please see attached pdf.

Review of “Unique Microphysical Properties of Small Boundary Layer Ice

Particles under Pristine Conditions on Dome C, Antarctica” by Adrian Hamel et al.

Overview of the paper :

This paper analyses ground-based, optical in situ measurements of falling ice crystals sampled by the PPD-2K during several weeks in Summer at Dome C. Number concentrations, spherical equivalent diameters, habits and morphological complexity are derived from the optical measurements using well-established algorithms and methods. These microphysical properties of ice crystals with size larger than 11 µm are statistically analyzed for different thermodynamic conditions measured at the station. A lidar is also used to identify specific “meteorological” events such as precipitation, diamond dust and ice (or liquid) fogs. The results of the paper begin with an interesting overview of the ice crystal habit fraction for different atmospheric conditions and hydrometeor sizes. The paper then focuses on a case study encompassing ice crystal precipitation and an ice fog event. The temporal evolution of the microphysical properties of the ice fog is presented clearly. The analysis shows that there are distinct properties between ice crystal precipition and ice fogs, which can be captured by the modes of the particle size distribution (PSD) measured by the PPD-2K. A statistical analysis is performed on the microphysical properties of almost all ice fog and cirrus precipitation or diamond dust events. The PSD, concentration and ice crystal morphology parameters are inter-compared and also compared to previous measurements carried out in Fairbanks. The differences in these properties are discussed as are the possible nucleation processes occurring in ice fogs. The paper concludes whith a discussion of the specific properties of ice fogs in inland Antarctica, emphasizing the necessity of new parametrization of ice crystal-related processes and properties in this region.

The unique instrumental setup deployed at Dome-C for this study allows for detailed and pertinent characterization of the microphysical properties of ice crystals. In my opinion, the strength of the paper lies in its analysis of the morphological properties (habit and complexity) of ice crystals for different thermodynamic conditions and “cloud types”. This approach is essential for improving our understanding and modeling of the microphysical processes that control the life cycle of ice crystals and precipitation in this region of the globe. The relevance of this study is clear, and the results are original and very interesting. While the description of the methods and results is of good quality, it could be improved by avoiding certain repetitions and providing clarifications, mainly on the sampling strategy/limitations and some of the results. I would recommend minor revisions before the manuscript can be considered for publication in ACP.

General comments :

1. Introduction/motivation : I think the motivation/storyline of this study could be improved to deliver a clearer message. It should state more clearly what is needed in models to better simulate the life cycle of ice fogs. Modeling and characterizing the properties of ice precipitation is also subject to significant uncertainty in Antarctica. For instance, I think there is a need to improve the modeling of the fall speed of ice crystals which depends on their size and shape. The authors also mention that their measurements are valuable for validation of satellite retrievals. I would suggest they clarify their statement by demonstrating how their measurements could improve the remote sensing retrievals (not just satellite also more importantly ground based lidar/radar retrievals). Strong assumptions are needed to retrieve basic properties such as ice number concentration and reliable retrievals are even more challenging near the ground, especially in Antarctica. Did the authors have in mind that their measurements could be used to compute look-up tables or parametrization relating ice water content to radar reflectivity or lidar backscattering for several ice crystal morphologies ? (please refer to the detailed comments below)

2. Sampling of ice crystals (Methods and results section) : I recognize the challenges involved in performing microphysical measurements at Dome C. Although the methods section is well described, i think some details are missing regarding the ice crystal sampling and the potential additional sizing and counting uncertainties arising from it. In particular, the impact of the pumping speed, the influence of wind speed and direction and underlying non isokinetic issues inside the instrument inlet should be addressed.

3. Liquid fog event (results section and discussion) : it is mentioned several times that a liquid fog was sampled during the campaign. The corresponding data are included in the “all event” dataset and seem to have a significant impact on the maximum measured particle concentration presented in Table 2 as well as on the PSD and habit fractions shown in Figure 7.I understand that the authors intend to present this event in a separate publication. However, to clarify some of the results, I would suggest not including the liquid fog event in the “all event” category. Fig7a shows a sharp increase in the concentration for small sizes, which could be attributed to the presence of supercooled water droplets representative of this liquid fog event only. The habit fractions also appearto be impacted by this event which makes the inter comparisons less meaningful. Additionally, the wind speed and direction could also affect the sampling of small water droplets and larger ice crystals differently. I would recommend either dismissing the liquid fog event from the analysis (in the all category) or investigating this event separately (by adding a new category to table 2 and figure 7). This analysis could also provide information on the nucleation mechanisms in ice fogs.

Specific comments :

1. Introduction

Line 16 : Could be more specific about your statement that “a fraction of about 40% clear-sky precipitation” ? What do the authors mean by clear-sky precipitation : diamond dust only ?

Line 16-17 : In Walden et al., 2003 it is said that falling ice crystals collected on a gridded glass slide at the South Pole Station were photographed through a microscope. They showed that the effective radius of diamond dust was 12-15 µm ; 11 µm for blowing snow and 24µm for snow grains. These values seem quite small as the effective size is not always representative of the actual size of ice crystals, especially for hexagonal columns or plates. Walden et al., 2003 also show that diamond dust sizes range from 2 µm to 1000 µm. It think that it would be wise to mention the variability in ice crystal size or to specify that this 10 µm value correspond to an effective size. Since the PPD-2K cannot in principle detect crystals small than 10 µm, readers may wonder how these new measurements will contribute to our understanding of microphysical properties when most of the crystals are smaller than 10 µm.

