This is a detailed study using WRF and 4 microphysics schemes for 8 snow events

during ICE-POP for the 2018 Winter Olympics, focusing on one of each of 3 types

of events, cold-low, warm-low and air-sea interaction. The inner domain is at

a relatively high resolution of 1 km. Observations used in addition to the 

surface AWS stations include disdrometers and radar from which particle types

were derived.

The paper distinguishes some large differences in particle types between the 

schemes and verifies them against observations. Some additional understanding

is gained by evaluating the importance of every process as a source and sink

for each particle type in each case. This is a large amount of data that is

presented and a good attempt is made to derive the most important points

and distinctions between the schemes from it.

I think the paper is acceptable after minor revisions. The level of detail

may appeal mostly to microphysics parameterization developers, and is

probably more than most would read through, but the conclusions are of more

general interest. I have itemized my Minor Points below,  the response to some 

of which may help to improve the paper.

Minor Points

1. line 33. What is meant by "inefficient melting"? Less efficient?

2. L52. convections -> convection here and several places. Common English error.

3. L76 Thompson.

4. L76 "snow efficiently affects precipitation efficiency for" rephrase to

not include efficient twice.

5. L114 "of precipitationi systems" typo.

6. Table 2. Refers to Morcrette. I am sure this is not the correct reference.

7. Table 4. WDN typo.

8. L213 and Figure 5. Case color code should be mentioned in the text too.

9. L216. How is a rate used for an accumulated amount in the whole period? It

says mm h-1.

10. L218 and Figure 5. Hard to interpret biases without absolute totals which 

vary from 6 mm in Case 3 to 57 mm in Case 4. Perhaps put totals from Table 1 on

Figure 5.

11. L222. Hard to tell from Figure 6 that these schemes have more liquid rain.

I would suggest finding a different way to show precip type. Either a 

separate plot of type, or shading by type and contouring amount.

12. Figure 8. It was hard to find qc because the dash length does not match the

key. Make the key pattern exactly match the plot. Also hard to see that qs is

a dot pattern in the key.

13. L262. Important to note that schemes with QCGEN have condensation mostly

there while those without combine condensation and evaporation in QCCON.

Is there much separate condensation in QCCON in the QCGEN schemes or is this

all just evaporation?

14. Figure 9a-d. Maybe QRWET should be QCWET in labels. Check all these against

Table 4 names.

15. Table 4. QRAUT in cloud section could be QCAUT? I am not sure about the

rules for naming when the same processes may have different names. QCACR

for example has the same name.

15. Figure 9, etc. Can a scaling number be put on these plots to show relative 

size? L270 points out an important scale difference that would not have

been seen in the Figure. For example, add what 100 equals in absolute terms.

16. L301. As in note 11 above, this is hard to see.

17. L372. Should be Fig. 7l.

18. L373. Westerly wind is weak. The model clearly has an onshore wind that

must be mainly northerly. This component should be mentioned.

Review of “Simulated microphysical properties of winter storms from bulk-type microphysics schemes and their evaluation in the WRF (v4.1.3) model during the ICE-POP 2018 field campaign” by Ko et al. 2022

Recommendation: Could be acceptable for publication following mandatory major revision

This study evaluates the performance of four bulk microphysics schemes for the simulation of snowstorms during the Pyeongchang 2018 Olympics and Paralympic Games. The analysis compares the amount of ice, snow, precipitation and cloud water predicted by the various schemes, and attributes the processes giving the production of the relevant hydrometeor categories. They conclude that melting is key for generating rain, and the bias in precipitation for war-low and cold-low biases can be mitigated through the use of inefficient melting in all schemes. Although I think the paper makes a contribution by testing these microphysical schemes in new meteorological situations, there are aspects of the presentation that should be improved before the paper is accepted for publication.

MAJOR COMMENTS

1. I found that the introduction was not overly focused. Although the stated goal of this study is to evaluate the performance of the microphysical schemes in the simulation of wintertime precipitation, much of the introduction compared how these schemes have previously been used to simulate convection. There should have been more focus on how the use of these schemes has been evaluated in simulations of wintertime storms, and also the understanding that has been gained from past observational and modeling studies of winter storms should be highlighted. There were many past studies of winter storms that were not referenced.

2. I am concerned about the resolution of the model that is being used as the highest horizontal grid spacing is 1 km. A lot of the convection and generating cells that commonly occur in winter time storms are on scales of much less than 1 km. Thus, the model is not able to represent well the spatial scales on which the evolution of these storms is occurring. This limitation should be clearly acknowledged (or run at finer resolution) and explain the caveats with the interpretation of this study because of this resolution difficulty.

3. The authors use four different microphysical schemes in their investigations, but do not use some of the most state-of-art microphysics schemes in their simulations. Why is this? For example, the P3 schemes (Morrison and Milbrandt 2015), Predicted Particle Properties scheme is the next generation parameterization scheme that uses a very different approach for representing ice, and would offer an interesting complement to the schemes that are presented here. It predicts bulk properties rather than predicting separate species, which eliminates unphysical conversion processes between traditional ice categories, and hence can be used for giving a better comparison against observations.

4. For the most part, the writing in the manuscript where the results of the different simulations and comparisons against observations is performed is overly qualitative. Terms such as “overestimate”, “matches well”, “substantial amount”, “abundant”, “insignificant”, “similar” and others are used excessively, with little information on what this actually means. The paper would read more clearly if these descriptions were made more quantitative with appropriate reference to the figures. There are a few places where this is done (e.g., approximately 10 times larger, magnitude is 5.5 g/kg), but this could be done much more effectively.