This technical note provides valuable information on the characteristics of thunderstorms and non-thunderstorms cells based on polarimetric radar and lightning observations over South China.

The manuscript is well written, the methodology is clear, and the results are correct and valuable. They provide quantitative data on the conditions that lead to the occurrence of lightning, and can be helpful to improve lightning forecasting systems. I have some minor concerns before the technical note can be published:

- General: The thunderstorms and non-thunderstorms cell analysed in this study have already been analysed by the authors in two previously published studies, as acknowledged here. In particular, the Z_H, the Z_DR, the content of graupel and the graupel shape has already been analysed in Zhao et al. (2022, GRL). In this study, the ice microphysics associated with graupel is studied by comparing different thunderstorms and non-thunderstorm cells instead of by only studying the evolution of particular cells. This is novel. However, I would appreciate a more detailed description of the novelty of this work with respect to the two previous studies.

In addition, I would appreciate more information about the analysis of these thunderstorms from the two previously published papers. For example, while reading this technical note I wondered if the content of aerosols could pay a role in this analysis, as they are not mentioned here. Later, I noted that the content of aerosols in these thunderstorms was analysed in Zhao et al. (2022, GRL). Mentioning this in this paper could help the reader understanding the analysed thunderstorms.

- Abstract: I think authors should define Z_DR

- Line 61: Mention in-cloud corona discharges when discussing the different types of lightning activity.

- Lins 69-74: Aerosols play an important role in cloud electrification. Please mention.

- Line 122: Plotting the analysed area could be helpful for the readers.

- Line 142: Please provide the coverage of LFEDA.

- Line 153: This has already been said before.

- Line 190: 98% of total FLF are IC. Could you compare this percentage with other studies? I do not say that the authors have to do this, but this could be interesting.

- Results: I miss an analysis of the significance of the differences between thunderstorms and non-thunderstorms cells, which was for example provided in Zhao et al. (2022, GRL).

- Figure 2: Why is the graupel volume larger in non-thunderstorm cells during the first stage? Could you provide an explanation?

- Line 259: "...in this volume These characteristics..." -> "...in this volume. These characteristics..."

General comments

The objective of this paper is to «better understand the ice microphysics associated with graupel within thunderstorms». To this aim, 57 isolated thunderstorms and 39 isolated non-thunderstorms over South China in the warm season (2016 and 2017) are studied using the Guangzhou S-pol radar with a hydrometeor classification algorithm, and 3 independant lightning location systems. The investigated storms are divided into 3 stages: first radar volume scanning (Z_H > 5 dBZ), intermediate radar volume scanning (between the first and third stages), radar volume scanning where the first lightning flash is detected (for thunderstorms) or where the most intense radar echo is detected (for non-thunderstorms).

This paper addresses the microphysical characteristics of the storm at the occurrence of the first lightning flash which is a relevant scientific question. However, technical notes should report «new developments, significant advances, or novel aspects of experimental and theoretical methods and techniques that are relevant for scientific investigations within the scope of the journal». In the submitted manuscript, I was not able to find which development, method or technique is new. Same data and tools as in Zhao et al. (2021) and Zhao et al. (2022) are used. The authors emphasize that their study is based on a large sample statistics of thunderstorms and non-thunderstorms (see their Table 1 where Mattos et al. (2016) is not referenced), and that they based their study on graupel particles inferred by a hydrometeor classification method (also done in previous papers by Zhao et al. among others). This point should be clarified. Also, I wonder why the authors did not use the 3D mapping of lightning flashes from LFEDA to investigate further the relationship between ice microphysics and the first lightning flash (triggering altitude, altitude of the charged regions...).

The article requires thorough proofreading. Abbreviations (e.g. Z_DR, Z_H) should be defined.

Most of the time, the number and quality of references are appropriate. But some references are missing for some statements. For exemple:

• lines 83-87: references are missing about coalescence-freezing rime as the main pathway for graupel formation in warm-base clouds
• lines 105-109: there are 2 different statements: studies about relationships between ice microphysics and lightning activity, and methods to predict the first lightning flash occurrence based on the riming electrification mechanism. Please separate the references associated with the 2 statements. The important work of Latham et al. (2007) should be cited.
• Concerning the non-inductive mechanism (lines 76-80), modelling studies also support this mechanism as the main contributor in charge separation conducive to lightning flash triggering in timescales relevant to storm duration  (e.g. Helsdon et al., 2001; Mansell et al., 2005; Barthe and Pinty, 2007).
• The reference Boggs et al. (2002) is not relevant for tropospheric lightning flashes; it rather deals with upward electrical discharges from thunderstorms.