Line 20-22 : The authors mention that the clean environment of Antarctica enables us to study cirrus formation and that several measurements have been performed in the Antarctic boundary layer. This is confusing as cirrus clouds are typically located in the high/free troposphere. I would recommend to use the term ice clouds or ice containing clouds.

Lines 24-29 : The authors state that previous instruments could not reliably determine the properties of ice crystals with sizes below 50 µm. While I agree with this, it is not consistent with what is written in lines 16-17. Furthermore, i would recommend that the authors elaborate on why this information on 50µm ice crystal is necessary for improving the estimation of the radiative impact in the boundary layer, providing some references.

Line 35 : Is the parametrization used in Girard and Blanchet, 2001a, the only one available and most commonly used ? Are there any other parametrization with lower number concentrations ?

Line 36: the sentence beginning with “During winter time falling diamond dust is simulated to increase the downward infrared flux by up to .....” is unclear. Increased relative to what ? Not taking diamond dust into account ?

Line 42 : What is the open question regarding particle morphology ?

Line 47 : it is stated that in situ measurements of the microphysical properties of ice crystals are valuable for validating of satellite data retrievals and atmospheric modelling. This is quite a general justification. I would recommend that the authors be more precise by indicating some of the shortcomings of remote sensing retrievals (not only satellite as i think your results might be more interesting for radar/lidar retrievals from ground-based systems). Regarding atmospheric modeling it would also be interesting to show which parametrization or process should be improved.

2. Methods

I would reorganize this section slightly, starting with a short paragraph mentioning the sampling procedure, and the instruments and tools that will be used in the study. The atmospheric instrumentation could come first but the authors need to explain why the lidar measurements and air parcel back trajectories are necessary. The sampling strategy could also be developed further, with a focus on the possible impact of wind on the measurements.

Lines 96-97 : Could you explain why and how the diffraction patterns with Mie fringes for spherical particles would increase the mean complexity ? Could you please elaborate on this ?

Line 111-112 : How was this uncertainty estimated ? Is a mass flow of 5 l /min sufficient to achieve accurate measurements ? In this respect, did the authors investigated the impact of the flow rate on the measured concentrations and on the size of ice crystals ? What is the sampling speed in the instrument’s inlet ? I’m wondering if you are working in isokinetic conditions ?

Line 132 : This sentence is unclear. Are the inlet and the metal tube the same thing? Could you please clarify this ?

Line 140-145 : The author mention that, for low wind speeds, the data is usually excluded from the analysis due to aerosol pollution. However, what is the effect of wind speed and direction on the ice crystal sampling ? Do you expect there to be a lot of undersampling of falling ice crystals when the wind speed is high ?

Figure 1 : What is the meaning of the color bar “time of flight” ?

3. Results

Line 181: Could you clarify this sentence : “the fraction of rough particle was with 21%the highest” ?

Line 190 : Do you mean that sublimating ice particles could be frozen droplets ? This is interesting. Do you have a method such as the analysis of lidar depolarization or PPD-2k to distinguish between these two types of hydrometeors ?

Line 202-203 : So is it plausible that the sublimating particles are frozen or quasi spherical droplets ? What level of confidence do you have in discriminating between the two types ?

Line 207 : I think it’s a bit disturbing to call the ice fog a “ground level thin cirrus”

Line 233 : How can you be sure that an increase in the ice crystal complexity is indicative of homogeneous freezing ? What about other mechanisms that could lead to an increase in complexity such as sublimation, riming, rapid growth …. ?

Line 285 : Is there a way to discriminate between cirrus precipitation and diamond dust ? Their formation processes should be different as well as their microphysical properties.

Line 306-308 : Can you really be sure that all events exhibit the ice fog mode at 11 µm given that the instrument cannot quantify ice crystals smaller than 11 µm ? It is also possible that the counting is overestimated in this size range. Is there typically a larger measurement uncertainty in this size range ?

Line 331 and Table 2 : The concentration of droplets in liquid fog is included in the analysis to derive a maximum concentration during the complete measurement period. I think that it would be more relevant to separate these events from the other cases and even more pertinent to present the PSD of this liquid fog event. Why did the author choose to consider liquid fogs in a separate paper ?

Line 332 : I think the comments on ice fog from Fairbanks have already been mention earlier.

Line 337-340 : It is indeed surprising that the habit fractions are so similar. Are the thermodynamic conditions similar too ? How would you explain this ? Once again, the low level supercooled liquid clouds should be excluded from this analysis and assigned to a specific liquid fog category (in table 2 and Fig 7a).

Line 357 : This sentence should be rewritten as “is with” is unclear. During cpdd events, 31% of ice crystals are pristine crystals .....

Line 363 : This sentence could be placed in the discussion section, as it has already been mentioned several times in the results section.

Line 376 : The maximum dimension can be 2.5 times the spherical equivalent diameter. I suggest placing this information the methods section or result section when presenting PPD-2K measurements.

Discussion : This section should include a brief description of the limitations of the measurements due to sampling and possible contamination from blowing snow, and how it could be improved.

Appendix : Appendix A and D could be summarized in the main text.