Review of "Distinct impacts on precipitation by aerosol radiative effect over three different megacity regions of eastern China" by Sun et al.

The manuscript "Distinct impacts on precipitation by aerosol radiative effect over three different megacity regions of eastern China" mainly studies the influence of aerosol on the start and peak time of precipitation over three different regions, the North China Plain (NCP), the Yangtze River Delta (YRD), and the Pearl River Delta (PRD). In general, the paper is well written and presented in a logical way. It is a timely and important piece of work, and of general interest for Atmospheric Chemistry and Physics related communities. I therefore recommend publication of this paper in Atmospheric Chemistry and Physics after minor revisions. My comments are listed as follows: 

Specific Comments:

1. Lines 158-160: If precipitation occurs in the troposphere and is more affected by aerosols below cloud bases, why is the column-integrated aerosol amount (AOD) not suitable but ground-based observations of PM₅ are more suitable?
2. Lines 160-165: Perhaps the authors’ opinion is that PM₁₀ is more suitable for studying larger particle aerosols such as dust, and there are fewer large particle aerosols in the three selected research areas, so PM₅ is more suitable than PM₁₀ in this study. A clearer description is needed here. And what does “100 nm” represent?
3. Lines 177-179: Why can the previous phenomenon suggest that “it is not suitable to use PM₁₀ mass concentration or AOD at a given moment to examine the influence of aerosol on precipitation”?
4. Line 180: Why do the authors select the 4-hours mean PM₅ mass concentration before precipitation to investigate the impact of aerosols on precipitation? The relationship between daily mean PM_2.5 and 4-hours mean PM_2.5 mass concentration before precipitation, similar to Figure 1, is needed.
5. Line 189: What is the full name of LTS?
6. Line 219: The authors regard PM₅ that is greater than 2 times the standard deviation as abnormal values and remove it, which could lead to mistakenly remove some heavy pollution conditions as abnormal values. Is it reasonable? And what is the proportion of the sample size that is eliminated as abnormal values in the total sample?
7. Lines 344-355: I think it suggests that the aerosol has caused the secondary crest of precipitation peak time 1 hour delayed.
8. Lines 422-424: The information of “the crests of the stratiform precipitation start time are at 19:00 and 17:00 LT under clean and polluted conditions in the afternoon, respectively” cannot be gotten from Figure 7f. It likely to be 20:00 and 18:00 LT under clean and polluted conditions, respectively.
9. Line 429: I think they are 14:00 and 16:00 in NCP from Figure 8a.
10. Lines 527-529: Please clearly indicate which figure and which situation are aimed at.

This manuscript reports on the distinct impacts on precipitation start/peak time by aerosol radiative effect over three different megacity regions of eastern China, which is found mainly caused by the different aerosol concentration and types over the three regions. The manuscript argues that the precipitation start time is 3 hours advanced in North China Plain due to high proportion of absorbing aerosol, 2 hours delayed in Pearl River Delta due to high proportion of scattering aerosol and negligible changed in Yangtze River Delta. The authors found that the period with the most occurrence frequency of precipitation start time is delayed and prolonged by aerosols over North China Plain, and discussed the response to precipitation to aerosol under different meteorological conditions. With the very interesting and valuable findings that include but are not limited to the parts I mention here, I believe this study is very important contribution to the science community regarding the aerosol-precipitation interaction.

Some minor comments

Line 60: “with the increase of the aerosol” should be “with the increase of aerosol”

Line 86-87: “in the initial stage” should be “at the initial stage”, “in the development stage” should be “at the development stage”

Line 114-115: I would suggest adding a reference for topographic rain effect.

Line 134-135: I would suggest changing the description to “at a vertical interval of 125 meters”

Lines 140-141: Please provide a brief description about the method to classify the convective, stratiform, and other precipitation types.

Line 174-181: The authors attempt to find suitable indicator as a proxy for CCN and they select 4-hours mean PM2.5 mass concentration before precipitation to investigate the impact of aerosols on precipitation. Why do not the authors choose 5-hours mean PM2.5 mass concentration before precipitation or the PM2.5 mass concentration during the precipitation to represent the CCN?

Line 187: I would suggest changing the description to “The low troposphere stability (LTS) can ...” to define LTS.

Line 199-200: I would suggest changing “have contributed to” to “have been used by”

Line 203: I would suggest changing “on different pressure levels” to “at different pressure levels”

Line 220-222: It seems the description here is wrong. I believe the correct description should be “Second, we rank the PM2.5 mass concentration observations from high to low, and define the top 1/3 of group C as polluted condition and the bottom 1/3 group C as clean condition.”

Line 240: “The diurnal variations” should be “The diurnal variation”

Line 250: I would suggest changing “make” to “making”

Line 288: I would suggest changing the description here from “the PDFs of the precipitation duration time and when the peak time occurs after start time” to “the PDFs of the precipitation duration time and the time difference between precipitation peak and start time”

Line 416: I would suggest adding “that” after “show” here.

Line 531: “are” should be “is”, corresponding to “response”.

Line 590: “which are essential for improve our understanding” should be “which are essential to improve our undesrtanding”