Specific comments

line 85: what is «supercooled temperature»?

line 109-110: add a reference for this statement

lines 171-173: a map showing the analysis area would be appreciated, even if it is already shown in Zhao et al. (2021, 2022)

line 204: the average duration between the 1st and the 3rd stage is 19 and 24 min, respectively, for thunderstorms and non-thunderstorms. However, the definition of the 3rd stage is different for these two types of storms. It would be interesting to compare the duration between the 1st stage and the occurrence of the maximum radar echo.

line 211: there is no «grey triangles» in Figure 1a

line 216: the -30°C isotherm is not shown on Figure 1

line 221: «deep convective clouds, indicated by thunderstorms, were formed when first lightning flashes occurred»: does it mean that those cloud are not considered as deep convective if there’s no lightning activity?

Line 225: «values… are slight». Give values to justify this statement.

line 239: the 0°C isotherm is not plotted in Figure 1

lines 247-250: this sentence and the causal link with the previous one are not clear

lines 251-254: not clear how the graupel volume is computed: please clarify. What is a height layer?

Lines 254-255: «graupel is rare in thunderstorms or non-thunderstorms during the first stage of cloud development». As stated in Lang and Rutledge (2011), «the existence of 30-dBZ echo above the freezing altitude is a “necessary” condition (in ~90% of cases) for lightning occurrence». This value is well above the 5 dBZ threshold used in this study to detect the 1st stage of the storm and can explain why graupel is rare in stage 1. From a modeling study of an isolated thunderstorm, Barthe and Pinty (2007) showed a delay ~ 20 min between the first occurrence of graupel and the first lightning flash (see their Figure 1). In this case study, this delay was attributed to the time for graupel and vapor-grown ice to locally gain charge through the non-inductive mechanism, and to the sedimentation of the different particles leading to macroscopic charge separation.

line 256: how to explain the larger graupel volume in non-thunderstorms during stage 1?

line 257: «is reached» → reaches

lines 263-264: «the black dots … in km³»: not clear, please rephrase.

Lines 271-278: the graupel volume is also found in upper layers in thunderstorms compared to non-thunderstorms. For non-thunderstorms, graupel volume is only found below 4 km above the melting layer, while it reaches altitudes well above the -38°C isotherm in thunderstorms. Mattos et al. (2016) observed different radar signatures in the glaciated part of thunderstorms and non-thunderstorms during the CHUVA field campaign. It could have been discussed.

Lines 289-301: add relevant references in this paragraph.

Line 304: what are «the average intensities of the Z_H and Z_DR»? How are they computed?

Line 327: the differences in graupel formation for winter snowstorms and warm-season thunderstorms should be make clear.

References

Barthe, C., and J.-P. Pinty: Simulation of electrified storms with comparison of the charge structure and lightning efficiency, J. Geophys. Res., 112, D19204, doi:10.1029/2006JD008241, 2007.

Boggs, L. D., Mach, D., Bruning, E., Liu, N., van der Velde, O. A., Montanyá, J., Cummer, S., Palivec, K., Chmielewski, V., MacGorman, D., and Peterson, M.: Upward propagation of gigantic jets revealed by 3D radio and optical mapping, Science Advances, 8, eabl8731, doi:45110.1126/sciadv.abl8731, 2022.

Helsdon Jr., J. H., W. A. Wojcik, and R. D. Farley: An examination of thunderstorm-charging mechanisms using a two-dimensional storm electrification model, J. Geophys. Res., 106(D1), 1165–1192, doi:10.1029/2000JD900532, 2001

Lang, T. J., and Rutledge, S. A.: A framework for the statistical analysis of large radar and lightning datasets: results from STEPS 2000, Monthly Weather Review, 139, 2536–2551, https://doi.org/10.1175/MWR-D-10-05000.1, 2011.

Mansell, E. R., D. R. MacGorman, C. L. Ziegler, and J. M. Straka: Charge structure and lightning sensitivity in a simulated multicell thunderstorm, J. Geophys. Res., 110, D12101, doi:10.1029/2004JD005287, 2005.

Mattos, E. V., Machado, L. A. T., Williams, E. R., and Albrecht, R. I.: Polarimetric radar characteristics of storms with and without lightning activity, Journal of Geophysical Research: Atmospheres, 121, 14201–14220, https://doi.org/10.1002/2016JD025142, 2016.

Mattos, E. V., Machado, L. A. T., Williams, E. R., Goodman, S. J., Blakeslee, R. J., and Bailey, J. C.: Electrification life cycle of incipient thunderstorms, Journal of Geophysical Research: Atmospheres, 122, 4670–4697, https://doi.org/10.1002/2016JD025772, 2017.

Zhao, C., Zheng, D., Zhang, Y. J., Liu, X., Zhang, Y., Yao, W., and Zhang, W.: Turbulence characteristics before the occurrence of the first flash in thunderstorms and non-thunderstorms, https://doi.org/10.1029/2021GL094821, 2021.

Zhao, C., Zhang, Y. J., Zheng, D., Liu, X., Zhang, Y., Fan, X., Yao, W., and Zhang, W.: Using polarimetric radar observations to characterize first echoes of thunderstorms and nonthunderstorms: A comparative study, Journal of Geophysical Research: Atmospheres, 127, e2022JD036671, https://doi.org/10.1029/2022JD036671, 2022.