifically, is influencing it in the other configurations?

L361: Please, change to read “A less”.

L388: I don’t understand the reference to this threshold in this sentence. Please, clarify.

L415: Please, remove “are”.

L449: Please, remove the second “changes”.

L545: Please, change to read “leading edge”.

L563: It would be beneficial to indicate that the colorbar is scaled differently (or scale them all to the same range).

• Reference:

Cha, T., and M. M. Bell, 2023: Three-Dimensional Variational Multi-Doppler Wind Retrieval over Complex Terrain. J. Atmos. Oceanic Technol., 40, 1381–1405, https://doi.org/10.1175/JTECH-D-23-0019.1.